Clinical Gastroenterology2008!06!03

www.cghjournal.
org
March 2008 Volume 6
Number 3
Issue Highlights Photodynamic Therapy for Cholangiocarcinoma Chronic Abdominal Pain in Adolescents Prognostic Factors in Acute Liver Failure Supplementation Versus Dietary Counseling in Chronic Pancreatitis
Ofcial Clinical Practice Journal of the
Image of the Month

Metastatic Malignant Melanoma of the Gastrointestinal Tract
KEELY R. PARISIAN,* JOEL E. MCFARLAND, and ASHOK N. SHAH
*Department of Internal Medicine, and Department of Gastroenterology, University of Rochester Medical Center, Rochester, New York
75-year-old man with a history of malignant melanoma presented with fatigue and dyspnea for 4 months and melena in the setting of increased nonsteroidal anti-inammatory drug use over the past 7 days for knee pain. The patient rst was found to have malignant melanoma of the right upper cheek and the right and left auricle in 2003, with a second occurrence in 2005 on the posterior neck diagnosed on excision as Clark level IV with a Breslow depth of 6 mm. Sentinel lymph node mapping and a positron emission tomography scan were negative for evidence of metastatic disease. He presented to our institution after an episode of presyncope. A review of systems was signicant only for anorexia and a weight loss of 20 pounds over 3 months. Relevant laboratory values on admission revealed a decreased hematocrit from 42% to 23% over 1 month, and an international normalized ratio of 2.5 while on anticoagulation for atrial brillation. Before esophagogastroduodenoscopy the patient received 4 units of fresh-frozen plasma and 5 units of packed red blood cells. Esophagogastroduodenoscopy revealed numerous ulcerations including a pigmented, umbilicated, polypoid ulcer in the body of the stomach (Figure A), as well as a protuberant, ulcerated, and pigmented mass in the bulb of the duodenum (Figure B). Histology of the duodenal biopsy specimen showed melanocytes throughout and nuclei with differences in size, shape, and staining intensity consistent with malignant melanoma (Supplementary Figure; supplementary material available online at www.cghjournal.org). Immunohistochemical stains showed that the neoplastic cells were strongly reactive for Melan-A and HMB-45, but were not reactive for S-100. Staging computerized tomography of the abdomen with contrast showed subcutaneous nodules and a nodule adjacent to the duodenal bulb consistent with esophagogastroduodenoscopy ndings in addition to multiple bilateral pulmonary nodules, bilateral large adrenal masses, and multiple liver lesions compatible with metastatic disease. There was no further evidence of bleeding and the patient was discharged home with anticipation of beginning treatment with chemotherapy or biologic agents after consultation with the oncology department. The risk of rebleeding outweighed the benet of maintaining anticoagulation for atrial brillation and therefore Warfarin was discontinued at the time of discharge. Malignant melanoma of the gastrointestinal tract is a rare neoplasm. Malignant melanoma of the skin is one of the most common malignancies to metastasize to the gastrointestinal tract.1 Studies have reported that
autopsies show gastrointestinal tract involvement in approximately 50% to 60% of patients with malignant melanoma; however, clinical diagnosis is made in only 4.7% of all patients with malignant melanoma.2 Discouragingly, many patients with resected metastatic melanoma will present with disease relapse years later. Several studies have reported relief of symptoms and prolonged survival with surgical resection in patients with localized or moderate spread of disease. Patients with extensive metastasis and poor performance status are considered inoperable, as in the case that we have presented here. Further, extraintestinal manifestations of disease have been shown in 50% of patients who present with gastrointestinal tract involvement.3 As a result, persistent vague abdominal pain, fatigue, melena, and anemia in patients with a history of malignant melanoma should prompt suspicion of metastasis to the gastrointestinal tract and aggressive diagnostic evaluation at the onset.
Supplementary Data
Note: To access the supplementary material accompanying this article, visit the online version of Clinical Gastroenterology and Hepatology at www.cghjournal.org. References
1. Liang KV, Sanderson SO, Nowakowski GS, et al. Metastatic malignant melanoma of the gastrointestinal tract. Mayo Clin Proc 2006;81:511516. 2. Panagiotou I, Brountzos EN, Bafaloukos D, et al. Malignant melanoma metastatic to the gastrointestinal tract. Melanoma Res 2002;12:169 173. 3. Schuchter LM, Green R, Fraker D. Primary and metastatic diseases in malignant melanoma of the gastrointestinal tract. Curr Opin Oncol 2000;12:181185.
The authors would like to give special thanks to Zhenhong Qu, MD, and Ellen Giampoli, MD, of the University of Rochester Medical Center for providing pathologic images. 2008 by the AGA Institute 1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.010 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2008;6:xxiv
xxiv.e1
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 6, No. 3
Supplementary gure.
Image of the Month

Transmural Perforation of the Stomach by a Fishbone
IGOR J. NASTASKIN, KANAT RANSIBRAHMANAKUL, and WALTER TRUDEAU
Division of Gastroenterology, University of California Davis Medical Center, Sacramento, California
formed and a sh bone penetrating the stomach wall (Figure B, arrow) with surrounding edema and ulcerations was noted. This was removed with a snare (Figure C). The patient was observed overnight and discharged home without complication. More than 80% of foreign bodies that reach the stomach will pass unimpeded through the gastrointestinal tract.1 However, those thicker than 2 cm and longer than 5 cm tend to lodge in the stomach.2 Objects that lodge in the gastric lumen can remain there for prolonged periods without adverse consequences.3 Complications of foreign body ingestion include obstruction (occurring particularly at the pyloric sphincter and ileocecal valve), perforation, and hemorrhage. Watchful waiting generally is justied and may include administration of emetics, laxatives, or spasmolytics, depending on the type and site of object. However, early endoscopic or surgical removal of large foreign bodies from the stomach and of objects causing complications is recommended.4
73-year-old Russian woman with a history of poorly controlled hypertension presented to the emergency department with a 2-day history of epigastric abdominal pain. One day before her visit she ate a soup called uha, which is made from sh. Four hours later she developed sharp epigastric pain radiating to the back. She denied fevers, nausea, vomiting, or diarrhea. The pain was not associated with food intake and there were no factors that improved or worsened her pain. Her physical examination was unremarkable except for mild epigastric tenderness to palpation without peritoneal signs. Basic laboratory test results were normal. Computerized tomography of the abdomen showed a radiodense linear foreign body measuring 3 cm, extending transmurally through the anterior wall of the stomach, associated with mesenteric stranding, and no free air (Figure A). An esophagogastroduodenoscopy was per-
References
1. Selivanov V, Sheldon GF, Cello JP, et al. Management of foreign body ingestion. Ann Surg 1984;199:187191. 2. Koch H. Operative endoscopy. Gastrointest Endosc 1977;24:65 68. 3. Roark GD, Subramanyam K, Patterson M. Ingested foreign material in mentally disturbed patients. South Med J 1983;76:1125 1127. 4. Velitchkov NG, Grigorov GI, Losanoff JE, et al. Ingested foreign bodies of the gastrointestinal tract: retrospective analysis of 542 cases. World J Surg 1996;20:10011005.
2008 by the AGA Institute
1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.024 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2008;6:xxvi
Image of the Month

Hairball in the Stomach: A Case of Gastric Trichobezoar
KOLA HISAMUDDIN* and CHRISTOPHER P. BRANDT
*Division of Gastroenterology and Department of General Surgery, MetroHealth Medical Center, Case Western Reserve University, Cleveland, Ohio
43-year-old woman presented to the gastroenterology clinic with symptoms of upper-abdominal discomfort, early satiety, and the sensation of a ball rolling within the abdomen for several months. She initially had consulted her obstetrician for these symptoms with concerns that she could be pregnant. The patient denied any history of gastrointestinal symptoms or abdominal surgery before the onset of her current symptoms. The abdominal examination was remarkable for a palpable mass in the epigastrium. A computerized tomography scan of the abdomen (Figure A) showed a foreign body (A) in the stomach. At laparotomy (Figure B), a large mass occupying the entire stomach and extending into the pylorus was noted. A longitudinal gastrostomy was performed and a large foreign body was removed (Figure B). This was a trichobezoar measuring 840 g that was removed intact from the stomach (Supplementary gure; see supplementary material online at www.cghjournal.org) without any complications. An upper-gastrointestinal endoscopy was not performed before surgery because the patients symptoms and CT of abdomen was suggestive of a foreign body in the stomach in keeping with a bezoar. Furthermore, there was some suggestion of the patient eating her hair in the past. However, it became more obvious during the follow-up evaluation that she had a history of trichotillomania and trichophagia that was most pronounced while she was pregnant 4 years before her initial presentation. The patient then was referred for a psychiatric assessment. The word bezoar is derived from the Persian language, which means protection from the poison. Historically, bezoars were believed to have the power of a universal antidote against any poison. Gastric bezoars are the end result of the accumulation of undigested, ingested foreign material that form masses in the stomach. They are classied as trichobezoar (hair), phytobezoar (fruits and vegetables), or pharmacobezoar (drugs) on the basis of composition. Bezoars are encountered in less than 1% of patients undergoing upper endoscopy.1 Trichobezoars develop in patients with underlying psychiatric problems, trichophagia, trichotillomania, and gastric surgery. Delayed gastric emptying was considered
as a predisposing factor, however, a study revealed no evidence of gastroparesis in patients with bezoars.2 These patients present with nonspecic symptoms such as abdominal pain, nausea, vomiting, early satiety, anorexia, and weight loss. Nonetheless, they usually are diagnosed as incidental nding. Patients can present with gastrointestinal bleeding related to the ulcers, which may be caused by peptic ulcer disease or pressure necrosis. Diagnosis is made by radiologic means such as plain radiograph, barium studies, and computerized tomography scan or upper-gastrointestinal endoscopy. Trichobezoars usually need to be removed either by endoscopy or surgery. Attempts to dissolve them with enzymes such as cellulose and papain usually are unsuccessful. Bezoars have a tendency to recur in up to 14% of patients.3 Therefore, once the bezoar is removed by whichever means, it is important to treat the underlying or predisposing condition. These include increased water intake, to alter the diet appropriately (eg, avoid persimmons and stringy vegetables) in case of phytobezoars, to chew food carefully, and to seek psychiatric evaluation for trichobezoars.
Supplementary Data
Note: To access the supplementary material accompanying this article, visit the online version of Clinical Gastroenterology and Hepatology at www.cghjournal.org. References
1. Kadian RS, Rose JF, Mann NS. Gastric bezoars: spontaneous resolution. Am J Gastroenterol 1978;70:79 82. 2. Calabuig R, Navarro S, Carrio I, et al. Gastric emptying and bezoars. Am J Surg 1989;157:287290. 3. Robles R, Parrilla C, Escamilla J, et al. Gastrointestinal bezoars. Br J Surg 1994;81:1000 1001.
1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.013 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2008;6:xxviii
xxviii.e1
IMAGE OF THE MONTH
Supplementary gure.
Image of the Month

Gastric Outlet Obstruction Caused by a Large Gallstone in the Duodenum (Bouverets Syndrome)
KATSUNARI KISHI, KAZUKI YAMADA, and TOSHIRO SUGIYAMA
Department of Gastroenterology, University of Toyama, Toyama, Japan
58-year-old woman presented with a 7-day history of sudden onset of severe vomiting followed by abdominal distention and an inability to tolerate oral intake beyond small sips of water. Computerized tomography examination showed pneumobilia, cholecystoduodenal stula, and a 4.5 3.5 cm impacted gallstone in the duodenal bulb. Endoscopic examination of the upper-gastrointestinal tract showed pyloric stenosis caused by the impacted stone (Figure A). The stulous stoma was visualized in the posterior wall of the duodenal bulb (Figure B, white arrow). An upper-gastrointestinal series showed a stula with the gallbladder and the common bile duct (Figure C, thin arrow). These ndings conrmed the diagnosis of an impacted gallstone (Figure C, wide arrow) in the duodenal bulb as the cause of gastric outlet obstruction. Because the endoscopic stone retrieval and transendoscopic shock-wave litho-
tripsy was unsuccessful, surgical lithotomy consisting of simple enterotomy was performed. Bouverets syndrome is a gastric outlet obstruction caused by a large gallstone impacted in the duodenal bulb, and was rst described by Bonnet in 1841 and Bouveret in 1896. It almost always presents with abdominal pain and vomiting and occurs most commonly in elderly women with a mean age of 68.6 years.1 Endoscopic retrieval has a low success rate (10%). Enterolithotomy or gastrotomy, with or without cholecystectomy and stula repair, are the most common surgical therapies.2 References
1. Frattaroli FM, Reggio D, Guadalaxara A, et al. Bouverets syndrome: case report and review of the literature. Hepatogastroenterology 1997;44:1019 1022. 2. Cappell MS, Davis M. Characterization of Bouverets syndrome: a comprehensive review of 128 cases. Am J Gastroenterol 2006; 101:2139 2146.
1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.027 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2008;6:e11
Image of the Month

Mirizzi With Pre-Bouverets Syndrome
ALI SHAKOURI and SHOUJIANG TANG
Department of Digestive and Liver Diseases, UT Southwestern Medical Center, Dallas, Texas
35-year-old woman was admitted with a 3-week history of recurrent biliary colic, jaundice, nausea, and nonbilious emesis after meals. Her serum bilirubin was 18 mg/dL (normal, 0.21.3 mg/dL), and her alkaline phosphatase was 1436 units/L (normal, 38 126 units/L). A right upper quadrant ultrasound showed common bile duct dilatation with evidence of cholelithiasis and masslike structure in the gallbladder fossa. Murphys sign was negative. A computed tomography scan of the abdomen/pelvis showed intrahepatic duct dilatation and common bile duct (CBD) dilatation up to 2.5 cm, just before the head of the pancreas (Figure A). There was a large hypodense lesion at this level with a hyperattenuating rim (Figure A, arrow), with abrupt duct narrowing thereafter. There was no pancreatic head mass. On endoscopic retrograde cholangiopancreatography (ERCP), a 3- to 4-cm lling defect was seen coming from the cystic duct and causing obstruction within the CBD, ie, Mirizzi syndrome (Figure B). On endoscopy, there was a large protruding bulge at the lateral wall in the rst portion of the duodenum (Figure C), corresponding to the large gallstone seen on cholangiogram. On uoroscopy, the tip of the endoscope was touching the gallstone when we placed tip of the endoscope on the duodenal bulge. The large gallstone moved appropriately on gentle poking of the duodenal bulge with the tip of the endoscope. After biliary sphincterotomy, a 10F 9-cm biliary stent was placed with good drainage. During cholecystectomy, a 4-cm gallstone was removed from the CBD, and there was almost complete obliteration of the cystic duct with formation of a stulous tract from the gallbladder to the CBD. Mirizzi syndrome is a rare cause of obstructive jaundice produced by the impaction of a gallstone either in the cystic duct or in the gallbladder, resulting in stenosis of the extrahepatic bile duct.1,2 Although she had Mirrizi syndrome, the patients presentation was also concerning for Bouverets syndrome, which describes gastric outlet obstruction caused by impacted gallstone in the pylorus or duodenum.3 A single gallstone of at least 2.5 cm in diameter is the most common underlying cause of Bouverets syndrome. The symptoms are epigastric pain, nausea, and emesis. Jaundice is an uncommon physical nding with Bouverets syndrome and is more commonly seen with Mirizzi syndrome. In patients with Mirizzi syndrome, a preoperative diagnosis and therapy by ERCP reduce the risk of biliary damage during surgery. Although enterotomy or gastrotomy with or without cholecystectomy and stula repair is the most common surgical therapy, there are reported cases of successful endoscopic treatment of Bouverets syndrome with intracorporeal electrohydraulic lithotripsy and mechanical lithotripsy.
1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.006 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2008;6:e12 e13
March 2008
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY
e13
References
1. Lai EC, Lau WY. Mirizzi syndrome: history, present and future development (review). ANZ J Surg 2006;76:251257. 2. Abou-Saif A, Al-Kawas FH. Complications of gallstone disease:
Mirizzi syndrome, cholecystocholedochal stula, and gallstone ileus (review). Am J Gastroenterol 2002;97:249 254. 3. Cappell MS, Davis M. Characterization of Bouverets syndrome: a comprehensive review of 128 cases (review). Am J Gastroenterol 2006;101:2139 2146.
Image of the Month

Bouverets Syndrome: Diagnosis and Endoscopic Treatment
LINCOLN E. V. V. C. FERREIRA, MARK D. TOPAZIAN, and TODD H. BARON
Department of Medicine, Division of Gastroenterology & Hepatology, Mayo Clinic, Rochester, Minnesota
n 82-year-old woman was admitted with a 1-day history of epigastric abdominal pain accompanied by nausea and vomiting. She denied fever, chills, or melena. Serum amylase and lipase levels were increased markedly. Abdominal computerized tomography scan showed a 3.6-cm lamellated gallstone lodged in the second portion of the duodenum (Figure A). The pancreatic duct as well as the extrahepatic and intrahepatic bile ducts were dilated, with air in the biliary tree. Free air was present in the right pericolic gutter. Abdominal exploration was undertaken, but signicant inammation prevented denitive intervention. Esophagogastroduodenoscopy revealed a large green pigmented stone impacted at the junction of the rst and second duodenum (Figure B). Endoscopic electrohydraulic lithotripsy was performed, excavating a shaft through the center of the stone. The stone was fragmented further with an endoscopic dilating balloon, additional endoscopic electrohydraulic lithotripsy, and mechanical lithotripsy, and the duodenal obstruction was relieved completely. A large cholecystoduodenal stula was seen (supplementary video; see supplementary material online at www.cghjournal.org). A subsequent endoscopic retrograde cholangiopancreatography revealed an additional bile duct stone that was removed after biliary sphincterotomy. The patient recovered fully and 5 months later remains asymptomatic. Bouverets syndrome is a rare condition dened as gastric outlet obstruction by a gallstone. Enterotomy or gastrotomy
with or without cholecystectomy and stula repair is the most common surgical therapy. Endoscopy therapy is increasingly common and different modalities of endoscopic treatment have been reported.13 Endoscopic therapy for large stones such as this one facilitate minimally invasive therapy.
Supplementary Data
Note: to access the supplementary material accompanying this article, visit the online version of Clinical Gastroenterology and Hepatology at www.cghjournal.org. References
1. Huebner ES, DuBois S, Lee SD, et al. Successful endoscopic treatment of Bouverets syndrome with intracorporeal electrohydraulic lithotripsy. Gastrointest Endosc 2007;66:183184. 2. Polistena A, Santi F, Tiberi R, et al. Endoscopic treatment of Bouverets syndrome. Gastrointest Endosc 2007;65:704 706. 3. Gemmel C, Weickert U, Eickhoff A, et al. Successful treatment of gallstone ileus (Bouverets syndrome) by using extracorporal shock wave lithotripsy and argon plasma coagulation. Gastrointest Endosc 2007;65:173175.
1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.055 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2008;6:e15
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2008;6:e17
ABSTRACTS FROM AROUND THE WORLD

Click on citations to link to these articles and additional papers of interest.
Additional Papers of Interest

Sagnelli E, Stroffolini T, Mele A, et al, and the Italian Hospitals Collaborating Group. Chronic hepatitis B in Italy: new features of an old diseaseapproaching the Universal prevalence of hepatitis B e antigen-negative cases and the eradication of hepatitis D infection. Clin Infect Dis 2008;46:110 113. Abraides JG, Villanueva C, Baares R, et al. Hepatic venous pressure gradient and prognosis in patients with acute variceal bleeding treated with pharmacologic and endoscopic therapy. J Hepatol 2008;48:229 236.
Del Piano M, Ballar M, Carmagnola S, et al. DPEJ placement in cases of PEG insertion failure. Dig Liv Dis 2008:40:140 143. Biagi F, Campanella J, Bianchi PI, et al. The incidence of coeliac disease in adult rst degree relatives. Dig Liv Dis 2008;40:97100. Veijola L, Oksanen A, Linnala A, et al. Persisting chronic gastritis and elevated Helicobacter pylori antibodies after successful eradication therapy. Helicobacter 2007;12:605 608. Allen KJ, Gurrin LC, Constantine CC, et al. Ironoverload-related disease in HFE hereditary hemochromatosis. N Engl J Med 2008;358:221230.
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2008;6:261262
CME ActivitiesExams 1 and 2

Miguel R. Arguedas, MD, Editor, CME Section
CME Credits:
The American Gastroenterological Association (AGA) Institute is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. The AGA Institute designates this educational activity for a maximum of 1.0 AMA PRA Category 1 Credit(s). Physicians should only claim credit commensurate with the extent of their participation in the activity.
Faculty Disclosure:
In accordance with the Accreditation Council for Continuing Medical Educations Standards for Commercial Support of Continuing Medical Education, all faculty and planning partners must disclose any relevant nancial relationship(s) or other relationship(s) held within the past 12 months. The AGA Institute implements a mechanism to identify and resolve all conicts of interest prior to delivering the educational activity to learners.
Instructions:
Category 1 credit can be earned by reading the relevant articles and taking these CME examinations online at http://www.cghjournal .org/content/cme. Answers can be obtained online after completing the exam(s).
Objectives:
Upon completion of these activities, participants should be able to demonstrate an increase in or afrmation of their knowledge of clinical medicine and evaluate the appropriateness of the clinical information as it applies to the provision of patient care.
262
CME ACTIVITIES
Exam 1: Pancreatic Cancer: A Review of the Evidence on Causation

Hart AR, et al, Authors
Test ID No.: 0053
Contact hours: 1.0
Expiration Date: March 31, 2009
Question 1:
Smoking accounts for approximately what percentage of pancreatic cancer cases? a. b. c. d. 10% 25% 50% 75%
Question 2:
All of the following risk factors and exposures have been positively associated with pancreatic cancer, except: a. b. c. d. High folate consumption History of cholecystectomy Helicobacter pylori infection Asthma
Question 3:
All of the following genetic conditions are associated with an increased risk of pancreatic cancer, except: a. b. c. d. Familial adenomatous polyposis Ataxia-telangiectasia Peutz-Jeghers syndrome Tylosis
Question 4:
Which of the following statements regarding Helicobacter pylori and pancreatic cancer is correct? a. b. c. d. H H H H pylori-induced gastritis reduces absorption of vitamin C. pylori infection results in decreased levels of gastrin. pylori infection results in decreased levels of secretin. pylori-induced increase in IL-1 levels.
Exam 2: What Is the Role of Serological Testing in Celiac Disease? A Prospective, Biopsy-Conrmed Study With Economic Analysis
Hopper AD, et al, Authors
Test ID No.: 0054
Contact hours: 1.0
Expiration Date: March 31, 2009
Question 1:
According to the present study, which strategy is associated with the highest yield of nding a case of celiac disease on endoscopy? a. b. c. d. tTG-positive EMA-positive Either tTG or EMA-positive TTG-positive, followed by EMA-positive
Question 2:
A high TTG titer is seen in the following conditions, except: a. b. c. d. SLE Advanced heart failure Diabetes mellitus Chronic liver disease
Question 3:
According to the results by Hopper et al, what percentage of patients on a gluten-free diet and with villous atrophy on follow-up biopsies have tTG levels 15U/mL? a. b. c. d. 10% 20% 40% 80%
Question 4:
According to the authors, which factor may affect the results of studies that measure the sensitivity and specicity of tTG and EMA, except? a. b. c. d. Prevalence of celiac disease Denition and ascertainment of the gold standard Study design (prospective versus retrospective) The use of automated assays
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2008;6:264 265

Visit CGH online at www.cghjournal.org to link to these articles and additional papers of interest.
Your Mom Was RightEat Your Fruits and Vegetables

Millen AE, Subar AF, Graubard BI, et al, and Ziegler RG for the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial Project Team. Fruit and vegetable intake and prevalence of colorectal adenoma in a cancer screening trial. Am J Clin Nutr 2007;86:1754 1764.
Summary. There continues to be intense interest in the relationship between diet and colorectal neoplasia. This study uses data from the prostate, lung, colorectal, and ovarian cancer screening trial (PLCO) where 3057 patients with prevalent adenoma were compared with 29,413 controls. All patients underwent sigmoidoscopy and if a polyp was found a full colonoscopy was recommended. A food frequency questionnaire was used to quantify fruit and vegetable intake 12 months before study entry. When comparing the lowest versus highest quintile of fruit intake, the risk of distal adenoma was less statistically signicant in the highest quintile. An inverse association between adenoma and total fruit intake was also detected, regardless of adenoma number and pathology. Total vegetable intake was not different based on adenoma although there were some differences based on the presence of specic vegetable types. A diet rich in fruit and deep yellow vegetables, dark green vegetables, and onions and garlic are modestly associated with a reduced risk of colorectal adenoma. Editors comment. As with all studies examining diet, one has to question the amount of food source necessary for consumption to achieve the end point to see if it is realistic. Additionally, how long has the food stuff been ingested prior to the determination of the endpoint? For the highest quintile, patients reported approximately 5 pyramid servings per day. This amount would likely far exceed what most of us consume. It seems difcult to believe that 1 year of such a diet prevents adenomas; it is more likely that such a diet, and perhaps other healthy life styles, were ongoing for some time. Nevertheless, given the other potential health benets of such a diet, even if you dont consume these food stuffs in the highest quintile and prevent an adenoma, maybe your mother is still correct. ............................................................
and remove adenomas which may progress to carcinoma. Little is known about the yield of capsule endoscopy (CE) in screening. Twenty-three patients with FAP underwent CE. Jejunal / ileal polyps were found in 7 patients (30%); duodenal polyps, however, were only seen in 4 of 11 (36%) and the ampulla of vater was not found. Proximal jejunal polyps detected by CE were conrmed and removed by push enteroscopy (PE) in 4 of 23 (23%) patients, and all removed polyps were tubular adenomas. No complications arose. Editors comment. In FAP, duodenal adenomas, particularly those involving the ampulla are at highest risk for development of carcinoma. Because these patients have already undergone colectomy, the distal small bowel can be easily surveyed at the time of surveillance ileoscopy. Therefore, the role for capsule endoscopy is likely limited in FAP. This study, while showing that more distal polyps can be identied, supports the continued use of side viewing endoscopy to examine the ampulla and push enteroscopy to survey for proximal jejunal lesions. ............................................................
Take Your Probiotic Before Your Antibiotic

Hickson M, DSouza AL, Muthu N, et al. Use of probiotic Lactobacillus preparation to prevent diarrhea associated with antibiotics: randomized double blind placebo controlled trial. BMJ 2007;335:80 85.
Summary. Diarrhea is a common accompaniment of broad spectrum antibiotic use leading to potential morbidity and precipitation of Clostridium difcile colitis. This randomized double-blind placebo-controlled trial studied 135 hospitalized patients receiving antibiotics. Patients consumed a 100 g cocktail containing Lactobacillus casei, Lactobacillus bulgaricus, and Streptococcus thermophilus twice daily for 1 week. Twelve percent of the probiotic group developed diarrhea compared with 34% in the placebo group (P.007). C difcile diarrhea was identied in 17% of the placebo group compared with 0 in the probiotic group (P.001) with an absolute risk reduction of 17% and the number needed to treat of 6. Editors comment. This well done trial suggests that a probiotic cocktail taken with broad-spectrum antibiotics in a hospital setting not only may reduce diarrhea but potentially decreases the likelihood of C difcile diarrhea which has become epidemic and more difcult to treat in the hospital environment. Further study is warranted to conrm these ndings, given the widespread implications. Until such data is available, it may be reasonable to adopt such a strategy.
Capsule Endoscopy Is Not the Answer for Surveillance in FAP

Iaquinto G, Fornasarig M, Quaia M, et al. Capsule endoscopy is useful and safe for small-bowel surveillance in familial adenomatous polyposis. Gastrointest Endosc 2008;67:61 67.
Summary. Screening of the small bowel is recommended in familial adenomatous polyposis (FAP) to detect
March 2008
265
............................................................
............................................................
Our Expanding Arsenal Against Hepatitis B Virus Infection

Chan HLY, Heathcote J, Marcellin P, et al, and The 018 Study Group. Treatment of hepatitis B e antigen-positive chronic hepatitis with telbivudine or adefovir. A randomized trial. Ann Intern Med 2007;147:745754. Lai CL, Gane E, Liaw YF, et al, and Brown NA for the Globe Study Group. Telbivudine versus lamivudine in patients with chronic hepatitis B. N Engl J Med 2007;357:2576 2588. Lampertico P, Vigan M, Manenti E, et al. Low resistance to adefovir combined with lamivudine: a 3-year study of 145 lamivudine-resistant hepatitis B patients. Gastroenterology 2007; 133:14451451.
Forget the Cookbook: Individualize Treatment Duration for Hepatitis C Genotype 1 Patients
Mangia A, Minerva N, Bacca D, et al. Individualized treatment duration for hepatitis C genotype 1 patients: a randomized controlled trial. Hepatology 2008;47:4350.
Summary. Hepatitis B virus (HBV) infection remains an extraordinarily important public health issue worldwide. Fortunately, our therapeutic armamentarium continues to expand, underscored by these recent publications. Chan et al compared 2 newer agents and showed that at 24 weeks of treatment, telbivudine had more consistent HBV DNA suppression than adefovir. At 52 weeks, patients who switched from adefovir to telbivudine also had a greater HBV DNA suppression; at week 24, 3 times as many telbivudine treated patients were HBV PCR negative. Lai et al compared telbivudine to lamivudine in chronic HBV infection with the primary endpoint in this noninferiority trial of a reduction in serum HBV DNA levels to fewer than 5 log 10 copies/mL. At 52 weeks, a signicantly higher proportion of HBeAg-positive patients receiving telbivudine, compared with those receiving lamivudine, had a therapeutic response (75% vs 67%) as well as histologic response (65% vs 56%). The mean time required for serum HBV DNA to become undetectable by PCR was also signicantly shorter in the telbivudine group, and overall, 60% vs 40% of the HBeAg-positive patients became PCR negative. Importantly, resistance developed in 5%, compared with 2.3% of the HBeAg-positive and HBeAg-negative patients who received telbivudine, compared with 11% and 10.7% who received lamivudine, respectively. Lastly, the report by Lampertico et al, suggest that for those receiving long-term combination adefovir and lamivudine, lamivudine resistant hepatitis B resistance is infrequent (4%) and the prevention of virologic and clinical breakthrough was durable. Editors comment. We continue to learn more about drug resistance with HBV treatment as we advance our therapies for HBV now using well tolerated oral agents. The results using telbivudine are encouraging and suggest that more widespread use of these agents will make a signicant impact in the treatment of HBV infection. Unfortunately many in the third world will not have the opportunity to receive these drugs nor may recognize they are infected. Indeed there are many similarities between the HIV and HBV epidemics. Further public health measures are warranted, and for diagnosis and prevention of HBV infection, the use of these drugs should be more rapidly expanded.
Summary. Interferon and ribavirin are prescribed currently as a standard 48-week treatment for HCV virus genotype 1 patients. This prospective randomized trial compared this standard regimen for 24, 48, or 72 weeks if HCV RNA was negative at weeks 4, 8 or 12, respectively (variable group). In the standard treatment group, 87.1%, 70.3%, and 38.1% of patients who rst achieved undetectable HCV RNA at 4, 8 or 12 weeks attained sustained virologic response (SVR), respectively. In the variable group, corresponding SVR rates were 77.2%, 71.9%, and 63.5%. Independent predictors of rapid virologic response (RVR) were HCV RNA levels 400,000 IU/mL and absence of advanced brosis. Patients with this RVR and baseline viremia 400,000 IU/mL achieved higher SVR when treated for 48 rather than 24 weeks (86.8 vs 73.1, P .14). Editors comment. This pragmatic study assesses therapy based on early response by persistent viremia. If there was persistent viremia (albeit at low levels) at 12 weeks, and one persisted with treatment to 48 weeks, an SVR still was achievable. Conversely, if there was a 2-log decline at 12 weeks, patients were considered nonresponders and treatment was discontinued. These results are similar to those who were viremic at 24 weeks. Using a patient directed duration of therapy, not only could one achieve better cure rates with longer duration of therapy, one could discontinue patients treatment earlier given persistent viremia. Thus, approximately a quarter of HCV genotype 1 patients could be cured by therapy in only 24 weeks and an approximately comparable rate of patients may require extended treatment for 72 weeks to enhance SVR. Such a tailored therapy would provide clear benet for patients given the cost, complexity, and potential side effects of current HCV therapy.
Additional Papers of Interest

Jeong SH, Lee DJ, Kim YB, et al. Diagnostic value of terminal ileum intubation during colonoscopy. J Gastroenterol Hepatol 2008;23:5155. Salem T, Molloy R, ODwyer P. Prospective, ve-year follow-up study of patients with symptomatic uncomplicated diverticular disease. Dis Colon Rectum 2007;50: 1460 1464. Perry CD, Hutter MM, Smith DB, et al. Survival and changes in comorbidities after bariatric surgery. Am J Ann Surg 2008;247:2127.
Editorial
Photodynamic Therapy: Standard of Care for Palliation of Cholangiocarcinoma?
alliation of malignant biliary obstruction in patients with unresectable hilar cholangiocarcinoma (CCA) is challenging for surgeons, interventional radiologists, endoscopists, and radiation oncologists alike. Compared with distal bile duct obstruction where palliation is relatively easily achieved operatively and non-operatively, hilar CCA poses particular difculties. The tumors involve bifurcation and spread progressively and inexorably in a proximal direction such that eventually too many branches of the intrahepatics are involved to permit stent placement to be effective. Indeed it, was Klatskin1 who described that these patients die of liver failure and cholangitis from biliary obstruction. Photodynamic therapy (PDT) is an ablative treatment for premalignant and malignant lesions. A photosensitizing drug is administered followed by application of a specic wavelength of light leading to intracellular activation of the drug and cellular injury.2 In addition, thrombosis of vessels and immune response may lead to tumor destruction.3 The wavelength of light necessary to cause activation is 630 nm and is provided by laser bers that emit light of that wavelength. When used in the esophagus, the bers are easily placed into the esophageal lumen because the endoscope is in a straightened position. The rst report of PDT for CCA was published more than 15 years ago.4 Since then there have been numerous studies using PDT for CCA both basic science57 and clinical. It is important to note that the patient selection, primary outcomes, types of bers and ber delivery vary between studies. It is also of importance to note that the passage of these bers through a duodenoscope poses special challenges, particularly in passage around the tight angle in the duodenum toward the bile duct. The landmark clinical study using PDT for CCA was published by Ortner et al, nearly 10 years ago.8 Patients with Bismuth type III and IV CCA who did not have a fall in bilirubin levels of at least 50% after bilateral stent placement were treated with PDT. Laser light was delivered cholangioscopically using small caliber bers. After treatment, a signicant fall in bilirubin levels was achieved, and a sustained improvement in indices of quality of life was noted. The same lead investigator then performed a randomized prospective trial of plastic biliary stents alone versus plastic biliary stents and PDT in a similar patient population that failed to have a fall in serum bilirubin levels following bilateral biliary stent placement.9 A signicant increase in survival was seen in the PDT group (median survival 493 days vs 98 days). PDT also improved biliary drainage and quality of life over stent placement alone. Other groups have published similar results.10 14 These studies from outside the US have used thin, exible, 400-m diameter bers. In addition, in some series, patients were re-treated 4 6 weeks later because of advanced Bismuth type IV tumors with residual segmental duct occlusions/stenoses or tumorpositive biopsies at a 1-month follow-up examination.10 There are limited publications using PDT for CCA from centers within the US. Indeed, nearly all reports have come from 1 center.1517 This group uses standard FDA-approved bers designed for use in the esophagus without the use of cholangioscopy. Fiber breakage occurs in about one third of patients. In this issue of Clinical Gastroenterology and Hepatology, Kahaleh et al,18 from Virginia publish their results using PDT for unresectable cholangiocarcinoma. Forty-eight patients were treated over a 5-year period. Of these, plastic biliary stents alone were placed in 29 patients; PDT and plastic biliary stents were placed in 19 patients. Laser delivery was achieved using standard rigid 2.5-cm bers that were preloaded into a 10F delivery system. One or 2 biliary segments were treated per session. Therapy was repeated every 3 months until death or withdrawal from the study. The degree of decline in bilirubin levels was similar in the 2 groups. KaplanMeier survival analysis showed a statistically signicant prolongation in survival in the PDT group (mean 16.2 2.4 months) compared with the stent only group (mean 7.4 1.6 months). The study, while one of the rst direct comparative survival studies from the US to show an improved survival using PDT, has a number of aws. It is a small, retrospective study. Although the majority of patients had Bismuth III and IV lesions, patients with Bismuth I and II lesions were included. Some patients in each group received chemoradiation therapy. With so many options available for the palliation of biliary obstruction in patients with inoperable hilar cholangiocarcinoma chemoradiation, brachytherapy, plastic and metal stents (endoscopically or percutaneously placed) what can we recommend for these patients? Is there enough data to say that PDT should be given to all patients? Lets start with the disadvantages of PDT. The treatment is not available at all centers. It requires expertise in both endoscopy and photodynamic therapy. The procedure is time consuming and can be quite prolonged based on number of segments treated; sufce it to say that one should allot at least 90 minutes for the procedure which entails stent removal, treatment, and stent replacement. The bers available in the US are suboptimal for endoscopic retrograde cholangiopancreatography use. They are stiff and prone to breakage. Because of the stiffness, treatment is generally limited to the main hepatic ducts since the ber does not bend around corners to reach intrahepatic branches. Finally, let us not forget the photosensitivity that occurs for 4 6 weeks after therapy which may limit quality of life. The advantages of PDT are that it is reasonably well tolerated and seems to be effective. It can be repeated without a ceiling dosage effect.10 PDT is the only treatment to date where there is evidence to support an improvement in survival over plastic stent placement alone for advanced CCA. In fact, one retrospective study suggested that the survival with PDT and stents was similar to those who underwent attempted curative, but incomplete surgical resection.19 Many questions about the role of PDT and CCA remain. In nearly all studies plastic biliary stents have been used. There is little experience with self expandable metal stents and PDT.13 The downside to metal stent placement is that light may not effectively reach the tumor through the interstices of the stent
March 2008
267
and/or be scattered during follow-up PDT, leading to inadequate treatment. Some authors, however, routinely combine metal stent placement and PDT, though it requires 10 cm parallel stents placed into the right and left systems following baseline PDT to have access to both sides in the future if bilateral PDT is repeated (Richard Kozarek, personal communication, November 2007). Most studies have used bilateral PDT and stent placement. Is this necessary? What about patients who have parenchymal atrophy on 1 side? Are we exposing those patients to a higher risk of cholangitis as compared to unilateral treatment? Do only patients with Bismuth III and IV lesions benet? Should type I and II patients be offered PDT? What are the effects of other treatments such as chemoradiation complementary or antagonistic? Does it make sense to treat patients with large masses when the treatment is relatively supercial? Finally, is cholangioscopic delivery of the laser light superior to uoroscopic delivery? In addition to time constraints and reimbursement, there are several reasons PDT availability in the US is limited. For endoscopists in the US to embrace PDT, smaller and more exible laser bers are needed. Furthermore, better delivery methods for the bers are also needed. For patients to embrace PDT, better photosensitizing agents are needed. Agents with a shorter duration of phototoxicity and more rapid onset (which would allow patients to receive the photosensitizing agent and treatment on the same day, rather than 48 hours apart) would be welcomed. Unfortunately, the number of patients with CCA is relatively small such that the commercial benet to companies is limited and thus there is not a nancial incentive for advancement in this area. So, should photodynamic therapy be considered standard of care for palliation of cholangiocarcinoma? The answer is a qualied yes. The data suggest that PDT is an excellent option for patients with unresectable CCA, especially for Bismuth III and IV lesions. However, there are no comparative trials with chemoradiation, and PDT availability is limited. Therefore, these patients could be managed with standard palliative care at their institution or referred to a specialized center with PDT availability. Further comparative trials are needed to determine the optimal regimen for palliation of obstructive jaundice in these patients.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
TODD H. BARON, MD Division of Gastroenterology and Hepatology Mayo Clinic Rochester, Minnesota
References
1. Klatskin G. Adenocarcinoma of the hepatic duct at its bifurcation within the porta hepatitis. An unusual tumor with distinctive clinical and pathological features. Am J Med 1965;38:241256. 2. Petersen BT, Chuttani R, Crofe J, et al. Photodynamic therapy for gastrointestinal disease. Gastrointest Endosc 2006;63:927 932. 3. Ortner MA, Dorta G. Technology insight: photodynamic therapy for
17.
18.
19.
cholangiocarcinoma. Nat Clin Pract Gastroenterol Hepatol 2006;3:459 467. McCaughan JS Jr, Mertens BF, Cho C, et al. Photodynamic therapy to treat tumors of the extrahepatic biliary ducts: a case report. Arch Surg 1991;126:111113. Wong Kee Song LM, Wang KK, Zinsmeister AR. Mono-L-aspartyl chlorin e6 (NPe6) and hematoporphyrin derivative (HpD) in photodynamic therapy administered to a human cholangiocarcinoma model. Cancer 1998;82:421 427. Kiesslich T, Berlanda J, Plaetzer K, et al. Comparative characterization of the efciency and cellular pharmacokinetics of Foscanand Foslip-based photodynamic treatment in human biliary tract cancer cell lines. Photochem Photobiol Sci 2007;6:619 627. Oertel M, Schastak SI, Tannapfel A, et al. Novel bacteriochlorine for high tissue-penetration: photodynamic properties in human biliary tract cancer cells in vitro and in a mouse tumour model. J Photochem Photobiol B 2003 15;71:110. Ortner MA, Liebetruth J, Schreiber S, et al. Photodynamic therapy of nonresectable cholangiocarcinoma. Gastroenterology 1998; 114:536 542. Ortner ME, Caca K, Berr F, et al. Successful photodynamic therapy for nonresectable cholangiocarcinoma: a randomized prospective study. Gastroenterology 2003;125:13551363. Berr F, Wiedmann M, Tannapfel A, et al. Photodynamic therapy for advanced bile duct cancer: evidence for improved palliation and extended survival. Hepatology 2000;31:291298. Wiedmann M, Berr F, Schiefke I, et al. Photodynamic therapy in patients with non-resectable hilar cholangiocarcinoma: 5-year follow-up of a prospective phase II study. Gastrointest Endosc 2004;60:68 75. Zoepf T, Jakobs R, Arnold JC, et al. Palliation of nonresectable bile duct cancer: improved survival after photodynamic therapy. Am J Gastroenterol 2005;100:2426 2430. Dumoulin FL, Gerhardt T, Fuchs S, et al. Phase II study of photodynamic therapy and metal stent as palliative treatment for nonresectable hilar cholangiocarcinoma. Gastrointest Endosc 2003;57:860 867. Zoepf T, Jakobs R, Arnold JC, et al. Photodynamic therapy for palliation of nonresectable bile duct cancerpreliminary results with a new diode laser system. Am J Gastroenterol 2001;96: 20932097. Rumalla A, Baron TH, Wang KK, et al. Endoscopic application of photodynamic therapy for cholangiocarcinoma. Gastrointest Endosc 2001;53:500 504. Harewood GC, Baron TH, Rumalla A, et al. Pilot study to assess patient outcomes following endoscopic application of photodynamic therapy for advanced cholangiocarcinoma. J Gastroenterol Hepatol 2005 Mar;20:415 420. Prasad GA, Wang KK, Baron TH, et al. Factors associated with increased survival after photodynamic therapy for cholangiocarcinoma. Clin Gastroenterol Hepatol 2007;5:743748. Kahaleh M, Mishra R, Shami VM, et al. Unresectable cholangiocarcinoma: Comparison of Survival in Biliary Stenting alone vs Stenting with Photodynamic Therapy. Clin Gatsroenterol Hepatol 2008;6:290 297. Witzigmann H, Berr F, Ringel U, et al. Surgical and palliative management and outcome in 184 patients with hilar cholangiocarcinoma: palliative photodynamic therapy plus stenting is comparable to r1/r2 resection. Ann Surg 2006;244:230 239.
doi:10.1016/j.cgh.2008.01.015
REVIEWS
Chronic Hepatitis B: Preventing, Detecting, and Managing Viral Resistance
EMMET B. KEEFFE,* DOUGLAS T. DIETERICH, JEANMICHEL PAWLOTSKY,, and YVES BENHAMOU
*Division of Gastroenterology and Hepatology, Department of Medicine, Stanford University School of Medicine, Stanford, California; Department of Medicine, The Mount Sinai Medical Center, New York, New York; French National Reference Center for Viral Hepatitis B, C and Delta and Virology Laboratory, Hpital Henri Mondor, Universit Paris 12, Crteil, France; INSERM U841, Crteil, France; and the Service dHpatologie Hpital Piti-Salptrire, Universit Paris 5, Paris, France
See Deng G et al on page 716 for companion article in the March 2008 issue of Gastroenterology.
Licensed oral agents for antiviral therapy in patients with chronic hepatitis B virus (HBV) infection include lamivudine, adefovir, entecavir, and telbivudine. Emtricitabine, tenofovir, and the combination of tenofovir plus emtricitabine in 1 tablet, which are licensed for the treatment of human immunodeciency virus infection, are additional off-label options for treating HBV infection. Preventing HBV antiviral drug resistance to nucleoside/nucleotide analogues and appropriate management when resistance occurs has become a major focus in the management of chronic hepatitis B. HBV antiviral drug resistance may be best prevented by using an agent or combination of agents with a high genetic barrier to resistance, and 2 potent nucleoside and nucleotide drugs with different resistance proles may prove to be the optimal rst-line treatment for chronic hepatitis B. Frequent assessment of quantitative serum HBV DNA remains the best approach to early detection of resistance, and antiviral therapy should be modied as soon as resistance is detected. Results from several clinical trials have shown that the addition or substitution of newer antiviral agents can restore suppression of viral replication, normalize alanine aminotransferase levels, and reverse histologic progression in patients with resistance to lamivudine, but little information exists regarding the longterm benets of second-line treatment regimens. Despite the substantial advances in treatment made to date, new agents with novel viral targets will be needed for patients who ultimately may fail second- or third-line therapy.
ertheless, because of a high HBV prevalence rate in many developing countries and population migrations to developed countries, chronic HBV infection is an important health problem in the United States and Europe. Antiviral therapy in patients with chronic hepatitis B is associated with improved outcomes.5,6 Over the past 10 years, oral antiviral agents available for patients with chronic HBV infection have included lamivudine, adefovir, entecavir, and telbivudine, which are licensed to treat chronic hepatitis B, and emtricitabine, tenofovir, and the combination of tenofovir plus emtricitabine, which has been licensed for treatment of patients with human immunodeciency virus (HIV) infection, but also is active against HBV infection.7 However, 20 years of experience with the treatment of HIV, as well as current HBV resistance data, point to the emergence of HBV antiviral drug resistance,8 as illustrated by the case of lamivudine, the nucleoside that has the longest history of use in patients with chronic HBV infection and the highest rate of resistance. The development of resistance is associated with a poorer long-term prognosis.9,10 Clinicians need to be aware of current approaches to resistance testing and of effective treatment strategies to minimize the emergence of resistant HBV strains, and they should be cognizant of regimens that may be effective after resistance has emerged.
Viral Replication and Selection of Resistant Strains

The genome of HBV consists of partially double-stranded, 3.2-kb, covalently closed, circular DNA (ccc DNA) comprising 4 overlapping open reading frames.11 The viral genome is transcribed into 4 major subgenomic viral messenger RNAs, under the control of specic enhancers,12 and is the template for the pregenomic RNA. The virus is encapsidated after binding of the polymerase and core to the pregenomic RNA in the cytoplasm. Nucleocapsids are enveloped by budding into the endoplasmic reticulum, after which they are secreted from the cell or return to the nucleus to amplify the cccDNA reservoir.13 Viral mutations occur spontaneously during HBV replication. Viral reverse transcriptases intrinsically are error prone and lack a
Abbreviations used in this paper: AASLD, American Association for the Study of Liver Diseases; ALT, alanine aminotransferase; HBeAg, hepatitis B e antigen; HBV, hepatitis B virus; HIV, human immunodeciency virus. 2008 by the AGA Institute 1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.043
t is estimated that 1.25 million people in the United States have chronic hepatitis B, and there were 60,000 new hepatitis B virus (HBV) infections in 2004.1 Acute HBV infection may be asymptomatic, result in self-limited or fulminant hepatitis, or progress to chronic hepatitis, which can lead to cirrhosis or hepatocellular carcinoma.2 Cirrhosis and hepatocellular carcinoma result in about 1 million deaths worldwide each year.3 In the United States, there has been a substantial decline in the incidence of acute hepatitis B over the past 15 to 20 years that has coincided with the national strategy to eliminate HBV transmission by vaccination and public health measures.4 Nev-
March 2008
CHRONIC HEPATITIS B AND ANTIVIRAL RESISTANCE
269
proofreading function, allowing for replication errors to occur. These replication errors result in the emergence of multiple HBV variant quasispecies that coexist and reach population densities in direct proportion to their relative replication tness.11 This phenomenon is responsible for the generation of signicant diversity; it has been shown that HBV genomes in one given patient displayed a rate of 1.4 to 3.2 105 nucleoside substitutions per year, a value approximately 104 times greater than DNA genomes and about 102 less than that for HIV.14 A chronic HBV carrier can produce up to 1013 virions per day, and, as a result, every nucleotide of the 3.2-kb HBV genome theoretically can be substituted within one patient every day.2 The dominant quasispecies is by denition the best adapted to its host environment, and, as expected, random mutations generally will impair its tness to a degree. Thus, the emergence of successful HBV variants resistant to an antiviral drug is affected by the mutation rate and viral load and is determined eventually by the replication tness of the mutant virus in relation to the antiviral potency of the drug and the number of mutations required to confer resistance (ie, the genetic barrier to resistance).15
Mutations in the Polymerase Gene and Antiviral Resistance

Mutations resulting in resistance to nucleoside/nucleotide analogues mainly involve the viral polymerase gene (Figure 1).11,2325 This gene contains 7 functional domains (AG), and mutations that give rise to nucleoside/nucleotide resistance are located essentially in domains A through E.23
Lamivudine Resistance
Lamivudine inhibits viral reverse-transcriptase activity as a competitive inhibitor of deoxycytidine triphosphate.23 High-level lamivudine resistance results from M204V and M204I mutations in the C domain.26 HBV variants with these mutations display reduced reverse-transcriptase activity and replication capacity, but compensatory mutations in the B domain (V173L and L180M) restore the replication capacity of virus with the M204V or M204I mutation.27,28 Results from 998 patients with HBeAg-positive compensated chronic hepatitis B who took lamivudine for up to 6 years indicated that the proportion of patients with documented lamivudine-resistant mutations increased from 23% in year 1 to 43%, 55%, 71%, and 65% in years 2 through 5, respectively. Patients with lamivudine-resistant mutations who were followed up for more than 4 years had signicantly higher risk for hepatic decompensation and liver diseaserelated severe adverse events.9
Mutations Associated With Resistance to Nucleoside/Nucleotide Analogues Dening the Consequences of Resistance
Treatment failure can be dened as primary or secondary. Primary treatment failure is the failure of a drug to reduce HBV DNA levels by 1 log10 IU/mL or greater within 3 months of initiation of therapy, and secondary failure is dened as a rebound of HBV replication by 1 log10 IU/mL or greater from nadir in patients in whom treatment initially produced a decrease in serum HBV DNA of 1 log10 IU/mL or greater.16 Secondary treatment failure due to resistance to antiviral therapy can result in a decreased rate of hepatitis B e antigen (HBeAg) seroconversion17; reversion of virologic, biochemical, and histologic improvement18,19; increased rate of disease progression5; severe exacerbations in the presence of cirrhosis20,21; and risk for graft loss and death after liver transplantation.22
Adefovir Resistance
Adefovir inhibits priming of reverse transcription by preventing the incorporation of deoxyadenosine triphosphate into the viral primer and inhibits viral-minus strand DNA elongation.29 Two mutations have been described that confer approximately 5- to 10-fold reduced susceptibility to adefovir in vitro: N236T in the D domain of viral polymerase and A181V in the B domain.23 Most recently, another rare mutation, I233V, has been suggested to result in primary resistance to adefovir, but that mutation has not been conrmed to be associated with a reduced susceptibility to adefovir in vitro.30 A study of adefovir in 125 patients who were followed up for up to 240 weeks indicated that the cumulative probabilities of mutations associated with a virologic breakthrough at 48, 96, 144, 192, and 240 weeks were 0%, 3%, 11%, 18%, and 29%, respectively. The respective values for mutations associated with virologic and biochemical breakthrough (alanine aminotransferase [ALT] increases) were 0%, 2%, 6%, 10%, and 11%.31,32
Entecavir Resistance
Entecavir is a deoxyguanosine analogue that is greater than 100-fold more potent against HBV in culture than either lamivudine or adefovir, and halts HBV DNA elongation after incorporation of a few bases more than lamivudine or adefovir.33 Analysis of results from 673 patients treated with entecavir indicated that 3% showed virologic rebound by 96 weeks of treatment. Three entecavir rebounds were attributable to lamivudine-resistant virus present at baseline, and none of the others was associated with either entecavir genotypic resistance or loss of entecavir susceptibility.34 After administration of entecavir to patients who were refractory to lamivudine, entecavir resistance was detected in 1% of patients after 48 weeks and an additional 9% after 96 weeks. Entecavir resistance was associated with lamivudine-resistance conferring mutations
Figure 1. Location of the major antiviral drug-resistant mutations associated with LMV, LdT, ADV, TDF, and ETV11 (Reprinted with permission). LMV, lamivudine; Ldt, telbuvidine; ADV, adefovir dipivoxil; TDF, tenofovir DF; ETV, entecavir. POL/RT, polymerase reverse-transcriptase.
270
KEEFFE ET AL
plus additional changes at T184, S202, and/or M250.35 Recent results have indicated that the cumulative probability of virologic breakthrough associated with entecavir resistance in nucleoside-naive patients was 0.8% over 4 years, and that for patients who received prior lamivudine therapy it was 39.5%, with 1% having resistant virus after 1 year, 10% after 2 years, 16% after 3 years, and 15% after 4 years of treatment.36
Telbivudine Resistance
Telbivudine targets the synthesis of positive-strand HBV DNA, which is thought to result in slower emergence of resistance than lamivudine, which targets negative-strand synthesis.37 Results from the GLOBE trial indicated that the 2-year cumulative rate of virologic breakthrough for HBeAg-positive patients was 21.6%, and for HBeAg-negative patients was 8.6%.38
Emtricitabine Resistance
Emtricitabine inhibits HBV polymerase in essentially the same manner as lamivudine, and the M204V mutation that confers resistance to lamivudine also results in resistance to emtricitabine.39,40 Emtricitabine is effective as monotherapy in patients with chronic HBV infection, reducing HBV DNA levels to less than 400 copies/mL in 54% of patients over 48 weeks.39 However, at 48 weeks, 9% to 16% of patients treated with 25 to 200 mg/day of emtricitabine developed resistance-conferring mutations (M204I or M204V with or without L180M or V173L).41
Figure 2. Serial changes in serum HBV DNA and ALT levels in association with emergence of antiviral-resistant HBV mutants. The rst manifestation of resistance is the detection of resistance-conferring mutations (ie, genotypic resistance).46 (Reprinted with permission). ULN, upper limit of normal.
Tenofovir Resistance
Tenofovir is an acyclic nucleotide analogue that retains signicant activity against HBV variants with mutations (M204V plus L180M), conferring resistance to lamivudine and N236T that results in resistance to adefovir.42,43 Tenofovir retains good activity in patients with lamivudine-resistant HBV variants, either as monotherapy or when added to continuing lamivudine treatment.44,45 HBV mutations that confer resistance to tenofovir have not yet been identied.
Detecting, Preventing, and Managing Hepatitis B Virus Resistance

Clinical evidence, as summarized earlier, indicates that resistance to antiviral monotherapy: (1) is common, particularly in patients treated with lamivudine; (2) increases with duration of treatment; and (3) is associated with poorer clinical outcomes. Many new antiviral agents, however, are available for initial therapy and as second-line treatment for those failing lamivudine. It is important that patients with lamivudine resistance be treated with agents that are not cross-resistant, such as the nucleotide analogues adefovir or tenofovir. The availability of genetic testing for identication of viral genotypes likely to show phenotypic resistance and the fact that different antiviral agents do not have completely overlapping resistance proles raises 3 questions: When should genotypic resistance testing be performed in patients with HBV infection? What approach(es) to initial treatment are least likely to result in the rapid emergence of resistance? What treatment alternatives are most useful in patients who have developed resistanceconferring mutations?
ical and/or clinical breakthrough. Phenotypic resistance is dened as virologic breakthrough (ie, an increase in serum HBV DNA by 1 log10 above nadir after achieving virologic response during continued treatment), or virologic rebound (ie, increase in serum HBV DNA to 20,000 IU/mL or to above pretreatment level after achieving virologic response during continued treatment). Biochemical relapse is characterized as an increase of ALT to above the upper limit of normal after achieving normalization with treatment and may be associated with a hepatitis are (Figure 2).46,47 Development of resistance can be detected readily by monitoring serum HBV DNA levels and is indicated by a greater than 1 log10 increase from the patients lowest level conrmed by measurement on 2 assays. Patients taking lamivudine should have HBV DNA levels evaluated with a sensitive assay every 3 to 6 months and those being treated with adefovir or entecavir should be monitored every 6 months after the rst year of treatment. Patients with advanced liver disease should be monitored every 3 months.47 Guidelines published by the American Association for the Study of Liver Diseases (AASLD) recommend serum HBV DNA monitoring every 3 to 6 months for patients receiving therapy for chronic hepatitis B.46 It also has been noted that monitoring the emergence of resistance-conferring mutations may be the most sensitive way to assess patients who remain viremic on current treatments.48
Assays for Detection of Resistance-Conferring Mutations

The reference method for detection of resistance-conferring mutations is population-based sequencing (ie, a direct sequence analysis of the HBV polymerase gene). Line-probe assays using probes for individual mutations are specic and reproducible and can detect a mutant representing as little as 5% of the viral population. However, this approach is limited in that it can detect only known mutations. It requires periodic updating with new probes specic to novel mutations because HBV strains resistant to newly developed antiviral drugs are isolated and characterized.49,50 Other powerful technologies are in development, including high-throughput systems capable of detecting polymorphisms in the entire HBV genome using gene chip technology.51 This method is limited by the fact that it, too, can detect only known polymorphisms and needs to be updated as resistance-conferring mutations are identied.50
Resistance Testing
Genotypic resistance precedes phenotypic resistance with virologic breakthrough, which in turn precedes biochem-
March 2008
271
Table 1. Treatment Recommendations of the AASLD and US Algorithm for Patients With Antiviral-Resistant HBV46,47
Resistance to LMV AASLD46 Add ADV or TDF; stop LMV and switch to FTC/TDF combination; stop LMV and switch to ETV (pre-existing LMV resistance mutations predispose to ETV resistance) Add LMV; stop ADV and switch to FTC/TDF combination; switch to or add ETV US Algorithm47 Add ADV (may be preferred over switch to ADV); switch to ETV (risk for subsequent ETV resistance); potential future management: add TDF or switch to FTC/TDF combination Add LMV (may be preferred over switch to LMV); switch to ETV (if no prior LMV resistance); potential future therapy: switch to FTC/TDF combination Add or switch to ADV or TDF
ADV
ETV LdT
Switch to or add ADV or TDF Add ADV or TDF; stop LdT and switch to FTC/TDF combination; stop LdT and switch to ETV (pre-existing LdT resistance mutations predispose to ETV resistance)
ADV, adefovir dipivoxil; ETV, entecavir; FTC, emtricitabine; LdT, telbivudine; LMV, lamivudine; TDF, tenofovir DF.
When to test. Clinicians might consider testing at the following times: (1) before selection of initial therapy, (2) routinely during the course of therapy, and (3) at the time of virologic breakthrough, before changing the therapeutic regimen. Evaluation of viral genetic mutations has been used extensively in patients with HIV infection, and it is useful to review current recommendations for HIV genotypic resistance testing. Testing before initiation of therapy. Current guidelines recommend HIV drug resistance testing before initiation of therapy.52 This recommendation is justied by repeated observations of transmission of viral strains resistant to one or more antiretroviral drugs53; however, it is much less clear whether resistance testing is necessary before the initiation of nucleoside/nucleotide antiviral therapy in patients with HBV infection. At this time, it is therefore reasonable to suggest that resistance testing before the initiation of nucleoside/nucleotide treatment is not warranted in patients with HBV infection unless they acquired HBV from an infected patient undergoing antiviral treatment. Testing during treatment. Although monitoring the emergence of resistance-conferring mutations may be the most sensitive way to monitor patients taking nucleosides/nucleotides for the treatment of HBV infection,48 this approach does not seem warranted on the basis of current information. Viral rebound occurs well before a biochemical breakthrough or hepatitis are,46 and monitoring of serum HBV DNA with a sensitive assay as recommended should permit clinicians to detect changes in viral load that signal loss of efcacy before increases in ALT levels or clinical deterioration. Notably, guidelines for management of HIV-infected patients make no recommendations for routine genotypic or phenotypic resistance testing in patients who maintain viral suppression during therapy,52 and the same can be said for HBV infection. Resistance testing at virologic breakthrough. HIV treatment guidelines recommend resistance testing before changing treatment regimens after virologic failure,52 and recommendations for the management of patients with HBV infection also recognize the importance of resistance testing when HBV DNA testing indicates a 1 log10 or greater increase in viral load. Resistance testing should be performed, if possible, before discontinuation of the failing regimen so that the results reect the latest detectable viral mutations. Results of resistance testing then can be used to guide selection of new treatment.46
Management of Patients With Resistant Virus

Both the AASLD guidelines and those published by Keeffe et al47 provide recommendations for changing therapy in patients with HBV antiviral drug resistance to lamivudine, adefovir, or entecavir. The AASLD guidelines also provide recommendations for new therapy in patients with resistance to telbivudine (Table 1).46,47 Clinical trials performed to date have focused on second-line therapy for patients who have developed resistance to lamivudine and have shown that adding on or switching to adefovir, or switching to entecavir, can resuppress viral replication, normalize ALT levels, and provide signicant histologic improvement.54 57 Tenofovir also has been shown to be effective in patients with HBV resistance to both lamivudine and adefovir.44 All data indicate that entecavir monotherapy should not be recommended in lamivudine-resistant patients because of the high risk of entecavir resistance and adding a second drug is always preferable to a switch. Rapti et al58 recently showed that the addition of adefovir to lamivudine in HBeAg-negative patients with lamivudine resistance effectively suppressed serum HBV DNA in the majority of patients without the development of adefovir resistance. In this study of patients with lamivudine resistance, 14 patients were switched to adefovir, and 28 patients had adefovir added to ongoing lamivudine therapy. Serum HBV DNA levels became undetectable and ALT levels normalized in 71% and 90% of patients, with no differences between the 2 groups. However, virologic and biochemical breakthrough occurred in 3 of 14 patients (21%) who were switched to adefovir monotherapy 15 to 18 months from the start of therapy, whereas suppression of HBV replication was persistent for up to 3 years in the group receiving combination adefovir plus lamivudine. Timing of the change in treatment also is important for patients who experience viral breakthrough while receiving nucleotide/nucleoside therapy. A prompt switch to new treatment results in more rapid viral suppression than does a delayed change in therapy. A study of HBeAg-negative patients with chronic hepatitis B and genotypic and phenotypic resistance to lamivudine indicated that adding adefovir when genotypic resistance was detected (3 6 log10 copies/mL of HBV DNA and normal ALT levels) was more effective than adding adefovir
272
KEEFFE ET AL
when phenotypic resistance was detected (6 log10 copies/mL of HBV DNA and high ALT levels).59 Three months after the change in treatment, serum HBV DNA was undetectable in all patients who had adefovir added when genotypic resistance was detected, versus 46% of those who received adefovir when phenotypic resistance was observed. The 2-year rates of virologic response were 100% and 78%, respectively, and ALT levels remained normal in all patients who were given early addition of adefovir. In contrast, ALT levels were normalized in 50% of patients at 6 months, 72% at 12 months, and 93% at 24 months in the group with late addition of adefovir.59
Avoiding Resistance Development: Selection of Initial Therapy

Current treatment of HIV infection is based on creating a high genetic barrier to resistance. This refers to the number of critical drug-resistant mutations required for the virus to overcome the anti-HIV activity of the drug regimen.60 Current guidelines for initial treatment of patients with HIV infection reect that view by recommending a combination of 3 or more agents from different classes having different mechanisms of action.52 Single-drug class monotherapy or dual therapy with nucleoside reverse-transcriptase inhibitors for the treatment of HIV has been shown to accelerate the development of resistance, cross-resistance, and virologic failure, and also limit future treatment options. Although data are not available for all antiviral agents used to treat HBV infection, it is becoming clear that single-drug therapy may not provide a sufciently high barrier to resistance. That is certainly the case for lamivudine9 and it also may be true for other agents. This has prompted consideration of 2-drug combinations as initial treatment in patients with chronic hepatitis B. Potential exceptions to the requirement of a 2-drug regimen to reduce or prevent resistance are entecavir and tenofovir, with the cumulative experience suggesting that HBV antiviral drug resistance occurs at very low rates of less than 1% over 4 years of treatment, in the case of entecavir.36 Several small-scale studies have assessed the efcacy of combination therapy in treatment-naive patients with HBV infection. Results from a 1-year comparison of 104 treatment-naive HBeAg-positive patients treated with telbivudine or lamivudine monotherapy or telbivudine plus lamivudine showed that telbivudine alone and in combination therapy was superior to lamivudine alone in reducing HBV DNA levels and achieving ALT normalization, and that there were no signicant differences between results obtained with telbivudine versus telbivudine plus lamivudine.61 Moreover, addition of telbivudine to lamivudine, as might be expected, did not decrease the occurrence of viral breakthrough (15.8% for lamivudine alone vs 12.2% for the combination treatment) or the emergence of resistance-associated mutations.61 The concept that combination therapy should use drugs that are not cross-resistant is exemplied by a study of chronic hepatitis B patients treated with lamivudine alone or combined with adefovir for 1 year.62 Initially, both treatments resulted in equally effective viral suppression, with 4 to 5 log10 HBV DNA reductions by week 16. However, by week 52, the combination therapy group remained virally suppressed, with a median reduction from baseline of 5.4 log10, whereas HBV DNA levels increased in the lamivudine monotherapy group. That increase largely was owing to the development of lamivudine resistance
conferring mutations, which occurred in 20 of the patients who received lamivudine alone versus 2% of those treated with lamivudine plus adefovir.62 Another study of 30 treatment-naive, HBeAg-positive patients showed that 48 weeks of treatment with adefovir plus emtricitabine resulted in greater antiviral activity than adefovir monotherapy; median HBV DNA reductions at 48 weeks were 3.48 log10 versus 2.22 log10, respectively.63 No information was provided regarding the emergence of viral resistance. The combination of lamivudine with an immune modulator, peginterferon alfa-2a, also has been shown to slow the emergence of lamivudine resistance in a 72-week study (48 weeks of active treatment and 24 weeks of additional follow-up evaluation) in which 537 treatment-naive HBeAg-negative patients with chronic hepatitis B were treated with peginterferon alfa-2a alone, lamivudine alone, or a combination of the two. At 48 weeks, mutations conferring resistance to lamivudine developed in 18% of patients treated with lamivudine alone versus 1% of those who received combination therapy.64 Results from longerterm follow-up evaluation of 36 HBeAg-negative, anti-HBe positive, treatment-naive patients showed that treatment with lamivudine plus interferon alfa-2b for 1 year followed by lamivudine alone for 3 additional years is an approach that also substantially may slow the emergence of lamivudine resistance. Cumulative rates of breakthrough viremia at the end of 1, 2, 3, and 4 years of treatment were 0%, 14%, 32%, and 59%, respectively.65 These values compare with 23%, 46%, 55%, and 71% for the rst 4 years of lamivudine monotherapy in another study.9 Further study of an induction/maintenance approach to HBV therapy with other antiviral combinations is needed.
Conclusions
The goal of therapy in patients with chronic hepatitis B is rapid viral suppression and long-lasting maintenance of undetectable levels of serum HBV DNA. Nucleoside/nucleotide analogues rapidly are becoming the treatment of choice for most patients with chronic hepatitis B. These agents generally require indenite administration but usually are well tolerated, result in rapid viral suppression, improve ChildPugh scores in cirrhotic patients, and improve overall survival. The major drawback of this treatment is the considerable risk of developing antiviral drug resistance, which occurs most frequently in lamivudine-treated patients, but also has been shown for other agents (eg, adefovir, emtricitabine, and telbivudine) evaluated in long-term studies. Preventing emergence of resistance and virologic breakthrough may best be achieved by using an agent or combination of agents with a high genetic barrier to resistance. This is clearly not the case for lamivudine, which requires the selection of only a single mutation to result in high-level resistance and should no longer be used as monotherapy in patients with chronic HBV infection, except possibly in patients with compensated liver disease who become HBV DNA negative early in therapy. Yuen et al66 recently reported that serum HBV DNA measurements at week 4 and week 16 were useful in predicting lamivudine response at year 5, with the week 4 measurement the most advantageous in allowing earlier prediction of response and/or failure to therapy owing to resistance. The combination of lamivudine with other agents may increase the barrier to resistance, but large-scale, long-term studies still are needed to document this effect. We also need more information about
March 2008
273
long-term viral suppression and emergence of resistance to newer agents (entecavir, telbivudine, tenofovir, and emtricitabine) when they are used as monotherapy or in combination. A combination of 2 potent nucleosides/nucleotides with different resistance proles may prove to be the optimal rstline treatment for chronic hepatitis B, but it remains to be shown in clinical trials. However, these studies are extremely difcult to conduct and may not be feasible. Drugs such as entecavir and tenofovir have such low rates of resistance that any study to determine improved outcomes, such as reduction in already very low rates of resistance and/or added efcacy, would need huge numbers of patients treated for an extended period of time before any difference between combination therapy and monotherapy might be shown. Thus, the decision to use combination therapy de novo might have to be made before appropriate information becomes available from formal randomized controlled trials. Further evaluation of the combination of nucleosides/nucleotides with a course of interferon treatment also is warranted, given short-term results suggesting that this approach may slow the emergence of resistance to lamivudine. Frequent, periodic assessment of viral load with a sensitive HBV DNA assay remains the best approach for early detection of resistance, and testing frequency must be tailored to the patients disease severity and treatment history. The treatment regimen should be modied as soon as resistance is detected; agents should be added or discontinued on the basis of known effects of the detected mutations on the efcacy of specic drugs. Addition or substitution of newer antiviral agents can restore suppression of viral replication after emergence of lamivudine resistance. Nevertheless, new agents with viral targets that differ from those of currently available drugs will be needed for patients who have failed or ultimately may fail second or third treatment regimens. References
1. Centers for Disease Control and Prevention (CDC). Viral hepatitis B. 2006. Available at: http://www.cdc.gov/ncidod/diseases/ hepatitis/b/fact.htm. Accessed: July 17, 2007. 2. Glebe D. Recent advances in hepatitis B virus research: a German point of view. World J Gastroenterol 2007;13:8 13. 3. World Health Organization. Fact sheet no. 204. Hepatitis B. 2000. Available at: http://www.who.int/mediacentre/factsheets/fs204/ en/. Accessed: July 17, 2007. 4. Wasley A, Miller JT, Finelli L, Centers for Disease Control and Prevention (CDC). Surveillance for acute viral hepatitisUnited States, 2005. MMWR Surveill Summ 2007;56:124. 5. Liaw YF, Sung JJ, Chow WC, et al. Lamivudine for patients with chronic hepatitis B and advanced liver disease. N Engl J Med 2004;351:15211531. 6. Arora G, Keeffe EB. Chronic hepatitis B with advanced hepatic brosis or cirrhosis: impact of antiviral therapy. Rev Gastroenterol Disord 2007;7:6373. 7. Zoulim F. In vitro models for studying hepatitis B virus drug resistance. Semin Liver Dis 2006;26:171180. 8. Bartholomeusz A, Locarnini SA. Antiviral drug resistance: clinical consequences and molecular aspects. Semin Liver Dis 2006;26: 162170. 9. Lok AS, Lai CL, Leung N, et al. Long-term safety of lamivudine treatment in patients with chronic hepatitis B. Gastroenterology 2003;125:1714 1722. 10. Andreone P, Gramenzi A, Cursaro C, et al. High risk of hepatocellular carcinoma in anti-HBe positive liver cirrhosis patients
11. 12.
13. 14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
developing lamivudine resistance. J Viral Hepat 2004;11: 439 442. Locarnini S. Molecular virology of hepatitis B virus. Semin Liver Dis 2004;24(Suppl 1):310. Norder H, Courouce AM, Magnius LO. Complete genomes, phylogenetic relatedness, and structural proteins of six strains of the hepatitis B virus, four of which represent two new genotypes. Virology 1994;198:489 503. Beck J, Nassal M. Hepatitis B virus replication. World J Gastroenterol 2007;13:48 64. Okamoto H, Imai M, Kametani M, et al. Genomic heterogeneity of hepatitis B virus in a 54-year-old woman who contracted the infection through materno-fetal transmission. Jpn J Exp Med 1987;57:231236. Bartholomeusz A, Locarnini S. Hepatitis B virus mutations associated with antiviral therapy. J Med Virol 2006;78(Suppl 1):S52 S55. Locarnini S, Hatzakis A, Heathcote J, et al. Management of antiviral resistance in patients with chronic hepatitis B. Antivir Ther 2004;9:679 693. Da Silva LC, Pinho JR, Sitnik R, et al. Efcacy and tolerability of long-term therapy using high lamivudine doses for the treatment of chronic hepatitis B. J Gastroenterol 2001;36:476 485. Wright TL. Clinical trial results and treatment resistance with lamivudine in hepatitis B. Semin Liver Dis 2004;24(Suppl 1): 3136. Dienstag JL, Goldin RD, Heathcote EJ, et al. Histological outcome during long-term lamivudine therapy. Gastroenterology 2003; 124:105117. Yuen MF, Kato T, Mizokami M, et al. Clinical outcome and virologic proles of severe hepatitis B exacerbation due to YMDD mutations. J Hepatol 2003;39:850 855. Si Ahmed SN, Tavan D, Pichoud C, et al. Early detection of viral resistance by determination of hepatitis B virus polymerase mutations in patients treated by lamivudine for chronic hepatitis B. Hepatology 2000;32:1078 1088. Mutimer D, Pillay D, Shields P, et al. Outcome of lamivudine resistant hepatitis B virus infection in the liver transplant recipient. Gut 2000;46:107113. Zoulim F. Mechanism and viral persistence and resistance to nucleoside and nucleotide analogs in chronic hepatitis B virus infection. Antiviral Res 2004;64:115. Hussain M, Lok AS. Mutations in the hepatitis B virus polymerase gene associated with antiviral treatment for hepatitis B. J Viral Hepat 1999;6:183194. Poch O, Sauvaget I, Delarue M, et al. Identication of four conserved motifs among the RNA-dependent polymerase encoding elements. EMBO J 1989;8:38673874. Allen MI, Deslauriers M, Andrews CW, et al. Identication and characterization of mutations in hepatitis B virus resistant to lamivudine. Hepatology 1998;27:1670 1677. Delaney WE 4th, Yang H, Westland CE, et al. The hepatitis B virus polymerase mutation rtV173L is selected during lamivudine therapy and enhances viral replication in vitro. J Virol 2003;77: 1183311841. Fu L, Cheng YC. Role of additional mutations outside the YMDD motif of hepatitis B virus polymerase in L()SddC (3TC) resistance. Biochem Pharmacol 1998;55:15671572. Seigneres B, Martin P, Werle B, et al. Effects of pyrimidine and purine analog combinations in the duck hepatitis B virus infection model. Antimicrob Agents Chemother 2003;47:18421852. Curtis M, Zhu Y, Borroto-Esoda K. HBV rtI233V polymerase variant remains sensitive to adefovir (abstr). J Hepatol 2007; 46(Suppl 1):S26 S27. Hadziyannis SJ, Tassopoulos NC, Heathcote EJ, et al. Longterm therapy with adefovir dipivoxil for HBeAg-negative chronic
274
KEEFFE ET AL
32.
33. 34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46. 47.
48. 49.
50. 51.
52.
hepatitis B for up to 5 years. Gastroenterology 2006;131: 17431751. Borroto-Esoda K, Arterburn S, Snow A, et al. Final analysis of virological outcomes and resistance during 5 years of adefovir dipivoxil monotherapy in HBeAg-negative patients (abstr). J Hepatol 2006;44(Suppl 2):S179 S180. Langley DR, Walsh AW, Baldick CJ, et al. Inhibition of hepatitis B virus polymerase by entecavir. J Virol 2007;81:3992 4001. Colonno RJ, Rose R, Baldick CJ, et al. Entecavir resistance is rare in nucleoside nave patients with hepatitis B. Hepatology 2006; 44:1656 1665. Tenney DJ, Rose RE, Baldick CJ, et al. Two-year assessment of entecavir resistance in lamivudine-refractory hepatitis B virus patients reveals different clinical outcomes depending on the resistance substitutions present. Antimicrob Agents Chemother 2007;51:902911. Colonno RJ, Rose RE, Pokornowski K, et al. Four year assessment of ETV resistance in nucleoside-nave and lamivudine refractory patients (abstr). J Hepatol 2007;46(Suppl 1):S294. Seifer M, Patty A, Dukhan D, et al. Telbivudine (LdT) preferentially inhibits second () strand HBV DNA synthesis (abstr). Gastroenterology 2005;128(Suppl 2):A742A743. Lai CL, Gane E, Hsu CW, et al. Two-year results from the GLOBE trial in patients with hepatitis B: greater clinical and antiviral efcacy for telbivudine (LdT) vs lamivudine (abstr). Hepatology 2006;44(Suppl 1):222A. Lim SG, Ng TM, Kung N, et al. A double-blind placebo-controlled study of emtricitabine in chronic hepatitis B. Arch Intern Med 2006;166:49 56. Yang H, Qi X, Sabogal A, et al. Cross-resistance testing of nextgeneration nucleoside and nucleotide analogues against lamivudine-resistant HBV. Antivir Ther 2005;10:625 633. Gish RG, Trinh H, Leung N, et al. Safety and antiviral activity of emtricitabine (FTC) for the treatment of chronic hepatitis B infection: a two-year study. J Hepatol 2005;43:60 66. Delaney WE 4th, Ray AS, Yang H, et al. Intracellular metabolism and in vitro activity of tenofovir against hepatitis B virus. Antimicrob Agents Chemother 2006;50:24712477. Lada O, Benhamou Y, Cahour A, et al. In vitro susceptibility of lamivudine-resistant hepatitis B virus to adefovir and tenofovir. Antivir Ther 2004;9:353363. van Bmmel F, Zollner B, Sarrazin C, et al. Tenofovir for patients with lamivudine-resistant hepatitis B virus (HBV) infection and high HBV DNA level during adefovir therapy. Hepatology 2006; 44:318 325. van der Eijk AA, Hansen BE, Niesters HG, et al. Viral dynamics during tenofovir therapy in patients infected with lamivudine-resistant hepatitis B virus mutants. J Viral Hepat 2005;12:364 372. Lok AS, McMahon BJ. Chronic hepatitis B. Hepatology 2007;45: 507539. Keeffe EB, Dieterich DT, Han SH, et al. A treatment algorithm for the management of chronic hepatitis B virus infection in the United States: an update. Clin Gastroenterol Hepatol 2006;4: 936 962. Tillmann HL. Antiviral therapy and resistance with hepatitis B virus infection. World J Gastroenterol 2007;13:125140. Lok AS, Zoulim F, Locarnini S, et al. Monitoring drug resistance in chronic hepatitis B virus (HBV)-infected patients during lamivudine therapy: evaluation of performance of INNO-LiPA HBV DR assay. J Clin Microbiol 2002;40:3729 3734. Zoulim F. New nucleic acid diagnostic tests in viral hepatitis. Semin Liver Dis 2006;26:309 317. Tran N, Berne R, Chann R, et al. European multicenter evaluation of high-density DNA probe arrays for detection of hepatitis B virus resistance mutations and identication of genotypes. J Clin Microbiol 2006;44:27922800. Department of Health and Human Services (DHHS). Guidelines
53. 54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
for the use of antiretroviral agents in HIV-1-infected adults and adolescents. October 10, 2006. Available at: http://AIDSinfo. nih.gov/contentles/Adultandadolescent6L.pdf. Accessed: July 17, 2007. Tang JW, Pillay D. Transmission of HIV-1 drug resistance. J Clin Virol 2004;30:110. Sherman M, Yurdaydin C, Sollano J, et al. Entecavir for treatment of lamivudine-refractory, HBeAg-positive chronic hepatitis B. Gastroenterology 2006;130:2039 2049. Peters MG, Hann HWH, Martin P, et al. Adefovir dipivoxil alone or in combination with lamivudine in patients with lamivudine-resistant chronic hepatitis B. Gastroenterology 2004;126:91101. Perrillo R, Hann HW, Mutimer D, et al. Adefovir dipivoxil added to ongoing lamivudine in chronic hepatitis B with YMDD mutant hepatitis B virus. Gastroenterology 2004;126:8190. Chang TT, Gish RG, Hadziyannis SJ, et al. A dose-ranging study of the efcacy and tolerability of entecavir in lamivudine-refractory chronic hepatitis B patients. Gastroenterology 2005;129:1198 1209. Rapti I, Dimou E, Mitsoula P, et al. Adding-on versus switching-to adefovir therapy in lamivudine-resistant HBeAg-negative chronic hepatitis B. Hepatology 2007;45:307313. Lampertico P, Vigano M, Manenti E, et al. Adefovir rapidly suppresses hepatitis B in HBeAg-negative patients developing genotypic resistance to lamivudine. Hepatology 2005;42:1414 1419. Luber AD. Genetic barriers to resistance and impact on clinical response. EJIAS: eJournal of the International AIDS Society 2005;7:69:111. Available at: http://www.medscape.com/ viewarticle/504524_print. Accessed: March 20, 2007. Lai CL, Leung N, Teo EK, et al. A 1-year trial of telbivudine, lamivudine, and the combination in patients with hepatitis B e antigen-positive chronic hepatitis B. Gastroenterology 2005;129: 528 536. Sung JJY, Lai JY, Zeuzem S, et al. A randomised double-blind phase II study of lamivudine (LAM) compared with lamivudine plus adefovir dipovoxil (ADV) for treatment nave patients with chronic hepatitis B (CHB): week 52 analysis (abstr). J Hepatol 2003;38(Suppl 2):2526. Lau G, Cooksley H, Ribeiro RM, et al. Randomised, double-blind study comparing adefovir dipivoxil plus emtricitabine combination therapy versus adefovir alone in HBeAg() chronic hepatitis B: efcacy and mechanisms of treatment response (abstr). Hepatology 2004;40(Suppl 1):272A. Marcellin P, Lau GK, Bonino F, et al. Peginterferon alfa-2a alone, lamivudine alone, and the two in combination in patients with HBeAg-negative chronic hepatitis B. N Engl J Med 2004;351: 1206 1217. Nikolaidis NL, Giouleme OI, Tziomalos KA, et al. Interferon/longterm lamivudine combination therapy in anti-HBe positive chronic hepatitis B patients. J Gastroenterol Hepatol 2005;20:1721 1725. Yuen MF, Wong DKH, Fung J, et al. Predictive value of HBV DNA levels at frequent time points during early and maintenance phase of 5-year lamivudine and mutational proles of reverse transcriptase (RT) and surface (S) genes (abstr). Hepatology 2006;44:557A.
Address requests for reprints to: Emmet B. Keeffe, MD, Division of Gastroenterology and Hepatology, Department of Medicine, Stanford University Medical Center, 750 Welch Road, Suite 210, Palo Alto, California 94304-1509. e-mail: ekeeffe@stanford.edu; fax: (650) 4985692. The development of this manuscript was supported by an unrestricted educational grant from Gilead Sciences (Foster City, CA). The authors wish to thank Filippo Cavalieri, PhD, for editorial contributions and assistance in the preparation of the manuscript.
Pancreatic Cancer: A Review of the Evidence on Causation

ANDREW R. HART,*, HUGH KENNEDY, and IAN HARVEY*
*Population Health, School of Medicine, Health Policy and Practice, University of East Anglia, Norwich, Norfolk, United Kingdom; Department of Gastroenterology, Norfolk & Norwich University Hospital National Health Service Trust, Colney, Norfolk, Norwich, United Kingdom
See CME exam on page 262.

Pancreatic cancer kills more than 250,000 people each year worldwide and has a poor prognosis. The aim of this article is to critically review the epidemiologic evidence for exposures that may either increase or decrease the risk. A Medline search was performed for epidemiologic studies and reviews published up to April 2007. Consistent evidence of a positive association was found for family history and cigarette smoking. Many studies documented a positive association with diabetes mellitus and chronic pancreatitis, although the etiologic mechanisms are unclear. Other associations were detected, but the results were either inconsistent or from few studies. These included positive associations with red meat, sugar, fat, body mass index, gallstones, and Helicobacter pylori, and protective effects of increasing parity, dietary folate, aspirin, and statins. There was no evidence linking alcohol or coffee consumption with an increased risk of pancreatic cancer. The associations with many exposures need to be claried from further epidemiologic work in which there is both precise measurement of risk factors, adjustment for potential confounders, and, for dietary studies, information recorded on the method of food preparation and pattern of consumption. Such work is important to reduce the incidence of this fatal disease. orldwide, adenocarcinoma of the pancreas causes more than a quarter of a million deaths annually, is the 13th most common cancer and the eighth most frequent cause of death from cancer.1 Survival rates are among the worst for any tumor, with a mortality to incidence ratio of 98%.1 Most patients present with advanced disease and management options include chemotherapy, pain control, and relief of jaundice by either surgical or endoscopic means. The cause of pancreatic cancer is largely unknown and there are currently no accepted models of the disease. This review aims to critically review the epidemiologic evidence for exposures and, in areas where there is sparse or conicting evidence, suggest studies that may provide clarication. In writing this article, the Medline database was searched for published articles between 1950 and April 2007. The principal initial search terms used, which were combined with pancreatic cancer, were as follows: family history, occupation, diabetes, pancreatitis, cholecystectomy, Helicobacter pylori, smoking, parity, meat, fat, sugar, fruits, vegetables, coffee, alcohol, physical activity, body mass index, aspirin, and statins. For several terms comprehensive reviews had been published on specic areas,
namely occupations, smoking, diabetes, coffee, and alcohol, and these are referenced. Two types of epidemiologic study design can be used to investigate exposures for pancreatic cancer, namely case-control and cohort studies, with both having advantages and disadvantages. Case-control studies recruit a wide selection of cases and use accurate measures of risk factors because the numbers of participants included are relatively small. However, a major problem can be recall bias, where it is difcult to ensure accurate information on risk factors before the onset of symptoms that potentially were involved in the cause of the disease. Another limitation is selection bias if control groups are chosen that are not representative of the general population. The second method of study is cohort studies, and their strength is that the exposure data collected truly represent that before the onset of symptoms. Also, there is less selection bias because both participants who develop the disease and those who remain well are drawn from the same population. In writing this review we decided that recall bias for many exposures such as illnesses, occupation, smoking, and medications is likely to be low and therefore included both study designs. However, for dietary exposures, recall bias for food items exists in casecontrol studies, which is removed by a cohort design. Therefore, in this review only dietary data from the latter type is included to ensure the recorded diet truly represents that associated with risk. However, we acknowledge that there are at least a further 55 case-control studies that have examined diet and pancreatic cancer.
Family History and Genetic Disorders

A family history of pancreatic cancer in a rst-degree relative is associated with an increased risk of pancreatic cancer of between 2.5 to 5.3 times.2 6 The risk increases the more relatives are affected, with a risk of 6.4 (95% condence interval [CI], 1.8 16.4) in those with 2 affected relatives, increasing to a risk of 32.0 (95% CI, 10.274.7) in those with 3 affected relatives.7 The increased risk in family members could be owing to genetic factors and/or lifestyle habits that may be similar in different generations, such as smoking. A case-control study conducted in Michigan addressed this possibility6 and reported that pancreatic cancer in a rst-degree relative (parent, sibling, offspring) was associated with an increased risk of the disease of 2.5 (95% CI, 1.3 4.7) times, after adjusting for smoking in the patients relative. Interaction existed between the 2 factors in
Abbreviations used in this paper: BMI, body mass index; CI, condence interval; RR, relative risk. 2008 by the AGA Institute 1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.041
276
HART ET AL
that the risk was 6.0 times greater (95% CI, 2.0 18.3) in those who had an affected relative who also smoked. A number of genetic disorders are associated with an increased risk of pancreatic cancer, although no common genetic abnormality has been identied in all these conditions. These include hereditary pancreatitis, familial adenomatous polyposis, familial atypical multiple mole melanoma, Peutz-Jeghers syndrome, hereditary nonpolyposis colorectal cancer, familial breast cancer, ataxia-telangiectasia, and cystic brosis. Patients with hereditary pancreatitis have a 53 (95% CI, 23105) times greater risk of disease with a cumulative lifetime risk of 40% (95% CI, 9%71%),8 and in those with Peutz-Jeghers syndrome the relative risk is 132 (95% CI, 44 261) with a cumulative lifetime risk of 36%.9 In this condition, germline mutations in the STK 11 gene have been identied,10 which acts as a tumor-suppressor gene in the earliest steps of the pathogenesis of hamartomas into adenocarcinomas. Finally, Maisonneuve et al11 collected reports of patients presenting with pancreatic cancer who also suffered from cystic brosis in North America and Europe. By using population data, the relative risk of cancer in individuals with cystic brosis was 5.3 (95% CI, 2.4 10.1), with an early age of onset at a median age of 35 years (range, 18 58 y). Furthermore, there is a case-report of a patient with cystic brosis and bilateral lung transplantation who underwent a Whipple procedure for pancreatic adenocarcinoma at the age of 12 years.12 Because all these genetic conditions are rare, the contributions they make to all cases of pancreatic cancer are small.
compared with nulliparous women (180 vs 212 ng/mL, P for trend .003).18 Secondly, during pregnancy, there is a marked physiologic reduction in total body iron stores. Two case-control studies have reported that increased iron consumption and increased serum iron concentrations are associated with an increased risk of pancreatic cancer.19,20 Free iron may be involved in carcinogenesis by inducing DNA damage by causing oxidative stress.21 Several epidemiologic studies have reported that parity reduces the risk of pancreatic cancer,2226 which may explain why women have a lower incidence of the disease. These studies recorded a reduction in risk of pancreatic cancer of at least 20% in women who had 4 or 5 children. Two Scandinavian studies did not report an association, although these did not control for the possible confounding effect of smoking.27,28 Finally, a cohort investigation from Japan, which did adjust for smoking, again did not show an effect.29 The epidemiologic evidence is inconsistent, despite the biological mechanisms, and should be claried by future studies considering all potential confounders.
Concomitant Illness
In patients with diabetes mellitus, a meta-analysis of 36 epidemiologic studies reported an odds ratio of 1.82 (95% CI, 1.66 1.89).30 The mechanism of the diabetes association needs to be studied, investigating possibilities that include a common causative agent, the metabolic consequences of diabetes, and, nally, residual confounding. A possible mechanism is that hyperglycemia promotes hyperinsulinemia and activation of insulin-like growth factor 1, which stimulates cell proliferation.31 A potential difculty with this association between diseases is the issue of reverse causality, in that destruction of islet cells by malignant tissue may precipitate diabetes, rather than the latter actually being a causative factor.32 However, this association is likely to be real because epidemiologic studies report an increased risk in patients diagnosed with diabetes more than 10 years before cancer onset.33 Acute and chronic pancreatitis have been shown consistently to be risk factors for pancreatic cancer.34 38 The biological mechanisms are again unknown, but possibly related to damage induced by the inammatory process. The data suggest that this is not a spurious association owing to pancreatic cancer being misdiagnosed initially as chronic pancreatitis. The association still persists when cases of cancer diagnosed more than 7 years after an initial diagnosis of pancreatitis are analyzed.37 A historical cohort study of 2015 subjects with chronic pancreatitis, who were followed up for a mean of 7.4 years, reported a standardized incidence ratio of 26.3 (95% CI, 19.9 34.2), with a cumulative 4% risk of pancreatic cancer at 20 years follow-up evaluation.38 Two Swedish cohort studies monitoring patients with pancreatitis34,36 reported relative risks of 2.2 (95% CI, 1.6 2.9) and 5.0 (95% CI, 4.1 6.1) for any form of pancreatitis. A study of American Veterans found an odds ratio of 3.42 (95% CI, 1.98 5.91) for pancreatic cancer in patients with all types of pancreatitis.37 The risks are lower after an episode of acute pancreatitis compared with chronic disease.34,36 Furthermore, the effect of chronic pancreatitis does not seem to be mediated through alcohol because the risk was similar36,37 in patients in whom the cause of pancreatitis was alcohol-related compared with nonalcohol-induced pancreatitis. Finally, data are required in investigating the risk in patients with chronic pan-
Occupational Exposures
The effect of occupational exposures was assessed in a comprehensive meta-analysis of data from 92 studies that reported effects for 23 agents.13 The results showed that chlorinated hydrocarbon solvents and related compounds had a meta-risk ratio of 1.4 (95% CI, 1.0 1.8), and for nickel and nickel compounds the meta-risk ratio was 1.9 (95% CI, 1.23.2). For other agents, small but nonsignicant risks were reported, including chromium compounds, polycyclic aromatic hydrocarbons, organochlorine insecticides, silica dust, and asbestos (meta-risk ratio, 1.1; 95% CI, 0.9 1.5). Although most studies have not conrmed an association with occupational asbestos exposure, there are ecologic data to support this hypothesis from communities with high levels of asbestos in the drinking water supplies.14 A second meta-analysis of 14 epidemiologic studies that assessed risk in occupations exposed to formaldehyde reported a pooled relative risk of 1.1 (95% CI, 1.0 1.3).15 However, this was limited to embalmers, anatomists, and pathologists (relative risk [RR], 1.3; 95% CI, 1.0 1.7). Currently, there is no plausible biological mechanism because most biological effects of formaldehyde are at the site of contact with the chemical.
Hormonal Factors
Increasing parity may prevent pancreatic cancer by reducing insulin-like growth factor levels and body iron stores. Insulin-like growth factors are involved in the development of cancer and promote cellular proliferation and inhibit apoptosis.16,17 Pregnancy can induce changes in the insulin-like growth factor axis and an analysis of plasma insulin-like growth factors showed a lower concentration in women with 4 or more births
March 2008
CAUSE OF PANCREATIC CANCER
277
creatitis who also have developed diabetes as a complication of pancreatitis. Gallstones and cholecystectomy have been reported as risk factors for pancreatic cancer with risks in the range 1.3 to 2.8 times,5,39 43 although other studies have not found any associations.44 49 A possible mechanism is that cholecystectomy increases the circulating levels of the gastrointestinal hormone, cholecystokinin-pancreozymin, which has trophic effects on the pancreas including pancreatic hyperplasia and hypertrophy.50,51 Why discrepancies exist in the epidemiologic literature is unknown, although possibilities include other causative factors associated with stone development, which have not been adjusted for in the analyses. In the US Nurse Health and Health Professionals Follow-up Studies, the risk decreased from 1.31 (95% CI, 0.931.83) to 1.11 (95% CI, 0.78 1.56) after correcting for other factors including body mass index (BMI) and physical activity.52 Other illnesses are associated with an increased risk of pancreatic cancer including asthma,53 familial atypical multiple mole melanoma,54 and other primary malignancies.55 Asthma doubled the risk of pancreatic cancer, which was not altered by adjustment for smoking, and, furthermore, other respiratory conditions, namely emphysema and chronic bronchitis, were not associated.53 Familial atypical multiple mole melanoma is an autosomal-dominant condition characterized by malignant melanoma of the skin and atypical precursor lesions. Germline mutations in the p16 (CDKN2A) gene have been reported in at least a quarter of patients and the cumulative risk of pancreatic cancer by age 75 years is 17%.54 Finally, the incidence of pancreatic cancer is increased secondary to other smoking-related malignancies including lung cancer (RR, 1.3; 95% CI, 1.0 1.6 in men; RR, 2.5; 95% CI, 1.9 3.2 in women), head and neck cancer in women (RR, 1.8; 95% CI, 1.22.5), and bladder cancer in women (RR, 1.5; 95% CI, 1.12.0).55 Interestingly, a marked decrease in risk after lymphoma in men has been reported (RR, 0.2; 95% CI, 0.0 0.8).55
Cigarette Smoking
Smoking is reported consistently as an environmental risk factor for pancreatic cancer and accounts for approximately 25% of all pancreatic cancers.65 Carcinogens reach the pancreas via the blood steam or reuxed bile, and nitrosamines, present in cigarettes, induce pancreatic tumors in animal models.66 Cigarette smoking approximately doubles the risk of pancreatic cancer and the effect is related to its duration and intensity.6775 The importance of these data prompted the International Agency for Research in Cancer to state that cigarette smoking is an important cause of pancreatic cancer.75 Whether passive smoking may be involved is unknown. A casecontrol study conducted in Canada reported that nonsmokers who had been exposed to such smoking, both as a child and as an adult, had a small but nonsignicant increased risk of the disease (odds ratio, 1.21; 95% CI, 0.60 2.44).76 Smoking always should be measured and adjusted for in etiologic epidemiologic studies of pancreatic cancer and encouraging nonsmoking should reduce the incidence of the disease.
Total Energy Intake

The effect of total energy intake has been reported, to the best of our knowledge, in only 2 prospective cohort studies. One from Finland, in men, found an inverse trend (P .05).77 However, a prospective investigation from Iowa, in women, found no association with total caloric intake.78 The relationship may be confounded by physical activity because there is a correlation between physical activity and energy intake, and physical activity itself may be protective against pancreatic cancer.79 81 Neither study adjusted for the effects of total physical activity. In most of the investigations into diet discussed in the following sections, risks associated with individual food types were adjusted for total energy intake. This is important because energy adjustment helps to control for several factors including body size, metabolic rate, and physical activity.
Meat H pylori Infection

There are several plausible biological mechanisms for how H pylori infection could increase the risk of pancreatic cancer. These are secondary to the H pyloriinduced gastritis including a reduced absorption of anti-oxidants such as vitamin C,56 and hypergastrinemia, which stimulates pancreatic cancer cell growth.57 Also, increased secretin levels, as a consequence of decreased somatostatin production, have trophic effects on the pancreatic ductal epithelium.58 Three epidemiologic studies, 1 cohort59 and 2 case-control investigations,60,61 have investigated this hypothesis. These reported a positive association, with the 2 largest recording an approximate doubling of the risk of pancreatic cancer in those infected with H pylori.59,60 To conrm consistency of the association, further cohort studies are required because such methodology ensures that the H pylori status truly represents whether infection is present or not before the development of cancer. The association has clinical implications in that individuals at high risk, such as those with a family history of diabetes mellitus, could be screened for infection. The H pylori hypothesis may help explain other clinical associations such as patients with a partial gastrectomy for ulcer disease having a greater risk of pancreatic cancer.62 64 The hypothesis that meat increases risk is based on experimental data that cooking meat at high temperatures, particularly red meat, produces heterocyclic amines and polycyclic aromatic hydrocarbons, which are carcinogenic in animals.82 To date, 7 cohort studies specically have investigated red meat consumption and pancreatic cancer, with 3 studies reporting a signicant positive association,69,83,84 3 studies reporting no association,67,77,85 and 1 study reporting a reduced risk.86 In studies reporting a positive association, a high intake of red meat approximately doubled the risk of disease.69,83,84 The inconsistency in the evidence may be related to the method of meat preparation in that cooking at higher temperatures, such as frying, produces more carcinogens than preparing meat at lower temperatures. Further clarication from epidemiologic studies therefore is required, particularly data on the cooking method. For chicken and poultry, most cohort studies did not nd any association with the disease.69,77,84,85 If red meat were involved, then it potentially could be switched in public health programs.
Sugar Intake
A high dietary sugar intake leads to hyperinsulinemia, which may be carcinogenic by altering the cell cycle, inhibiting
278
HART ET AL
apoptosis, and down-regulating insulin-like growth factor binding protein 1. Prospective studies involving biomarkers have reported double the risk of pancreatic cancer in subjects with both higher fasting serum glucoses and fasting insulin concentrations.87,88 However, cohort investigations using food-frequency data, have either found no association with total sucrose89 or total sugar intake.90 One study reported a positive nonsignicant association with the highest quintile of glycemic load (1.53; 95% CI, 0.96 2.45),89 and another study found no relationship to this measure.90 Glycemic load is a measure that allows the carbohydrate content of individual foods to be classied according to their postprandial effects and their effects on insulin levels. The source of sugar and the pattern of consumption may be important. An investigation of 77,797 Swedish men and women reported a relative risk of 1.95 (95% CI, 1.10 3.46) for the highest versus the lowest consumption of added sugar, with a signicant trend across quartiles (P .03).91 In the United States, women consuming the highest intake of sugarsweetened soft drinks had a 1.57 (95% CI, 1.022.41) times greater risk of disease,92 although no association was found in men. Soft drinks are the leading source of added sugar in the US diet contributing to a high glycemic load. In summary, there is supportive evidence for an effect of sugar from experimental and biomarker data and for added sugar consumption, although questionnaire data on total sugar intake is unsupportive. Clarication is required from further cohort investigations investigating the intake and pattern of sugar consumption from different sources.
both pancreatic carcinogenesis in animal models and the growth of human pancreatic cancer cell lines. Consistent evidence exists for a protective effect of dietary folate. Folate is important in DNA synthesis and repair through its ability to donate methyl groups. The Swedish cohort study reported a highly protective effect, reducing the risk of pancreatic cancer to 0.25 (95% CI, 0.11 0.59) in those consuming the highest quintile of folate.97 Similar ndings also were reported in the male Finnish smokers investigation98 (hazard ratio, 0.52; 95% CI, 0.31 0.87 for highest vs lowest quintile of dietary folate) and in combined data from the Nurses and Health Professionals follow-up studies (RR, 0.66; 95% CI, 0.421.03 for highest vs lowest quartile).99 In a nested case-control analysis within the Finnish study, men in the highest tertile of serum folate intake had a statistically signicant (55%; 95% CI, 0.18 0.76) reduction in risk.100 Before conclusions can be reached regarding fruit and vegetable intake, more data again are required from other cohort studies, particularly on individual foods.
Beverages
Reviewing the studies on the potential carcinogenicity of alcohol and coffee, The International Agency for Research on Cancer concluded that there was little evidence to support casual associations.101,102 Since these statements, similar data have been reported in the US Health Professionals Follow Up Study,103 The Nurses Health Study,103 The Japan Collaborative Cohort Study for Evaluation of Cancer Risk,104 and The Swedish Twin Registry Cohort Study.10 However, The Iowa Womens Health Study, reported a nonsignicant positive association with increased alcohol intake (P trend .11) and a doubling of the risk with an increased coffee consumption (RR, 2.2; 95% CI, 1.1 4.3).78 The data that the risk of pancreatic cancer is similar in patients with alcohol-related and nonalcohol-related chronic pancreatitis argues against a role for the direct effects of alcohol in causes.
Dietary Fat
Dietary fat entering the duodenum initiates the release of cholecystokinin, which stimulates pancreatic hyperplasia, increasing its susceptibility to carcinogens.93 Fat intake has been investigated in 3 cohort studies with no consistent associations detected.77,84,85 In the Finnish male smokers study, energy-adjusted saturated fat intake was associated positively with pancreatic cancer (highest vs lowest quintile: hazard ratio, 1.60; 95% CI, 0.96 2.64; P trend .02).77 No signicant associations were detected with total fat, polyunsaturated fatty acids, or monounsaturated fatty acids. In the Multiethnic Cohort Study84 in Hawaii and Los Angeles no associations with the percentage of total energy from fat, saturated fat, or cholesterol were detected. Similarly, in the US Nurses Health Study, no associations were reported with the intake of total fat, different types of fat, and cholesterol.85 Other prospective cohort studies investigating dietary fat are needed.
Anthropometric Measurements and Physical Activity

Anthropometric measures including BMI, waist circumference, and waist-hip ratio are dependent on several factors including diet and physical activity. A meta-analysis of 6 casecontrol and 8 cohort studies of 6391 cases of pancreatic cancer reported an increased risk of 2% (95% CI, 1%3%) for each unit increase in BMI.105 Since this meta-analysis in 2003, 8 prospective studies have been published with conicting results. The European Prospective Investigation Into Cancer and Nutrition Study found a nonsignicant increased cancer risk with increasing BMI (RR, 1.09; 95% CI, 0.951.24 per 5 kg/m2, and a signicant effect of both a larger waist-to-hip ratio (RR/0.1, 1.24; 95% CI, 1.04 1.48) and waist circumference (RR/10 cm, 1.13; 95% CI, 1.011.26).80 Positive associations also were reported for increasing BMI and waist circumference in 2 Swedish population-based cohorts106 and in The American Cancer Society Cancer Prevention Study II Nutrition Cohort.107 In the US Multiethnic Cohort Study, an increased risk was found in men, but a protective association was found in women.108 However, in 2 Japanese cohorts, one reported no effect of BMI at baseline109 and in the second reported a protective effect that was found in men and no association in women.110 In the Iowa
Fruit and Vegetable Consumption

Fruits and vegetables contain anti-oxidants that have anticarcinogenic properties. Cohort studies have not reported any protective associations between pancreatic cancer and total fruit or vegetable intake,39,67,69,77,94 96 although, again, small associations could be undetected because of measurement error in the dietary assessment methods. However, in a Swedish cohort study a signicant inverse association was observed with the cruciferous vegetable cabbage (1 serving/wk vs never consumption: hazard ratio, 0.62; 95% CI, 0.39 0.99).96 This protective effect may be related to the high content of glucosinolates, which are degraded into isothiocyanates, which inhibit
March 2008
279
Womens Health Study, no effect of increasing BMI was found on the risk for pancreatic cancer. Although most data suggest a positive association, inconsistencies need to be claried that may be related to the prevalence and denition of obesity and variations in the cut-off points used for analysis of BMI in the different studies. Physical activity may affect the risk of pancreatic cancer through its effects on glucose metabolism and insulin levels. Nine cohort studies in this area were identied, with 6 showing no effect,107112 1 showing a nonsignicant protective effect,80 and 2 showing a protective effect of increased physical activity.53,81 An investigation from the United States reported a protective effect of moderate physical activity, with those in the highest quintile having a signicant 55% reduction in the risk of pancreatic cancer with a signicant trend across quintiles (P .001).81 This investigation in male health professionals and female nurses used frequency of exercise questionnaires validated against diaries of physical activity and physiologic measurements such as pulse rate. Of the investigations that showed no effect, several only had limited questions on physical activity,107,109,110,112 and 2 lacked generalizability.111,112 Before a conclusion on physical activity and disease risk can be reached, further investigations are required using validated questionnaires measuring all forms of physical activity in representative populations.
were statistically not signicant. The role of aspirin therefore is unclear and needs to be claried in epidemiologic studies with detailed measures of the drug dose and consideration of confounders. A second group of medications that could reduce the risk of pancreatic cancer are the statins (3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors). These drugs reduce serum cholesterol by inhibiting the rate-limiting step in cholesterol synthesis. Statins also have effects on many cellular processes that may protect against the development and progression of cancer. In addition to cholesterol there are several other products derived from 3-hydroxy-3-methylglutaryl coenzyme A reductase including farnesyl pyrophosphate and geranylgeranyl pyrophosphate. These affect guanosine triphosphatase signaling proteins whose functions can inuence cell proliferation. A cohort study of nearly half a million US veterans reported that statin use of more than 6 months was associated with a reduction in pancreatic cancer risk, with an odds ratio of 0.33 (95% CI, 0.26 0.41; P .01).124
Conclusions
Smoking and family history are unequivocal risk factors for pancreatic cancer. The data are consistent for diabetes mellitus and chronic pancreatitis, but importantly the biological mechanisms are unknown and need further investigation. For most other exposures the evidence is either inconsistent or derived from relatively few studies. Clarication is required in which there is, rst, accurate measurement of exposures, and, second, consideration adjustment for other risk factors. Such work is important to reduce the incidence of this highly fatal cancer. References
1. Parkin D, Bray F, Ferlay J, et al. Global cancer statistics. CA Cancer J Clin 2005;55:74 108. 2. Falk R, Pickle L, Fontham E, et al. Life-style risk factors for pancreatic cancer in Louisiana: a case-control study. Am J Epidemiol 1988;128:324 336. 3. Ghadirian P, Liu G, Gallinger S, et al. Risk of pancreatic cancer among individuals with a family history of cancer of the pancreas. Int J Cancer 2002;97:807 810. 4. Fernandez E, La Vecchia C, DAvanzo B, et al. Family history and the risk of liver, gallbladder, and pancreatic cancer. Cancer Epidemiol Biomarkers Prev 1994;3:209 212. 5. Silverman D, Schiffman M, Everhart J, et al. Diabetes mellitus, other medical conditions and familial history of cancer as risk factors for pancreatic cancer. Br J Cancer 1999;80:1830 1837. 6. Schenk M, Schwartz A, O Neal E, et al. Familial risk of pancreatic cancer. J Natl Cancer Inst 2001;93:640 644. 7. Klein A, Brune K, Petersen G, et al. Prospective risk of pancreatic cancer in familial pancreatic cancer kindreds. Cancer Res 2004;64:2634 2683. 8. Lowenfels A, Maisonneuve P, DiMagno E, et al, and The International Heriditary Pancreatitis Study Group. Hereditary pancreatitis and the risk of pancreatic cancer. J Natl Cancer Inst 1997;89:442 446. 9. Giardiello F, Brensinger J, Tersmette A, et al. Very high risk of cancer in familial Peutz-Jeghers syndrome. Gastroenterology 2000;119:14471453. 10. Jenne D, Reimann H, Nezu J, et al. Peutz-Jeghers syndrome is caused by mutations in a novel serine threonine kinase. Nat Genet 1998;18:38 43.
Potential Effects of Drugs

Aspirin inhibits cyclooxygenase enzymes, which may prevent carcinogenesis by enhancing cellular immune responses, inhibiting prostaglandin synthesis, and inuencing apoptosis, tumorigenesis, and angiogenesis.113 Aspirin also either could increase or decrease the risk of pancreatic cancer by its ability to inuence different lipoxygenase enzymes, which can have either anticarcinogenic or procarcinogenic properties.114 The only trial that has assessed aspirin in the incidence of pancreatic cancer was conducted in the Womens Health Initiative in the United States.115 In the Womens Health Initiative, nearly 40,000 women were randomized to receive either 100 mg aspirin every other day or placebo, with an average follow-up time of 10.1 years. No statistically signicant difference was noted between test and placebo groups, with a reported relative risk of 1.42 (95% CI, 0.812.49) for pancreatic cancer. However, the dose of aspirin was relatively small, was taken intermittently, and only 73% of aspirin users took at least two thirds of the study medication. Whether higher doses of aspirin, taken more often, would be chemopreventive in other populations is unknown. A meta-analysis of aspirin use and disease risk, using data from 7 epidemiologic studies, plus the data from the earlier-reported trial, reported a summary estimate of 0.98 (95% CI, 0.86 1.13).116 However, this conclusion on aspirin may be difcult to interpret because not all assessed the daily use of aspirin115,117,118 and several did not adjust for smoking.117119 The latter probably is important because aspirin use and smoking may be associated negatively and consequently the latter could mask any protective effects of aspirin use itself. However, 4 observational epidemiologic studies did assess the daily use of aspirin and adjusted for smoking.120 123 A cohort study in postmenopausal women reported a signicant reduction in the incidence of pancreatic cancer with a relative risk of 0.40 (95% CI, 0.20 0.82) in those taking aspirin more than 6 times per week.120 In the other 3 studies either a positive,122 negative,121 or no effect123 was found, all of which
280
HART ET AL
11. Maisonneuve P, Marshall B, Lowenfels A. Risk of pancreatic cancer in patients with cystic brosis. Gut 2007;56:1327 1328. 12. Petrowsky H, Schuster H, Irani S, et al. Pancreatic cancer in cystic brosis after bilateral lung transplantation. Pancreas 2006;33:430 432. 13. Ojajarvi I, Partanen T, Ahlbom A, et al. Occupational exposures and pancreatic cancer: a meta-analysis. Occup Environ Med 2000;57:316 324. 14. Marsh GM. Critical review of epidemiological studies related to ingested asbestos. Environ Health Perspect 1983;53:49 56. 15. Collins J, Esmen N, Hall T. A review and meta-analysis of formaldehyde exposure and pancreatic cancer. Am J Ind Med 2001; 39:336 345. 16. Yu H, Rohan T. Role of the insulin-like growth factor family in cancer development and progression. J Natl Cancer Inst 2000; 92:14721489. 17. Jones J, Clemmons D. Insulin-like growth factors and their binding proteins: biological actions. Endocr Rev 1995;16:334. 18. Holmes M, Pollak M, Hankinson SE. Life-style correlates of plasma insulin-like growth factor 1 and insulin-like growth factor binding protein 3 concentrations. Cancer Epidemiol Biomarkers Prev 2002;11:862 867. 19. Silverman D, Swanson C, Gridely G, et al. Dietary and nutritional factors and pancreatic cancer: a case-control study based on direct interviews. J Natl Cancer Inst 1998;90:1710 1719. 20. Friedman G, Van Den Eeden S. Risk factors for pancreatic cancer: an explanatory study. Int J Epidemiol 1993;22:30 37. 21. Toyokuni S. Iron-induced carcinogenesis: the role of redox regulation. Free Radic Biol Med 1996;20:533566. 22. Bueno de Mesquita H, Maisonneuve P, Moerman C, et al. Anthropometric and reproductive variables and exocrine carcinoma of the pancreas: a population-based case-control study in The Netherlands. Int J Cancer 1992;52:24 29. 23. Teras L, Patel A, Rodriguez C, et al. Parity, other reproductive factors, and risk of pancreatic cancer mortality in a large cohort of U.S. women (United States). Cancer Causes Control 2005; 16:10351040. 24. Fernandez E, La Vecchia C, DAvanzo B, et al. Menstrual and reproductive factors and pancreatic cancer risk in women. Int J Cancer 1995;62:1114. 25. Kreiger N, LaCroix J, Sloan M. Hormonal factors and pancreatic cancer in women. Ann Epidemiol 2001;11:563567. 26. Skinner HG, Michaud D, Colditz G, et al. Parity, reproductive factors, and the risk of pancreatic cancer in women. Cancer Epidemiol Biomarkers Prev 2003;12:433 438. 27. Karlson B, Wuu J, Hsieh C, et al. Parity and the risk of pancreatic cancer: a nested case-control study. Int J Cancer 1988;77: 224 227. 28. Kvale G, Heuch I, Nilssen S. Parity in relation to mortality and cancer incidence: a prospective study of Norwegian women. Int J Epidemiol 1994;23:691 699. 29. Lin Y, Kikuchi S, Tamakoshi A, et al. Association of menstrual and reproductive factors with pancreatic cancer risk in women: ndings of the Japan Collaborative Cohort Study for evaluation of cancer risk. J Gastroenterol 2006;41:878 883. 30. Huxley R, Ansary-Moghaddam A, Berrington de Gonzalez A, et al. Type-II diabetes and pancreatic cancer: a meta-analysis of 36 studies. Br J Cancer 2005;92:2076 2083. 31. Ohmura E, Okada M, Onoda N, et al. Insulin-like growth factor I and transforming growth factor alpha as autocrine growth factors in human pancreatic cell growth. Cancer Res 1990;50: 103107. 32. Permert J, Ihse I, Jorfeldt L, et al. Improved glucose metabolism after subtotal pancreatectomy for pancreatic cancer. Br J Surg 1993;80:10471050. 33. Zendehdel K, Nyren O, Ostenson C, et al. Cancer incidence in
34.
35. 36.
37. 38.
39. 40. 41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
patients with type 1 diabetes mellitus: a population-based cohort study in Sweden. J Natl Cancer Inst 2003;95:1797 1800. Ekbom A, McLaughlin J, Karlsson B-M, et al. Pancreatitis and pancreatic cancer: a population-based study. J Natl Cancer Inst 1994;86:625 627. Malka D, Hammel P, Maire F, et al. Risk of pancreatic adenocarcinoma in chronic pancreatitis. Gut 2002;51:849 852. Karlson B-M, Ekbom A, Josefsson S, et al. The risk of pancreatic cancer following pancreatitis: an association due to confounding? Gastroenterology 1997;113:587592. Bansal P, Sonnenberg A. Pancreatitis is a risk factor for pancreatic cancer. Gastroenterology 1995;109:247251. Lowenfels A, Maisonneuve P, Calallini G, et al. Pancreatitis and the risk of pancreatic cancer. N Engl J Med 1993;328:1433 1437. Shibata A, Mack TM, Paganini-Hill A, et al. A prospective study of pancreatic cancer in the elderly. Int J Cancer 1994;58:46 49. Cuzick J, Babiker A. Pancreatic cancer, alcohol, diabetes mellitus and gall-bladder disease. Int J Cancer 1989;43:415 421. Ekbom A, Yuen J, Karlsson BM, et al. Risk of pancreatic and periampullar cancer following cholecystectomya populationbased cohort study. Dig Dis Sci 1996;41:387391. Kalapothaki V, Tzonou A, Hsieh C, et al. Tobacco, ethanol, coffee, pancreatitis, diabetes mellitus, and cholelithiasis as risk factors for pancreatic carcinoma. Cancer Causes Control 1993;4:375382. Chow W, Johansen C, Gridley G, et al. Gallstones, cholecystectomy and risk of cancers of the liver, biliary tract and pancreas. Br J Cancer 1999;79:640 644. Gullo L. Risk of pancreatic cancer and periampullary cancer following cholecystectomy. Ann Oncol 1999;10(Suppl 4):127 128. Ye W, Lagergren J, Nyren O, et al. Risk of pancreatic cancer after cholecystectomy: a cohort study in Sweden. Gut 2001;49:678 681. Haines AP, Moss AR, Whittemore A, et al. A case-control study of pancreatic carcinoma. J Cancer Res Clin Oncol 1982;103: 9397. Bueno de Mesquita H, Maisonneuve P, Moerman C, et al. Aspects of medical history and exocrine carcinoma of the pancreas: a population-based case-control study in the Netherlands. Int J Cancer 1992;52:1723. Maringhini A, Moreau J, Melton L, et al. Gallstones, gallbladder cancer, and other gastrointestinal malignancies. Ann Intern Med 1987;107:30 35. Ichimiya H, Kono S, Ikeda M, et al. Cancer mortality among patients undergoing cholecystectomy for benign biliary diseases. Jap J Cancer Res 1986;77:579 583. Rosenberg L, Duguid W, Brown RA. Cholecystectomy stimulates hypertrophy and hyperplasia in the hamster pancreas. 1984; 37:108 111. Watanapa P, Williamson R. Experimental pancreatic hyperplasia and neoplasia: effects of dietary and surgical manipulation. Br J Cancer 1993;67:877 884. Schernhammer E, Michaud D, Leitzmann M, et al. Gallstones, cholecystectomy, and the risk for developing pancreatic cancer. Br J Cancer 2002;86:10811084. Stolzenberg-Solomon R, Pietinen P, Taylor P, et al. A prospective study of medical conditions, anthropometry, physical activity, and pancreatic cancer in male smokers (Finland). Cancer Causes Control 2002;13:417 426. Vasen H, Gruis N, Frants R, et al. Risk of developing pancreatic cancer in families with familial atypical multiple mole melanoma associated with a specic 19 deletion of p16 (p16-Leiden). Int J Cancer 2000;87:809 811. Neugut A, Ahsan H, Robinson E. Pancreas cancer as a second
March 2008
281
56.
57.
58.
59.
60.
61.
62.
63.
64. 65.
66.
67.
68.
69.
70. 71. 72.
73.
74.
75. 76.
77.
primary malignancy. A population-based study. Cancer 1995; 76:589 592. Annibale B, Capurso G, Delle Fave G. Consequences of Helicobacter pylori infection on the absorption of micronutrients. Dig Liver Dis 2002;34(Suppl 2):S72S77. Jensen RT. Involvement of cholecystokinin/gastrin-related peptides and their receptors in clinical gastrointestinal disorders. Pharmacol Toxicol 2002;91:333350. Risch HA. Etiology of pancreatic cancer with a hypothesis concerning the role of N-nitroso compounds and excess gastric acidity. J Natl Cancer Inst 2003;95:948 960. Stolzenberg-Solomon R, Blaser MLP, Perez-Perez G, et al. Helicobacter pylori seropositivity as a risk factor for pancreatic cancer. J Natl Cancer Inst 2001;93:937941. Raderer M, Wrba F, Kornek G, et al. Association between Helicobacter pylori infection and pancreatic cancer. Oncology 1998; 55:16 19. Nilsson H, Stenram U, Ihse I, et al. Re: Helicobacter pylori seropositivity as a risk factor for pancreatic cancer. J Natl Cancer Inst 2002;94:632 633. Tascilar M, van Rees BP, Sturm PDJ, et al. Pancreatic cancer after remote peptic ulcer surgery. J Clin Pathol 2002;55:340 345. Mack TM, Yu MC, Hanisch R, et al. Pancreas cancer and smoking, beverage consumption and past medical history. J Natl Cancer Inst 1986;76:49 60. Ross AHM, Smith MA, Anderson JR, et al. Late mortality after surgery for peptic ulcer. N Engl J Med 1982;307:519 522. Lowenfels A, Maisonneuve P. Epidemiology and risk factors for pancreatic cancer. Best Pract Res Clin Gastroenterol 2006;20: 197209. Rivenson A, Hoffmann D, Prokopczyk B, et al. Induction of lung and exocrine pancreas tumours in F344 rats by tobacco-specic and areca-derived N-nitrosamines. Cancer Res 1988;48:6912 6917. Coughlin S, Calle E, Patel A, et al. Predictors of pancreatic cancer mortality among a large cohort of United States adults. Cancer Causes Control 2000;11:915923. Doll R, Peto R, Wheatley K, et al. Mortality in relation to smoking: 40 years observation on male British doctors. BMJ 1994; 309:901911. Zheng W, McLaughlin J, Gridley G, et al. A cohort study of smoking, alcohol consumption, and dietary factors for pancreatic cancer (United States). Cancer Causes Control 1993;4: 477 482. MacMahon B, Yen S, Trichopoulos D, et al. Coffee and cancer of the pancreas. N Engl J Med 1981;304:630 633. Wynder EL, Hall NEL, Polansky M. Epidemiology of coffee and pancreatic cancer. Cancer Res 1983;43:3900 3906. Whittemore AS, Paffenbarger RS, Anderson K, et al. Early precursors of pancreatic cancer in college men. J Chron Dis 1983; 36:251256. Lin Y, Tamakoshi A, Kawamura T, et al. A prospective cohort study of cigarette smoking and pancreatic cancer in Japan. Cancer Causes Control 2002;13:249 254. Silverman D, Dunn J, Hoover R, et al. Cigarette smoking and pancreas cancer: a case-control study based on direct interviews. J Natl Cancer Inst 1994;86:1510 1516. International Agency for Research on Cancer. Tobacco smoking. IARC Monogr Eval Carcinog Risks Hum 1986;38:83126. Villeneuve P, Johnson K, Mao Y, et al. Environmental tobacco smoke and the risk of pancreatic cancer: ndings from a Canadian population-based case-control study. Can J Public Health 2004;95:3237. Stolzenberg-Solomon R, Pietinen P, Taylor P, et al. Prospective study of diet and pancreatic cancer in male smokers. Am J Epidemiol 2002;155:783792.
78. Harnack L, Anderson K, Zheng W, et al. Smoking, alcohol, coffee, and tea intake and incidence of cancer of the exocrine pancreas: The Iowa Womens Health Study. Cancer Epidemiol Biomarkers Prev 1997;6:10811086. 79. Stolzenberg-Solomon R, Pietinen P, Taylor P, et al. A prospective study of medical conditions, anthropometry, physical activity, and pancreatic cancer in male smokers (Finland). Cancer Causes Control 2002;13:417 426. 80. Berrinton de Gonzalez A, Spencer E, Bueno-de-Mesquita HB, et al. Anthropometry, physical activity, and the risk of pancreatic cancer in the European Prospective Investigation into Cancer and Nutrition. Cancer Epidemiol Biomarkers Prev 2006;15: 879 885. 81. Michaud D, Giovannucci E, Willett W, et al. Physical activity, obesity, height, and the risk of pancreatic cancer. JAMA 2001; 286:921929. 82. Yoshimoto M, Tsutsumi M, Iki K, et al. Carcinogenicity of heterocyclic amines for the pancreatic duct epithelium in hamsters. Cancer Lett 1999;143:235239. 83. Larsson S, Hakanson N, Permert J, et al. Meat, sh, poultry and egg consumption in relation to risk of pancreatic cancer; a prospective study. Int J Cancer 2006;118:2866 2870. 84. Nothlings U, Wilkens L, Murphy SP, et al. Meat and fat intake as risk factors for pancreatic cancer: the multiethnic cohort study. J Natl Cancer Inst 2005;97:1458 1465. 85. Michaud DGE, Willett W, Colditz G, et al. Dietary meat, dairy products, fat, and cholesterol and pancreatic cancer risk in a prospective study. Am J Epidemiol 2003;157:11151125. 86. Isaksson B, Jonsson F, Pedersen N, et al. Lifestyle factors and pancreatic cancer risk: a cohort study from the Swedish Twin Registry. Int J Cancer 2002;98:480 482. 87. Stolzenberg-Solomon R, Graubard B, Chari S, et al. Insulin, glucose, insulin resistance, and pancreatic cancer in male smokers. JAMA 2005;294:28722878. 88. Jee S, Ohrr H, Sull J, et al. Fasting serum glucose level and cancer risk in Korean men and women. JAMA 2005;293:194 202. 89. Michaud D, Liu S, Giovannucci E, et al. Dietary sugar, glycaemic load, and pancreatic cancer risk in a prospective study. J Natl Cancer Inst 2002;94:12931300. 90. Silvera S, Rohan T, Jain M, et al. Glycaemic index, glycaemic load, and pancreatic cancer risk (Canada). Cancer Causes Control 2005;16:431 436. 91. Larsson S, Bergkvist L, Wolk A. Consumption of sugar and sugar-sweetened foods and the risk of pancreatic cancer in a prospective study. Am J Clin Nutr 2006;84:11711176. 92. Schernhammer E, Hu F, Giovannucci E, et al. Sugar-sweetened soft drink consumption and risk of pancreatic cancer in two prospective cohorts. Cancer Epidemiol Biomarkers Prev 2005; 14:2098 2105. 93. Roebuck B, Kaplita P, Edwards B, et al. Effects of dietary fats and soybean protein on azaserine-induced pancreatic carcinogenesis and plasma cholecystokinin in the rat. Cancer Res 1987;47:13331338. 94. Mills PK, Beeson WL, Abbey DE, et al. Dietary habits and past medical history as related to fatal pancreas cancer among Adventists. Cancer 1988;61:2578 2585. 95. Sauvaget C, Nagano J, Hayashi M, et al. Vegetables and fruit intake and cancer mortality in the Hiroshima/Nagasaki Life Span Study. Br J Cancer 2003;88:689 694. 96. Larsson S, Hakansson N, Naslund I, et al. Fruit and vegetable consumption in relation to pancreatic cancer risk: a prospective study. Cancer Epidemiol Biomarkers Prev 2006;15:301305. 97. Larsson S, Hakansson N, Giovannucci E, et al. Folate intake and pancreatic cancer incidence: a prospective study of Swedish women and men. J Natl Cancer Inst 2006;98:407 413. 98. Stolzenberg-Solomon R, Pietinen P, Barrett M, et al. Dietary and
282
HART ET AL
99.
100.
101.
102. 103.
104.
105.
106.
107.
108.
109.
110.
111.
other methyl-group availability factors and pancreatic cancer risk in a cohort of male smokers. Am J Epidemiol 2001; 153:680 687. Skinner HG, Michaud D, Giovannucci E, et al. A prospective study of folate intake and the risk of pancreatic cancer in men and women. Am J Epidemiol 2004;160:248 258. Stolzenberg-Solomon R, Albanes D, Nieto F, et al. Pancreatic cancer risk and nutrition-related methyl-group availability indicators in male smokers. J Natl Cancer Inst 1999;91:535541. International Agency for Research on Cancer. Coffee, tea, mate, methylxanthines and methylglyoxal. IARC Monogr Eval Carcinog Risks Hum 1991;51:1513. International Agency for Research on Cancer. Alcohol drinking. IARC Monogr Eval Carcinog Risks Hum 1988;44:1378. Michaud D, Giovannucci E, Willett W, et al. Coffee and alcohol consumption and the risk of pancreatic cancer in two prospective United States Cohorts. Cancer Epidemiol Biomarkers Prev 2001;10:429 437. Lin Y, Tamakoshi A, Kawamura T, et al. Risk of pancreatic cancer in relation to alcohol drinking, coffee consumption and medical history: ndings from the Japan Collaborative cohort study for evaluation of cancer risk. Int J Cancer 2002;99:742 746. Berrington de Gonzalez A, Sweetland S, Spencer E. A metaanalysis of obesity and the risk of pancreatic cancer. Br J Cancer 2003;89:519 523. Larsson S, Permert J, Hakansson N, et al. Overall obesity, abdominal obesity, diabetes and cigarette smoking in relation to the risk of pancreatic cancer in two Swedish populationbased cohorts. Br J Cancer 2005;93:1310 1315. Patel A, Rodriguez C, Bernstein L, et al. Obesity, recreational physical activity, and risk of pancreatic cancer in a large US Cohort. Cancer Epidemiol Biomarkers Prev 2005;14:459 466. Nothlings U, Wilkens L, Murphy S, et al. Body mass index and physical activity as risk factors for pancreatic cancer: the Multiethnic Cohort Study. Cancer Causes Control 2007;18:165 175. Lin Y, Kikuchi S, Tamakoshi A, et al. Obesity, physical activity and the risk of pancreatic cancer in a large Japanese cohort. Int J Cancer 2007;120:26652671. Luo J, Iwasaki M, Inoue M, et al. Body mass index, physical activity and the risk of pancreatic cancer in relation to smoking status and history of diabetes: a large-scale population-based cohort study in Japanthe JPHC study. Cancer Causes Control 2007;18:603 612. Lee I, Sesso H, Oguma Y, et al. Physical activity, body weight, and pancreatic cancer mortality. Br J Cancer 2003;88:679 683.
112. Sinner P, Schmitz K, Anderson K, et al. Lack of association of physical activity and obesity with incident pancreatic cancer in elderly women. Cancer Epidemiol Biomarkers Prev 2005;14: 15711573. 113. Dannenberg A, Altorki N, Boyle J, et al. Cyclo-oxygenase 2: a pharmacological target for the prevention of cancer. Lancet Oncol 2001;2:544 551. 114. Ding X, Kuszynski C, El-Metwally T, et al. Lipoxygenase inhibition induced apoptosis, morphological changes, and carbonic anhydrase expression in human pancreatic cancer cells. Biochem Biophys Res Commun 1999;266:392399. 115. Cook N, Lee I, Gaziano JM, et al. Low-dose aspirin in the primary prevention of cancer: the womens health study: a randomised controlled trial. JAMA 2005;294:4755. 116. Larsson S, Giovannucci E, Bergkvist L, et al. Aspirin and nonsteroidal anti-inammatory drug use and risk of pancreatic cancer: a meta-analysis. Cancer Epidemiol Biomarkers Prev 2006; 1:25612564. 117. Schreinemachers D, Everson RB. Aspirin use and lung, colon and breast cancer incidence in a prospective study. Epidemiology 1994;5:138 146. 118. Ratnasinghe L, Graubard B, Kahle L, et al. Aspirin use and mortality from cancer in a prospective cohort study. Anticancer Res 2004;24:31773184. 119. Friis S, Sorensen H, McLaughlin J, et al. A population-based cohort study of the risk of colorectal and other cancers among users of low-dose aspirin. Br J Cancer 2003;88:684 688. 120. Anderson K, Johnson T, Lazovich D, et al. Association between nonsteroidal anti-inammatory drugs and the incidence of pancreatic cancer. J Natl Cancer Inst 2002;94:1168 1171. 121. Menezes R, Huber K, Mahoney M, et al. Regular use of aspirin and pancreatic cancer risk. BMC Public Health 2002;2:18. 122. Schernhammer E, Kang J, Chan A, et al. A prospective study of aspirin use and the risk of pancreatic cancer in women. J Natl Cancer Inst 2004;96:2228. 123. Jacobs E, Connell C, Rodriquez C, et al. Aspirin use and pancreatic cancer mortality in a large United States cohort. J Natl Cancer Inst 2004;96:524 528. 124. Khurana V, Sheth A, Caldito G, et al. Statins reduce the risk of pancreatic cancer in humans: a case-control study of half a million veterans. Pancreas 2007;34:260 265.
Address requests for reprints to: Dr Andrew R. Hart, MB, CHB. Population Health, School of Medicine, Health Policy and Practice, University of East Anglia, Norwich, Norfolk, NR4 7TJ, United Kingdom. e-mail: a.hart@uea.ac.uk; fax: (0044) 1603-593752.
CLINICAL IMAGING
Computerized Tomography Enterography and Its Role in Small-Bowel Imaging
JOEL G. FLETCHER,* JAMES HUPRICH,* EDWARD V. LOFTUS, JR, DAVID H. BRUINING, and JEFF L. FIDLER*
*Department of Radiology, and Division of Gastroenterology & Hepatology, Mayo Clinic Rochester, Rochester, Minnesota
Computerized tomography enterography is being adopted rapidly by many institutions as the primary technique used to image the small bowel. The purpose of this article is to summarize how the examination is performed, and to specify appropriate indications and alternatives, performance characteristics, and anticipated future developments. omputerized tomography (CT) enterography is a CT of the abdomen and pelvis designed to image the lumen and wall of the small bowel. Such optimization requires luminal distention (with a larger than normal volume of ingested oral contrast material), high-resolution imaging, and the appropriate phase of intravenous contrast enhancement. Although positive (high attenuation) oral contrast can be used, the oral contrast used for CT enterography is generally a neutral enteric contrast, meaning that its CT attenuation is similar to water. When used with iodinated intravenous contrast agents at CT enterography, neutral oral contrast agents maximize the conspicuity of hypervascular pathologies such as inamed bowel loops or masses. For a CT enterography examination, 1500 to 2000 mL of oral contrast is ingested over approximately 60 minutes before the examination to distend the small bowel. Polyethylene glycol solution and low-concentration barium suspension (which contains sorbitol) are the most commonly used oral contrast agents because they result in superior small-bowel distention relative to water, but may cause self-limited diarrhea or gas shortly after the examination.1 Although the examination is highly robust in an outpatient setting, gastroenterologists should be aware that hospitalized patients or patients who are unable to drink large volumes of contrast usually cannot undergo CT enterography successfully. Contraindications also include severe contrast dye allergies and signicant renal insufciency. To adequately visualize the small-bowel lumen and wall at CT enterography, slice thicknesses of 3 mm or less are used in an overlapping fashion. Because of the increased number of images generated during a CT enterography examination, gastroenterologists will nd it most helpful to review images in an electronic fashion, with routine coronal images generated, to facilitate visualization of tortuous small-bowel loops and stulae. Maximal mural enhancement of the normal small bowel occurs 50 seconds after intravenous contrast material injection or 14 seconds after peak aortic enhancement2; CT imaging normally is initiated at this time. However, CT scanning up to 70 seconds after contrast injection permits routine hepatic
phase imaging, but does not compromise the ability to identify Crohns disease.3,4 Imaging of the small bowel may be performed in multiple vascular phases in the setting of occult gastrointestinal (GI) bleeding (see later).
Normal Findings at Computerized Tomography Enterography

Gastroenterologists should be familiar with the normal appearance of the small bowel at CT enterography. Although the jejunum possesses the valvulae conniventes, which are seen as thin, closely spaced folds resulting in a feathery appearance, the ileum possesses few, if any, folds. The small-bowel wall should be less than 3 mm in thickness when the lumen is distended by uid. When imaging is performed in the enteric phase of maximal small-bowel enhancement, the jejunum normally will have greater attenuation than the ileum, with a gradual decrease in attenuation of the enhancing small-bowel wall as one moves toward the terminal ileum. Intramural fat can be a normal nding in the terminal ileum, but is a sign of chronic inammation elsewhere in the small bowel. In Crohns disease, it often is accompanied by mucosal hyperenhancement, pointing to both active and chronic inammation. The jejunum occasionally is collapsed at enterography, which can be a normal nding in the minority of cases, but the ileum almost always is distended adequately.5 Notwithstanding this weakness, CT enterography is acceptable for most small-bowel indications, particularly given the fact that many small-bowel diseases, such as Crohns, or masses will mildly obstruct the lumen as a result of transmural inammation, edema, or involvement. Moreover, mural and mesenteric pathologies often are occult at luminal imaging such as uoroscopy or capsule endoscopy.
Imaging Findings at Computerized Tomography Enterography

Although multiple indications exist for CT enterography (Table 1), the overwhelming indication is known or suspected Crohns disease. In this setting, the goal of CT enterography is to identify active Crohns disease, CT ndings
Abbreviations used in this paper: CT, computerized tomography; GI, gastrointestinal; MR, magnetic resonance. 2008 by the AGA Institute 1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.049
284
FLETCHER ET AL
Table 1. Indications for CT Enterography

Type of examination Single-phase CT enterography Indication Rule out Crohns disease Stage Crohns disease activity Rule out small-bowel mass Assess for complications of celiac sprue Unexplained diarrhea Occult GI bleeding
Multiphase CT enterography
indicating bowel wall and adjacent mesenteric inammation, stulizing disease (including abscess), small-bowel obstruction (if present), and any extraintestinal manifestations. Findings of active inammatory Crohns disease include mural hyperenhancement, thickening, and stratication, in addition to peri-enteric fat stranding and the comb sign. Mural hyperenhancement is segmental attenuation greater than adjacent small-bowel loops and is the most sensitive CT nding for active inammatory Crohns disease.6 Mural thickening is considered present when the small-bowel wall is greater than 3 mm in thickness and the lumen is lled with oral contrast. Although segmental mural hyperenhancement alone is a nonspecic nding for Crohns disease, asymmetric mural enhancement and thickening are pathognomonic for Crohns disease (Figure 1). Mural stratication refers to a laminated appearance to the bowel wall and is seen with more advanced small-bowel inammation (Figure 2).7 With mural stratication, the mucosa is hyperenhancing, with the submucosa being of lower attenuation representing edema (water attenuation) or inammatory inltrate (soft-tissue attenuation). Because the transmural inammation of Crohns disease extends into the mesentery, there is stranding in the peri-enteric fat, which is associated with increased serum concentrations of C-reactive protein.8 Fibro-fatty proliferation is distinct from increased fat density and refers to the surgical and imaging ndings of fatty proliferation along the mesenteric side of the bowel, which may participate in the Crohns inammatory response, and typically reects chronic Crohns inammation.9 The comb sign heralds signicant inammation in the small bowel or colon and refers to engorged vasa recta (or the vessels supplying the bowel).10 When these vessels become enlarged during acute inammation, they resemble the teeth of a comb (Figure 2). The comb sign is associated with increased serum C-reactive protein levels and length of hospital stay during Crohns ares.11 Penetrating disease is seen frequently at CT enterography in the Crohns disease population (up to 20% of the time12). Penetrating complications, particularly entero-enteric and enterocolic stulas, frequently are unsuspected clinically yet identied incidentally during CT enterography (Figure 2). Fistulas generally appear as hyperenhancing extra-enteric tracks with or without air or uid.13 Although routine CT can detect stulae and abscess,14 CT enterography can describe the type of stula (ie, its connections), its size, whether or not it is associated with proximal obstruction, whether it arises from active inammatory disease, or whether it is associated with abscess.13 In our experience, smaller stulas associated with active disease will resolve with medical treatment. Perianal stulas or stulas arising in the postoperative setting unassociated with Crohns disease may not be hyperenhancing, potentially because of their lack of acute inammatory response.
In a large retrospective review of patients, Bruining et al12 found that extraintestinal manifestations of Crohns disease were seen in about 12% to 15% of patients, including sacroiliitis, nephrolithiasis, and portal or mesenteric vein thrombosis. Moreover, noninammatory bowel diseaserelated abnormalities including pulmonary nodules, adrenal nodules, lymphoma, and other malignancies were detected. Certainly, patients with Crohns disease also have an increased susceptibility to enteric and hepatic malignancies and lymphoma.15 CT enterography also can be used for a variety of nonCrohns indications. Besides active Crohns disease, the most common nding at CT enterography is a mass. Masses can include carcinoid tumors, adenocarcinomas, lymphomas, GI stromal tumors, and Meckels diverticula (and their complications). Carcinoid tumors are the most frequently seen tumors in our practice, and are seen as hyperenhancing polypoid or carpet lesions that can invade the mesentery, with characteristic softtissue attenuation mesenteric metastases (Figure 3). The appearance of other small-bowel tumors and celiac disease at CT enterography has been described elsewhere.6
Performance of Computerized Tomography Enterography for Active Crohns Disease

Before 2000, sensitivity estimates of CT enterography for active inammatory Crohns disease were based largely on surgical or uoroscopic series without a reliable endoscopic
Figure 1. Asymmetric mural hyperenhancement in Crohns disease. Note additional presence of intramural fat, indicating there also has been chronic inammation.
March 2008
CT ENTEROGRAPHY
285
Figure 2. (A) Enterocolic stula (arrows) arising from the top of an inamed jejunal loop (large arrowhead) and connecting to the hepatic exure. (B) The proximal jejunum shows mural stratication (large arrow), with the distal jejunum being decompressed (arrowhead). (B) Engorged vasa recta (or the comb sign; small arrows) supply the (C) mildly inamed proximal jejunal loop (arrows), which contains an enterolith (arrowhead).
reference standard. Since that time, most assessments have shown a sensitivity for active inammatory Crohns disease on the order of 80% to 90%,3,4,16,17 with some showing sensitivity around 60%.18,19 When quantitative and visual measures of segmental hyperenhancement are compared with endoscopy and small-bowel biopsy, the odds ratio for active disease is greater than 10, and the sensitivity for histologic inammation is about 80%, similar to endoscopic assessment alone.3 We recently conducted a head-to-head trial comparing CT enterography, capsule endoscopy, ileocolonoscopy, and small-bowel
follow-through in 42 patients with suspected or known Crohns disease. The sensitivity of capsule endoscopy and CT enterography was almost identical (83% and 82%, respectively), with the specicity of CT being signicantly higher (89% vs 53%; P .02).16 In addition, although we excluded patients with symptoms of obstructive disease, 17% of the patients had asymptomatic strictures, which precluded capsule assessment. In a secondary analysis of 27 patients who received both capsule endoscopy and CT enterography, capsule endoscopy identied 2 additional patients with active Crohns disease who had no
286
FLETCHER ET AL
endoscopy, the capsule failed to reach the colon in 25% of the patients. In the Voderholzer et al18 experience, only 1 of 41 patients had a retained capsule. In summary, CT enterography is moderately sensitive and highly accurate for active Crohns disease and can exclude signicant strictures that may result in retained capsules. Nevertheless, capsule endoscopy may identify some patients with mild active inammatory disease occult at CT imaging.
Clinical Contribution of Computerized Tomography Enterography

CT enterography provides objective evidence of active Crohns disease or small-bowel mass proximal to the reach of the endoscope, as well as providing visualization of the mesentery and liver. Studies examining the clinical benet of CT enterography are beginning to emerge.12,20 Higgins et al20 recently had gastroenterologists review clinical information before the revelation of CT enterography ndings, to make a clinical assessment on whether or not patients would benet from corticosteroid treatment. Subsequent to these predictions, CT enterography information was provided. Their study showed that CT enterography did not replicate original clinical impressions, and changed the impression of corticosteroid benet in 61% of these patients. Similarly, in our experience, approximately half of the patients with penetrating Crohns disease had no symptoms or signs of a stula at their initial clinical assessment.21 Detection of unsuspected stulizing Crohns disease resulted in either escalation of medical therapy or surgical/procedural intervention in the majority of these patients. Prospective trials detailing the effect of CT enterography results on clinical decision making are needed.
Multiphase Computerized Tomography Enterography for Occult Gastrointestinal Bleeding

Although earlier radiologic literature was disappointing for the ability of CT to detect the causes of occult GI bleeding,22 recent work with multiphase CT enterography has shown that CT has an important role to play in this clinical setting. Because of the requirement for rapid multiphase imaging, this technique is best performed on a 64-channel (or higher) CT scanner.23 Neutral oral contrast and iodinated intravenous contrast are the same as for a single-phase examination. Scanning is performed from the diaphragm to the symphysis pubis during each of 3 phases, with scanning initiated when a region-ofinterest attenuation threshold in the aorta is reached. The rst phase is a bolus-triggered arterial phase, followed by an enteric phase (20 seconds after trigger), and a delayed phase (70 seconds after trigger). Once the region-of-interest trigger threshold is reached by iodinated contrast arriving in the aorta, the arterial phase is initiated after a short 6-second delay. During the arterial phase, vascular ectasias such as AVMs and their early draining veins are seen. The enteric phase often highlights enhancing tumors, with delayed images showing that iodinated contrast is accumulating in the small-bowel lumen, indicating active bleeding (Figure 4). Huprich et al23 showed the benet of CT enterography in identifying multiple small-bowel tumors and vascular lesions detected and undetected at capsule endoscopy, with a larger prospective study underway. In the
Figure 3. Ileal carcinoid with typical mesenteric metastasis. (A) Axial image shows an ileal loop with focal mural thickening and enhancement (arrowhead) associated with a soft-tissue mesenteric mass with excentric calcications and stranding to adjacent small-bowel loops (arrow). (B) Coronal image shows similar ndings (bowel loops, arrowheads; mesenteric metastasis, arrow). The surgical specimen showed a cluster of mural carcinoids with metastases to local nodes.
evidence of Crohns disease at CT imaging, but again showed signicantly lower specicity (53% vs 100%). Although a recent trial by Voderholzer et al18 showed superior performance of capsule endoscopy versus CT enteroclysis (ie, with a nasojejunal tube) for patients undergoing both tests, 27% of their study population was excluded from undergoing capsule endoscopy because of small-bowel strictures at CT enteroclysis. Moreover, in the 41 patients who underwent CT enterography and capsule
March 2008
CT ENTEROGRAPHY
287
Figure 4. Active bleeding at multiphase CT enterography. (A) Arterial, (B) enteric, and (C) delayed phase images show progressive focal contrast accumulation (arrows in A, B, and C) in an ileal angiodysplasia conrmed at intraoperative endoscopy. (C) Additional focus (arrows) of contrast on delayed image is also presumed to be active bleeding.
setting of occult GI bleeding, CT enterography and capsule endoscopy are not competitive imaging modalities but complementary ones, because ndings from each imaging study may be nonspecic.
Appropriate Use of Cross-Sectional Alternatives to Computerized Tomography Enterography

Table 2 lists alternative tests preferred over CT enterography in certain clinical settings. Because CT and magnetic resonance (MR) enteroclysis (after nasojejunal tube placement) permit the rapid infusion of large volumes of enteric contrast at faster rates, these examinations better show small-bowel caliber changes reecting low-grade obstruction. Similarly, because CT
enterography can result in collapsed jejunal loops, we perform CT or MR enteroclysis to investigate further nonspecic capsule endoscopy ndings in the jejunum. CT enteroclysis is not performed for patients with known or suspected Crohns disease or occult GI bleeding because of the acceptable performance characteristics of CT enterography, patient objections to, and time required for nasojejunal tube placement. Because of the marked increase in signal differences between pathology and anatomic structures near the anus, we perform MR imaging to stage perianal disease or investigate potential J-pouch complications after ileal pouchanal anastomosis. MR enterography is preferred to CT enterography when imaging to determine if inammed bowel loops or stulas have responded to therapy (owing to lack of radiation), and in patients with contraindica-
288
FLETCHER ET AL
Table 2. Cross-Sectional Alternatives to CT Enterography and Suggested Indications

Type of examination MR or CT enteroclysis Indication Suspected low-grade partial small-bowel obstruction Questionable or nonspecic capsule endoscopy ndings, particularly if jejunal Staging of perianal stulas Suspected complications after ileal pouchanal anastomosis Monitoring response to therapy Fluoroscopic, dynamic characterization of stenotic or adhesive lesions Contraindications to CT
Table 3. Relative Strengths and Weaknesses of MR vs CT Enterography

MR Strengths Visualize bowel wall Bowel wall signal Mesenteric signal to noise ratio Spatial resolution Temporal resolution Ionizing radiation Dynamic characterization of stenotic lesions Amount of data Length of examination CT
Pelvic MR
MR enterography
NOTE. Relative strength: ; weekness: .
tions to CT. In addition, there are several advantages to MR imaging not seen at CT enterography, which include the ability to perform MR uoroscopy to observe bowel peristalsis in real time and dynamically characterize stenotic lesions.24 Table 3 lists the relative strengths of CT versus MR enterography. In our experience, CT enterography is used frequently as a rstline test given its speed (scanning takes seconds rather than 45 60 min), robustness (it rarely fails to generate high-quality images), spatial and temporal resolution (which are important
for visualizing small masses and stulizing disease), and CT scanner access.
The Future
We believe that CT enterography will continue to be incorporated into wider clinical measures of Crohns disease, particularly given the promise that objective CT ndings such as mural hyperenhancement can be quantitated.3,25 Moreover, we anticipate that continuing technical developments in CT image reconstruction will substantially reduce the radiation
Figure 5. Potential for dramatic radiation dose reduction at CT enterography shown using dual-energy CT enterography. (A) Normal dose CT enterography is achieved by combining the output from 2 radiograph tubes (each with half the normal dose), and shows mild wall thickening and mural hyperenhancement in the ascending colon (arrow) compared with the normal enhancement in the descending colon (arrowhead). (B) The 140-kV tube and (C) the 80-kV tube also are diagnostic, but were performed at half the radiation dose. Note the increased conspicuity of the iodine signal showing mural enhancement at the lower tube energy.
March 2008
CT ENTEROGRAPHY
289
dose at CT enterography, which is already the same or less than routine abdominal CT. The use of dual-source CT systems will permit wider use of low-energy CT scanning, which will increase the conspicuity of hypervascular inammation and permit further radiation dose reduction (Figure 5). Developments in MR technology will facilitate rapid growth of MR enterography. Finally, we believe that there will be routine correlation of CT enterography and capsule endoscopy ndings, and that these examinations will be used in combination in more patients, combining the exquisite mucosal visualization of capsule endoscopy with the mural and mesenteric images of CT enterography. References
1. Young B, Fletcher J, Booya F, et al. Head-to-head comparison of oral contrast agents for cross-sectional enterography: small bowel distention, timing and side effects. J Comput Assist Tomogr 2008;32:3238. 2. Schindera ST, Nelson RC, DeLong DM, et al. Multi-detector row CT of the small bowel: peak enhancement temporal window initial experience. Radiology 2007;243:438 444. 3. Bodily KD, Fletcher JG, Solem CA, et al. Crohn disease: mural attenuation and thickness at contrast-enhanced CT enterography correlation with endoscopic and histologic ndings of inammation. Radiology 2006;238:505516. 4. Booya F, Fletcher JG, Huprich JE, et al. Active Crohn disease: CT ndings and interobserver agreement for enteric phase CT enterography. Radiology 2006;241:787795. 5. Wold P, Fletcher J, Johnson C, et al. Assessment of small bowel Crohn disease: noninvasive perioral enterography compared with other imaging methods and endoscopy-feasibility study. Radiology 2003;229:275281. 6. Paulsen SR, Huprich JE, Fletcher JG, et al. CT enterography as a diagnostic tool in evaluating small bowel disorders: review of clinical experience with over 700 cases. Radiographics 2006;26: 641 662. 7. Macari M, Megibow AJ, Balthazar EJ. A pattern approach to the abnormal small bowel: observations at MDCT and CT enterography. AJR Am J Roentgenol 2007;188:1344 1355. 8. Colombel JF, Solem CA, Sandborn WJ, et al. Quantitative measurement and visual assessment of ileal Crohns disease activity by computed tomography enterography: correlation with endoscopic severity and C reactive protein. Gut 2006;55:15611567. 9. Desreumaux P, Ernst O, Geboes K, et al. Inammatory alterations in mesenteric adipose tissue in Crohns disease. Gastroenterology 1999;117:73 81. 10. Meyers MA, McGuire PV. Spiral CT demonstration of hypervascularity in Crohns disease: vascular jejunization of the ileum or the comb sign. Abdom Imaging 1995;20:327332. 11. Lee S, Ha H, Yang S, et al. CT of prominent pericolic or perienteric vasculature in patients with Crohns disease: correlation with
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24. 25.
clinical disease activity and ndings on barium studies. AJR Am J Roentgenol 2002;179:1029 1036. Bruining DH, Fletcher JG, Tremaine WJ, et al. Prevalence of penetrating disease and extraintestinal manifestations of Crohns disease with CT enterography. Gastroenterology 2007; 132:650. Vogel J, da Luz Moreira A, Baker M, et al. CT Enterography for Crohns disease: accurate preoperative diagnostic imaging. Dis Colon Rectum 2007;50:17611769. Maconi G, Sampietro GM, Parente F, et al. Contrast radiology, computed tomography and ultrasonography in detecting internal stulas and intra-abdominal abscesses in Crohns disease: a prospective comparative study. Am J Gastroenterol 2003;98: 15451555. Bernstein CN, Blanchard JF, Kliewer E, et al. Cancer risk in patients with inammatory bowel disease: a population-based study. Cancer 2001;91:854 862. Solem CA, Loftus LV, Fletcher JG, et al. Small bowel (SB) imaging in Crohns disease (CD): a prospective, blinded, 4-way comparison trial. Gastroenterology 2005;128(Suppl 2):A-74. Hassan C, Cerro P, Zullo A, et al. Computed tomography enteroclysis in comparison with ileoscopy in patients with Crohns disease. Int J Colorectal Dis 2003;18:121125. Voderholzer WA, Beinhoelzl J, Rogalla P, et al. Small bowel involvement in Crohns disease: a prospective comparison of wireless capsule endoscopy and computed tomography enteroclysis. Gut 2005;54:369 373. Hara A, Leighton J, Sharma V, et al. Small bowel: preliminary comparison of capsule endoscopy with barium study and CT. Radiology 2004;230:260 265. Higgins PD, Caoili E, Zimmermann M, et al. Computed tomographic enterography adds information to clinical management in small bowel Crohns disease. Inamm Bowel Dis 2007;13:262268. Booya F. CT enterography ndings in stulizing Crohns disease (abstract 1982 GI-e). In: Radiological Society of North America Scientic Assembly and Annual Meeting, Chicago, IL, 2004:768. Hara AK, Leighton JA, Sharma VK, et al. Small bowel: preliminary comparison of capsule endoscopy with barium study and CT. Radiology 2004;230:260 265. Huprich JE, Fletcher JG, Alexander J, et al. Obscure GI bleeding: 64-channel multiphase, multiplanar CT enterography has a role. Radiology 2008;246:562571. Fidler J. MR imaging of the small bowel. Radiol Clin North Am 2007;45:317331. Loftus E. Objective measures of disease activity: alternatives to symptom indices. Rev Gastroenterol Dis 2007;7:S8 S16.
Address requests for reprints to: J. G. Fletcher, MD, Department of Radiology, 200 First St. SW, Mayo Building, West 2-B, Rochester, Minnesota 55905. e-mail: etcher.joel@mayo.edu; fax: (507) 2664609.
ENDOSCOPY CORNER
Unresectable Cholangiocarcinoma: Comparison of Survival in Biliary Stenting Alone Versus Stenting With Photodynamic Therapy
MICHEL KAHALEH,* RAJNISH MISHRA,* VANESSA M. SHAMI,* PATRICK G. NORTHUP,* CARL L. BERG,* PENNY BASHLOR,* PETRA JONES,* KRISTI ELLEN,* GEOFFREY R. WEISS, CHRISTIANA M. BRENIN, BARBARA E. KURTH, TYVIN A. RICH, REID B. ADAMS, and PAUL YEATON*
*Digestive Health Center, Cancer Center, Departments of Radiation Oncology and Surgery, University of Virginia Health System, Charlottesville, Virginia
See Editorial on page 266. Background & Aims: Photodynamic therapy (PDT) for unresectable cholangiocarcinoma is associated with improvement in cholestasis, quality of life, and potentially survival. We compared survival in patients with unresectable cholangiocarcinoma undergoing endoscopic retrograde cholangiopancreatography (ERCP) with PDT and stent placement with a group undergoing ERCP with stent placement alone. Methods: Forty-eight patients were palliated for unresectable cholangiocarcinoma during a 5-year period. Nineteen were treated with PDT and stents; 29 patients treated with biliary stents alone served as a control group. Multivariate analysis was performed by using Model for End-Stage Liver Disease score, age, treatment by chemotherapy or radiation, and number of ERCP procedures and PDT sessions to detect predictors of survival. Results: KaplanMeier analysis demonstrated improved survival in the PDT group compared with the stent only group (16.2 vs 7.4 months, P<.004). Mortality in the PDT group at 3, 6, and 12 months was 0%, 16%, and 56%, respectively. The corresponding mortality in the stent group was 28%, 52%, and 82%, respectively. The difference between the 2 groups was signicant at 3 months and 6 months but not at 12 months. Only the number of ERCP procedures and number of PDT sessions were signicant on multivariate analysis. Adverse events specic to PDT included 3 patients with skin phototoxicity requiring topical therapy only. Conclusions: ERCP with PDT seems to increase survival in patients with unresectable cholangiocarcinoma when compared with ERCP alone. It remains to be proved whether this effect is attributable to PDT or the number of ERCP sessions. A prospective randomized multicenter study is required to conrm these data.
holangiocarcinoma is the second most common malignancy arising within the liver and is associated with signicant morbidity and mortality.1 The majority of patients are found to be unresectable on presentation2 4; their survival is approximately 3 months without intervention and 4 6 months with biliary decompression.510 Successful palliation of biliary
obstruction remains the main goal for reducing morbidity and mortality in these patients with unresectable disease.6,11 Hepaticojejunostomy is associated with a 30-day postoperative mortality rate between 7% and 24%,1216 and signicant quality of life after surgery is only improved in a minority because of time needed to recover.9,17,18 Endoscopy with placement of biliary stents has become the standard to palliate jaundice in this population, with less morbidity and mortality than that associated with surgery.13,18 23 Limitations of this approach are related to (1) the ability to decompress affected proximal segments and (2) recurrent stent occlusion, because these stents offer no ability to remodel malignant tissues.6,7,12,24 26 Photodynamic therapy (PDT) is an evolving therapy that involves the intravenous administration of a photosensitizing agent followed by its activation by using light illumination of a specic wavelength, resulting in ischemic necrosis2729 proportional to tissue oxygenation.30,31 PDT was demonstrated to reduce xenografted human cholangiocarcinoma tumor volume by 60% in a mouse model.32 In uncontrolled human studies in which pormer sodium based PDT was combined with stenting, there was improvement in cholestasis and survival, and few complications related to pormer sodium were observed.3336 A prospective, randomized, controlled trial conrmed a signicant advantage attributable to PDT in relief of jaundice, quality of life, and survival.37 The improvement of survival in the PDT group was such that it was considered unethical to continue the study after the rst 39 patients were randomized (20 received PDT). This study was criticized because enrollment was limited to those patients in whom technically successful stenting did not result in successful drainage.38 We reported our experience with PDT in treatment of unresectable cholangiocarcinoma and compared its efcacy and results with patients palliated with only endoscopic biliary drainage.
Abbreviations used in this paper: ERCP, endoscopic retrograde cholangiopancreatography; MELD, Model for End-Stage Liver Disease; PDT, photodynamic therapy. 2008 by the AGA Institute 1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.004
March 2008
PDT IN CHOLANGIOCARCINOMA
291
Table 1. Pretreatment Clinical Characteristics of All the Patients in the Study, Showing Comparison Between the Groups With PDT and Stent Versus Stent Only
PDT group (n 19) Stent group (n 29) P value .61 .08 .09 .03
Age (y) (mean) 65.3 67.4 Gender (male, female) 11, 8 13, 16 Mean bilirubin at 8.3 mg/dL 12.8 (6.6) mg/dL presentation Mean bilirubin after 90 days 3.5 mg/dL 6.3 mg/dL Mean MELD on presentation 14.6 18.3 Tumor extension Bismuth IV 9 10 Bismuth III 7 10 Bismuth II 1 8 Bismuth I 2 1 Chemotherapy (n 22) 11 11 Radiotherapy (n 19) 9 10 Mortality 3 mo 0% 27.60% 6 mo 16% 51.70% 12 mo 56% 82.10% Cholangitis 7 10 No. of interventions, median 3.0 (18) 2.0 (113) (range) Patient alive at the end of 8 2 study Mean & median survival 16.2 (8) 7.4 (5) (mo)
Pathologic diagnosis was established in 69% of cases, which is in accordance with published studies.39,40 Tissue diagnosis was established by endoscopic retrograde cholangiopancreatography (ERCP) with triple sampling41 in 20 of 26 (77%) cases, by endoscopic ultrasound with ne-needle aspiration in 11 of 14 (78%) cases, percutaneous liver biopsy in 8 of 8 cases, and by laparotomy/laparoscopy in 4 of 4 cases. Staging was performed with computed tomography and/or magnetic resonance imaging. Resectability was dened according to the criteria of Vauthey and Blumgart.42
Materials and Techniques

Duodenoscopes (TJF-140, TJF-160, and TJF-160VF; Olympus America, Center Valley, PA) were used for all procedures. Biliary cannulation was performed with triple lumen sphincterotomes. After biliary sphincterotomy, a cholangiogram dened the extent of ductal involvement. Selective decompression of all opacied, dilated segments was attempted, with bougie and balloon dilation to assist in the placement of polyethylene stents (7F, 8.5F, and 10F diameter [Figures 1 4]). All procedures were performed by 1 of 2 dedicated pancreaticobiliary endoscopists (M.K. or P.Y.), each performing in excess of 500 ERCPs annually.
.011 .01 .636 .25 .053
Photodynamic Therapy
.003
Materials and Methods Patients

Sixty-four consecutive patients were referred to our institution with cholangiocarcinoma between December 2001 and November 2006. Sixteen patients (25%) underwent resection. The remaining 48 patients were palliated with endoscopic biliary stents, and of these, 19 received PDT after its availability at the University of Virginia in December 2004. All patients were captured in a dedicated database and followed prospectively. Patient pretreatment characteristics are summarized in Table 1. Therapy with PDT was offered to all patients with unresectable cholangiocarcinoma and those with resectable lesions deemed inoperable. No patients had contraindications to pormer sodium, such as compromised kidney or hepatic function, leukopenia or thrombocytopenia, or evidence of cancer of another organ.
Staging and Tissue Diagnosis

All patients had clinical and radiologic features characteristic for cholangiocarcinoma. Bismuth classication was documented for all patients (Type I, tumors below the conuence of the left and right hepatic ducts; Type II, tumors reaching the conuence but not involving the left or right hepatic ducts; Type III, tumors occluding the common hepatic duct and either the right (IIIa) or left (IIIb) hepatic duct; Type IV, tumors that are multicentric or that involve the conuence and both the right and left hepatic ducts).
Each patient to whom PDT was offered underwent a specic, detailed educational process by a dedicated team member (P.B., P.J., or K.E.), after which informed consent was obtained. Pormer sodium (Photofrin; Axcan Pharma Inc, Quebec, Canada) was used as a photo sensitizing agent, administered intravenously at a dose of 2 mg/kg body weight 48 hours before illumination. A diode laser system (InGaAIP Laser Diode; Diomed Inc, Andover, MA) with a maximum power output of 2000 mW and a wavelength of 633 3 nm was used as a light source, delivered through a 3.0-m length ber having a 2.5-cm-long cylindrical diffuser at its distal end (Pioneer Optics, Windsor Locks, CT). During ERCP, the biliary anatomy was dened, and appropriate locations for therapy were identied, after which guidewire access was obtained and dilation was used as necessary to introduce the diffuser within the malignant stricture. The diffuser was inserted into a 10F sheath of a plastic stent delivery system (MAJ-1419; Olympus America), placed at the level of the stricture to be treated (Figures 5 and 6). The sheet was positioned at the appropriate level, after being advanced over a wire. Photoactivation was performed at 620 nm with a light dose of 180 J/cm2, uence of 0.250 W/cm2, and irradiation time of 750 seconds. One or 2 segments were treated at the discretion of the endoscopist. After photoactivation, only plastic stents were inserted to decompress opacied radicals proximal to the treated lesion. PDT was repeated at 3-month intervals (Figure 7) at which time all stents were replaced; stents were exchanged earlier in the case of premature occlusion or migration to maintain optimal decompression. All patients received periprocedure antibiotic prophylaxis.
Denition of Events
Successful therapy was dened by relief of cholangitis, jaundice, and pruritus with decrease of bilirubin to less than 75% of the pretreatment value within 30 days.43 Complications
292
KAHALEH ET AL
related to ERCP were dened following consensus criteria44 and differentiated from complications related to PDT.
Follow-Up
Those patients undergoing therapy including PDT underwent ERCP with PDT every 3 months until death or withdrawal from the study. All patients were followed in clinic with laboratory values at least 1 month after intervention and every 3 months thereafter or earlier in case of complications. In patients who received plastic stents without PDT, repeat ERCP was performed if indicated (recurrent cholestasis, cholangitis, and/or pruritus) until patient refusal or death.
Statistical Analysis
Statistical analysis was performed with SAS, version 9.1 (SAS Institute, Cary, NC). Statistical signicance was assumed to be at an alpha of .05, and all statistical tests were two-sided. The Fisher exact test or 2 tests were used to compare proportions, and the Student t test or Wilcoxon signed rank test was used for continuous data. Multivariable logistic regression models were constructed with 6-month and 1-year survival end points. Univariate time-to-event survival models were constructed by using the Kaplan-Meier method and compared between groups with the log-rank test. Multivariate survival was analyzed with a Cox proportional hazards model and the method of maximum likelihood estimates. Variables were included in the multivariate model if they achieved a P value of less than .20 in the univariate comparison (Model for End-Stage Liver Disease [MELD], number of sessions), or if they were clinically expected to be important in the disease process (age, radiation, or chemotherapy). MELD and bilirubin are not independent variables, so both were not included in the nal model. It was believed that MELD was a better representation of
Figure 2. After placement of 2 stents on each side, selective access of segments VI and VII is achieved with a guidewire (patient A).
the subjects severity of disease and was therefore chosen to be in the nal model instead of total bilirubin. This study was approved by the University of Virginia Institutional Review Board for Health Sciences Research.
Results
Forty-eight patients (24 male), mean age, 66.6 years (range, 26 94 years) were palliated. A total of 19 (40%) patients received ERCP and PDT, whereas 29 (60%) underwent ERCP with stenting alone. The pretreatment clinical characteristics of the study population are summarized in Table 1. Mean and median pretreatment carbohydrate antigen 19-9 (n 41/48) was 3203 and 227 U/mL (range, 359,100 U/mL), respectively. Mean and median pretreatment carcinoembryonic antigen (n 36/48) was 42.9 and 3.5 (range, 0.7 650), respectively. Nineteen patients were found to have Bismuth IV cancers (40%), 17 had Bismuth III (35%), 9 had Bismuth II (19%), and 3 patients had Bismuth I (6%) (Table 1). All patients had clinical and cross-sectional imaging compatible with cholangiocarcinoma. No patients were excluded because of a large mass. The median number of ERCP procedures was 3.0 (range, 1 8) in the PDT group, including a mean number of 1.6 (range, 13) PDT sessions. The stent only group underwent a median of 2.0 (range, 113) ERCP procedures to maintain biliary decompression. There was no signicant difference between age, gender, preprocedure total bilirubin, carcinoembryonic antigen, and carbohydrate antigen 19-9 levels in each group (Table 1). The stent only group had a statistically signicant (P .03) higher preprocedure MELD score (18.3 vs 14.6) compared with the PDT group.
Figure 1. Cholangiogram showing a Bismuth III lesion (patient A).
March 2008
293
Table 2. Multivariate Analysis of Factors Predictive of Increased Survival

Variable PDT Age Radiation Chemotherapy MELD ERCP sessions P value .0062 .9001 .185 .9285 .6947 .0002 Hazard ratio 0.27 0.998 0.308 1.074 1.016 0.65 95% Condence interval 0.1050.689 0.9691.028 0.0541.757 0.2285.059 0.9381.102 0.5170.818
abscess with prolonged cholangitis, requiring percutaneous drainage. Other complications included hemobilia (2) and cholecystitis (2). Both patients with cholecystitis were managed nonsurgically. Among the adverse events specic to PDT, 3 patients experienced skin phototoxicity, one World Health Organization grade 3; all were managed with topical therapy.
Survival
At the time of analysis, 10 patients were living, with 8 being in the PDT group. Cause of death included tumor progression (n 30), cholangitis (n 2), pulmonary embolus (n 1), multiorgan failure (n 4), and myocardial infarction (n 1). Overall mean and median survival was 8.5 7.2 and 7 months (range, 0.2536 months), respectively. Kaplan-Meier survival analysis (Figure 9) showed statistically signicant (P .003) prolonged survival in the PDT group (mean, 16.2 2.4 months, compared with the stent only group (mean, 7.4 1.6 months). Overall survival at 3, 6, and 12 months was 83%, 33%, and 27%, respectively. In the PDT group the 3-, 6-, and 12-month mortality rates were 0%, 16%, and 56%, respectively. The corresponding mortality rates in the stent only group were 28%, 52%, and 82%, respectively. This difference was statistically signicant at 3 (P .01) and 6 months (P .01) but not at 12 months (P .08). On analysis of maximum likelihood estimates, factors predictive of survival were treatment with PDT and number of ERCP sessions. There was no signicance related to age, MELD score, chemotherapy, or radiation (Table 2). The proportional hazards survival model was re-run with an interaction term between exposure to PDT and number of ERCP sessions, and there was a trend toward statistical signicance of the interaction term (P .08). In this new model, the inuence of PDT exposure was persistently statistically signicant (P .004), and the remaining conclusions of the study were unchanged.
Biliary Decompression
In both groups, serum bilirubin was successfully decreased (75% of the original value) (Figure 8). After 3 months, a greater than 50% reduction of bilirubin was noted in 26 patients. In both groups the decrease in bilirubin was signicant when comparing pretreatment versus post-treatment bilirubin (Figure 8) (P .008 in the PDT group and P .0001 in the stent only group). However, when comparing the degree of decrease between the 2 groups, no signicant difference was observed (P .78).
Chemotherapy and Radiation

A total of 22 (46%) patients received chemotherapy with various combinations of gemcitabine and capecetabine. Eleven (58%) were in the PDT group, and 11 (38%) were in the stent only group (P .17). A total of 19 patients also were treated with concomitant extracorporeal radiation. Of these, 9 (47%) were included in the PDT group and 10 (34%) in the stent only group (P .37).
Complications and Adverse Events

Complications in the stent only group included 10 patients developing cholangitis after endoprosthetic therapy, with 2 patients dying as a consequence. Post-ERCP pancreatitis was observed in 4 patients and duodenal perforation in one. Other adverse events in the stent group included hepatic abscess (1), Billroth II perforation (1), and non-ST elevation myocardial infarction (2). In the PDT group, 7 patients (37%) developed cholangitis treated with antibiotics alone. One patient developed a hepatic
Discussion
Almost 80% of patients with cholangiocarcinoma are diagnosed at an unresectable stage.3,8,45 Effective palliation is essential, because biliary drainage and prevention of cholestasis are crucial to prevent pruritus, cholangitis, and death. The approach to palliative biliary decompression has evolved from surgery with its attendant morbidity and mortality16,22
Table 3. Table Comparing Studies Performed by Using ERCP With PDT With Photofrin Sodium for Palliation of Cholangiocarcinoma
Median TB before and after PDT (mg/dL) 18.6 6 11.2 1.1 2.7 1.3 13.3 2.6 (mean) N/A (decreased) 11.8 3.1 7.7 1.1 N/A (decreased) 6.1 3.5 (mean) Mean PDT sessions (range) 1.5 (12) 3 (15) 2.3 (12) 12 2.4 (15) 1.5 (14) 2 (15) 2 (16) 1.6 (14) Median survival (mo) 14.6 11.1 6 9.9 16.4 14.2 9.2 12 13.4 Adverse events: phototoxicity, cholangitis 1 (11%), 0 (0%) 3 (13%), 8 (35%) 2 (33%), 2 (33%) 2 (8%), 5 (21%) 2 (10%), 5 (25%) 3 (10%), 6 (19%) 2 (25%), 2 (25%) 8 (12%), 38 (56%) 1 (4%), 2 (8%)
Study Ortner et al35 Berr et al36 Rumalla et al34 Dumoulin et al50 Ortner et al37 Ortner et al37 Harewood et al57 Witzigmann et al39 Prasad et al40
Year 1998 2000 2001 2003 2003 2003 2005 2006 2007
N (M/F) 9 23 6 24 20 31 8 68 25
Study type Single arm Single arm Single arm Single arm Randomized Nonrandomized Single arm Single arm Single arm
TB, total bilirubin; N/A, not available.
294
KAHALEH ET AL
and more recently from percutaneous to endoscopic management; the latter has been shown superior to a percutaneous approach in prevention of cholestasis and improved mortality.46 Unfortunately, the benet of ERCP with stent placement is often obviated by proximal tumor obstruction.6,7,10,12 To address this issue, multiple studies have been conducted to investigate the effect of combining bile duct stent insertion and PDT.3336,47,48 PDT is thought to destroy cancer and neovascular cells and to induce tumor thrombosis through formation of cytotoxic reaction products, including singlet oxygen radicals.27,28,30 Understanding that survival in cholangiocarcinoma is related to the efcacy of biliary decompression, PDT offers a logical mechanism to use. The efcacy of stenting is limited by stent patency; unlike benign conditions, no tissue remodeling is affected by stenting tumor. PDT offers the possibility of
Figure 4. Fluoroscopic image after drainage of all dilated segments (patient A).
remodeling tumoral mass,49 which might enhance or prolong the decompressive effect. Accepting this hypothesis, PDT could improve cholestasis and survival in the setting of incomplete decompression with stents by opening previously inaccessible segments, an observation noted by ourselves and others. In most prospective trials (Table 3) and in our series, PDT was associated with a signicant reduction in bilirubin and increase in survival compared with historical data.34 36,39,40,43,50 This study compared the efcacy of PDT with biliary stenting with a group receiving only biliary plastic stenting. Notable differences with most previous published reports were that multiple sessions of ERCP were offered to both groups, and PDT was not restricted to a single session. Our results seem
Figure 3. Selective access of segments V and VIII is then achieved by using a sphincterotome preloaded with a guidewire (patient A).
Figure 5. Malignant stricture (Bismuth I) before PDT treatment (patient B).
March 2008
295
Figure 6. PDT application at the level of the malignant stricture (patient B).
Figure 8. Mean change in bilirubin at entry and after 3 months of treatment. When comparing the PDT group with the stent only group, there is no statistical difference in either the pretreatment or post-treatment values (P .09), and there is not any statistical difference in the degree of decrease between the 2 groups (P .78). In both groups, decompression was successful by criteria dened in Materials and Methods.
similar to those of Ortner et in which a dramatic increase in median survival after PDT (16.4 months) was observed, compared with 3.3 months in the patients receiving stent placement alone. Unlike the study by Ortner et al, patients who underwent successful stenting were not excluded from our study, and bilirubin values in both groups were signicantly decreased (Figure 8). Our aggressive approach to biliary decompression, as advocated by Prat et al,51 appeared to inuence survival independently of PDT administration (Table 2). The design of this study does not resolve the possibility that an additive effect would be observed combining PDT and stenting. In a recent study by Zoepf et al,47 in which a hematoporphyrin derivative (Photosan-3; Scotia Pharmaceuticals, Guildford,
al,37
UK) was used as a photosensitizer, patients were randomized to stents only or PDT with stents. Stents were changed in both groups every 3 months. Survival in the PDT group was 21 months, compared with 7 months in the stent only group. Although this suggests an additive effect of PDT, there was no multivariate analysis of the number of ERCP sessions as a surrogate of decompression, and by design, those living longer would have had more ERCP sessions. An alternative explanation of the observed results would be improved efcacy with Photosan-3 compared with Photofrin. A recent retrospective study40 reported 25 patients with unresectable cholangiocarcinoma treated with PDT; on multivariate analysis, the presence of a visible mass on imaging and increasing time between diagnosis and PDT predicted a poorer
Figure 7. Cholangiogram at 3-month follow-up (patient B).
Figure 9. Cumulative survival of patients treated with PDT and stent (solid line) versus stent only group (broken line). KaplanMeier analysis showed estimated mean survival of 16.2 2.4 months (95% condence interval, 727 months) after PDT vs 7.4 1.6 months (95% condence interval, 37 months), respectively, which was statistically signicant (P .003).
296
KAHALEH ET AL
Table 4. Comparison of Costs Between ERCP With Biliary Stent Placement and PDT
43268 ERCP with biliary stent Physician charge: $1440 Medicare payment: $375.74 Highest commercial payment: $631.18 96570 PDT Physician charge: $125.00 Medicare payment: $53.46 Highest commercial payment: $91.65 7.
8.
9.
10.
survival rate after PDT. Therefore, early treatment with PDT might preserve hepatic function. We used the MELD score to stratify disease severity in our study population. Thirty-four of 48 patients (71%) presented with a MELD 14. Of these 34 patients, 23 (68%) were in the stent only group; despite this observation, MELD was not an independent predictor of survival by multivariate analysis (Table 2). Two groups have attributed improved survival in this population to the use of chemotherapy and radiation.52,53 We offered chemoradiation to all patients; 22 patients (11 PDT, 11 stent only) received chemotherapy, of whom 19 also received external beam radiation (9 PDT, 10 stent only). This therapy conferred no additional survival advantage by multivariate analysis, conrming the results of other studies that show only limited response in cholangiocarcinoma.54,55 Finally, as seen in several other studies (Table 3), direct adverse effects of PDT were minor and largely related to phototoxicity.34,36,56 The incidence of bacterial cholangitis was similar in both the groups (35%) and is compatible with the published results of 25%35%.34,50,57 Cholangitis accounted for only 5% of all deaths in our series, again suggesting the value of aggressive biliary decompression. Costs related to ERCP and PDT are provided in Table 4. In summary, ERCP with PDT appears to improve survival compared with ERCP with biliary stenting alone in unresectable cholangiocarcinoma. Although our results are comparable with other published results, the effect we observed is not independent of the number of ERCP sessions, raising a complex issue related to adequacy of stenting. In addition, the uncontrolled design of this study prevented denitive conclusions from being drawn. To denitely prove that PDT confers an additional benet compared with ERCP with stenting alone, a prospective, randomized, controlled trial is required with strict criteria dening successful stenting (Figures 39 and Table 4). References
1. Ahrendt SA, Nakeeb A, Pitt HA. Cholangiocarcinoma. Clin Liver Dis 2001;5:191218. 2. de Groen PC, Gores GJ, LaRusso NF, et al. Biliary tract cancers. N Engl J Med 1999;341:1368 1378. 3. Jarnagin WR, Fong Y, DeMatteo RP, et al. Staging, resectability, and outcome in 225 patients with hilar cholangiocarcinoma. Ann Surg 2001;234:507519. 4. Reed DN Jr, Vitale GC, Martin R, et al. Bile duct carcinoma: trends in treatment in the nineties. Am Surg 2000;66:711715. 5. Liu CL, Lo CM, Lai EC, et al. Endoscopic retrograde cholangiopancreatography and endoscopic endoprosthesis insertion in patients with Klatskin tumors. Arch Surg 1998;133:293296. 6. Polydorou AA, Cairns SR, Dowsett JF, et al. Palliation of proximal
11.
12. 13.
14.
15.
16.
17.
18.
19.
20. 21. 22.
23.
24.
25.
26.
27.
28.
malignant biliary obstruction by endoscopic endoprosthesis insertion. Gut 1991;32:685 689. Ducreux M, Liguory C, Lefebvre JF, et al. Management of malignant hilar biliary obstruction by endoscopy: results and prognostic factors. Dig Dis Sci 1992;37:778 783. Farley DR, Weaver AL, Nagorney DM. Natural history of unresected cholangiocarcinoma: patient outcome after noncurative intervention. Mayo Clin Proc 1995;70:425 429. Baer HU, Stain SC, Dennison AR, et al. Improvements in survival by aggressive resections of hilar cholangiocarcinoma. Ann Surg 1993;217:20 27. Deviere J, Baize M, de Toeuf J, et al. Long-term follow-up of patients with hilar malignant stricture treated by endoscopic internal biliary drainage. Gastrointest Endosc 1988;34:95101. Gibson RN, Yeung E, Hadjis N, et al. Percutaneous transhepatic endoprostheses for hilar cholangiocarcinoma. Am J Surg 1988; 156:363367. Lai EC, Tompkins RK, Mann LL, et al. Proximal bile duct cancer: quality of survival. Ann Surg 1987;205:111118. Havlik R, Sbisa E, Tullo A, et al. Results of resection for hilar cholangiocarcinoma with analysis of prognostic factors. Hepatogastroenterology 2000;47:927931. Kapoor VK, Pradeep R, Haribhakti SP, et al. Intrahepatic segment III cholangiojejunostomy in advanced carcinoma of the gallbladder. Br J Surg 1996;83:1709 1711. Chaudhary A, Dhar P, Tomey S, et al. Segment III cholangiojejunostomy for carcinoma of the gallbladder. World J Surg 1997;21: 866 871. Guthrie CM, Banting SW, Garden OJ, et al. Segment III cholangiojejunostomy for palliation of malignant hilar obstruction. Br J Surg 1994;81:1639 1641. Figueras J, Llado L, Valls C, et al. Changing strategies in diagnosis and management of hilar cholangiocarcinoma. Liver Transpl 2000;6:786 794. Blom D, Schwartz SI. Surgical treatment and outcomes in carcinoma of the extrahepatic bile ducts: the University of Rochester experience. Arch Surg 2001;136:209 215. Bismuth H, Castaing D, Traynor O. Resection or palliation: priority of surgery in the treatment of hilar cancer. World J Surg 1988; 12:39 47. Cotton PB. Endoscopic methods for relief of malignant obstructive jaundice. World J Surg 1984;8:854 861. Classen M, Hagenmuller F. Biliary drainage. Endoscopy 1983; 15(Suppl 1):221229. Smith AC, Dowsett JF, Russell RC, et al. Randomised trial of endoscopic stenting versus surgical bypass in malignant low bileduct obstruction. Lancet 1994;344:16551660. Soehendra N, Reynders-Frederix V. Palliative bile duct drainage: a new endoscopic method of introducing a transpapillary drain. Endoscopy 1980;12:8 11. Schmassmann A, von Gunten E, Knuchel J, et al. Wallstents versus plastic stents in malignant biliary obstruction: effects of stent patency of the rst and second stent on patient compliance and survival. Am J Gastroenterol 1996;91:654 659. Born P, Rosch T, Bruhl K, et al. Long-term outcome in patients with advanced hilar bile duct tumors undergoing palliative endoscopic or percutaneous drainage. Z Gastroenterol 2000;38:483 489. Davids PH, Groen AK, Rauws EA, et al. Randomised trial of self-expanding metal stents versus polyethylene stents for distal malignant biliary obstruction. Lancet 1992;340:1488 1492. Hsi RA, Rosenthal DI, Glatstein E. Photodynamic therapy in the treatment of cancer: current state of the art. Drugs 1999;57: 725734. Webber J, Herman M, Kessel D, et al. Current concepts in gastrointestinal photodynamic therapy. Ann Surg 1999;230: 1223.
March 2008
297
29. Oleinick NL, Evans HH. The photobiology of photodynamic therapy: cellular targets and mechanisms. Radiat Res 1998;150: S146 S156. 30. Nelson JS, Liaw LH, Orenstein A, et al. Mechanism of tumor destruction following photodynamic therapy with hematoporphyrin derivative, chlorin, and phthalocyanine. J Natl Cancer Inst 1988;80:1599 1605. 31. Pass HI. Photodynamic therapy in oncology: mechanisms and clinical use. J Natl Cancer Inst 1993;85:443 456. 32. Wong Kee Song LM, Wang KK, Zinsmeister AR. Mono-L-aspartyl chlorin e6 (NPe6) and hematoporphyrin derivative (HpD) in photodynamic therapy administered to a human cholangiocarcinoma model. Cancer 1998;82:421 427. 33. McCaughan JS Jr, Mertens BF, Cho C, et al. Photodynamic therapy to treat tumors of the extrahepatic biliary ducts: a case report. Arch Surg 1991;126:111113. 34. Rumalla A, Baron TH, Wang KK, et al. Endoscopic application of photodynamic therapy for cholangiocarcinoma. Gastrointest Endosc 2001;53:500 504. 35. Ortner MA, Liebetruth J, Schreiber S, et al. Photodynamic therapy of nonresectable cholangiocarcinoma. Gastroenterology 1998; 114:536 542. 36. Berr F, Wiedmann M, Tannapfel A, et al. Photodynamic therapy for advanced bile duct cancer: evidence for improved palliation and extended survival. Hepatology 2000;31:291298. 37. Ortner ME, Caca K, Berr F, et al. Successful photodynamic therapy for nonresectable cholangiocarcinoma: a randomized prospective study. Gastroenterology 2003;125:13551363. 38. Gores GJ. A spotlight on cholangiocarcinoma. Gastroenterology 2003;125:1536 1538. 39. Witzigmann H, Berr F, Ringel U, et al. Surgical and palliative management and outcome in 184 patients with hilar cholangiocarcinoma: palliative photodynamic therapy plus stenting is comparable to r1/r2 resection. Ann Surg 2006;244:230 239. 40. Prasad GA, Wang KK, Baron TH, et al. Factors predicting survival in patients with cholangiocarcinoma treated with photodynamic therapy. Clin Gastroenterol Hepatol 2007;5:743748. 41. Jailwala J, Fogel EL, Sherman S, et al. Triple-tissue sampling at ERCP in malignant biliary obstruction. Gastrointest Endosc 2000; 51:383390. 42. Vauthey JN, Blumgart LH. Recent advances in the management of cholangiocarcinomas. Semin Liver Dis 1994;14:109 114. 43. De Palma GD, Pezzullo A, Rega M, et al. Unilateral placement of metallic stents for malignant hilar obstruction: a prospective study. Gastrointest Endosc 2003;58:50 53. 44. Cotton PB, Lehman G, Vennes J, et al. Endoscopic sphincterotomy complications and their management: an attempt at consensus. Gastrointest Endosc 1991;37:383393. 45. Bismuth H, Nakache R, Diamond T. Management strategies in resection for hilar cholangiocarcinoma. Ann Surg 1992;215: 3138.
46. Speer AG, Cotton PB, Russell RC, et al. Randomised trial of endoscopic versus percutaneous stent insertion in malignant obstructive jaundice. Lancet 1987;2:57 62. 47. Zoepf T, Jakobs R, Arnold JC, et al. Palliation of nonresectable bile duct cancer: improved survival after photodynamic therapy. Am J Gastroenterol 2005;100:2426 2430. 48. Wiedmann M, Berr F, Schiefke I, et al. Photodynamic therapy in patients with non-resectable hilar cholangiocarcinoma: 5-year follow-up of a prospective phase II study. Gastrointest Endosc 2004;60:68 75. 49. Berr F, Tannapfel A, Lamesch P, et al. Neoadjuvant photodynamic therapy before curative resection of proximal bile duct carcinoma. J Hepatol 2000;32:352357. 50. Dumoulin FL, Gerhardt T, Fuchs S, et al. Phase II study of photodynamic therapy and metal stent as palliative treatment for nonresectable hilar cholangiocarcinoma. Gastrointest Endosc 2003;57:860 867. 51. Prat F, Chapat O, Ducot B, et al. Predictive factors for survival of patients with inoperable malignant distal biliary strictures: a practical management guideline. Gut 1998;42:76 80. 52. Foo ML, Gunderson LL, Bender CE, et al. External radiation therapy and transcatheter iridium in the treatment of extrahepatic bile duct carcinoma. Int J Radiat Oncol Biol Phys 1997;39: 929 935. 53. Morganti AG, Trodella L, Valentini V, et al. Combined modality treatment in unresectable extrahepatic biliary carcinoma. Int J Radiat Oncol Biol Phys 2000;46:913919. 54. Kelley ST, Bloomston M, Serani F, et al. Cholangiocarcinoma: advocate an aggressive operative approach with adjuvant chemotherapy. Am Surg 2004;70:743749. 55. Goldstein RM, Stone M, Tillery GW, et al. Is liver transplantation indicated for cholangiocarcinoma? Am J Surg 1993;166:768 772. 56. Ortner M. Photodynamic therapy in the biliary tract. Curr Gastroenterol Rep 2001;3:154 159. 57. Harewood GC, Baron TH, Rumalla A, et al. Pilot study to assess patient outcomes following endoscopic application of photodynamic therapy for advanced cholangiocarcinoma. J Gastroenterol Hepatol 2005;20:415 420.
Address requests for reprints to: Michel Kahaleh, MD, Digestive Health Center Box 800708, University of Virginia Health System, Charlottesville, VA 22908-0708. e-mail: mk5ke@virginia.edu; fax: 434-9240491. Part of this work was presented as an oral presentation during Digestive Diseases Week 2007, American Society for Gastrointestinal Endoscopy Topic Forum #140, Washington, DC, May 21, 2007, with reference: Gastrointest Endosc 2007;65:AB96. The authors acknowledge the strong support by the technical staff of the Digestive Health Centers Clinic and Endoscope Staff.
Performance Characteristics of the Suspected Blood Indicator Feature in Capsule Endoscopy According to Indication for Study
JONATHAN M. BUSCAGLIA, SAMUEL A. GIDAY, SERGEY V. KANTSEVOY, JOHN O. CLARKE, PRISCILLA MAGNO, ELAINE YONG, and GERARD E. MULLIN
Division of Gastroenterology and Hepatology, Johns Hopkins University School of Medicine, Johns Hopkins Hospital, Baltimore, Maryland
Background & Aims: The suspected blood indicator (SBI)

feature of wireless capsule endoscopy (WCE) was developed for rapid screening of intestinal lesions with bleeding potential. Our aim was to assess the accuracy and performance characteristics of the SBI according to the indications for study in a large cohort of patients. Methods: We reviewed collected data on all WCE studies performed at Johns Hopkins Hospital from January 2006 to June 2007. Study indications were as follows: anemia of unknown origin (n 53), obscure gastrointestinal bleeding (n 112), suspected Crohns disease (n 122), and other (n 4). Concordant and discordant ndings between gastroenterologists readings and SBI were recorded for each patient. Results: A total of 221 lesions with bleeding potential was detected. The overall sensitivity, specicity, positive predictive value, and negative predictive value for the SBI were 56.4%, 33.5%, 24.0%, and 67.3%, respectively. For actively bleeding lesions, the SBI sensitivity and positive predictive value were only 58.3% and 70%, respectively. The sensitivity was highest (64%) in patients undergoing WCE for suspected Crohns disease, with a negative predictive value of 80.4%. The sensitivity was only 58.3% and 41.3%, respectively, in studies performed for obscure gastrointestinal bleeding and anemia. Conclusions: Performance characteristics of the currently available SBI feature in WCE are suboptimal and insufcient to screen for lesions with bleeding potential. Even in patients with active intestinal bleeding, the sensitivity of SBI was less than 60%, which is lower than previously reported. However, in patients with suspected Crohns disease, the high sensitivity and negative predictive value of SBI may make it a useful tool for the detection of large areas of abnormal mucosa. ireless capsule endoscopy (WCE) was rst developed for advanced imaging of the small intestine.13 Although its clinical efcacy is most proven in the diagnosis of obscure gastrointestinal bleeding,4 6 it also has been helpful as an adjunct to radiologic studies for patients with suspected Crohns disease, celiac disease, small-bowel tumors, anemia of unknown origin, chronic abdominal pain, and other indications.711 The WCE system (Given Imaging, Yoqneam, Israel) in combination with RAPID software (Duluth, Georgia) has a suspected blood indicator (SBI) feature that initially was developed to facilitate the screening of possible sites of active bleeding, thereby signicantly reducing the reading time spent by a physician-reviewer.12 The SBI function works by identifying redcolored pixels on the viewing screen, and then automatically
annotating the study scroll bar with a red hash mark; thus indicating to the reader that a potential bleeding lesion or other site of pathology may be present. Previously reported studies on the accuracy of the SBI either have been dedicated solely to gastrointestinal bleeding,13 or were limited by a small number of enrolled patients.12,14 The aim of our study was to assess the accuracy and performance characteristics of the SBI according to the indications for the study in a large cohort of patients.
Methods
Consecutive patients undergoing WCE at our institution from January 2006 through June 2007 were reviewed for the study. Permission to review patient records was granted by the Johns Hopkins University Institutional Review Board. All patients were asked to refrain from eating or drinking at least 8 hours before swallowing the Given M2A video capsule endoscope. Laxative bowel preparation was not used. Patients were allowed to eat and drink 4 hours after the start of their study. Each WCE study was interpreted by 1 of 5 board-certied/board-eligible gastroenterologists (J.M.B., S.A.G., P.M., J.O.C., and E.Y.). All 5 readers had experience reviewing greater than 50 cases each. Images were reviewed at a speed of 8 to 15 frames per second. Captured thumbnail images and summary reports were re-examined and veried by a separate, boardcertied gastroenterologist (G.E.M.) who had experience with at least 250 cases. There was greater than 95% concordance between the verifying reader (G.E.M.) and each of the 5 initial reviewers. After each study, the interpreting physician was asked to record all endoscopic ndings within our WCE database. The following pathologic lesions were considered signicant as it pertains to the SBI function of the Given software: ulcers, erosions, arteriovenous malformations, red spots, varices, venous ectasias, blood, and blood clots. Physicians were asked to document whether the SBI accurately predicted the lesions discovered within each patients study. Performance characteristics of the SBI were dened by concordance between what was positively sensed by the Given software system, and what was positively detected by the physician-reviewer. Cases in which a signicant lesion was found by the reviewer, but not detected by the SBI, also were documented. Those cases in which SBI
Abbreviations used in this paper: GI, gastrointestinal; NPV, negative predictive value; PPV, positive predictive value; SBI, suspected blood indicator; WCE, wireless capsule endoscopy. 2008 by the AGA Institute 1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.029
March 2008
PERFORMANCE OF THE SUSPECTED BLOOD INDICATOR
299
Table 1. Patient Demographics and Study Indications

Anemia of unknown origin (n 53) 25 28 53.7 Obscure GI bleeding (n 112) 64 48 58.9 Suspected Crohns disease (n 122) 36 86 46.3
Table 3. Concordance Between Reader and the SBI Function

Anemia of Suspected unknown Obscure GI Crohns origin bleeding disease Total Total number of analyzed recordings SBI present, signicant lesion present (true positive), n SBI present, signicant lesion absent (false positive), n SBI absent, signicant lesion absent (true negative), n SBI absent, signicant lesion present (false negative), n 53 7 112 21 122 16 287 44
Indication for WCE Males Females Mean age, y
Total (%) (n 287) 125 (43.6) 162 (56.4) 53.0
24
55
60
139
correctly predicted the ndings in one part of the study, but failed to detect the lesions in another part of the study, were recorded as well. The indications for each WCE study also were documented for all patients.
12
21
37
70
10
15
34
Results
A total of 291 patient WCE recordings were studied between January 2006 and June 2007. When the total number of studies was divided according to clinical indications, the single most common indication was suspicion of small-bowel Crohns disease in patients with diarrhea and abdominal pain (n 122; 41.9%). Obscure gastrointestinal (GI) bleeding was the second most common indication (n 112; 38.5%), followed by anemia of unknown origin (n 53; 18.2%). Four additional studies were performed for other indications (1.4%). These 4 studies were not included in the analysis, making the nal number of WCE studies equal to 287 (Table 1). There were 125 males (43.6%) and 162 females (56.4%), with a mean age of 53 years. A total of 221 pathologic lesions were identied in 287 studies by 5 interpreting physicians (Table 2). Mucosal red spots were the most common lesions recognized (69 of 221; 31.2%), followed by erosions (55 of 221; 24.9%), arteriovenous malformations (41 of 221; 18.6%), ulcers (41 of 221; 18.6%), venous ectasias (11 of 221; 5%), and intestinal varices (1 of 221; 0.5%). There were 15 cases in which blood was recognized, and 12 separate lesions were documented to account for the bleeding: 4 ulcers, 3 erosions, and 5 arteriovenous malformations. In 3 cases (1.4%), blood clots or active bleeding was documented in the small bowel by the interpreting physician, yet no discernable lesion was recognized (Table 2).
There was a similar proportion of overall patients in the obscure GI bleeding group and the anemia group with positive ndings detected by the interpreting physician: 32.1% (36 of 112) and 32.1% (17 of 53), respectively. In the suspected smallbowel Crohns disease group, 20.5% (25 of 122) of patients had positive ndings. Table 3 highlights the concordance between the reader and the SBI function of the software. Performance characteristics of the SBI did not differ among the 5 different readers. Of a total of 287 studies, 44 patients had a signicant lesion with an associated SBI (true positive). In 139 patients, an SBI was present without any discernable disease (false positive). The study was negative in 70 patients without an SBI present (true negative), and 34 patients had a lesion recognized by the phy-
Table 2. Pathologic Lesions With Bleeding Potential Found in 287 WCE Studies
Without active bleeding (%) 37 (16.7) 52 (23.5) 36 (16.3) 69 (31.2) 1 (0.5) 11 (5.0) 206 (93.2) Actively bleeding (%) 4 (1.8) 3 (1.4) 5 (2.3) 0 0 0 3 (1.4) 15 (6.8)
Lesion Ulcers Erosions Arteriovenous malformations Red spots Varices Venous ectasias Blood clots and/or active bleeding Total
Total (%) 41 (18.6) 55 (24.9) 41 (18.6) 69 (31.2) 1 (0.5) 11 (5.0) 3 (1.4) 221 (100)
Figure 1. Example of a false-negative SBI in a patient with obscureovert GI bleeding found to have lymphonodular hyperplasia on resection of the terminal ileum.17 PillCam (Given Imaging, Yogneam, Israel)
300
BUSCAGLIA ET AL
Table 4. Performance Characteristics of the SBI for the Prediction of Pathologic Lesions With Bleeding Potential
Indication for WCE Sensitivity Specicity PPV NPV Anemia of unknown origin 41.3 (19.466.6) 33.3 (19.151.1) 22.6 (10.341.5) 54.5 (32.774.9) Obscure GI bleeding 58.3 (40.174.0) 27.6 (18.339.3) 27.6 (18.339.3) 58.3 (40.174.0) Suspected Crohns disease 64.0 (42.681.3) 38.1 (28.648.6) 21.1 (12.932.2) 80.4 (65.690.1) Total 56.4 (44.767.4) 33.5 (27.240.4) 24.0 (18.231.0) 67.3 (57.376.0)
NOTE. Percentages (95% condence intervals) shown.
sician-reviewer without an associated SBI (false negative, Figure 1). Concordance values based on each of the 3 major indications for the study also are listed in Table 3. Table 4 lists the performance characteristics of the SBI in all 287 WCE studies, as well as performance characteristics according to each clinical indication. The overall sensitivity, specicity, positive predictive value (PPV), and negative predictive value (NPV) of the SBI were 56.4%, 33.5%, 24.0%, and 67.3%, respectively. In patients undergoing WCE for anemia of unknown origin, the sensitivity, specicity, PPV, and NPV of the SBI were 41.3%, 33.3%, 22.6%, and 54.5%, respectively (Table 4). In patients undergoing WCE for obscure GI bleeding, the sensitivity, specicity, PPV, and NPV of the SBI were 58.3%, 27.6%, 27.6%, and 58.3%, respectively. When performed for the suspicion of smallbowel Crohns disease, the sensitivity, specicity, PPV, and NPV were 64.0%, 38.1%, 21.1%, and 80.4%, respectively. Factoring in only those cases with active bleeding, the overall sensitivity of the SBI was 58.3%, with a PPV of 70.0% (Table 5). When accounting for all patients studied, the sensitivity of the SBI was highest (64%) in those undergoing WCE for suspicion of small-bowel Crohns disease. The specicity of the SBI was poor for all 3 indications, with the highest value being 38.1% in the suspected Crohns group (Table 4). The NPV also was highest in this same group, with a value of 80.4%. The PPV was poor in all groups, with a peak value of 27.6% in patients with obscure GI bleeding.
Discussion
Since the invention of capsule endoscopy, it has been used most often and most effectively for the diagnosis of obscure GI bleeding.15,16 More recently, however, its indications have broadened with its usefulness shown in the diagnosis of inammatory bowel disease and the evaluation of chronic abdominal pain.711 Early in its development, the WCE software system (Given Imaging) offered an SBI feature with each of its recordings. This small red line appearing at the top of the time tracing initially was designed as a rapid screening tool, allowing physicians to identify areas of active bleeding quickly.12 Although the SBI feature has been available for more than 5 years, reports of its clinical usefulness are relatively sparse, usually completed on a Table 5. Sensitivity and PPV of the SBI in 15 WCE Studies With Blood Clots or Active Bleeding
Sensitivity PPV 58.3 (28.683.6) 70.0 (35.491.9)
NOTE. Percentages (95% condence intervals) shown.
limited number of patients, and showing signicant variations in performance characteristics.1214 Furthermore, other studies have not reported the accuracy of SBI when capsule endoscopy is used for indications other than active bleeding (ie, suspected small-bowel Crohns disease, anemia of unknown origin, and so forth). In 2003, Liangpunsakul et al12 rst reported the accuracy of SBI in 24 patients undergoing WCE. Eighteen studies were performed for occult GI bleeding or iron-deciency anemia, and 6 for abdominal pain. Overall, the sensitivity of the SBI in the detection of small-bowel lesions with bleeding potential was poor at 25.7%. However, if only actively bleeding lesions in the intestine were considered, the sensitivity of the SBI was significantly better (81.2%). Signorelli et al13 also studied the performance characteristics of the SBI in 100 consecutive patients. The overall sensitivity was poor (40.9%), and it only increased to 60.9% in the group of patients with red blood or active bleeding in the small bowel. DHalluin et al14 looked at a larger cohort of patients (n 156) with obscure GI bleeding. The overall sensitivity of the SBI for detecting low-risk or high-risk lesions in the small bowel still was low (37%), and it increased to only 57% for the detection of active bleeding. The aim of our study was to verify the SBI accuracy in a large cohort of patients (n 287), and to examine the SBI performance characteristics according to the indications for the study. Altogether, our ndings suggest that when accounting for all possible lesions with bleeding potential in the small bowel, the sensitivity of the SBI may be higher than originally reported (56.4% vs 25% 41%). More importantly, however, the sensitivity in our study was not improved signicantly when the SBI was used only in patients with active bleeding (58.3% vs 60% 81% in previous studies). This nding has signicant clinical relevance, showing that the current version of the SBI cannot be a substitute for a complete and detailed review of the entire WCE study; therefore, it should not serve as the primary interpretation modality by those physicians who do not have time to read the entire report from mouth to cecum. The sensitivity of the SBI in our study was highest in patients undergoing WCE for the evaluation of diarrhea and abdominal pain with a clinical suspicion of small-bowel Crohns disease (64%). Previous publications showed signicantly lower overall sensitivity values of 25% to 41% in studies performed for a similar indication.1214 It appears from our study that with an NPV of 80.4%, the use of an SBI to aid in the detection of larger areas of abnormal mucosa in patients with suspected inammatory bowel disease warrants further investigation. The main limitation of our study was its retrospective design. A large prospective study will be necessary to verify the SBI accuracy and performance characteristics shown by our retro-
March 2008
PERFORMANCE OF THE SUSPECTED BLOOD INDICATOR
301
spective review. Improvements in the existing Given software system could increase the diagnostic yield of the SBI, and thus will make it a reliable screening tool for patients undergoing WCE. Until then, there is no substitution for an expert physician reviewing the study in full. In conclusion, our large retrospective study showed that the performance characteristics of the currently available SBI feature in WCE are suboptimal and insufcient for adequate detection of lesions with bleeding potential. Even in patients with active small intestinal bleeding, the sensitivity of the SBI still is less than 60%, which is lower than previously reported (60% 81%). However, in patients with suspected small-bowel Crohns disease, our ndings indicate that the high sensitivity and NPVs of SBI warrant further investigation of this tool as a clinical aid in the detection of large areas of abnormal mucosa. References
1. Iddan G, Meron G, Glukhovsky A, et al. Wireless capsule endoscopy. Nature 2000;405:417. 2. Meron GD. The development of the swallowable video capsule M2A. Gastrointest Endosc 2000;52:817 819. 3. ASGE Technology Assessment Committee. Wireless capsule endoscopy. Gastrointest Endosc 2002;56:621 624. 4. Saurin JC, Delvaux M, Vahedi K, et al. Clinical impact of capsule endoscopy compared to push enteroscopy: 1-year follow-up study. Endoscopy 2005;37:318 323. 5. Hartman D, Schmidt H, Bolz G, et al. A prospective two-center study comparing wireless capsule endoscopy with intraoperative enteroscopy in patients with obscure GI bleeding. Gastrointest Endosc 2005;61:826 832. 6. Ell C, Remke S, May A, et al. The rst prospective controlled trial comparing wireless capsule endoscopy with push enteroscopy in chronic gastrointestinal bleeding. Endoscopy 2002;34:685 689. 7. Eliakim R, Fischer D, Suissa A, et al. Wireless capsule video endoscopy is a superior diagnostic tool in comparison to barium follow-through and computerized tomography in patients with suspected Crohns disease. Eur J Gastroenterol Hepatol 2003;15: 363367. 8. Costamagna G, Shah SK, Riccioni ME, et al. A prospective trial comparing small bowel radiographs and video capsule endoscopy for suspected small bowel disease. Gastroenterology 2002; 123:999 1005.
9. Herreiras JM, Caunedo A, Rodriguez-Tellez M, et al. Capsule endoscopy in patients with suspected Crohns disease in negative endoscopy? Endoscopy 2003;35:15. 10. Bardan E, Nadler M, Chowers Y, et al. Capsule endoscopy for the evaluation of patients with chronic abdominal pain. Endoscopy 2003;35:688 689. 11. Vasquez-Iglesias J, Gonzalez-Conde B, Estevez-Prieto E, et al. A prospective study of COX-2 inhibitors versus nonspecic NSAIDs induced small bowel lesions using video capsule endoscopy. Endoscopy 2003;35:A183. 12. Liangpunsakul S, Mays L, Rex DK. Performance of Given suspected blood indicator. Am J Gastroenterol 2003;98:2676 2768. 13. DHalluin PN, Delvaux M, Lapalus MG, et al. Does the suspected blood indicator improve the detection of bleeding lesions by capsule endoscopy? Gastrointest Endosc 2005;61:243249. 14. Signorelli C, Villa F, Rondonotti E, et al. Sensitivity and specicity of the suspected blood identication system in video capsule enteroscopy. Endoscopy 2005;37:1170 1173. 15. Soussan BE, Antonietti M, Herve S, et al. Diagnostic yield and therapeutic implications of capsule endoscopy in obscure gastrointestinal bleeding. Gastroenterol Clin Biol 2004;28:1068 1073. 16. Gupta R, Lakhtakia S, Tandan M, et al. Capsule endoscopy in obscure gastrointestinal bleedingan Indian experience. Indian J Gastroenterol 2006;25:188 190. 17. Buscaglia J, Carroll C, Daniels J, et al. Benign lymphoid hyperplasia: a rare cause of obscure overt bleeding in an adult. Gastrointest Endosc 2007;66:1248 1250.
Address requests for reprints to: Jonathan M. Buscaglia, MD, Johns Hopkins Hospital, 1830 E. Monument Street, Room 7100-A, Baltimore, Maryland 21205. e-mail: jbuscaglia@jhmi.edu; fax: (410) 9552108. Jonathan Buscaglia, MD, initiated the study design, conducted the data analysis, and prepared the manuscript. Samuel Giday, MD, and Sergey Kantsevoy, MD, PhD, aided in the data analysis and the manuscript preparation. John Clarke, MD, Priscilla Magno, MD, and Elaine Yong, MD, collectively interpreted greater than 60% of the WCE studies, and each edited the manuscript draft. Gerard Mullin, MD, MHS, reviewed all WCE studies, contributed to the study design and data analysis, and edited the manuscript draft.
ORIGINAL ARTICLESALIMENTARY TRACT

Patient Predictors of Esophageal Stricture Development After Photodynamic Therapy
PATRICK YACHIMSKI, WILLIAM P. PURICELLI, and NORMAN S. NISHIOKA
Gastrointestinal Unit and Wellman Center for Photomedicine, Massachusetts General Hospital, Boston, Massachusetts
See Curvers W et al on page 670 for companion article in the March 2008 issue of Gastroenterology. Background & Aims: The most common signicant adverse event after photodynamic therapy (PDT) with pormer sodium is esophageal stricture formation. This study assessed whether pretreatment variables, including prior endoscopic therapy for Barretts esophagus, are associated with post-PDT stricturing. Methods: Data from all patients who had undergone PDT with pormer sodium for Barretts esophagus with high-grade dysplasia, intramucosal carcinoma, or T1 cancer at our institution since 1997 were reviewed. Results: One hundred sixteen patients underwent 160 courses of PDT. The incidence of stricture formation after index PDT was 16% (19/116). For all PDT courses, the overall incidence of stricture was 23% (37/160). Stricture rate was signicantly higher after a second PDT course compared with index PDT (43% vs 16%, P .0007). There was no association between post-PDT stricture development and age, gender, body mass index, or prior endoscopic mucosal resection. Patients who developed a stricture had a longer length of Barretts esophagus before treatment than those who did not develop a stricture (7.7 vs 5.7 cm for index PDT only, P .025; 7.4 vs 5.7 cm for all PDT courses, P .007). Length of Barretts esophagus, multiple PDT courses, and presence of intramucosal carcinoma on pretreatment pathology were independent predictors of post-PDT stricture in a stepwise logistic regression analysis controlling for treatment variables, including treatment length. Conclusions: An increased risk of stricture development was seen after multiple courses of PDT. An association between post-PDT stricture and length of Barretts esophagus but not treatment length was also found. Endoscopic mucosal resection did not appear to inuence the likelihood of stricture development after pormer sodium based PDT. he term Barretts esophagus is used to describe metaplasia of the distal esophageal mucosa where columnar intestinaltype mucosa replaces squamous mucosa. This metaplasia is a result of chronic distal esophageal acid exposure1 and is typically encountered in adults with gastroesophageal reux dis-
ease. Barretts esophagus is a precursor lesion to esophageal adenocarcinoma.2 The sequence from Barretts esophagus to esophageal adenocarcinoma is believed to occur by stepwise neoplastic progression. Barretts esophagus progresses to low-grade dysplasia (LGD); LGD progresses to high-grade dysplasia (HGD); HGD progresses to adenocarcinoma. Among individuals with longsegment Barretts esophagus (3 cm), 1 in 200 per year (or 0.5% per year) will develop adenocarcinoma.3 Esophageal adenocarcinoma now has the fourth highest incidence of all gastrointestinal tract cancers diagnosed in the United States4 and is increasing at a rapid rate. Distal esophagectomy is considered by most physicians to be the standard of care for patients with Barretts esophagus and HGD. For patients undergoing esophagectomy for HGD, operative mortality rates of approximately 2% have been reported from high-volume centers.5,6 However, morbidity rates are considerably higher.7 Elderly patients and patients with medical comorbidities are often rejected as surgical candidates, whereas other patients decline surgical management. These patients might instead benet from nonoperative treatment of HGD, and this has prompted the development of endoscopic means of ablating Barretts mucosa. Photodynamic therapy (PDT) is a Food and Drug Administrationapproved endoscopic treatment for Barretts esophagus with HGD.8 During PDT, patients receive a photosensitizing agent followed by illumination with red laser light to achieve ablation of Barretts mucosa. The most common clinically signicant adverse effect of PDT is esophageal stricture formation.9 In some published series, more than 30% of patients treated with PDT developed esophageal stricture.10,11 The reason strictures form after PDT with pormer sodium is not known. It has been hypothesized that the deep, circumferential tissue injury and resulting inammatory reaction produced by pormer sodium PDT incites a brotic response that produces stricturing. However, post-treatment oral steroids do not appear to reduce the likelihood of stricture formation or recurrence.12 Potential endoscopic treatment parameters that might be associated with the formation of strictures after PDT
Abbreviations used in this paper: ALA, aminolevulinic acid; BMI, body mass index; CI, condence interval; EMR, endoscopic mucosal resection; HGD, high-grade dysplasia; LGD, low-grade dysplasia; OR, odds ratio; PDT, photodynamic therapy. 2008 by the AGA Institute 1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.001
March 2008
PATIENT PREDICTORS OF ESOPHAGEAL STRICTURE DEVELOPMENT AFTER PDT
303
include light dose,13 centering balloon length,11,12, multiple treatments of the same esophageal segment,10 and pre-treatment of nodular mucosa.10 The purpose of this study was to determine whether any pretreatment variables including prior endoscopic therapy are associated with an increased likelihood of esophageal stricture development after PDT.
Patients and Methods

Records were reviewed for all patients who underwent PDT in a single large urban teaching hospital between 1997 and 2006. This retrospective review was approved by the Partners Institutional Review Board. Patients who underwent pormer sodium based PDT for treatment of Barretts esophagus with HGD, intramucosal carcinoma, or T1 adenocarcinoma were included in this analysis. All patients in the analysis were required to have undergone at least 1 follow-up endoscopy at our institution after PDT. Patients who underwent PDT for palliative treatment of esophageal adenocarcinoma stage T2 or greater (6 patients), esophageal squamous cell carcinoma (3 patients), or gastric cancer (3 patients) were excluded. Sources of data included endoscopy reports, pathology reports, clinic notes, and, when available, endoscopy unit nursing ow sheets. All records were reviewed by a single investigator (P.Y.), who was unaware of clinical outcome (stricture versus no stricture) at the time of data extraction. Length of Barretts esophagus was determined by the distance between the squamocolumnar junction and the top of the gastric folds. All patients underwent a biopsy protocol consisting of 4-quadrant jumbo forceps biopsies every 2 cm along the length of Barretts esophagus to stage and grade the extent of dysplasia. Histopathologic diagnosis was determined by an experienced gastrointestinal pathologist at our institution. For cases in which diagnostic endoscopy and biopsy had been performed at a referring institution, pathology specimens were sent to our institution for diagnostic conrmation before PDT. All patients underwent endoscopic ultrasound before PDT. High-frequency probe endoscopic ultrasound was not routinely performed. If there was evidence of a focal or nodular lesion on initial endoscopy and if endoscopic ultrasound excluded stage T2 or greater disease, endoscopic mucosal resection (EMR) could be performed. No additional or alternative endoscopic treatment was offered before PDT. High-dose proton pump inhibitor therapy (omeprazole 40 mg twice daily or equivalent) was prescribed for all patients before PDT and continued indenitely after PDT. Some patients underwent esophageal pH studies if they experienced heartburn or reux symptoms on proton pump inhibitor therapy; however, pH studies were not routinely performed. A standardized PDT protocol was used that consisted of intravenous pormer sodium (Photofrin; Axcan Pharma, MontSaint-Hilaire, Quebec, Canada) administration on day 0. A weight-based dose of 2 mg/kg was administered, except for a period of time before 1998 when the maximum dose was capped at 150 mg. Patients returned to the endoscopy unit on day 2 for initial light exposure. The majority of cases used a cylindrical diffusing ber that was passed through the working channel of a standard diagnostic gastroscope. Fiber lengths of 1, 2.5, or 5 cm were used on the basis of the desired treatment length. Centering balloons were used in some patients between
1998 and 2005. The target esophageal mucosa was exposed to laser light at a wavelength of 630 nm and a total energy of 150 J/cm. Any deviation from this protocol was documented in the endoscopy report. A single course of PDT can have 1 or 2 light exposures separated by 48 hours. Patients returned to the endoscopy unit on day 4 for a possible second light exposure. This second light exposure could be used to either extend the length of treatment and/or re-treat the previously treated area if the mucosal PDT effect appeared to be uneven. Day 4 light exposure was withheld only if the extent of mucosal injury was judged to be severe, if the patient was experiencing signicant symptoms of dysphagia, odynophagia, or chest discomfort, or if the patient otherwise exhibited poor tolerance to PDT after initial light exposure. All patients returned for surveillance endoscopy 3 months after PDT. Endoscopy was performed sooner when post-PDT stricture was suspected on the basis of dysphagia. Stricture was dened as esophageal luminal narrowing preventing passage of a standard gastroscope. All dilations were performed with a through-the-scope esophageal balloon dilator. The majority of patients receiving PDT at our institution undergo all endoscopic follow-up at our institution. However, in certain cases such as extensive travel, patients underwent additional endoscopic evaluation at an outside institution. In these cases endoscopic reports were retrieved and reviewed for this analysis.
All statistical analyses were performed with SAS version 9.1 software (SAS Institute, Cary, NC). For univariate analyses, Student t test or Wilcoxon rank sum testing was performed for analysis of continuous variables. Chi-square test or Fisher exact test was used for analyses of binary or ordinal variables. After univariate analysis, stepwise logistic regression analysis was performed. Multiple treatment variables were incorporated into the model, including pormer sodium dose, light dose, number of light exposures, treatment length, and use of a centering balloon versus cylindrical diffusing ber. Patient characteristics analyzed included age, gender, body mass index (BMI), length of Barretts esophagus, pathology (treated as a binary variable [HGD versus intramucosal and/or submucosal carcinoma]), history of prior stricture, and history of prior EMR. The threshold for statistical signicance was dened as a two-sided P value less than .05. Because of the exploratory nature of the analysis, Bonferroni correction was not performed.
Results Pretreatment Patient Characteristics

One hundred sixteen patients with Barretts esophagus underwent 160 courses of PDT. Seventy-nine patients underwent a single course of PDT, whereas 31 patients underwent 2 PDT courses, 5 patients underwent 3 PDT courses, and 1 patient underwent 4 PDT courses. Seven patients were part of a multicenter study and were included in a previous report.10 Eighty-one percent of patients were male, and 99% were white, with a mean age of 70.4 years at the time of index PDT. Mean BMI was 28.6 kg/m2 at the time of index PDT.
304
YACHIMSKI ET AL
105 days in the group without post-PDT stricture, P .25). There was no difference in the mean number of EMR resections in patients who did and did not develop post-PDT stricture (1.7 for both groups).
Photodynamic Therapy Variables

The majority of patients received 2 light exposures, with a maximum light dose of 150 J/cm each session and without the use of a centering balloon. Documentation of pormer sodium dose was unavailable for 15 PDT courses, and documentation of light dose was missing for 13 courses.
Stricture Rate
Figure 1. Patients undergoing EMR before PDT.
History of esophageal stricture was present in 6 of 116 patients (5%). Five patients had undergone esophageal surgery before PDT. Three had undergone Nissen fundoplication, and 2 had undergone partial esophagectomy for treatment of esophageal cancer and were referred for PDT for recurrent cancer. Of these latter 2 patients, one had also been previously treated for esophageal cancer with chemotherapy and radiation. No other patients reported a history of esophageal or mediastinal radiation. The location of the squamocolumnar junction and gastric folds was documented in 96% of endoscopy reports. In the remainder of reports, only total length of Barretts esophagus was described. Before index PDT, the mean length of Barretts esophagus was 6.0 cm. Fifty-nine patients (51%) had been diagnosed with HGD as the most advanced pathology on pre-PDT biopsy specimens, whereas the remaining 57 patients (49%) had been diagnosed with intramucosal carcinoma.
The overall incidence of stricture was 23% (37/160). The incidence of stricture after index PDT was 16% (19/116). The stricture rate after a second PDT course was 43% (16/37). The increase in stricture rate after a second course versus index PDT was statistically signicant (43% vs 16%, P .0007; risk ratio, 2.64; 95% condence interval [CI], 1.52 4.59) (Figure 2). Stricture rates of 33% (2/6) after 3 PDT courses and 0% (0/1) after 4 PDT courses were observed. Of the 19 patients who developed a stricture after index PDT, 6 underwent a second PDT course. Three of these 6 patients (50%) developed recurrent stricture after the second course of PDT. Median number of dilations required to treat post-PDT stricture was 3 (range, 0 to 20). In 1 patient, a stricture was traversed with the gastroscope without use of a balloon dilator. There were no observed complications of stricture dilation.
Univariate and Logistic Regression Analysis

In univariate analyses of patients undergoing index PDT only, factors associated with the development of post-PDT stricture were length of Barretts esophagus (7.7 4.2 cm in the group with post-PDT stricture vs 5.7 3.1 cm in group without post-PDT stricture, P .025) and PDT treatment length (6.5 1.4 cm in the group with post-PDT stricture and 5.6 1.6 cm in the group without post-PDT stricture, P .03) (Table 1). There was no apparent association between post-PDT stricture and any other treatment variable including number of light exposures, light dose, and use of a diffusing ber rather than centering balloon.
Endoscopic Mucosal Resection

Thirty patients underwent a total of 31 EMR sessions before index PDT (Figure 1). EMR was performed by using either a saline-assisted cap and snare technique or a band ligator device (Duette; Cook Medical, Bloomington, IN) followed by snare resection. The mean number of EMR specimens resected per session was 1.7 (range, 1 6). Nine patients underwent more than 1 resection in a given session, whereas the remainder underwent a single resection. Two patients experienced post-EMR bleeding requiring immediate further endoscopic therapy. There were no other observed complications of EMR. Pathologic examination of EMR specimens revealed nondysplastic Barretts esophagus in 2 patients, Barretts esophagus with LGD in 1 patient, Barretts esophagus with HGD in 4 patients, intramucosal carcinoma in 18 patients, and carcinoma with submucosal involvement in 3 patients. In 1 patient, the EMR specimen was not available for review, and in 1 patient the EMR specimen contained no epithelial tissue. No patient was excluded from PDT on the basis of the EMR ndings. Median time between EMR and index PDT was 61 days (range, 16 762 days). There was no association between mean time from EMR to index PDT and development of post-PDT stricture (119 days in the group with post-PDT stricture versus
Figure 2. Stricture rate: index PDT vs second PDT course.
March 2008
305
In analyses of all 160 PDT courses, an association persisted between development of post-PDT stricture and length of Barretts esophagus (7.4 3.7 cm in the group with postPDT stricture vs 5.7 3.2 cm in the group without post-PDT stricture, P .007). However, no association was seen with treatment length (Table 2). A trend toward signicance (P .08) was observed for the association between light dose and post-PDT stricture. Among patients with prior stricture who underwent PDT, 22% (8/37) developed a post-PDT stricture, as compared with a 7% (9/123) post-PDT stricture rate among patients without a history of esophageal stricture (P .03). The unadjusted risk ratio for post-PDT stricture in a patient with history of prior esophageal stricture was 2.32 (95% CI, 1.27 4.23), compared with a patient without a prior stricture. In addition, patients with intramucosal or submucosal carcinoma appeared more likely than patients with HGD to develop post-PDT stricture on univariate analysis (31% [23/75] vs 16% [14/85]; unadjusted risk ratio, 1.86; 95% CI, 1.033.35). Stepwise logistic regression analyses were performed for both index PDT alone and for all PDT courses. For patients undergoing index PDT, the only variable associated with postPDT stricture was length of Barretts esophagus (odds ratio [OR], 1.23; 95% CI, 1.04 1.44) (Table 3). In analyses of all PDT courses, length of Barretts esophagus was conrmed to be associated with post-PDT stricture (OR, 1.27; 95% CI, 1.10 1.47). Additional variables associated with post-PDT stricture in this inclusive analysis were multiple PDT courses (OR, 3.15; 95% CI, 1.43 6.94) and presence of intramucosal or submucosal carcinoma on pretreatment pathology (OR, 4.62; 95% CI, 1.66 12.89) (Table 3).
Table 2. Univariate Predictors of Stricture After PDT (All Treatment Courses)

Stricture No stricture P value
N 37 123 Age 69.2 11.3 y 70.4 10.1 y Male gender 28/37 (76%) 99/123 (80%) 28.9 6.8 28.3 6.0 BMI (kg/m2) Prior EMR 11/37 (30%) 26/123 (21%) Prior stricture 8/37 (22%) 9/123 (7%) Pormer sodium 169.5 39.9 mg 169.2 39.5 mg dose Light dose 300 J/cm 5/35 (14%) 5/112 (4%) 260 J/cm 1/35 (3%) 0/112 (0%) 225 J/cm 0/35 (0%) 1/112 (1%) 150 J/cm 28/35 (80%) 98/112 (88%) 130 J/cm 1/35 (3%) 8/112 (7%) Light exposures 2 30/37 (81%) 111/123 (90%) 1 7/37 (19%) 12/123 (10%) Light delivery Diffusing ber 33/37 (89%) 102/122 (84%) Centering balloon 4/37 (11%) 20/122 (16%) Length Barretts 7.4 3.7 cm 5.7 3.2 cm esophagus Treatment length 6.2 1.5 cm 5.7 1.6 cm Pathologic diagnosis HGD 14/37 (38%) 71/123 (58%) Intramucosal 23/37 (62%) 52/123 (42%) carcinomaa
aIncludes
.54 .53 .77 .28 .03 .94
.08
.15
.41 .007 .23
.03
3 patients with submucosal (T1sm) disease.
Table 1. Univariate Predictors of Stricture After Index PDT

Stricture No stricture P value
Discussion
PDT is an established endoscopic technique for ablating Barretts esophagus with HGD and/or early-stage intraepithelial neoplasia. Patients with medical comorbidities that preclude surgical management and patients who decline surgical management might be appropriate candidates for PDT. In some patients, multiple courses of PDT might be required. With close biopsy surveillance, many patients might not exhibit progression to invasive cancer during extended post-PDT follow-up. Other patients might experience downgrading of dysplasia or complete ablation of Barretts epithelium. In the latter instance, glandular Barretts mucosa buried beneath neosquamous epithelium is a concern and has been associated with disease recurrence.14,15 Table 3. Predictors of Stricture After PDT: Logistic Regression Analysis
P value OR 95% CI
N 19 97 Age 67.9 12.4 y 70.9 9.7 y Male gender 15/19 (79%) 79/97 (81%) 29.6 7.5 28.4 6.4 BMI (kg/m2) Prior EMR 7/19 (37%) 23/97 (24%) Prior stricture 2/19 (11%) 4/97 (4%) Pormer sodium 173.1 41.6 mg 170.7 41.6 mg dose Light dose 300 J/cm 0/19 (0%) 4/92 (4%) 150 J/cm 18/19 (95%) 83/92 (90%) 130 J/cm 1/19 (5%) 5/92 (5%) Light exposures 2 15/19 (79%) 87/97 (90%) 1 4/19 (21%) 10/97 (10%) Light delivery Cylindrical ber 18/19 (95%) 83/97 (86%) Centering balloon 1/19 (5%) 14/97 (14%) Length Barretts 7.7 4.2 cm 5.7 3.1 cm esophagus Treatment length 6.5 1.4 cm 5.6 1.6 cm Pathologic diagnosis HGD 8/19 (42%) 51/97 (53%) Intramucosal 11/19 (58%) 46/97 (47%) carcinomaa
aIncludes
.24 .76 .58 .26 .25 .86
.65
.24
.46 .025 .03
.40
Index PDT Length Barretts esophagus All PDT courses Length Barretts esophagus Multiple PDT courses Pathology (Intramucosal carcinomaa vs HGD)
aIncludes
.01 .001 .004 .003
1.23 1.27 3.15 4.62
1.041.44 1.101.47 1.436.94 1.6612.89
306
YACHIMSKI ET AL
Esophageal stricture occurred at a rate of 16% after index PDT and 23% after all PDT courses in this study. A minimum of 3 months of endoscopic follow-up at our institution was required for inclusion in this study. The majority of post-PDT strictures develop within 3 weeks after treatment,13,16 so it is therefore likely that this analysis, although retrospective, accurately captures and reports the true incidence of post-PDT stricture in this cohort. Post-PDT stricture should be suspected in any patient who reports dysphagia after PDT. Strictures are typically supercial and might be effectively dilated with standard endoscopic accessories. Either through the scope balloon dilators or Savary (Cook Medical Inc, Bloomington, IN) dilators16 might be used. Complications including esophageal perforation, while not observed in our study, are a possibility in patients undergoing dilation of post-PDT stricture. As evident in our cohort, multiple dilations might be required to achieve stricture resolution. Intralesional steroid injection, stent placement, or home dilation techniques have been used by other investigators in a minority of patients with refractory strictures. Patients should be instructed to adhere to a liquid or soft-solid diet, with high calorie oral supplements as necessary to maintain adequate nutrition, until dysphagia has resolved. Overall, relatively few patients experience signicant weight loss as a consequence of post-PDT stricture.10 Among patients who develop post-PDT stricture, the majority will achieve complete stricture resolution. In recently published follow-up data from the original PDT with pormer sodium (PORPDT) study, 94% of subjects who developed postPDT stricture were stricture-free at the completion of the initial 2-year study phase; and only 3 subjects developed recurrent stricture between years 2 and 5 of follow-up.17 Previous studies have identied treatment variables that inuence the likelihood of stricture development, including light dose, centering balloon length, and pre-treatment of nodular mucosa. Panjehpour et al13 demonstrated that a light dose of 115 J/cm led to a 15.3% incidence of severe stricture (dened as requiring 6 or more dilations), as compared with a 5% 6% severe stricture rate in patients treated with 85 J/cm or 95 J/cm. In a longitudinal series of 100 patients, stricture rate was 52% after use of a 3-cm centering balloon, as compared with 28% when 5-cm or 7-cm centering balloons were used.11 However, the reverse trend was observed in a smaller series evaluating the effect of oral steroid therapy on post-PDT stricture, with a 31% stricture rate in patients treated with 7-cm centering balloons versus a 7% stricture rate in patients treated with 5-cm centering balloons.12 Preliminary analysis of the PORPDT phase III trial data identied associations between nodule pre-treatment, as
well as repeated treatments on the same esophageal segment and stricture development.10 Our data conrm the ndings of prior studies in identifying multiple treatment courses as a factor associated with development of post-PDT stricture (Table 4). Less attention has been focused on treatment-independent patient variables that predict stricture formation. This study identies an association between length of Barretts esophagus and development of post-PDT stricture after multivariable analysis. This association was independent of treatment length. Our data also suggest an association between intramucosal or submucosal carcinoma on pretreatment pathology and development of post-PDT stricture. In this study, dysplasia/tumor stage was assessed by forceps biopsy and endosonography in all patients, whereas only one fourth of patients had undergone EMR. EMR is superior to endoscopic ultrasound in detecting submucosal tumor invasion.18 It is conceivable that some patients in our cohort might have had undetected submucosal carcinoma, which might have been more accurately staged if they had undergone EMR. However, although this study was not designed to assess the efcacy of PDT in our cohort, long-term follow-up of these patients has failed to detect the presence of such carcinomas in the vast majority of patients. A pathophysiologic basis for the associations between length of Barretts esophagus and depth of involvement, respectively, and post-PDT stricture might be explained by the distribution of pormer sodium. If there is preferential uptake of pormer sodium by neoplastic tissue, including dysplastic Barretts epithelium, then this tissue might be subject to a greater extent or depth of tissue injury after light exposure. If this is the case, then stricture rate would be expected to be lower after pormer sodium PDT of nondysplastic Barretts or esophageal squamous mucosa. However, this hypothesis cannot be assessed on the basis of available evidence, because there are limited published experience with PDT ablation of nondysplastic Barretts epithelium and no data on the in vivo effect of PDT on esophageal squamous epithelium. Additional limitations of this study include the relatively small sample size, which limits statistical power. This is a factor in virtually all single-center studies of PDT for Barretts esophagus, even when considering data from relatively high-volume centers. A second limitation of this study is its retrospective nature. The associations reported in this study should undergo validation in a prospective analysis. Furthermore, because PDT protocols vary among institutions, the external validity of our ndings is uncertain. It is also possible that post-PDT stricture might be inuenced by additional variables not included in this analysis. A
Table 4. Stricture Rate After Pormer Sodium PDT for Barretts Esophagus
Stricture rate Author, year Overholt et al, 20038 Overholt et al, 200510 Prasad et al, 200719 Keeley et al, 200729 Present study
aIncludes
Maximum light dose 100250 J/cm 130 J/cm 130200 J/cm 300400 J/cm 130300 J/cm
Light exposures 12 2 12 2 12
Patients (N) 82 132 131a 50 116
Index PDT 18% 12% 23% 30% 16%
All PDT courses 30% 36% 27% 42% 23%
26 patients who received hematoporphyrin derivative photosensitizer.
March 2008
307
detailed medication, tobacco, and alcohol history had not been reliably documented in a signicant portion of our patients, and these variables were therefore not incorporated into the analysis. Because esophageal pH studies were not routinely performed to assess degree of acid suppression, it is conceivable that differential degrees of acid suppression might have inuenced stricture development in individual patients. During the period of this study, pormer sodium PDT was offered to patients with Barretts esophagus and HGD or intramucosal carcinoma, as well as some patients with T1 disease involving the submucosa (T1sm). The risk of post-PDT stricture was discussed with all patients before PDT. On the basis of the ndings of this analysis, we offer the following additional counseling: (1) patients with very long segments (6 7 cm) of Barretts esophagus might be at increased risk for stricture; (2) patients with intramucosal carcinoma or submucosal tumor involvement might be at increased risk for stricture; and (3) patients can safely undergo EMR for either staging or therapeutic intent before PDT, without increased risk of post-PDT stricture. This last nding is in direct contrast with recently reported results in which an association between prior EMR and postPDT stricture was found.19 This discrepancy might be due to differences in the patient cohort and/or PDT protocol. For example, the incidence of prior esophageal stricture was much higher in the Mayo group of patients (24%) than in our cohort (5% before index PDT). Treated patients might therefore have been more prone to recurrent esophageal stricture. In addition, the typical light dose used in the Mayo protocol was higher (200 J/cm vs 150 J/cm), and an association between intensity of light dose and stricture has been demonstrated previously. In addition, little detail is available regarding the comparative performance of EMR in the 2 studies. EMR might be used for focal resection of a single esophageal nodule or for wider debulking of dysplastic Barretts epithelium. EMR of greater than 75% of total esophageal circumference has been associated with post-EMR stricture.20 It is therefore conceivable that the extent of EMR might impact the likelihood of stricture development after EMR and PDT combination therapy. Time interval and esophageal healing between EMR and PDT might theoretically impact the likelihood of post-PDT stricture. Our data do not support this hypothesis, although the number of patients who underwent both EMR and PDT limits power to detect differences. The presence or absence of an association between EMR and post-PDT stricture is clinically important because EMR volume has increased with the assistance of dedicated and easy-use EMR devices.21,22 Appropriate, safe, and continued use of this combination therapy for patients with Barretts esophagus should be contingent on its long-term efcacy, as well as its short- and long-term side effect prole. As an increasing proportion of patients undergo EMR (for either therapeutic or staging purposes) before PDT (Figure 1), the impact of EMR on PDT outcomes might become further elucidated. For patients at high risk for post-PDT stricture, alternative endoscopic ablation techniques might provide alternative treatment options. Esophageal stricture is generally not observed when aminolevulinic acid (ALA) is used as the photosensitizer. In addition, ALA is administered orally, and photosensitivity is limited to 24 48 hours. High rates of both pathologic response and disease-free survival have been reported during multiple-
year follow-up in patients undergoing ALA-PDT.23,24 However, a signicant proportion of patients undergoing ALA-PDT report procedural discomfort including chest pain.24,25 ALA-PDT can be associated with profound nausea and vomiting, hypotension, and arrhythmia, and a fatality has been reported.25 Emerging endosopic ablation techniques for Barretts esophagus include radiofrequency ablation and cryoablation. A multicenter study of a circumferential balloon-based radiofrequency ablation device (HALO360 System; BRRX Medical, Inc, Sunnyvale, CA) in 100 subjects with nondysplastic Barretts reported high remission rates, no buried glands in more than 4300 biopsies, and no post-treatment strictures.26 Similar ndings have recently been reported in abstract form for patients with HGD.27 Preliminary data suggest that cryoablation with liquid nitrogen spray can achieve a histopathologic response in patients with HGD or intramucosal carcinoma.28 In this respect, the endoscopic management of Barretts esophagus serves as a paradigm for the future of endoscopic therapy. Endoscopic therapeutic options include PDT, EMR, radiofrequency ablation, or a combination of these modalities. Ideally, choice among these techniques should be determined by data regarding patient outcomes, as opposed to provider preferences or available local expertise. Determining which patients are likely to experience which outcomes (positive and negative) of which therapeutic modality merits further attention in endoscopic research. Prospective, multicenter, controlled trials will be required to answer these questions. In summary, our experience with PDT for Barretts esophagus identies an association between length of Barretts esophagus and development of post-PDT stricture. This association was independent of treatment length. Patients with intramucosal or submucosal carcinoma also appear to be at increased risk for post-PDT stricture when compared with patients with HGD. The risk of stricture increases with multiple PDT courses. Prior EMR did not inuence the likelihood of post-PDT stricture. No complications were observed in patients undergoing dilation of post-PDT stricture. References
1. Singh P, Taylor RH, Colin-Jones DG. Esophageal motor dysfunction and acid exposure in reux esophagitis are more severe if Barretts metaplasia is present. Am J Gastroenterol 1994;89: 349 356. 2. Falk GW. Barretts esophagus. Gastroenterology 2002;122: 1569 1591. 3. Shaheen NJ, Crosby NA, Bozymski EM, et al. Is there publication bias in the reporting of cancer risk in Barretts esophagus? Gastroenterology 2000;119:587589. 4. US Cancer Statistics Working Group. United States cancer statistics: 1999 2002incidence and mortality web-based report. Atlanta: US Department of Health and Human Services, Centers for Disease Control and Prevention and National Cancer Institute, 2005. Available at: www.cdc.gov/cancer/npcr/uscs. 5. Ferguson MK, Mannheim KS. Resection for Barretts mucosa with high grade dysplasia: implications for prophylactic photodynamic therapy. J Thorac Cardiovasc Surg 1997;114:824 829. 6. Reed MF, Tolis G Jr, Edil BH, et al. Surgical treatment of esophageal high-grade dysplasia. Ann Thorac Surg 2005;79:1110 1115. 7. Chang LC, Oelschleger BK, Quiroga E, et al. Long-term outcome of esophagectomy for high-grade dysplasia or cancer found during surveillance for Barretts esophagus. J Gastrointest Surg 2006; 10:341346.
308
YACHIMSKI ET AL
8. Overholt BF, Panjehpour M, Halberg DL. Photodynamic therapy for Barretts esophagus with dysplasia and/or early stage carcinoma: long-term results. Gastrointest Endosc 2003;58:183 188. 9. Overholt BF, Panjehpour M. Photodynamic therapy for Barretts esophagus. Gastrointest Endosc Clin N Am 1997;7:207220. 10. Overholt BF, Lightdale CJ, Wang KK, et al. Photodynamic therapy with pormer sodium for ablation of high-grade dysplasia in Barretts esophagus: international, partially blinded, randomized phase III trial. Gastrointest Endosc 2005;62:488 498. 11. Overholt BF, Panjehpour M, Haydek JM. Photodynamic therapy for Barretts esophagus: follow-up in 100 patients. Gastrointest Endosc 1999;49:17. 12. Panjehpour M, Overholt BF, Haydek JM, et al. Dysplasia and early cancer in Barretts esophagus and effect of oral steroids on stricture formation. Am J Gastroenterol 2000;95:21772184. 13. Panjehpour M, Overholt BF, Phan MN, et al. Optimization of light dosimetry for photodynamic therapy of Barretts esophagus: efcacy vs incidence of stricture after treatment. Gastrointest Endosc 2005;61:1318. 14. Ban S, Mino M, Nishioka N, et al. Histologic aspects of photodynamic therapy for dysplasia and early adenocarcinoma arising in Barretts esophagus. Am J Surg Pathol 2004;28:1466 1473. 15. Van Laethem JL, Peny MO, Salmon I, et al. Intramucosal carcinoma arising under squamous re-epithelialisation of Barretts oesophagus. Gut 2000;46:574 577. 16. Overholt BF. Photodynamic therapy strictures: who is at risk? Gastrointest Endosc 2007;65:67 69. 17. Overholt BF, Wang KK, Burdick JS, et al. Five-year efcacy and safety of photodynamic therapy with Photofrin in Barretts highgrade dysplasia. Gastrointest Endosc 2007;66:460 468. 18. Larghi A, Lightdale CJ, Memeo L, et al. EUS followed by EMR for staging of high-grade dysplasia and early cancer in Barretts esophagus. Gastrointest Endosc 2005;62:16 23. 19. Prasad GA, Wang KK, Buttar NS, et al. Predictors of stricture formation after photodynamic therapy for high-grade dysplasia in Barretts esophagus. Gastrointest Endosc 2007;65:60 66. 20. Katada C, Muto M, Manabe T, et al. Esophageal stenosis after endoscopic mucosal resection of supercial esophageal lesions. Gastrointest Endosc 2003;57:165169. 21. Pacico R, Wang KK, Wongkersong LM, et al. Combined endoscopic mucosal resection and photodynamic therapy versus
22.
23.
24.
25.
26.
27.
28.
29.
esophagectomy for management of early adenocarcinoma in Barretts esophagus. Clin Gastroenterol Hepatol 2003;1:252 257. Wolfson HC, Hemminger LL, Raimondo M, et al. Photodynamic therapy and endoscopic mucosal resection for Barretts dysplasia and early esophageal adenocarcinoma. South Med J 2004; 97:827 830. Mackenzie G, Selvasekar C, Jamieson N, et al. Low incidence of esophageal adenocarcinoma following optimal regimen of ALA PDT for high grade dysplasia in Barretts esophagus. Gastrointest Endosc 2007;65:AB132. Pech O, Gossner L, May A, et al. Long-term results of photodynamic therapy with 5-aminolevulinic acid for supercial Barretts cancer and high-grade intraepithelial neoplasia. Gastrointest Endosc 2005;62:24 30. Hage M, Siersema PD, van Dekken H, et al. 5-aminolevulinic acid photodynamic therapy versus argon plasma coagulation for ablation of Barretts oesophagus: a randomized trial. Gut 2004;53: 785790. Sharma VK, Wang KK, Overholt BF, et al. Balloon-based circumferential, endoscopic radiofrequency ablation of Barretts esophagus: 1-year follow-up. Gastrointest Endosc 2007;65:185195. Gondrie JJ, Peters F, Curvers WL, et al. Radiofrequency ablation of Barretts esophagus containing high-grade dysplasia. Gastrointest Endosc 2007;65:AB135. Dumot JA, Vargo JJ, Zuccaro G, et al. Preliminary results of cryotherapy ablation for esophageal high grade dysplasia (HGD) or intra-mucosal cancer (IMC) in high risk non-surgical patients. Gastrointest Endosc 2007;52:AB110. Keeley SB, Pennathur A, Gooding W, et al. Photodynamic therapy with curative intent for Barretts esophagus with high-grade dysplasia and supercial esophageal cancer. Ann Surg Oncol 2007; 14:2406 2410.
Address requests for reprints to: Patrick Yachimski, MD, Blake 4 Gastrointestinal Unit, Massachusetts General Hospital, 55 Fruit St, Boston, MA 02115. e-mail: pyachimski@partners.org; fax: 617-7246832. A version of these data was presented in abstract form at Digestive Diseases Week, May 19 24, 2007, Washington, DC.
Peptic Ulcerations Are Related to Systemic Rather Than Local Effects of Low-Dose Aspirin
MARTIJN G. H. VAN OIJEN,*, JEANNE P. DIELEMAN, ROBERT J. F. LAHEIJ,*, MIRIAM C. J. M. STURKENBOOM,, JAN B. M. J. JANSEN,* and FREEK W. A. VERHEUGT
*Department of Gastroenterology, and Department of Cardiology, Radboud University Nijmegen Medical Center, Nijmegen, The Netherlands; and the Department of Medical Informatics, and the Department of Epidemiology & Biostatistics, Erasmus MC, University Medical Center, Rotterdam, The Netherlands
Background & Aims: Effervescent calcium carbasalate is a

calcium-salt of acetylsalicylic acid causing less local gastric damage than acetylsalicylic acid at high doses in healthy controls. The aim of the study was to investigate the incidence of peptic ulcers in a population-based cohort using bioequivalent low-dose acetylsalicylic acid (80 mg) or effervescent calcium carbasalate (100 mg). Methods: Incident acetylsalicylic acid or effervescent calcium carbasalate users were identied from the Integrated Primary Care Information database. The study cohort comprised 19,819 subjects: 11,891 on acetylsalicylic acid and 7928 on effervescent calcium carbasalate. Incidence rates for documented peptic ulcer disease conrmed by endoscopy were calculated and time-dependent adjusted Cox regression analysis was used to compare the risk of peptic ulcers for patients using acetylsalicylic acid or effervescent calcium carbasalate. Results: During an average 1.85 years of follow-up evaluation, 115 ulcers were found. The risk for developing a peptic ulcer during drug use was: 3.07 per 1000 person-years for acetylsalicylic acid and 4.31 for effervescent calcium carbasalate. The risk of peptic ulcers was not statistically signicantly higher in patients using effervescent calcium carbasalate than in acetylsalicylic acid users (adjusted hazard ratio, 1.39; 95% condence interval, 0.922.12). Conclusions: The incidence rate of peptic ulcer disease is similar in patients using low-dose effervescent calcium carbasalate compared with regular low-dose acetylsalicylic acid. This implicates that peptic ulcers seem to be related to systemic rather than to local effects of low-dose acetylsalicylic acid. he benecial effect of platelet aggregation inhibitors for secondary prevention of cardiovascular diseases is well established.13 The most frequently used drugs for this purpose are low-dose salicylates. In the Netherlands acetylsalicylic acid and effervescent calcium carbasalate (a calcium salt of salicylic acids) account for about 4.5 million prescriptions (2,707,800 and 1,767,100, respectively) per year.4 Both drugs work through irreversible inhibition of cyclooxygenase-1, thereby preventing the conversion of arachidonic acid into thromboxane A2, a platelet agonist. Salicylates also are known for their local and systemic side effects in the upper-gastrointestinal tract, even at low daily dosages of 80 mg. The formation of prostaglandins required for normal gastric mucosal integrity is blocked as a result of irreversible inhibition of cyclooxygenase-1.5,6 The main difference between acetylsalicylic acid and effervescent calcium carbasalate lies in the local gastrotoxic effects. Acetylsalicylic acid is a weak
acid, whereas effervescent calcium carbasalate is a soluble calciumsalt. Poor disintegration of the acetylsalicylic acid tablet may cause drug concentrations to be higher in certain parts of the stomach, leading to a direct irritating effect on the gastric mucosa, whereas calcium carbasalate should not have this problem.7 The effect of both medications on the occurrence of peptic ulcers has been studied in a randomized clinical cross-over trial wherein 20 healthy volunteers were administered high equipotent doses of acetylsalicylic acid (650 mg 3 times/day) or effervescent calcium carbasalate (826.8 mg 3 times/day) for 5 days.8 Effervescent calcium carbasalate caused fewer gastroduodenal mucosal erosions than acetylsalicylic acid. Currently, no data are available comparing the effect of long-term use of acetylsalicylic acid and effervescent calcium carbasalate on the occurrence of clinically symptomatic peptic ulcers in standard low doses for the prevention of cardiovascular disease in the general population. We therefore assessed and compared the incidence of peptic ulcers among users of lowdose acetylsalicylic acid and users of effervescent calcium carbasalate in a population-based primary care setting.
Methods Setting
All data were retrieved from the Integrated Primary Care Information project, a general practice research database containing data from electronic patient records of a group of about 150 general practitioners in the Netherlands. Details of the database have been described elsewhere.9,10 Briey, the database contains the complete medical records of more than 500,000 patients. The electronic records contain coded and anonymous data on patient demographics, reasons for visit (in free text), diagnoses (using the International Classication for Primary Care11 and free text) from general practitioners and specialists, referrals, laboratory ndings, hospitalizations, and drug prescriptions, including indications and dosage regimen. To maximize completeness of the data, general practitioners participating in the Integrated Primary Care Information project are not allowed to maintain a system of paper-based records aside from the electronic medical records. The system complies with European Union guidelines on the use of medical data for medical research and has been proven valid for pharmacoepidemiologic research.12 The Scientic and Ethical AdviAbbreviation used in this paper: CI, condence interval. 2008 by the AGA Institute 1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.018
310
VAN OIJEN ET AL
sory Board of the Integrated Primary Care Information project approved the study.
Source Population and Exposure Cohorts

The study cohort started from January 1, 1995, and was analyzed to September 30, 2003. The source population comprised all subjects with at least 1 year of valid history available in the database. This meant that patients had to be registered with the general practitioner for at least 1 year and the general practitioner had to be participating in the Integrated Primary Care Information project for at least 1 year. This year was needed to be able to identify incident use of acetylsalicylic acid or effervescent calcium carbasalate and to assess the presence of comorbidity at baseline. From the source population we identied all new users of either low-dose acetylsalicylic acid or effervescent carbasalate calcium (80 100 mg/day). These patients were followed up from start of prescription until the earliest of transferal out of the practice, death, end of the study period, or occurrence of a peptic ulcer. In this exposure cohort, the duration of use of individual medication was calculated by dividing the prescribed quantity by the prescribed daily dosage. From this we determined periods of exposure, which were classied as acetylsalicylic acid, effervescent calcium carbasalate, or nonuse. Person-time of exposure was accumulated during the follow-up period for calculation of incidence rates. Subjects were allowed to switch between acetylsalicylic acid and effervescent calcium carbasalate and have treatment interruptions. Hence, subjects could contribute person-time to multiple exposure categories. Switching was considered in the analysis by including previous use of another type of salicylic acid as a covariate.
cium carbasalate users. Pearsons chi-squared statistics were used to compare distributions of categoric variables between the cohorts and the Student t test was used to compare the age distributions. Crude incidence rates of documented peptic ulcer disease conrmed by endoscopy were calculated by dividing the number of peptic ulcer cases by the corresponding person-years of the exposure category in which they occurred. Ninety-ve percent condence intervals (CIs) were calculated based on a Poisson distribution. Cox regression analysis was used to calculate hazard ratios and 95% CIs for the risk of documented peptic ulcer conrmed by endoscopy during use of effervescent calcium carbasalate compared with acetylsalicylic acid (used as a reference 1). Adjusted Cox regression then was performed to calculate the risk (hazard ratios) of documented and conrmed peptic ulcer occurrence in our study population, adjusted for important (univariate P .10) covariates. Because switching between exposure categories might occur during the course of follow-up evaluation we analyzed exposure as a timedependent factor. Additional stratied adjusted analyses were conducted to explore the effect of a history of peptic ulcer disease and acid-suppressive medication use as effect modiers. All analyses were performed using SAS statistical software (version 8.2, SAS Institute Inc, Cary, NC). Statistical signicance was accepted as a 2-sided P value of less than .05.
Results
The source population comprised 364,683 patients who had on average 2.7 years of follow-up evaluation. Within this study population, 11,891 patients used low-dose acetylsalicylic acid and 7928 used low-dose effervescent calcium carbasalate at baseline. The average follow-up period of these patients was 1.85 years. Baseline characteristics of the study population are presented in Table 1. The mean age (SD) of the total population of platelet aggregation inhibitor users in this population was 67.5 years (13.5 y) and was comparable between acetylsalicylic acid and effervescent calcium carbasalate users (P .12). Concomitant use of nonsteroidal anti-inammatory drugs, calcium channel blockers, and acid-suppressive medication was common in our population (19%, 21%, and 14%, respectively). Effervescent calcium carbasalate users used signicantly more acid-suppressive medication and coumarin derivates (P .05). Users of effervescent calcium carbasalate also had a higher prevalence of heart failure, rheumatoid arthritis, and osteoarthritis (Table 1). The total duration of acetylsalicylic acid use was 3,068,033 days (8399.82 y) and the mean duration of acetylsalicylic acid use per patient was 691.19 days (SD, 582.31). For effervescent calcium carbasalate the total duration of use in our population was 2,426,427 days (6643.20 y), with a mean duration of use per patient of 653.73 days (SD, 590.21). During the follow-up period we identied 115 documented peptic ulcers conrmed by endoscopy. The overall incidence rate of peptic ulcer disease during acetylsalicylic acid use was 3.07 per 1000 person-years, which was comparable with effervescent calcium carbasalate users (incidence rate, 4.31 per 1000 person-years; P .68) (Table 2). A total of 871 patients (7%) switched from acetylsalicylic acid to effervescent calcium carbasalate and 1124 patients (14%) switched from effervescent calcium carbasalate to acetylsalicylic acid (P .01). These switchers were more likely to
Identication and Ascertainment of Peptic Ulcers

The rst occurrence of documented peptic ulcer disease for each individual was identied through searches on diagnoses and free text indicators for peptic ulcers. The medical records of all potential cases were reviewed manually by a physician unaware of the research question and classied as peptic ulcers, if conrmed by upper-gastrointestinal endoscopy in secondary care. In addition, the date of referral for endoscopy for rst occurrence of (symptoms for) a peptic ulcer was veried in the patient record. Cases were attributed to the exposure category in which they occurred.
Covariates
As covariates in this cohort analysis, we considered age, sex, patients history of peptic ulcer disease, other comorbidity including the indication for use and the use of interfering medication. As important concomitant medications we considered nonsteroidal anti-inammatory drugs, oral steroids, calcium antagonists, anticoagulates other than acetylsalicylic acid, selective serotonin reuptake inhibitors, and diuretics because of their gastrointestinal side effects. The use of gastric acid secretioninhibiting medication, such as proton pump inhibitors and H2-receptor antagonists, was considered as a covariate because of their benecial and preventive effect on peptic ulcers.
Data Analysis
Standard descriptive statistics were used to describe the characteristics of the acetylsalicylic acid and effervescent cal-
March 2008
PEPTIC ULCER IN LOW-DOSE ASPIRIN USERS
311
Table 1. Baseline Characteristics

Acetylsalicylic acid (n 11,891) Demographics Males, n (%) Mean age, y (SD) 65 y, n (%) Concurrent drug use, n (%) Nonsteroidal antiinammatory drugsa Oral steroids Calcium channel blocker Acid-suppressive medicationb Selective serotonin reuptake inhibitors Coumarin derivates Comorbidity, n (%) History of validated peptic ulcer disease Ischemic heart disease Heart failure Diabetes mellitus Rheumatoid arthritis Osteoarthritis Effervescent calcium carbasalate (n 7928)
Table 3. Unadjusted and Adjusted Hazard Ratios for Documented Peptic Ulcers Conrmed by Endoscopy
Hazard ratios for documented peptic ulcers (95% CI) Unadjusted Acetylsalicylic acidb Effervescent calcium carbasalate Age 65 y Female sex History of peptic ulcers Use of nonsteroidal antiinammatory drugs Use of acid suppressors Use of selective serotonin reuptake inhibitors Use of coumarin derivatives Medication switchc
aAdjustment
6,575 (55) 67.6 (13.1) 7270 (61) 2216 (19) 533 (4) 2532 (21) 1574 (13) 347 (3) 526 (4) 104 (1) 10,565 (89) 572 (5) 1461 (12) 281 (2) 548 (5)
4,347 (55) 67.3 (14.1) 4864 (61) 1561 (20) 344 (4) 1600 (20) 1218 (15) 227 (3) 439 (6) 90 (1) 7128 (90) 516 (7) 987 (12) 329 (4) 501 (6)
Adjusteda Reference 1.39 (0.922.12) 1.97 (1.183.30) 0.74 (0.481.15) 10.1 (5.2319.6) 1.27 (0.792.02) 1.33 (0.822.14) 3.76 (1.977.16) 3.44 (1.856.39) 6.82 (4.3410.7)
Reference 1.70 (1.132.57) 2.06 (1.283.33) 0.70 (0.471.05) 15.4 (8.9026.6) 1.06 (0.681.64) 2.39 (1.603.57) 3.78 (2.156.64) 4.26 (2.387.62) 9.43 (6.1714.4)
NOTE. Bolded entries: P .05. aNonselective nonsteroidal anti-inammatory drugs and selective cyclooxygenase-2 inhibitors. bH -receptor antagonists and proton pump inhibitors. 2
for age, sex, history of peptic ulcer, medication switch, and concurrent medication use. bAcetylsalicylic acid is used only as a reference for comparison with effervescent calcium carbasalate. cSwitch from acetylsalicylic acid to effervescent calcium carbasalate or vice versa.
experience a peptic ulcer (hazard ratio, 9.4; 95% CI, 6.214.4) compared with all nonswitchers. Other factors associated with the risk of peptic ulcers in our population were age older than 65 years (hazard ratio, 2.06; 95% CI, 1.28 3.33), patients history of peptic ulcer disease (hazard ratio, 15.4; 95% CI, 8.9 26.6), use of acid-inhibiting medication (hazard ratio, 2.39; 95% CI, 1.60 3.57), use of selective serotonin reuptake inhibitors (hazard ratio, 3.78; 95% CI, 2.15 6.64), use of coumarin derivatives (hazard ratio, 4.26; 95% CI, 2.38 7.62), and the use of effervescent calcium carbasalate compared with the use of acetylsalicylic acid (hazard ratio, 1.70; 95% CI, 1.132.57). From Table 3 we can conclude that, after adjustment for all known risk factors for peptic ulcer disease, effervescent calcium carbasalate shows a similar risk for peptic ulcer disease as compared with acetylsalicylic acid (hazard ratio, 1.39; 95% CI, 0.922.12). Stratied analyses for patients with and without a documented history of peptic ulcer disease showed that the relative risk was higher in the group without peptic ulcer disease in the past 1.75 (95% CI, 1.132.72) than in the group with a history of peptic ulcer disease in the past 0.90 (95% CI, 0.26 3.09). The unadjusted relative risk in the group using
gastric acid inhibitors (proton pump inhibitor or histamine2receptor antagonists) was not signicantly higher (hazard ratio, 1.95; 95% CI, 0.90 4.22) compared with patients who did not use this cotherapy (hazard ratio, 1.57; 95% CI, 0.96 2.56). After adjustment for age, sex, history of peptic ulcer, and concurrent medication use these relative risks changed to 1.40 (95% CI, 0.623.18) for gastric acid inhibitor users and remained at 1.57 (95% CI, 0.96 2.57) for the nonusers.
Discussion
This study showed that patients using low-dose effervescent calcium carbasalate have a similar risk of endoscopically documented peptic ulcer disease compared with persons using a similar low dose of acetylsalicylic acid, after adjusting for other known risk factors for peptic ulcer disease such as age, history of peptic ulcer, and concomitant drug use. A possible explanation for the higher than expected risk of peptic ulcers in effervescent calcium carbasalate users could be the result of differences in chemical properties of the drugs. As a weak acid, acetylsalicylic acid is less soluble in the gastric, low-pH environment and therefore may cause local damage and be less absorbed than effervescent calcium carbasalate. As a consequence equivalent molar oral doses of salicylic acid that are formulated differently could result in
Table 2. Incidence Rates per 1000 Person-Years for Both Acetylsalicylic Acid and Effervescent Calcium Carbasalate
Number of patients Acetylsalicylic acid Effervescent calcium carbasalate 11,891 7928 Person-years exposed 21,167 11,594 No. of events (% of patients) 65 (0.55) 50 (0.63) Incidence rate (per 1000 patient-years) 3.07 4.31 P value .68
312
VAN OIJEN ET AL
different plasma levels of salicylic acid and therefore result in systemic toxic effects. Previous studies on differently formulated acetylsalicylic acids have been conducted. Several studies have been undertaken to study the toxicity of high-dose buffered and entericcoated aspirin.13,14 Enteric-coated aspirin signicantly reduces (local) gastric toxicity, but buffered aspirins were comparable with plain aspirin.1315 To our knowledge, the only study that compared both effervescent carbasalate calcium and plain acetylsalicylic acid was the study by Murray et al8 in 20 healthy volunteers. They concluded that the high doses of the drugs are equivalent therapeutically, but that effervescent calcium carbasalate caused less gastric mucosal damage than acetylsalicylic acid. Being an observational study in an existing database of complete patient records, misclassication and confounding should be taken into consideration while interpreting the results. Confounding by (contra-)indication could be a point of concern. Because of its gastroprotective image, physicians may prescribe effervescent calcium carbasalate preferably to patients with known upper-gastrointestinal symptoms (ie, channeling). In our population about 10% of the patients switched between both forms of medication studied. Patients switched equally often from plain acetylsalicylic acid to effervescent calcium carbasalate and vice versa. However, there was no evidence of strong channeling because concurrent medications used for or known to cause gastrointestinal disease were not different between both groups. If any, the small differences in the prevalence of concomitant disease and drugs cannot explain the similarity of peptic ulcer disease occurrence during effervescent calcium carbasalate or acetylsalicylic acid use. However, patients who switched between both formulations were at higher risk for developing peptic ulcer compared with patients who did not switch. This may be contributed to the underlying reason for switching. If a patient will be admitted to a hospital, or presents to the prescribing physician with intolerance (eg, dyspepsia) for either effervescent calcium carbasalate or acetylsalicylic acid, the treating facility may choose to switch the patients aspirin formulation. Unfortunately, in our database no data were available regarding the reason for switching. Misclassication of the outcome may have occurred because we restricted the analysis to endoscopically conrmed ulcers only to avoid inclusion of false-positive peptic ulcers. We thus may have underestimated the incidence rate of peptic ulcer disease. This is only a point of concern if the misclassication is differential, which may be the case if physicians are more likely to request an endoscopy for patients using one salicylic acid rather than the other. Because of the gastroprotective image of effervescent calcium carbasalate we believe that, if any, differential misclassication of the outcome would only reduce the risk estimate because endoscopies probably are less likely to be performed in effervescent calcium carbasalate users. The primary outcome of this study was the occurrence of documented peptic ulcer disease, conrmed by endoscopy. We used free text to nd the combination of peptic ulcer and endoscopy, but unfortunately the exact anatomic location or pathologic reports were not documented in the studied general practitioner database. For additional insights into the etiology of gastroduodenal damage by different low-dose aspirin formu-
lations, an endoscopy study should be performed comparing these formulations, including the studied acetylsalicylic acid and effervescent calcium carbasalate. The combined effects of aspirin and clopidogrel have been studied widely in large cardiovascular randomized clinical trials. In our adjusted analysis, we adjusted for anticoagulants (eg, coumarin derivates), as covariates in our analysis. However, during the period of inclusion in the study cohort, the Dutch Cardiology guidelines and medicine reimbursement schemes did not advise or reimburse the long-term use of clopidogrel as cotherapy of acetylsalicylic acid or effervescent calcium carbasalate. Only recently, patients will be reimbursed for a half-year of double therapy with aspirin and clopidogrel after an acute coronary event or elective percutaneous coronary intervention or coronary artery bypass graft surgery intervention. For this reason we did not have adequate data about clopidogrel use and could not include that in our analysis. In conclusion, the incidence rate of endoscopically diagnosed peptic ulcers is similar in patients using low-dose effervescent calcium carbasalate compared with low-dose acetylsalicylic acid. This suggests that systemic effects of low-dose aspirin may be more important than local effects, which implicates that any formulation of aspirin should be used cautiously in high-risk patients. References
1. Pearson TA, Blair SN, Daniels SR, et al. AHA guidelines for primary prevention of cardiovascular disease and stroke: 2002 update. Circulation 2002;106:388 391. 2. Verheugt FWA. Aspirin, the poor mans statin. Lancet 1998;351: 227228. 3. Verheugt FWA. In search of a super-aspirin for the heart. Lancet 1997;349:1409 1410. 4. Facts and gures 2005. The Netherlands: Foundation for Pharmaceutical Statistics, 2006. 5. Garcia Rodriguez LA, Hernandez-Diaz S. Risk of uncomplicated peptic ulcer among users of aspirin and nonaspirin nonsteroidal antiinammatory drugs. Am J Epidemiol 2004;159:2331. 6. Derry S, Loke YK. Risk of gastrointestinal haemorrhage with long term use of aspirin: meta-analysis. BMJ 2000;321:1183 1187. 7. Cryer B, Feldman M. Effects of very low dose daily, long-term aspirin therapy on gastric, duodenal, and rectal prostaglandin levels and on mucosal injury in healthy humans. Gastroenterology 1999;117:1725. 8. Murray FE, Hudson N, Atherton JC, et al. Comparison of effects of effervescent calcium carbasalate and aspirin on gastroduodenal mucosal damage in human volunteers. Gut 1996;38: 1114. 9. Vlug AE, van der Lei J, Mosseveld BM, et al. Post-marketing surveillance based on electronic patient records: the IPCI project. Methods Inf Med 1999;38:339 344. 10. van der Lei J, Duisterhout JS, Westerhof HP, et al. The introduction of computer-based patient records in The Netherlands. Ann Intern Med 1993;119:1036 1041. 11. Lamberts H, Wood M, Hofmans-Okkes IM. International primary care classications: the effect of fteen years of evolution. Fam Pract 1992;9:330 339. 12. WHO Collaborating Centre for Drug Statistics Methodology. ATC index with DDDs. Oslo, Norway: WHO Collaborating Centre for Drug Statistics Methodology, 2002. 13. Lanza FL, Royer GL Jr, Nelson RS. Endoscopic evaluation of the effects of aspirin, buffered aspirin and enteric-coated
March 2008
PEPTIC ULCER IN LOW-DOSE ASPIRIN USERS
313
aspirin on gastric and duodenal mucosa. N Engl J Med 1980;303:136 138. 14. Kelly JP, Kaufman DW, Jurgelon JM, et al. Risk of aspirinassociated major upper-gastrointestinal bleeding with entericcoated or buffered product. Lancet 1996;348:14131416. 15. Garcia Rodriguez LA, Hernandez-Diaz S, de Abajo FL. Association between aspirin and upper gastrointestinal complications: systematic review of epidemiological studies. Br J Clin Pharm 2001;52:563571.
Address requests for reprints to: Martijn van Oijen, MSc, Department of Gastroenterology and Hepatology, Radboud University Nijmegen Medical Center, PO Box 9101, 6500 HB Nijmegen, The Netherlands. e-mail: M.vanOijen@MDL.umcn.nl; fax: (31) 243540103. The authors want to express special thanks to Fabian van der Sluis, MD, Erasmus MC, Rotterdam, for his contribution to the validation process.
What Is the Role of Serologic Testing in Celiac Disease? A Prospective, Biopsy-Conrmed Study With Economic Analysis
ANDREW D. HOPPER,* MARIOS HADJIVASSILIOU, DAVID P. HURLSTONE,* ALAN J. LOBO,* MARK E. McALINDON,* WILLIAM EGNER, GRAEME WILD, and DAVID S. SANDERS*
*Department of Gastroenterology, Department of Neurology, Royal Hallamshire Hospital, Shefeld, United Kingdom; and the Department of Immunology Northern General Hospital Shefeld, United Kingdom
See CME exam on page 263. Background & Aims: The optimal serologic tests for the detection of celiac disease and follow-up assessment remains controversial. Our aim was to evaluate all current immunologic assays for diagnosing celiac disease using the gold standard of duodenal biopsy. We also assessed whether tissue transglutaminase (tTG) antibody is a quantitative marker for histologic severity. Methods: Consecutive adult patients referred for gastroscopy without a previous known diagnosis of celiac disease were recruited (group 1). Concurrently, patients with a known diagnosis of celiac disease on a gluten-free diet for more than 1 year undergoing repeat duodenal biopsy were identied (group 2). All patients had duodenal biopsies and serologic analysis performed for immunoglobulin(Ig) A and antibodies to human immunoglobulin (Ig)A-tTG, IgA-gliadin, IgG-gliadin, and IgA-endomysial antibody. Results: Two thousand patients were recruited in the rst group. Seventyseven (3.9%) patients were diagnosed with new celiac disease. The sensitivity, specicity, positive predictive value, and negative predictive value for IgA tTG were 90.9%, 90.9%, 28.6%, and 99.6%. When adopting a 2-step approach using tTG rst and then EMA the sensitivity, specicity, positive predictive value, and negative predictive value was 85.7%, 98.6%, 71.7%, and 99.7%, respectively. The use of nondeamidated IgA/IgG gliadin antibodies conferred no additional diagnostic benet when considering the detection of adult celiac disease. In the second group 48 patients with celiac disease on a gluten-free diet were identied. Sixteen of 48 of these patients had persisting villous atrophy, but 7 of 16 (44%) had a normal tTG level. Conclusions: IgA tTG alone is a sensitive marker for celiac disease. A normal tTG level does not predict recovery of villous atrophy in patients with celiac disease on a gluten-free diet.
he prevalence of celiac disease in the United States, Europe, and the United Kingdom is between 0.75% and 1%.1 4 In addition, we now accept that there may be many associated conditions and symptoms that warrant serologic testing for celiac disease (adopting a case-nding approach). Ultimately, the diagnosis of celiac disease still requires a smallbowel biopsy showing villous atrophy.5 8 There are a number of serologic tests that have been reported to be accurate in identifying patients who then should be referred for a duodenal biopsy. However, the optimal serologic
test or test strategy remains controversial. Previously, anti endomysial antibody (EMA) had been reported as an accurate test with a sensitivity greater than 90% and a specicity greater than 98%.9 11 However, it is recognized that EMA requires a subjective immunouorescence method and has limited substrate resources (either monkey esophagus or umbilical cord).12 Recently, the introduction of antitissue transglutaminase (tTG) antibody testing using either guinea pig or human recombinant tTG has led investigators to suggest that human recombinant tTG may have higher sensitivities than EMA.12,13 An advantage of automated tTG testing is higher throughput of samples and also the opportunity to obtain a quantitative titer using the enzyme-linked immunosorbent assay method. However, there are more false-positive results associated with tTG testing.14,15 This lower specicity has led some investigators to describe a 2-step method (tTG rst and, if positive, followed by EMAif there is a positive EMA result then proceed to biopsy) to avoid patients undergoing unnecessary duodenal biopsies.12,13 Despite the high accuracy of EMA and tTG, seronegative celiac disease still occurs. This has been reported to account for 6.4% (8 of 126) of all cases of celiac disease,16 and appears to occur more often with lesser degrees of atrophy.17,18 For this reason duodenal biopsy in patients with a high suspicion of celiac disease still is recommended even if the serologic testing is negative.19 Currently published multicenter trials with large cohorts that have tried to compare EMA and tTG accuracy have taken 1 of 2 approaches. Either they have performed a duodenal biopsy in patients who were antibody positive or they have recruited patients known to have a high risk of celiac disease, for example, iron-deciency anemia.16,20,21 However, we were unable to identify any study that evaluated all of the current antibody tests in a large series of low-risk adult patients with concomitant duodenal biopsy performed in all patients. Once a diagnosis of celiac disease has been made, there is no single method that allows for assessment of compliance. Questioning patients for symptom response to a gluten-free diet and a dietary assessment can be used as a noninvasive marker, but this may not correlate with intestinal damage.22 However, villous recovery on a gluten-free diet may take up to 24 months or longer in adults.23,24 There may be a role for serology in the
Abbreviations used in this paper: EMA, endomysial antibody; IEL, intraepithelial lymphocyte; IG, immunoglobulin; NPV, negative predictive value; PPV, positive predictive value; tTG, tissue transglutaminase. 2008 by the AGA Institute 1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.008
March 2008
EVALUATION OF CURRENT SEROLOGIC TESTS IN CELIAC DISEASE
315
assessment of compliance but, to date, EMA has been shown to be a poor predictor of villous recovery.25 For these reasons the aim of our study was to assess which serologic tests or test strategy is optimal for both the recognition and follow-up evaluation of patients with celiac disease.
Methods Patient Recruitment

Patient recruitment took place in a single endoscopy department at the Royal Hallamshire Hospital in Shefeld (United Kingdom). Consecutive adult patients referred for gastroscopy were recruited by a single endoscopist over a 26-month period (patient group 1 was recruited from January 2004 to April 2006). The Department of Gastroenterology currently uses a policy of taking 4 duodenal biopsy specimens routinely as part of the endoscopic examination. Patient consent was obtained and the gastroscopy examination was performed with 4 biopsy specimens taken from the second part of the duodenum. At the same time a blood sample was obtained from each patient and analyzed for total immunoglobulin A (IgA), IgAgliadin, IgG-gliadin, IgA-tTG, and IgA-EMA. Patients were excluded from this cohort if they had a known diagnosis of celiac disease, a coagulopathy (international normalized ratio 1.3 or platelet count of 80), active gastrointestinal bleed or a suspected carcinoma observed during the examination, or refused to participate in the study. Patients then found to have villous atrophy (on duodenal biopsy) with supportive serology and symptoms were classied as having celiac disease. Patients with villous atrophy (conrmed on a second review of the sample to ensure a well-oriented sample) and a negative antibody prole were classied as having seronegative celiac disease after further assessment. To further support the diagnosis of antibody-negative celiac disease alternative causes of villous atrophy were excluded (such as Giardia and Helicobacter pylori infection, selective IgA deciency) and a human lymphocyte antigen prole was checked. To conrm the diagnosis of seronegative celiac disease these patients were required to have the DQ2 or DQ8 pattern consistent with celiac disease and a clinical and histologic response to a gluten-free diet.26,27
sections at 3 levels were stained with H&E. If changes suggestive of celiac disease were present in a biopsy specimen it was graded according to the modied Marsh criteria using the most severe lesion present28: Marsh 0, normal appearance; Marsh 1, normal morphology with raised intraepithelial lymphocytes (IELs); Marsh 2, raised IEL with crypt hyperplasia; Marsh 3a, partial villous atrophy; Marsh 3b, subtotal villous atrophy; and Marsh 3c, total villous atrophy.29 Total IgA was measured on a Behring BN2 nephelometer (Siemens Healthcare, Frankfurt, Germany). IgA gliadin, IgG gliadin, and human tTG antibodies were assayed on enzyme-linked immunosorbent assay kits from Aeskulisa (AESKU. DIAGNOSTICS, Wendelsheim, Germany). Serologic samples with a titer greater than 15 U/mL were taken as positive. IgA EMA was detected by immunouorescence on primate esophagus sections from The Binding Site (Birmingham, UK). Ethical approval was obtained from the South Shefeld research and ethics committee. Statistical analysis was performed using SPSS version 10.0 (SPSS Inc, Chicago, IL). All comparisons between sensitivities of tTG and EMA were made using the Fisher exact test, and the mean values of tTG levels and age groups were compared with an independent sample t test.
Results Assessment of the Clinical Performance of Available Serology Tests

In group 1 there were 2000 patients recruited (1167 [58.3%] females; mean age, 55.8 y; range, 16 94 y). From this group a total of 77 patients were diagnosed with new celiac disease, giving a prevalence for celiac disease in all patients attending for gastroscopy of 3.9% (prevalence previously reported by our group19). The histologic grading of villous atrophy for these 77 patients was as follows: 29 with Marsh 3a, 30 with Marsh 3b, and 18 with Marsh 3c lesions. During the recruitment period a total of 2220 patients had a gastroscopy, and 220 patients were excluded (previous diagnosis of celiac disease, 48; gastrointestinal bleed, 101; probable carcinoma, 12; unable to tolerate gastroscopy, 8; coagulopathy, 12; follow-up gastroscopy during study period, ie, already recruited, 36; refused to be included in the trial, 3). The indications for performing the gastroscopy for all the examinations and in patients found to have celiac disease are shown in Figure 1. In group 1, we compared the prevalence of symptoms in patients found to have celiac disease (n 77) with those who did not (n 1923). Patients with celiac disease had a significantly higher prevalence of weight loss (15.6% vs 5.3%) and diarrhea (42.9% vs 5.2%). Patients without celiac disease had a signicantly higher prevalence of dyspepsia (17.3% vs 1%), reux (13.8% vs 1%), and dysphagia (7.2% vs 0%). The celiac disease group was signicantly younger (mean age, 48.0 vs 56.1 y) and contained a higher percentage of females (70.1% vs 57.9%) (signicance is dened as P .05 for all comparisons). Table 1 shows the sensitivity, specicity, positive predictive value (PPV), and negative predictive value (NPV) for the individual antibodies and their combinations using tTG and EMA simultaneously or in a 2-step method. Also shown in Table 1 is
Correlation of Tissue Transglutaminase Antibody Titer With Degree of Villous Atrophy

To assess the relationship between tTG and histology we studied both patients diagnosed with celiac disease from group 1 and any patients not in that study but who were newly diagnosed with celiac disease during the same time frame and had the same antibody assessment (in our center). A second cohort of patients (group 2) formed a known celiac disease group. These were patients undergoing gastroscopy and duodenal biopsy for assessment of histologic remission. These patients had been on a gluten-free diet for greater than 1 year. Serology was obtained in an identical manner to that described previously.
Serology and Histology

All duodenal biopsy specimens were xed in buffered formalin and embedded in parafn wax. Standard, 3-mthick
316
HOPPER ET AL
Figure 1. Referral symptoms for gastroscopy. (A) Patients symptoms that prompted referral for gastroscopy. The total was 2114 as a result of some patients (n 2000) having more than 1 symptom. (B) Patients symptoms that prompted referral for gastroscopy and were found to have celiac disease. The total was 96 as a result of some patients (n 77) having more than 1 symptom.
the number of patients who would require a duodenal biopsy depending on which serologic strategy is used and the corresponding number of new celiac disease cases that potentially would be missed by taking the different approaches. Table 2 shows a corresponding economic analysis for the different serologic and biopsy strategies. When specically considering serology-negative celiac disease there were 6 patients in the cohort of 2000 who had celiac disease but were tTG and EMA negative. All 6 were IgA-gliadin negative; 1 of the 6 patients (who presented with anemia) had an IgA deciency but, despite this, this patient also had a negative IgG gliadin. Seven celiac disease patients were tTG negative, but 1 of these 7 patients had a positive EMA. The overall prevalence of IgA deciency in our series was 0.7% (14 of 2000), of which only 1 patient had celiac disease.
Correlating Serology With Histologic Severity (Newly Diagnosed Patients With Celiac Disease Not on a Gluten-Free Diet)
During the study period we identied a total of 114 patients with newly diagnosed celiac disease. Seventy-seven patients were obtained from group 1 (77 of 2000) and a further 37 who were found to have celiac disease in the department but outside of group 1 recuitment due to being on other endoscopists lists. The mean age of the whole group (n 114) was 47.3 years (range, 17 85 y), and 69.2% (79 of 114) were female. The ndings on histology showed that 37.7% (43 of 114) of these patients had a Marsh 3a lesion, 38.6% (44 of 114) had a Marsh 3b lesion, and 23.7% (27 of 114) had a Marsh 3c lesion. Patients with partial or subtotal villous atrophy (Marsh 3a or 3b lesion) had a signicantly lower average (mean) tTG titer
March 2008
317
Table 1. Evaluation of Different Serologic Strategies That Could Be Used for Testing for Celiac Disease
Resulting number of patients undergoing Missed cases of duodenal biopsy celiac disease per 2000 out of 77 245 104 92 92 257 114 238 51 7 10 11 11 6 40 39 49
Serologic tests used to refer for biopsy Only tTG positive Only EMA positive If tTG positive and then EMA positive (2-step) Both tTG positive and EMA positive Either tTG positive or EMA positive IgG gliadin positive IgA gliadin positive Both IgG gliadin and IgA gliadin positive
Sensitivity, (%)
Specicity, (%)
PPV, (%)
NPV, (%)
90.9 (82.494.5) 90.9 (89.592.1) 28.6 (23.334.5) 99.6 (99.299.8) 87.0 (77.792.8) 98.0 (97.498.6) 64.4 (54.973.0) 99.4 (99.099.7) 85.7 (76.291.8) 98.6 (98.099.0) 71.7 (61.879.9) 99.4 (99.499.0) 85.7 (76.291.8) 98.6 (98.099.0) 71.7 (61.879.9) 99.4 (99.499.0) 92.2 (84.096.4) 90.3 (88.991.6) 27.6 (22.533.4) 99.7 (99.399.8) 48.1 (37.359.0) 95.8 (94.999.6) 31.6 (23.940.5) 97.9 (97.198.4) 49.4 (38.560.2) 89.6 (88.290.1) 16.0 (11.921.2) 97.8 (97.098.4) 36.4 (26.547.5) 98.8 (98.299.2) 54.9 (41.467.7) 97.4 (96.798.1)
NOTE. Individual serology is listed and the practice of using both tTG and EMA together or sequentially is shown. If the serology results only are relied on (as the indication for biopsy) the last 2 columns show how many biopsies would be performed in our cohort and how many cases potentially would be undetected. The 95% condence intervals are shown in parentheses. The Only tTG positive data were reported previously by our group.19
(168.1 U/mL and 165.0 U/mL, respectively) than patients with total villous atrophy (255 U/mL) (P .05). From group 1, when considering patients with changes in the duodenal biopsy suggestive of potential celiac disease, 39 were found to have raised IELs (Marsh 1) and 2 patients had a lesion consistent with a Marsh grade 2. Patients with Marsh 1 or 2 lesions also had a signicantly lower average tTG titer (27.7 U/mL and 23.0 U/mL, respectively) than patients with villous atrophy (P .05) (Figure 2). The sensitivity of EMA increased from 79% in patients with partial villous atrophy to 100% in total villous atrophy (P .01). A similar observation was made for tTG sensitivity (using the standard cut-off level of 15), with 86.0% in Marsh 3a to 100% in Marsh 3c (P .05) (Figure 3).
ndings on histology showed that 43.8% (21 of 48) had a Marsh 0 lesion, 12.5% (6 of 48) had a Marsh 1 lesion, 10.4% (5 of 48) had a Marsh 2 lesion, 18.8% (9 of 48) had a Marsh 3a lesion, 10.4% (5 of 48) had a Marsh 3b lesion, and 4.2% (2 of 48) had a Marsh 3c lesion. Figure 4 shows the percentage of patients with a positive tTG or EMA result for each grade of villous atrophy. The percentage of patients with a positive EMA or tTG both were signicantly higher in patients with villous atrophy (EMA, 7 of 16; tTG, 7 of 16) than in those without villous atrophy (EMA, 3 of 32; tTG, 4 of 32) (EMA, P .02; tTG, P .05).
Discussion
This study performed concurrent serologic testing and a duodenal biopsy in all adult patients referred for endoscopy (n 2000). Previous investigators have used either serologic testing as a means of determining which patients should undergo a biopsy or have performed a routine duodenal biopsy without serologic testing in tandem. Our results suggest that tTG is the appropriate rst serologic test but thereafter whether a 2-step approach then is taken (with EMA) is likely to remain
Correlating Serology With Histologic Severity (Repeat Duodenal Biopsy in Patients With Celiac Disease on a Gluten-Free Diet for More Than 1 Year)
In group 2 there were 48 patients with known celiac disease (gluten-free diet for 1 y). The mean age was 52.7 years (range, 2178 y), with 68.8% (33 of 48) female patients. The
Table 2. Proposed Cost of Different Strategies Using tTG and EMA as Listed in Table 1
Resulting number of patients undergoing duodenal biopsy per 2000 245 104 92 257 Celiac disease cases missed out of a possible 77 7 (1 in 11) 10 (1 in 8) 11 (1 in 7) 6 (1 in 13)
Serologic test used to refer for biopsy Only tTG positive Only EMA positive If tTG positive and then EMA positive Either tTG positive or EMA positive
Celiac disease cases identied per 2000 70 67 66 71
Proposed total cost for 2000, $/ 53,880/28,210 38,260/20,032 40,200/21,051 81,950/42,906
Cost per celiac disease diagnosis, $/ 780/403 570/299 610/319 1150/604
NOTE. The cost of a duodenal biopsy in our hospital is converted to US dollars ($110/58), and a tTG and EMA antibody test are $13/7 each.
318
HOPPER ET AL
subject to individual clinical judgment. If a single antibody test is used in isolation tTG will detect most cases of celiac disease (tTG sensitivity, 90.9%), however, this is offset by the expense of a greater number of endoscopies with duodenal biopsy and the potential discomfort to those patients. This is reected by tTG having a PPV of only 28.6%. For this reason a 2-step approach may be cost effective (Table 2), but this will result in clinicians missing some cases of celiac disease as a result of the lower sensitivity of EMA. Because of this discordance between EMA and tTG positive results, one could test for both concurrently and opt to biopsy all patients who have any positive antibody. However, in our data set, 1 in 13 cases of undetected celiac disease still would be missed as a result of serology-negative celiac disease. On this basis we cannot make a recommendation on which test is the most cost effective or accurate for different clinical presentations; however, we still would recommend duodenal biopsy in any cases in which a clinician is suspicious of celiac disease irrespective of a negative antibody result. What is apparent from our evaluation is the poor performance of the gliadin antibodies. In this study, the gliadins gave no extra diagnostic benet when used specically to detect adult celiac disease. However, better results for gliadins have been shown in children.20 Variation may occur as a result of the numerous antigliadin assay kits. In addition, they may be a marker of extraintestinal manifestations of gluten sensitivity in the absence of enteropathy (eg, neurologic manifestations).30 Recently, with the emergence of newly developed IgA synthetic gliadin derived deamidated peptide, investigators have shown promising accuracy when used on its own or combined with tTG.31,32
Figure 3. Sensitivity of tTG (cut off 15 U/mL) and EMA antibody for different degrees of villous atrophy in newly diagnosed patients with potential and conrmed celiac disease. Sensitivity for both EMA and tTG was signicantly higher in Marsh grade 3c than in Marsh grade 3a (P .01 for EMA and P .05 for tTG).
We have reported a sensitivity for tTG of 90.9% and for EMA of 87.0%. These sensitivities are at the lower end of the spectrum when compared with previously reported studies.10 Perhaps the reasons for this are as follows: rst, our study was a prospective evaluation and other prospective studies also consistently have shown lower sensitivities for tTG and EMA.21,26
Figure 2. Graph showing patients with newly diagnosed potential and conrmed celiac disease. The tTG titer at presentation was grouped according to the Marsh grade of the duodenal biopsy. The black bars represent the average (mean) tTG values. The average titer for Marsh 3c was signicantly higher than that of 3a or 3b (P .01), and the average for Marsh grade 1 and 2 were signicantly lower than the average titers for Marsh grades 3a c. One dot can represent more than 1 patient if the tTG level was identical. tTG, IgA tissue transglutaminase antibody.
Figure 4. Percentage of positive IgA tTG (cut off 15 U/mL) and EMA results in patients with celiac disease on a gluten-free diet (1 y) grouped according to Marsh grade. The percentage of patients with a positive EMA or tTG were both signicantly higher in patients with villous atrophy (EMA, 7 of 16; tTG, 7 of 16) than in those without villous atrophy (EMA, 3 of 32; tTG, 4 of 32) (EMA, P .02; tTG, P .05).
March 2008
319
Second, our study was performed in a low (or lower)-risk group than other biopsy-conrmed studies, which reduced our ascertainment bias. Finally, unlike most previous studies all of our patients underwent a biopsy (irrespective of antibody status), therefore we believe our data truly assesses the prevalence of seronegative celiac disease. It is not surprising that by adopting this strategy this has resulted in a reduced sensitivity by comparison with previously published data. The PPV of IgA tTG in our study was 28.6% (Table 1). Although disappointing by comparison with other published studies (range, 21.8% 67%),3335 there may be several reasons for this. A high tTG titer has been shown to occur in patients with liver disease, diabetes,15,36 and in up to 40% of patients with end-stage heart failure.37 This might be reected in our cohort, which aimed to be relatively low risk for celiac disease (overall prevalence of celiac disease, 3.9%). However, our cohort might be considered an unwell group (with other comorbidities) because they were coming to the hospital for gastroscopy. There have been a few reports of a reduced sensitivity for both EMA and tTG in patients with lesser degrees of villous atrophy.17,18,38 We found similar results in our cohort of celiac disease patients. The sensitivity of tTG and EMA decreased from 100% and 100%, respectively, in patients with total villous atrophy (Marsh 3c) to 85% and 79%, respectively, in patients with partial villous atrophy (Marsh 3a) (Figure 3). However, despite our observation that the tTG titer in Marsh 3c lesions is signicantly higher than in lesser degrees of villous atrophy (Figure 2), conversely others have reported that a high tTG titer can occur in patients without villous atrophy.39 If we used a higher cut-off value than 15 U/mL (manufacturers recommendation), for example, higher than 100 U/mL, this increased the PPV in our series to 75% (with a corresponding sensitivity, specicity, and NPV of 20%, 97%, and 72.7%, respectively). We opted for the presence of a Marsh grade 3 lesion to be the gold standard for validating the serologic tests. This approach is concurrent with the majority of previous reports,16,20,26,35,38 although there have been a few studies that also incorporated Marsh 2 lesions.13,40 The concept of celiac disease or potential celiac disease without villous atrophy is well described.28 Minor mucosal changes (Marsh 1 and 2) may normalize with a glutenfree diet and improvement of symptoms. However, in our study there were only 2 patients with Marsh 2 lesions and thus this would have had a negligible effect on our tTG validation. There are other causes of raised IELs (Marsh 1) in the duodenum (apart from celiac disease) for which we did not assess.28 When trying to detect adult celiac disease using either tTG or EMA, the inclusion of Marsh 1 lesions (by expanding our diagnostic criteria for celiac disease; Figures 2 and 3) in our data set would reduce the sensitivity further. It is unclear which patients with a raised IEL require follow-up evaluation but a repeat biopsy or biopsy after a gluten-free challenge and HLA testing has been suggested.41 It would perhaps be an option to exclude these patients with a Marsh 1 (n 39) or Marsh 2 (n 2) biopsy result completely from both our celiac and control groups and classify them as having potential or unknown celiac disease: in doing so the resulting sensitivity, specicity, PPV, and NPV for tTG from our series would be 90.9%, 91.3%, 30.0%, 99.6%, respectively, and for EMA it would be 87.0%, 98.2%, 66.3%, and 99.5%, respectively. We consider that tTG offers clear advantages owing to automation (allowing higher throughput of samples) and it is a
quantitative test (using enzyme-linked immunosorbent assay). All patients in our study had a duodenal biopsy; thus, the reduced PPV is perhaps a reection of real clinical practice. In view of this observation we believe that EMA should not be regarded as an obsolete test. We did not assess IgG tTG in our patients because it was not available in our laboratory at the start of the study. Initial reports of the new IgG-tTG assays have been encouraging.21,42 Our data for the role of tTG in the assessment of histologic remission of celiac patients on a gluten-free diet are concurrent with previous reports.43 45 We would not recommend that tTG be used in isolation as a marker of histologic severity or remission. Forty-four percent (7 of 16) of celiac patients at follow-up evaluation (on a gluten-free diet for 1 y) who had villous atrophy (Marsh 3a c) also had a tTG of less than 15 U/mL. We would recommend that remission should be based on repeat duodenal biopsy, symptom response, dietary questioning, and serologic status as a composite assessment.46 In conclusion, IgA tTG alone is a sensitive marker for detecting celiac disease, but because of its poor PPV in clinical practice it has not superseded the use of IgA-EMA testing. The use of nondeamidated IgA/IgG gliadin antibodies confers no additional diagnostic benet when specically considering the detection of adult celiac disease. Finally, a normal tTG is unable to predict recovery of villous atrophy in patients with celiac disease (on a gluten-free diet for 1 y) and we would suggest that a duodenal biopsy always should be considered if patients still are symptomatic. References
1. Fasano A, Berti I, Gerarduzzi T, et al. Prevalence of celiac disease in at-risk and not-at-risk groups in the United States: a large multicenter study. Arch Intern Med 2003;163:286 292. 2. Bingley PJ, Williams AJ, Norcross AJ, et al. Undiagnosed coeliac disease at age seven: population based prospective birth cohort study. BMJ 2004;328:322323. 3. West J, Logan RF, Hill PG, et al. Seroprevalence, correlates, and characteristics of undetected coeliac disease in England. Gut 2003;52:960 965. 4. Sanders DS, Patel D, Stephenson TJ, et al. A primary care crosssectional study of undiagnosed adult coeliac disease. Eur J Gastroenterol Hepatol 2003;15:407 413. 5. Revised criteria for diagnosis of coeliac disease. Report of Working Group of European Society of Paediatric Gastroenterology and Nutrition. Arch Dis Child 1990;65:909 911. 6. United European Gastroenterology Working Group. When is a coeliac a coeliac? Report of a working group of the United European Gastroenterology Week in Amsterdam, 2001. Eur J Gastroenterol Hepatol 2001;13:11231128. 7. National Institutes of Health. Consensus Development Conference Statement on Celiac Disease, June 28-30, 2004. Gastroenterology 2005;128(Suppl 1):S1S9. 8. AGA Institute Medical Position Statement on. the Diagnosis and Management of Celiac Disease. Gastroenterology 2006;131: 19771980. 9. Lewis NR, Scott BB. Systematic review: the use of serology to exclude or diagnose coeliac disease (a comparison of the endomysial and tissue transglutaminase antibody tests). Aliment Pharmacol Ther 2006;24:4754. 10. Hill ID. What are the sensitivity and specicity of serologic tests for celiac disease? Do sensitivity and specicity vary in different populations? Gastroenterology 2005;128(Suppl 1):S25S32. 11. Rostom A, Dube C, Cranney A, et al. The diagnostic accuracy of
320
HOPPER ET AL
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27. 28. 29.
30.
31.
serologic tests for celiac disease: a systematic review. Gastroenterology 2005;128(Suppl 1):S38 S46. Hill PG, McMillan SA. Anti-tissue transglutaminase antibodies and their role in the investigation of coeliac disease. Ann Clin Biochem 2006;43:105117. Hill PG, Forsyth JM, Semeraro D, et al. IgA antibodies to human tissue transglutaminase: audit of routine practice conrms high diagnostic accuracy. Scand J Gastroenterol 2004;39:1078 1082. Volta U, De Franceschi L, Molinaro N, et al. Frequency and signicance of anti-gliadin and anti-endomysial antibodies in autoimmune hepatitis. Dig Dis Sci 1998;43:2190 2195. Clemente MG, Musu MP, Frau F, et al. Antitissue transglutaminase antibodies outside celiac disease. J Pediatr Gastroenterol Nutr 2002;34:3134. Collin P, Kaukinen K, Vogelsang H, et al. Antiendomysial and antihuman recombinant tissue transglutaminase antibodies in the diagnosis of coeliac disease: a biopsy-proven European multicentre study. Eur J Gastroenterol Hepatol 2005;17:8591. Abrams JA, Diamond B, Rotterdam H, et al. Seronegative celiac disease: increased prevalence with lesser degrees of villous atrophy. Dig Dis Sci 2004;49:546 550. Tursi A, Brandimarte G, Giorgetti GM. Prevalence of antitissue transglutaminase antibodies in different degrees of intestinal damage in celiac disease. J Clin Gastroenterol 2003;36:219 221. Hopper AD, Cross SS, Hurlstone DP, et al. Pre-endoscopy serological testing for coeliac disease: evaluation of a clinical decision tool. BMJ 2007;334:729. Tonutti E, Visentini D, Bizzaro N, et al. The role of antitissue transglutaminase assay for the diagnosis and monitoring of coeliac disease: a French-Italian multicentre study. J Clin Pathol 2003;56:389 393. Reeves GE, Squance ML, Duggan AE, et al. Diagnostic accuracy of coeliac serological tests: a prospective study. Eur J Gastroenterol Hepatol 2006;18:493501. Ciacci C, Cirillo M, Cavallaro R, et al. Long-term follow-up of celiac adults on gluten-free diet: prevalence and correlates of intestinal damage. Digestion 2002;66:178 185. Tursi A, Brandimarte G, Giorgetti GM, et al. Endoscopic and histological ndings in the duodenum of adults with celiac disease before and after changing to a gluten-free diet: a 2-year prospective study. Endoscopy 2006;38:702707. Grefte JM, Bouman JG, Grond J, et al. Slow and incomplete histological and functional recovery in adult gluten sensitive enteropathy. J Clin Pathol 1988;41:886 891. Dickey W, Hughes DF, McMillan SA. Disappearance of endomysial antibodies in treated celiac disease does not indicate histological recovery. Am J Gastroenterol 2000;95:712714. Feighery C, Weir DG, Whelan A, et al. Diagnosis of gluten-sensitive enteropathy: is exclusive reliance on histology appropriate? Eur J Gastroenterol Hepatol 1998;10:919 925. Shidrawi RG, Przemioslo R, Davies DR, et al. Pitfalls in diagnosing coeliac disease. J Clin Pathol 1994;47:693 694. Dickson BC, Streutker CJ, Chetty R. Coeliac disease: an update for pathologists. J Clin Pathol 2006;59:1008 1016. Oberhuber G, Granditsch G, Vogelsang H. The histopathology of coeliac disease: time for a standardized report scheme for pathologists. Eur J Gastroenterol Hepatol 1999;11:11851194. Hadjivassiliou M, Grunewald RA, Davies-Jones GA. Gluten sensitivity as a neurological illness. J Neurol Neurosurg Psychiatry 2002;72:560 563. Agardh D. Antibodies against synthetic deamidated gliadin pep-
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
tides and tissue transglutaminase for the identication of childhood celiac disease. Clin Gastroenterol Hepatol 2007;5:1276 1281. Sugai E, Vazquez H, Nachman F, et al. Accuracy of testing for antibodies to synthetic gliadin-related peptides in celiac disease. Clin Gastroenterol Hepatol 2006;4:11121117. Bardella MT, Trovato C, Cesana BM, et al. Serological markers for coeliac disease: is it time to change? Dig Liver Dis 2001;33: 426 431. Chan AW, Butzner JD, McKenna R, et al. Tissue transglutaminase enzyme-linked immunosorbent assay as a screening test for celiac disease in pediatric patients. Pediatrics 2001;107:E8. Lock RJ, Stevens S, Pitcher MC, et al. Is immunoglobulin A anti-tissue transglutaminase antibody a reliable serological marker of coeliac disease? Eur J Gastroenterol Hepatol 2004; 16:467 470. Vecchi M, Folli C, Donato MF, et al. High rate of positive antitissue transglutaminase antibodies in chronic liver disease. Role of liver decompensation and of the antigen source. Scand J Gastroenterol 2003;38:50 54. Peracchi M, Trovato C, Longhi M, et al. Tissue transglutaminase antibodies in patients with end-stage heart failure. Am J Gastroenterol 2002;97:2850 2854. Rostami K, Kerckhaert JP, Tiemessen R, et al. The relationship between anti-endomysium antibodies and villous atrophy in coeliac disease using both monkey and human substrate. Eur J Gastroenterol Hepatol 1999;11:439 442. Freeman HJ. Strongly positive tissue transglutaminase antibody assays without celiac disease. Can J Gastroenterol 2004;18:2528. Bazzigaluppi E, Roggero P, Parma B, et al. Antibodies to recombinant human tissue-transglutaminase in coeliac disease: diagnostic effectiveness and decline pattern after gluten-free diet. Dig Liver Dis 2006;38:98 102. Mahadeva S, Wyatt JI, Howdle PD. Is a raised intraepithelial lymphocyte count with normal duodenal villous architecture clinically relevant? J Clin Pathol 2002;55:424 428. Korponay-Szabo IR, Dahlbom I, Laurila K, et al. Elevation of IgG antibodies against tissue transglutaminase as a diagnostic tool for coeliac disease in selective IgA deciency. Gut 2003;52: 15671571. Gillett HR, Freeman HJ. Comparison of IgA endomysium antibody and IgA tissue transglutaminase antibody in celiac disease. Can J Gastroenterol 2000;14:668 671. Vahedi K, Mascart F, Mary JY, et al. Reliability of antitransglutaminase antibodies as predictors of gluten-free diet compliance in adult celiac disease. Am J Gastroenterol 2003;98:1079 1087. Kaukinen K, Sulkanen S, Maki M, et al. IgA-class transglutaminase antibodies in evaluating the efcacy of gluten-free diet in coeliac disease. Eur J Gastroenterol Hepatol 2002;14:311315. Kaukinen K, Peraaho M, Lindfors K, et al. Persistent small bowel mucosal villous atrophy without symptoms in coeliac disease. Aliment Pharmacol Ther 2007;25:12371245.
Address requests for reprints to: Dr Andrew D. Hopper, 15 Nairn Street, Shefeld, S10 1UL, United Kingdom. e-mail: andydhopper@ aol.com; fax: (44) 114-2712692. Some data were described previously as part of an alternative study reported in the British Medical Journal 2007;334:729 733. Any reference to this subset of results has been denoted throughout the article.
Symptom Severity but Not Psychopathology Predicts Visceral Hypersensitivity in Irritable Bowel Syndrome
PATRICK P. J. VAN DER VEEK,* YANDA R. VAN ROOD, and AD A. M. MASCLEE*
Departments of *Gastroenterology and Hepatology and Psychiatry, Leiden University Medical Center, Leiden, The Netherlands
Background & Aims: Visceral hypersensitivity is a hallmark of irritable bowel syndrome (IBS), but the relationship with clinical symptoms and psychological factors has not been fully established. We aimed to (1) evaluate these variables in a large cohort of IBS patients, recruited from both hospital and general practice, and in healthy controls and (2) assess which of these factors predicts the occurrence of visceral hypersensitivity in IBS. Methods: Rectal compliance and perception (intensity, perception thresholds; visual analogue scale, 0 100 mm) were assessed by a rectal barostat study (ramp distention) in 101 IBS patients and 40 healthy volunteers. IBS symptom severity was scored by using a 14-day 5-item diary. Anxiety, depression, somatization, vigilance, pain coping, dysfunctional cognitions, psychoneuroticism, and quality of life were assessed with psychometric questionnaires. Results: Rectal compliance was signicantly reduced in IBS patients compared with controls (P < .01), as were thresholds for pain (27 15 vs 35 8 mm Hg; P < .01) and urge (P < .05). Levels of anxiety, depression, neuroticism, somatization, and dysfunctional cognitions were signicantly increased in IBS patients versus controls, whereas pain coping and quality of life were signicantly worse. Hypersensitivity to rectal distention occurred in 33% of patients and was associated with increased symptom severity (P .016), but not with demographic characteristics or psychological disturbances. Conclusions: Hypersensitivity to balloon distention occurs in 33% of IBS patients and is predicted by symptom severity but not by psychological or demographic characteristics. rritable bowel syndrome (IBS) is characterized by recurrent abdominal discomfort or pain and disturbed bowel habits.1 Several pathophysiologic mechanisms have been suggested in symptom generation, including altered intestinal motility,2 autonomic dysfunction,3,4 inammation,5,6 and immune system alterations.6 8 Particularly, visceral hypersensitivity appears to play an important role9,10 and has been proposed as a biologic marker of IBS.11 Visceral hypersensitivity might result from disturbances at different levels of the brain-gut axis, in which peripheral sensitization of intestinal nerve endings,12 hyperexcitability of spinal dorsal horn neurons,13 and altered central processing of visceral afferent information14 are implicated. Abnormalities in regional brain activation, especially in areas involved in pain processing such as the anterior cingulate cortex and thalamus, have been reported in IBS patients in response to rectal balloon distention.15 These regions belong to the emotional limbic system and are involved in psychological and cognitive events.16,17
IBS symptomatology is associated with psychological factors, and these might affect clinical outcome.18 For instance, psychological distress is more prevalent among IBS patients who seek health care.19 Little is known about the relationship between psychological variables and visceral hypersensitivity. Such information is relevant because it might provide a better understanding of the pathogenesis of IBS and thereby improve its treatment. The few studies that explored this relationship have been criticized because of methodologic shortcomings such as sample size and patient selection (tertiary referrals).9,11,19 The aims of the present study were to (1) explore in a large cohort of IBS patients the prevalence of rectal hypersensitivity, levels of psychological distress, and IBS symptom severity and (2) assess which demographic, clinical, and psychological variables predict the occurrence of visceral hypersensitivity in IBS.
Methods Participants
This study was part of a large randomized controlled trial of psychological treatment in IBS, the results of which were recently published.20 IBS patients between 18 and 65 years of age were invited to participate. Baseline evaluation included detailed psychological assessment, rectal barostat measurements, and IBS symptom severity scores. To obtain a representative sample from the IBS population, patients were recruited from both the hospital IBS population (patients referred to the outpatient Department of Gastroenterology of the Leiden University Medical Center) and the general population through local advertisement. Healthy volunteers were recruited through advertisement for comparison with the patient sample. All eligible participants were screened by one of the investigators (P.vd.V). Each patient met Rome II criteria for IBS.1 Exclusion criteria were organic disease, previous abdominal surgery (except cholecystectomy and appendectomy), and pregnancy. Use of antispasmodics, laxatives, bulking agents, and occasional use of analgesics was permitted. We used the Mini International Neuropsychiatric Interview (Dutch version 5.0.0)21 to exclude patients with severe psychopathology
Abbreviations used in this paper: CSFBD, Cognitive Scale for Functional Bowel Disorders; IBS, irritable bowel syndrome; MMPI, Minnesota Multiphasic Personality Inventory; NVM, Netherlands version of MMPI; PCCL, Pain Coping and Cognition List; SAS, Somatosensory Amplication Scale; SCL-90, Symptom Checklist 90; SD, standard deviation; SF-36, Short Form 36; VAS, visual analogue scale. 2008 by the AGA Institute 1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.005
322
VAN DER VEEK ET AL
(psychosis or risk of suicide). Informed consent was obtained from each participant. The Leiden University Medical Center ethics committee had approved the study protocol.
Barostat
An electronic barostat (Synectics Visceral Stimulator; Synectics Medical, Stockholm, Sweden) was used to assess rectal compliance and perception. This device measures rectal motor activity as volume changes in a rectal balloon, in which constant pressure is maintained by injecting air when the rectal wall relaxes and aspirating air during rectal contraction. Intrabag pressure is directly measured via a separate lumen. Maximal airow is 38 mL/s. Pressure and volume are continuously monitored and recorded on a personal computer (Polygram for Windows SVS module; Synectics Medical).
Visceroperception
Perception of urge to defecate and abdominal pain during rectal distention was quantied on a 100-mm visual analogue scale (VAS). End points ranged from none to intolerable.
measure patients levels of dysfunctional cognitions concerning their IBS.25 Pain coping. Pain coping was measured by 1 of 4 subscales of the Pain Coping and Cognition List (PCCL). This inventory has been widely used in The Netherlands and awaits future validation. Patients were asked to rate the extent to which they agreed with 11 statements concerning pain coping on a 7-point scale, ranging from I completely disagree to I completely agree. Somatic symptoms. The SCL-90 was also used to record nonIBS-related somatic symptoms. There are 12 items concerning general complaints, including headache, vertigo, backache, myalgia, difculties with breathing, intolerance for high or low temperatures, dysphagia, etc. Quality of life. Quality of life was assessed by using the validated Short Form 36 (SF-36) questionnaire.26 This survey measures quality of life in 8 domains, ie, physical functioning, social functioning, role limitations due to physical problems and emotional problems, mental health, vitality, bodily pain, and general health.
Demographic Characteristics
The demographic group characteristics of interest were age, sex, and level of health care (general practice or referral).
Experimental Design
A small standardized, low caloric breakfast was permitted at 8:00 AM on the day of the barostat recordings. After arrival at our department at 10:00 AM, subjects lled out all questionnaires consecutively. Each participant was allowed the necessary time to complete the questionnaires, which took 80 90 minutes on average. After completion, the rectum was evacuated by using a tap water enema. Participants were then placed in a hospital bed, and with the subject in the left lateral position, a lubricated, tightly folded, highly compliant polyethylene bag (maximum capacity, 1000 mL) tied to the end of a multilumen tube (19F) was inserted through the anus and positioned in the rectal ampulla. Bag position was checked by manual ination of 150 mL of air and subsequent retraction of the catheter until prevented by the external anal sphincter. After balloon deation, the catheter was introduced an additional 2 cm, secured to the subjects upper leg by a piece of tape, and connected to the barostat. The hospital bed was placed in a 15-degree recumbent supine position (Trendelenburg) to avoid interference of abdominal mass with barostat measurements. Barostat measurements commenced approximately 4 hours after the light breakfast. The experimental protocol consisted of a slow ramp distention to assess rectal compliance. Intrabag pressure was increased at a rate of 1 mm Hg/min, starting at 5 mm Hg, until a maximum of 30 mm Hg. Patients rated the urge to defecate and level of abdominal pain on the 100-mm VAS scale at all even pressures (6, 8, . . . 30 mm Hg). After the experiment had ended, the rectal balloon was deated and removed, and each participant was provided with a 14-day symptom diary card and a stamped envelope to return the diary. Subjects were instructed to start lling out their symptom diary on the day after the experiment.
Symptom Severity
Patients and controls rated the severity of any abdominal discomfort, abdominal pain, constipation, diarrhea, and bloating daily for 14 days on a 5-point Likert scale (0, no symptoms; 1, mild; 2, moderate; 3, severe; 4, very severe symptoms) by using a symptom diary card. A composite score was computed by summing up the 14-day mean scores for each symptom (range, 0 20).
Psychological Assessment
A battery of questionnaires was administered to both IBS patients and control subjects to determine the following psychological characteristics of each group. Anxiety and depression. We used the Symptom Checklist 90 (SCL-90) to measure levels of anxiety (10 items) and depression (16 items). The SCL-90 is a validated survey and consists of 90 items addressing a range of physical and psychological problems.22 Psychoneuroticism. The level of psychoneuroticism was determined by summing up all 90 items of the SCL-90. Somatization. We used the abridged Dutch version (NVM) of the Minnesota Multiphasic Personality Inventory (MMPI) to measure somatization, which is 1 of 5 subscales on this questionnaire.23 The role of the abovementioned psychological factors in IBS has been studied previously.9,10,19 In addition, we considered the following psychological variables relevant, because they might confound the abovementioned determinants. Vigilance. We used the previously validated 10-item Somatosensory Amplication Scale (SAS)24 to determine the extent to which an individual is likely to report enhanced perception of physical symptoms (ie, lower cognitive perception thresholds). Cognitions. The recently developed 31-item Cognitive Scale for Functional Bowel Disorders (CSFBD) was used to
Barostat Analysis
Dynamic compliance was assessed by calculating volume increments for each individual pressure step in each study participant. Compliance was dened by the largest volume increment (ie, the steepest slope of the pressure-volume curve) for each participant and averaged over groups. Perception
March 2008
PREDICTORS OF VISCERAL HYPERSENSITIVITY IN IBS
323
scores were expressed as the mean score at each pressure step. Perception thresholds were dened as the rst pressure level at which perception scores exceeded 10 mm.
Visceral Hypersensitivity
Patients with a pain perception threshold 2 standard deviations (SDs) below the mean threshold in controls were considered to be hypersensitive to balloon distention.
We aimed to enroll at least 40 subjects in each group to be able to detect a 5-mm Hg difference in mean pain threshold, which we considered clinically relevant, with a power of 0.80 and SD of 8 mm Hg on the basis of previous studies by our group. All statistical analyses were carried out with SPSS for Windows, version 11.0.1 (SPSS Inc, Chicago, IL). Demographic characteristics were compared between groups by Student t test, Mann-Whitney, or 2 analysis as appropriate. Differences in rectal compliance and visceroperception were analyzed for statistical signicance by using mixed models, with patient numbers as indicator for repeated measurements. One model analyzed pressure, volume, and pressure by volume interaction as separate contributors to the model; a second model did the same for pressure, visceral perception, and pressure by perception interaction. Compliance, perception of urge and pain at maximum rectal pressure (30 mm Hg), and perception thresh-
olds for urge were compared by MannWhitney (patients versus controls) or KruskalWallis analysis (IBS subgroups). Because the pain threshold during ramp distention was not reached in all participants, the best estimates for the mean pain threshold and SD were obtained by maximum likelihood estimation by using software for parametric survival models. Normal distribution for the pain scores was assumed. These estimates were compared by log-rank analysis. Finally, binary logistic regression and backward stepwise analysis (method, likelihood ratio; entry at 0.05 probability, removal at 0.10 probability) was performed to identify demographic, clinical (symptom severity), and psychological characteristics that predict the occurrence of visceral hypersensitivity. Age, gender, health care level, predominant bowel habit, postinfectious symptom onset, rectal compliance, symptom severity, anxiety, depression, somatic symptoms, psychoneuroticism, dysfunctional cognitions regarding functional bowel disorders, vigilance, pain coping, somatization, and quality of life (general health subscale) were entered in the analysis as separate predictors. Data were expressed as mean SD. The level of signicance was set at P .05.
Results Subject Characteristics

We screened 130 patients, 26 of whom did not meet Rome II criteria, and 40 healthy volunteers. Two patients de-
Table 1. Baseline Demographic, Clinical, and Psychological Characteristics of IBS Patients and Healthy Controls
Characteristic Demographics Age (y) Female sex (%) Bowel habit (%) Diarrhea Constipation Alternating Not specied Normal (controls) Symptoms IBS symptom score (020) Psychological prole Anxiety (1050) Depression (1680) Somatic symptoms (1260) Psychoneuroticism (90450) Dysfunctional cognitions (31217) Vigilance (040) Pain coping (61) Somatization (02) Quality of life (0100) Physical functioning Role limitationsphysical Bodily pain Mental health Role limitationsemotional Social functioning Vitality General health
aP
IBS patients (N 101) 42.0 13.9 73 34 35 24 8 0 4.4 2.5a 13.4 4.6b 22.5 6.9a 18.3 5.6a 123.8 31.9a 110.3 35.8a 9.7 5.8 3.4 1.0b 0.6 0.4a 82.0 20.4a 60.0 42.0a 62.1 19.6a 75.2 16.3 80.8 35.3 73.2 23.7a 58.5 16.9a 61.2 18.8a
Healthy controls (N 40) 39.7 15.0 63 0 0 0 0 100 0.43 0.57 12.2 3.7 20.7 8.3 15.0 3.7 113.3 30.7 85.7 37.3 7.7 4.7 3.7 0.8 0.3 0.3 94.1 10.5 87.2 28.6 90.3 16.1 78.5 13.4 91.0 26.8 90.9 14.3 70.8 15.8 75.1 14.6
NOTE. Score ranges from best to worst are indicated after each parameter. Data are expressed as mean SD. .01 vs healthy controls. bP .05 vs healthy controls.
324
VAN DER VEEK ET AL
Table 2. Psychological Prole of Patients Recruited From the Tertiary Referral Center and From the General Population
Psychological prole Anxiety (1050) Depression (1680) Somatic symptoms (1260) Psychoneuroticism (90450) Dysfunctional cognitions (31217) Vigilance (040) Pain coping (61) Somatization (02) Referral center (N 31) 12.9 3.4 22.2 5.1 18.6 4.0 122.1 22.5 109.6 34.2 8.2 4.2 3.5 1.0 0.7 0.3 General population (N 70) 13.7 5.0 22.7 7.6 18.2 6.3 124.6 35.7 110.6 36.7 10.4 6.3 3.4 1.0 0.6 0.4
NOTE. Score ranges from best to worst are indicated after each parameter. Data are expressed as mean SD.
clined to participate in the barostat study, and 1 patient was diagnosed with conversion disorder. All healthy volunteers and 101 patients provided informed consent and were included in the nal analysis. Thirty-one patients (31%) were recruited through the outpatient department, and 70 patients (69%) were
recruited through advertisement. All patients in the latter group had previously consulted a physician and had been evaluated for their abdominal symptoms. Healthy controls were also recruited through advertisement. Demographic, clinical, and psychological characteristics of patients and controls are listed in Table 1. Mean age and male to female ratio were not different between groups. Symptom severity and levels of anxiety, depression, psychoneuroticism, somatization, other somatic symptoms, and dysfunctional cognitions were all slightly but signicantly increased in IBS patients compared with healthy controls. Pain coping scores were signicantly reduced in IBS. Compared with controls, patients had impaired quality of life on 6 of 8 SF-36 subscales. Psychological measures were not different between patients from the tertiary referral center and those from the general population (Table 2).
Rectal Compliance and Perception

Rectal compliance was signicantly reduced in the IBS group compared with healthy control subjects (29.7 13 vs 41.8 18 mL/mm Hg, P .0001) (Figure 1A). Subgroup analysis showed that rectal compliance was particularly reduced in patients with a diarrhea-predominant bowel habit (IBS-D; P .04) and those with alternating bowel habit (IBS-A; P .05) compared with constipation-predominant IBS (IBS-C) (Figure 1B).
Figure 1. (A) Dynamic rectal compliance (mL/mm Hg) in 101 IBS patients (squares) and 40 controls (triangles). Compliance was signicantly decreased in patients compared with controls. Data are expressed as mean standard error of the mean. (B) Dynamic rectal compliance (mL/mm Hg) in IBS-D, IBS-C, and IBS-A patients and 40 controls. Compliance was signicantly increased in IBS-C compared with IBS-D and IBS-A, but similar compared with controls. Data are expressed as mean standard error of the mean.
Figure 2. Intensity of urge perception in 101 IBS patients (squares) and 40 controls (triangles). Urge did not differ between patients and controls. Data are expressed as mean standard error of the mean.
March 2008
325
Figure 3. Intensity of pain perception in 101 IBS patients (squares) and 40 controls (triangles). Pain perception was signicantly increased in patients compared with controls (pressure by group interaction, P .0001). Data are expressed as mean standard error of the mean.
Figure 4. Individual pain thresholds in IBS patients and healthy controls. Signicantly more patients (N 55, 54%) compared with controls (N 10, 25%) reached the pain threshold before the end of the ramp distention (dotted line, 30 mm Hg).
Urge perception at high rectal pressure distention (30 mm Hg) was not signicantly different between IBS patients (6.6 2.7 cm) and controls (6.1 2.6 cm) (P .30). The pressure-urge curves were also not signicantly different between patients and controls (pressure by group interaction, P .82; Figure 2). In contrast, pain perception at high rectal pressure was significantly increased in IBS patients compared with controls (2.5 2.7 vs 1.0 1.4 cm, P .003), and the pressure-pain curves differed signicantly between groups (pressure by group interaction, P .0001; Figure 3). No differences between IBS subgroups were found (Table 3).
was signicantly reduced in IBS patients (27.5 15 mm Hg) compared with controls (35.3 8.2 mm Hg) (P .0009) but did not differ between IBS subgroups (Table 3).
Visceral Hypersensitivity
The threshold for hypersensitivity to balloon distention was set at 18.9 mm Hg (35.3 minus 16.4 [ie, 2 8.2 SD] mm Hg). Thirty-three IBS patients (33%), compared with 0 controls, were identied as hypersensitive to balloon distention (2 17.06, P .0001) (Table 4). Thus, pain thresholds fell outside the range of control subjects in approximately 1 in 3 IBS patients.
Perception Thresholds
Urge thresholds were reached in all participants, but they were somewhat reduced in IBS patients (15.6 6.1 mm Hg) compared with controls (18 6.0 mm Hg) (P .042). No differences were found between IBS subgroups (Table 3). In contrast, only 10 of 40 control subjects (25%), compared with 55 of 101 IBS patients (54%), reached the threshold for rectal pain during balloon distention (2 10.01, P .002) (Figure 4). Maximum likelihood estimation of the mean pain threshold and SD in each group and subsequent log-rank analysis showed that the threshold
Predictors of Visceral Hypersensitivity

Of all tested variables, only IBS symptom severity remained as a predictor of visceral hypersensitivity in the logistic regression analysis (odds ratio, 1.25; 95% condence interval, 1.04 1.50; P .016). Table 5 lists demographic, clinical, and psychological characteristics in hypersensitive and normosensitive patients. IBS symptom scores were signicantly higher in
Table 3. Rectal Compliance and Perception in IBS Patients, IBS Subgroups, and Healthy Controls
IBS patients IBS-D (N 34) 27.2 11 6.5 2.9 2.2 2.6 16.9 6.6 31.3 18 IBS-C (N 35) 35.2 14b 6.7 2.6 2.9 2.7 14.2 5.6 23.6 13 IBS-A (N 24) 26.6 11 6.7 2.9 2.6 3.0 15.6 6.1 29.6 15 All patients (N 101) 29.7 13a 6.6 2.7 2.5 2.7a 15.6 6.1c 27.5 15a Controls (N 40) 41.8 18 6.1 2.6 1.0 1.4 18.0 6.0 35.3 8.2
Measure Compliance (mL/mm Hg) Urge at 30 mm Hg (cm) Pain at 30 mm Hg (cm) Threshold urge (mm Hg) Threshold pain (mm Hg)
NOTE. Data for the group with unknown bowel habit are not shown because of the small number of patients (N 8). Data are expressed as mean SD. aP .01 compared with controls. bP .05 compared with IBS-D and IBS-A. cP .05 compared with controls.
326
VAN DER VEEK ET AL
Table 4. Visceral Hypersensitivity in IBS Patients and Healthy Controls

Hypersensitive IBS (N 101) Controls (N 40)
aP
Normosensitive 68 (67%) 40 (100%)
33 (33%)a 0 (0%)
.001 compared with controls.
hypersensitive compared with normosensitive patients (5.4 2.5 vs 4.0 2.4, P .007). No other differences were found.
Discussion
The present study showed that (1) visceral hypersensitivity is an important feature of IBS but is not present in all patients, and (2) hypersensitivity to rectal balloon distention is predicted by IBS symptom severity but not by demographic or psychological characteristics. Our data conrmed previous ndings that rectal compliance and pain thresholds are reduced and that the intensity of pain perception is increased in IBS patients when compared with healthy controls. Urge intensity at any given pressure was similar in patients and controls, with slightly lower thresholds for urge in IBS patients. Our observation that pain perception rather than urge is increased is consistent with other reports demonstrating decreased perception thresholds in IBS only for noxious stimuli and not for stool.11 It is presumed that a phasic distention protocol (ie, rapid balloon ination to predened pressure levels) is the preferred procedure to test visceral hypersensitivity, because this would elicit rectal sensations at lower volumes or pressures compared with slow ramp distention.27,28 However, we chose to perform only ramp distentions because we considered rectal compliance to be an important factor in the model on predictors of visceral hypersensitivity, and compliance is best measured by means of slow ramp distention.28 Phasic distentions were not performed because assessment of sensory thresholds during phasic distentions after preceding ramp distention might introduce perceptual response bias, and phasic distentions before ramp distention might affect subsequent rectal compliance measurements. The pain thresholds we observed during ramp distention are similar to those reported by others using phasic distentions,10,29,30 which supports previous ndings that the type of distention procedure (phasic, ramp, etc) does not affect perception.31 One of our main ndings is that hypersensitivity to balloon distention was less likely to occur in patients with milder symptoms. This challenges the view that visceral hyperalgesia is a biologic marker of IBS,11 because hypersensitivity might be absent in Rome IIpositive patients with mild symptoms. The difference in the proportion of hypersensitive patients in the study by Mertz et al11 (95%) and our study (33%) might in part be due to the use of different parameters to dene visceral hypersensitivity. Mertz et al11 used 3 parameters to score rectal perception simultaneously (ie, perception thresholds, intensity of sensations, and altered viscerosomatic referral), whereas we only identied patients having decreased pain thresholds and not those having decreased discomfort thresholds or altered pain referral patterns. It is, of course, essential to use equal denitions of visceral hypersensitivity when comparing its prevalence between studies. Because no accepted denition of vis-
ceral hypersensitivity is currently available, we decided to use a statistical point of view and consider patients with a pain perception threshold 2 SDs below the mean threshold in healthy controls as hypersensitive to rectal balloon distention. In general, this method is accepted to dene outliers. Although this cutoff is arbitrary, our data suggested that hypersensitivity to rectal distention is not a suitable biologic marker to identify patients with IBS. The pathophysiology of visceral hyperalgesia in IBS remains poorly understood. Recent evidence suggested that disturbances might occur at different levels of the brain-gut axis. First, sensitization of peripheral nerve endings at the intestinal level might occur during or after acute inammation,12,13 leading to higher excitability and/or increased ring of these neurons. Second, some studies suggested that alterations in the spinal dorsal horn neurons might provide an explanation for the extended viscerosomatic referral pattern that is often seen in IBS.11,12 Third, altered processing of afferent visceral information in the brain, particularly in the prefrontal cortex, anterior cingulated cortex, and thalamus, has repeatedly been demonstrated in IBS patients.15,32 These regions are not only involved in pain processing but are also part of the emotional limbic system and are therefore involved in numerous psychological and cognitive events.16,17 Because nociception (becoming
Table 5. Demographic, Clinical, and Psychological Characteristics of Hypersensitive and Normosensitive IBS Patients
Characteristic Age (y) Female sex (%) Recruitment (%) advertisement Bowel habit (%) Diarrhea Constipation Alternating Not specied/normal Postinfectious (%) Dynamic compliance IBS composite score Discomfort Pain Constipation Diarrhea Bloating General health Anxiety Depression Somatic symptoms Psychoneuroticism Dysfunctional cognitions Vigilance Pain coping Somatization Antispasmodics (%) Laxatives or bulking agents (%) Hypersensitive (N 33) 40.7 12.4 73 68 33 46 18 3 11 31.8 14.9 5.4 2.5a 1.38 0.8b 1.34 0.95c 0.73 0.64c 0.45 0.86 1.37 0.79c 62.7 16.4 13.9 5.0 23.1 6.5 19.0 4.5 126.5 32.2 106.8 35.3 9.2 5.3 3.5 1.1 0.6 0.4 15 30 Normosensitive (N 68) 42.6 14.5 74 71 34 29 27 10 13 28.6 11.3 4.0 2.4 1.17 0.62 0.98 0.72 0.37 0.56 0.48 0.69 1.01 0.75 60.5 19.9 13.2 4.4 22.3 7.1 18.0 6.1 122.5 32.0 111.9 36.1 9.9 6.1 3.3 0.9 0.6 0.4 12 31
aP .007 vs normosensitive patients (range, 0 [no symptoms] to 20 [worst imaginable]). bP .072 vs normosensitive patients. cP .02 vs normosensitive patients.
March 2008
327
aware of a painful stimulus) and emotional pain management both occur in similar regions of the brain, we hypothesized that psychological disturbances are related to visceral hypersensitivity. However, our results did not support this hypothesis, because none of the psychological variables we studied predicted the occurrence of hypersensitivity to balloon distention. These ndings substantiated previous observations that psychological characteristics such as anxiety, somatization, and neuroticism do not correlate with sensory thresholds.9,11,19 Similar results were obtained in a recent study in which multivariate analysis demonstrated that abdominal pain and bloating were signicantly associated with altered rectal perception, whereas psychological symptoms such as anxiety were not.33 Our data also showed that rectal hyperalgesia is not associated with other psychological factors (vigilance, dysfunctional cognitions, pain coping), demographic characteristics (age, gender), quality of life, or predominant bowel habit. Previously Whitehead et al34 proposed a model for psychological factors that inuence pain perception in IBS. It was suggested that low pain thresholds in IBS are inuenced by 2 related cognitive traits, ie, selective attention to gut sensations and a tendency to interpret these sensations as symptoms of disease. Our data showed that neither vigilance (selective somatic attention) nor cognitions regarding functional bowel disorders (interpretation of normal sensations as symptoms of disease) were different between hypersensitive and normosensitive IBS patients. These ndings suggested that hypersensitive patients do not perceive or manage their symptoms differently from normosensitive patients. Although vigilance and cognitions on functional bowel disorders differed signicantly between patients and controls, these parameters were not associated with increased rectal sensitivity. We aimed to obtain a representative sample from the IBS population by recruiting patients both from the outpatient clinic and by advertisement. Levels of psychological distress were low and did not differ signicantly between groups. One might argue that low levels of psychopathology explained why we found no correlation between psychological variables and visceral hypersensitivity, because a certain degree of parameter variability was required for correlations to be detected. Although some studies found signicantly more psychological disturbances in IBS patients recruited from tertiary care,18,19,35 one of these studies found no relation between psychological distress and visceral hypersensitivity in clinic patients with IBS,19 supporting our nding that visceral hypersensitivity was not affected by psychopathology, regardless of level of health care. Allowing patients to take antispasmodics, laxatives, and, occasionally, analgesics during barostat measurements is a limitation of this study because it might interfere with visceral sensitivity and affect sensory thresholds in general. Although use of these medications was similar in hypersensitive and normosensitive patients, prohibiting the use of these medications might have further increased the number of patients with hypersensitivity to balloon distention in both groups. In conclusion, we found that patients with IBS had impaired rectal compliance and reduced sensory thresholds to rectal distention compared with controls. Visceral hypersensitivity was present in one third of our IBS population and was associated with increased symptom severity. Although psychological parameters did not predict the occurrence of visceral hypersen-
sitivity, this does not exclude a common neuropsychological basis in the pathophysiology of IBS. Future studies should focus on the role of the brain-gut axis in the development of IBS. References
1. Thompson WG, Longstreth GF, Drossman DA, et al. Functional bowel disorders and functional abdominal pain. Gut 1999; 45(Suppl 2):II43II47. 2. Chey WY, Jin HO, Lee MH, et al. Colonic motility abnormality in patients with irritable bowel syndrome exhibiting abdominal pain and diarrhea. Am J Gastroenterol 2001;96:1499 1506. 3. Aggarwal A, Cutts TF, Abell TL, et al. Predominant symptoms in irritable bowel syndrome correlate with specic autonomic nervous system abnormalities. Gastroenterology 1994;106:945 950. 4. Van der Veek PP, Swenne CA, Van de Vooren CA, et al. Viscerosensory-cardiovascular reexes: altered baroex sensitity in irritable bowel syndrome. Am J Physiol 2005;289:R970 976. 5. Rodriguez LA, Ruigomez A. Increased risk of irritable bowel syndrome after bacterial gastroenteritis: cohort study. BMJ 1999; 318:565566. 6. Gwee KA, Collins SM, Read NW, et al. Increased rectal mucosal expression of interleukin 1beta in recently acquired post-infectious irritable bowel syndrome. Gut 2003;52:523526. 7. Gonsalkorale WM, Perrey C, Pravica V, et al. Interleukin 10 genotypes in irritable bowel syndrome: evidence for an inammatory component? Gut 2003;52:9193. 8. Van der Veek PP, van den Berg M, Kroon YE, et al. Role of tumor necrosis factor-alpha and interleukin-10 gene polymorphisms in irritable bowel syndome. Am J Gastroenterol 2005;40:2510 2516. 9. Whitehead WE, Holtkotter B, Enck P, et al. Tolerance for rectosigmoid distention in irritable bowel syndrome. Gastroenterology 1990;98:11871192. 10. Bouin M, Plourde V, Boivin M, et al. Rectal distention testing in patients with irritable bowel syndrome: sensitivity, specicity, and predictive values of pain sensory thresholds. Gastroenterology 2002;122:17711777. 11. Mertz H, Naliboff B, Munakata J, et al. Altered rectal perception is a biological marker of patients with irritable bowel syndrome. Gastroenterology 1995;109:40 52. 12. Mayer EA, Gebhart GF. Basic and clinical aspects of visceral hyperalgesia. Gastroenterology 1994;107:271293. 13. Olivar T, Cervero F, Laird JM. Responses of rat spinal neurones to natural and electrical stimulation of colonic afferents: effect of inammation. Brain Res 2000;866:168 177. 14. Verne GN, Himes NC, Robinson ME, et al. Central representation of visceral and cutaneous hypersensitivity in the irritable bowel syndrome. Pain 2003;103:99 110. 15. Silverman DH, Munakata JA, Ennes H, et al. Regional cerebral activity in normal and pathological perception of visceral pain. Gastroenterology 1997;112:64 72. 16. Bishop S, Duncan J, Brett M, et al. Prefrontal cortical function and anxiety: controlling attention to threat-related stimuli. Nat Neurosci 2004;7:184 188. 17. Bush G, Luu P, Posner MI. Cognitive and emotional inuences in anterior cingulate cortex. Trends Cogn Sci 2000;4:215222. 18. Drossman DA, McKee DC, Sandler RS, et al. Psychosocial factors in the irritable bowel syndrome: a multivariate study of patients and nonpatients with irritable bowel syndrome. Gastroenterology 1988;95:701708. 19. Guthrie E, Creed F, Fernandez L, et al. Cluster analysis of symptoms and health seeking behaviour differentiates subgroups of patients with severe irritable bowel syndrome. Gut 2003;52: 1616 1622.
328
VAN DER VEEK ET AL
20. van der Veek PP, van Rood YR, Masclee AA. Clinical trial: shortand long-term benet of relaxation training for irritable bowel syndrome. Aliment Pharmacol Ther 2007;26(6):943952. 21. Sheehan DV, Lecrubier Y, Sheehan KH, et al. The Mini-International Neuropsychiatric Interview (M.I.N.I.): the development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. J Clin Psychiatry 1998;59(Suppl 20):2233. 22. Derogatis LR, Rickels K, Rock AF. The SCL-90 and the MMPI: a step in the validation of a new self-report scale. Br J Psychiatry 1976;128:280 289. 23. Dahlstrom GW, Welsh GS, Dahlstrom LE. An MMPI handbook: volume I clinical interpretation. Minneapolis: University of Minnesota Free Press, 1972. 24. Speckens AE, Spinhoven P, Sloekers PP, et al. A validation study of the Whitely Index, the Illness Attitude Scales, and the Somatosensory Amplication Scale in general medical and general practice patients. J Psychosom Res 1996;40:95104. 25. Toner BB, Stuckless N, Ali A, et al. The development of a cognitive scale for functional bowel disorders. Psychosom Med 1998; 60:492 497. 26. Brazier JE, Harper R, Jones NM, et al. Validating the SF-36 health survey questionnaire: new outcome measure for primary care. BMJ 1992;305:160 164. 27. Sun WM, Read NW, Prior A, et al. Sensory and motor responses to rectal distention vary according to rate and pattern of balloon ination. Gastroenterology 1990;99:1008 1015. 28. Whitehead WE, Delvaux M. Standardization of barostat procedures for testing smooth muscle tone and sensory thresholds in the gastrointestinal tract: the Working Team of Glaxo-Wellcome Research, UK. Dig Dis Sci 1997;42:223241. 29. Chang L, Munakata J, Mayer EA, et al. Perceptual responses in patients with inammatory bowel disease. Gut 2000;47:497 505.
30. Naliboff BD, Munakata J, Fullerton S, et al. Evidence for two distinct perceptual alterations in irritable bowel syndrome. Gut 1997;41:505512. 31. Hammer HF, Phillips SF, Camilleri M, et al. Rectal tone, distensibility, and perception: reproducibility and response to different distensions. Am J Physiol 1998;274:G584 G590. 32. Ringel Y, Drossman DA, Turkington TG, et al. Regional brain activation in response to rectal distension in patients with irritable bowel syndrome and the effect of a history of abuse. Dig Dis Sci 2003;48:1774 1781. 33. Posserud I, Syrous A, Lindstrm L, et al. Altered rectal perception in irritable bowel syndrome is associated with symptom severity. Gastroenterology 2007, doi:10.1053/j.gastro.2007.07.024. 34. Whitehead WE, Palsson OS. Is rectal pain sensitivity a biological marker for irritable bowel syndrome: psychological inuences on pain perception. Gastroenterology 1998;115:12631271. 35. Longstreth GF, Hawkey CJ, Mayer EA, et al. Characteristics of patients with irritable bowel syndrome recruited from three sources: implications for clinical trials. Aliment Pharmacol Ther 2001;15:959 964.
Address requests for reprints to: Dr P. P. J. van der Veek, Department of Gastroenterology and Hepatology, Leiden University Medical Center, Building 1, C4-P, PO Box 9600, 2300 RC, Leiden, The Netherlands. e-mail: P.P.J.van_der_Veek@lumc.nl; fax: 31-71-5248115. This study was supported by a grant from the Dutch Digestive Diseases Foundation (NVGE). We thank Saskia le Cessie of the Department of Medical Statistics for statistical advice and our colleagues at the Department of Gastroenterology and Hepatology for assistance in the barostat measurements.
Chronic Abdominal Pain and Depressive Symptoms: Analysis of the National Longitudinal Study of Adolescent Health
NADER N. YOUSSEF,* KATHERINE ATIENZA,* ANNETTE L. LANGSEDER,* and RICHARD S. STRAUSS
*Center for Pediatric Irritable Bowel & Motility Disorders, Goryeb Childrens Hospital at Atlantic Health, Morristown, New Jersey; and Johnson and Johnson, Product Development Unit, Titusville, New Jersey
Background & Aims: Abdominal pain is common in

adolescence. The aim of this study was to determine the prevalence of depressive symptoms in a large cohort of patients with frequent abdominal pain. Methods: A prospective, cross-sectional, nationally representative sample of children aged 13 to 18 years (mean age, 16.2 1.7 y; 49% male) completed in-home interviews and separate in-school questionnaires for the National Longitudinal Study in Adolescent Health (the Add Health Study). Depressed mood was assessed with the Center for Epidemiologic Studies Depression Scale. Subjective measures of abdominal pain were reported by 20,745 adolescents from wave 1 of the Add Health Study. Frequency of abdominal pain over the previous 1 year was rated as rare (0 1 episode/wk), moderate (23 episodes/wk), or daily (>4 episodes/wk). Results: Daily pain is reported in 3.2% of adolescents, with an additional 14% reporting pain as moderate in frequency. Sixteen percent of all adolescents are at risk for developing depression. The risk for depression goes from 16% to 45% (P < .001) when the pain is daily. Compared with rare pain, children with daily pain were more likely to miss school 10 or more times per year (46% vs 19%, P < .001), cry (12.1% vs 1%, P < .001), feel sad (25.2% vs 5.3%, P < .001), and lonely (25.2% vs 6.4%, P < .001). Children with daily pain were likely to consider life a failure versus those with no pain (10.2% vs 3.3%, P < .001). Conclusions: Adolescents with frequent abdominal pain are at increased risk for depressive symptoms, social isolation, and missing school. requent abdominal pain is a common complaint among adolescents in the United States, with 17% of high school students in small community, school-based samples experiencing abdominal pain once a week.1 The exact prevalence of abdominal pain is unknown but does increase with age through adolescence and affects more females. Frequent abdominal pain in children and adolescents has been associated with social withdrawal, school absenteeism, and a poor quality of life for patients and their parents.2 Other associated restrictions include decreased appetite, not playing with friends, and less time with hobbies.3 Suffering from frequent abdominal pain earlier in life often leads to increased health care use as adults and also is more likely to result in an anxiety disorder later in life.4 Psychiatric morbidities such as anxiety and depression are common in adults who suffer from frequent abdominal pain. Recent literature has examined the relationship of suicidal ideation and depression in adults who suffer from both organic and nonorganic abdominal pain. Adult patients with chronic pain have higher rates of suicidal ideation.5 A recent study by
Smith et al5 suggested that abdominal pain in adults may be an independent risk factor for suicidal ideation. Moreover, suicidal ideation increases with symptom severity and perceived interference with life in adults with irritable bowel syndrome.6 The lifetime prevalence of depression through adolescence is 20%.7 One study noted the rate of weekly depressive symptoms in 15-year-olds to be 49% in females and 34% in males.7 The aim of this study was to examine on a national level the association between frequent abdominal pain in adolescents and depressive symptoms.
Methods
A prospective, cross-sectional, nationally representative sample of adolescents in the United States aged 13 to 18 years (mean age, 16.2 1.7 y; 49% male) completed both in-house interviews and at-school questionnaires for the National Longitudinal Study in Adolescent Health (the Add Health Study). Initiated in 1994 under a grant from the National Institute of Child Health and Human Development with co-funding from 17 other federal agencies, the Add Health Study is the largest, most comprehensive survey of adolescents ever undertaken.8 Data at the individual, family, school, and community levels were collected in 2 waves between 1994 and 1996. In addition to baseline demographic data, measures of health status, healthrelated behaviors, and social function including school activities, friends, and home activities were collected. A parental respondent gave information on household income and parental education. The sample population of adolescents was stratied according to region, school type, ethnicity, and urban areas. The current study used data from the in-home interview sample of 20,745 adolescents (wave 1 in AprilDecember 1995). Additional information on the Add Health Study design, sampling strategy, and methodology are available at www.cpc.unc.edu/addhealth. Depressed mood was assessed with the Center for Epidemiologic Studies Depression Scale (CES-D). The CES-D initially was designed to screen adults for symptoms of depression during the previous week. It is not a psychiatric diagnostic tool for clinical depression but identies individuals at risk for a clinical diagnosis of depression. The Add Health Study contains questions about depressive symptoms that correspond to those
Abbreviations used in this paper: Add Health, National Longitudinal Study in Adolescent Health; CES-D, Center for Epidemiologic Studies Depression Scale. 2008 by the AGA Institute 1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.019
330
YOUSSEF ET AL
found in the 20-question CES-D. Depressive symptoms measured included feeling sad, feel like crying, feeling lonely, low energy, and life as failure. The modied the Add Health Study CES-D scale is valid for use by adolescents.9 A cut-off score of 16 has been used to dene cases of depression in adults, but estimates the prevalence of depression in adolescents to be more than 50%. Roberts et al10 determined that scores of 24 in females and 22 in males maximized the sensitivity and specicity of the CES-D for predicting major depressive disorder among adolescents. We thus dened risk of depression in this study as a CES-D score of 23 or greater in adolescents, as other investigators have.10 The Add Health Study in-home questionnaire asked students how often they had a stomachache or upset stomach in the past 12 months. Respondents did not have the opportunity to characterize or quantify the duration of abdominal pain complaints during the day. Frequency of abdominal pain over the past year was rated as rare if there were 0 to 1 episodes per week, occasional if there were 2 to 3 episodes per week, and daily if there were daily or almost-daily episodes of pain per week.
Data Analysis
Variables in our analysis were selected on the basis of factors previously described as relevant to depressive disorders and adolescent mental health. These items included sociodemographics, health (subjective self-rating of abdominal pain), mental health issues (sadness, life is a failure), and school absenteeism. Data were analyzed with SPSS version 12.0 (Chicago, IL) statistical software. Bivariate analyses of outcomes and sociodemographic factors were conducted with chi-squared analysis. We analyzed potential predictive variables with multivariate logistic regression for the outcome of abdominal pain and persistent moderate/severe depressive symptoms (CES-D, 23) among adolescents. Logistic regression models were constructed in a forward fashion, adding variables bivariate analyses and stepwise modeling. Signicant variables (P .05) were retained in the model, and additional demographic variables that were signicant such as school absenteeism and participation in social activities were retained in the nal model.
Figure 1. Effect of abdominal pain on school attendance. Adolescents who suffered from daily abdominal pain reported an increased incidence of missing school more than 10 days per year as compared with those with minimal or no abdominal pain (46% vs 19%, *P .001).
Effect of Abdominal Pain on Social Activities

Adolescents who suffered from daily abdominal pain reported less participation in social activities as compared with those with minimal or no abdominal pain (0.6% vs 1.9%, P .04) (Figure 2). Social activities included hobbies, sports, and hanging out with friends. Children with daily abdominal pain were less likely to participate in active sports (19% vs 30%, P .001) or to participate in school activities (P .002).
Associated Depressive Symptoms

Adolescents with daily abdominal pain were more likely to feel tired, cry, feel sad, and feel lonely as compared with those with no abdominal pain (P .001). Adolescents with daily abdominal pain were more likely to consider their lives a failure versus those with no abdominal pain (10.7% vs 3.3%, P .001) (Table 1). Children with daily pain were more likely to have changed their primary residence in the prior 5 years (P .05). Frequency of abdominal pain or risk of depression was not related to family socioeconomic status.
Results
Daily pain was reported in 3.2% of the study population with an additional 14% reporting abdominal pain as moderate in frequency. In this representative population, 16% of adolescents had signicant depressive symptoms as dened by a CES-D score greater than 23. Adolescents with daily abdominal pain were signicantly more likely to have increased CES-D scores of 23 or greater compared with those with rare abdominal pain (45% vs 3.2%, P .001). Of those at risk for depression with daily abdominal pain, females outnumbered males 2:1.
Discussion
Utilizing the National Longitudinal Study of Adolescent Health, a prospective US population based registry, the relationship between frequent abdominal pain and depressive symptoms was examined. Adolescents who experience abdominal pain more than 4 times a week have an increased prevalence of school absenteeism (10 days per year), fatigue, crying, and feelings of sadness and loneliness as compared with healthy adolescents who had abdominal pain less than once a week. Also, adolescents with frequent abdominal pain were more
Effect of Abdominal Pain on School Attendance

Adolescents who suffered from daily abdominal pain reported an increased prevalence of missing school 10 or more days per year as compared with those with minimal or no abdominal pain (46% vs 19%, P .001) (Figure 1).
March 2008
ADOLESCENT ABDOMINAL PAIN AND DEPRESSION
331
Figure 2. Effect of abdominal pain on social activities. Adolescents who suffered from daily abdominal pain reported less participation in social activities as compared with those with minimal or no abdominal pain (0.6 vs 1.9, *P .04).
likely to restrict their daily activities and consider their lives a failure compared with children with rare episodes of abdominal pain. From the Add Health Study population, 16% of adolescents are at risk for depression as dened by the CES-D score, which dramatically increases to 45% when those patients experience frequent abdominal pain. About 40% of children with frequent abdominal pain meet criteria for depression.11 More than 15% of the Add Health Study responders had abdominal pain more than 2 times per week. This was a large and diverse sample of children followed up prospectively for 1 year to assess abdominal pain complaints. Previous studies estimating abdominal pain prevalence in children have been limited by their smaller single-community samples. Ghandour et al12 reported that 20% of adolescent females experience abdominal pain more than once a week, which is similar to 17% reported by Hyams et al1 in high school students. Numerous articles have shown that children with chronic abdominal pain were more likely to have increased health service use, restrictions in daily living, and suffer from functional impairments in adulthood.3,4,11,13 In this cohort of adolescent patients identied as having daily abdominal pain, we have conrmed the signicant negative impact on quality of life as evidenced by the increased days of school missed and diminished participation in social activities. School absenteeism, not meeting friends, loss of appetite, and sleep disturbances are some of the restrictions of daily living in children and adolescents with chronic abdominal pain that may need to be addressed in future intervention studies. These restriction in daily quality of life may indeed contribute to the most worrisome nding in our study that more than 10% of adolescents with daily abdominal pain believed their life was a failure. In adults, chronic abdominal pain is an independent risk factor for suicidal ideation.5 Adults who experience abdominal pain from IBS are thought to have suicidal ideations because of
the belief that their disorder is hopeless. Also, their perceived disease severity independently increases suicidal ideations.6 Recognition of frequent abdominal pain as a chronic pain syndrome allows for identication of adolescents at risk for depression and self-injurious behavior. Little et al14 were able to document positive depression screens in children with chronic abdominal pain and nongastrointestinal somatic complaints. Specic somatic complaints such as musculoskeletal pain 3 times a week in boys and girls and weekly headaches in girls may be clues for depression.15 Children with chronic abdominal pain have more somatic complaints compared with healthy children, which includes headache, stomachache, backache, and morning fatigue, possibly as a response to stress.4,13 Although it was not an aim of this study, it would have been interesting to measure the frequency of musculoskeletal pain and headaches among adolescents with frequent abdominal pain and possibly to determine an association with depressive symptoms as well. Although the Add Health Study relies on self-reported symptoms, the overall quality of the data is considered quite high. In this study, measures that can be validated externally such as self-reported height and weight were correlated highly with measured height and weight.16 In addition, self-reported friendship nominations also have been validated externally.17 This study shows an association with abdominal pain and depressive symptoms in adolescents, which is more signicant when the abdominal pain occurs daily. The limitations of this study were that the Add Health study did not provide extensive medical information on the respondents, particularly regarding whether any interventions were made or treatment was started. Also, it has been more than 10 years since the data gathered in the Add Health Study were collected and published. One may question whether the data are reective of current adolescents. The recent publication of the Youth Risk Behavior Surveillance System has conrmed that the rates of depressive symptoms and suicide attempts among adolescents have not changed signicantly from 1991 to 2005.18 In addition, with the recent insights of Smith el al5 suggesting that abdominal pain in adults may be an independent risk factor for suicidal ideation and may be related to both symptom severity and duration, the data reported here have potential important public health concerns for adolescents who suffer similarly from chronic abdominal pain. It would have been ideal to have a control group for this large cohort to strengthen the association of frequent abdominal pain in adolescents and increased risk for depressive symptoms, but this study analyzed previously gathered data. The prevalence of frequent abdominal pain increases with age into adolescence, which is similar to depression. The association of frequent abdominal pain and anxiety in children is
Table 1. Abdominal Pain and Associated Depressive Symptoms

Rare pain Feel Feel Feel Feel Feel sad like crying lonely low energy life is a failure 5.3% 1.0% 6.4% 4.8% 3.3% Daily pain 25.3% 12.1% 25.2% 19.1% 10.7% P value .001 .001 .001 .001 .001
332
YOUSSEF ET AL
well established.1,2,13,15 More than 80% of children with frequent abdominal pain have a comorbid anxiety disorder and approximately 40% have depressive symptoms.13 In this current study there was a dramatic increase from those patients who suffered only rarely from abdominal pain versus those with daily abdominal pain at risk for depression as dened by the CES-D score. Adolescent females with frequent abdominal pain appear to be the target population at highest risk for depression.13 In the current study we found a similar prevalence of frequent abdominal pain among females compared with males, with a ratio of 2:1. In a recent systematic review of the literature exploring suicidal behavior and irritable bowel syndrome, Spiegel et al conrmed that chronic abdominal pain is an independent predictor of suicidal behavior in adults, particularly females.19 Early recognition of hopelessness and other symptoms suggestive of depression in adolescents who present for evaluation of frequent abdominal pain is a key nding and offers an opportunity to provide more effective coping strategies to be offered earlier to these patients. An effective strategy for abdominal pain early in adolescence may have a positive effect on quality of life and decrease the associated psychosocial comorbidities associated with the chronic abdominal pain syndromes seen in adults. References
1. Hyams JS, Burke G, Davis PM, et al. Abdominal pain and irritable bowel syndrome in adolescence: a community-based study. J Pediatr 1996;129:220 226. 2. Youssef NN, et al. Quality of life for children with functional abdominal pain: a comparison study of patients and parents perceptions. Pediatrics 2006;117:54 59. 3. Roth-Isigkeit, et al. Pain among children and adolescents: restrictions in daily living and triggering factors. Pediatrics 2005;115: e152 e162. 4. Campo JV, Comer DM, Jansen-McWilliams L, et al. Recurrent pain, emotional distress, and health service use in childhood. J Pediatr 2002;141:76 83. 5. Smith MT, Edwards RR, Robinson RC, et al. Suicidal ideation, plans, and attempts in chronic pain patients: factors associated with increased risk. Pain 2004;111:201208. 6. Miller V, Hopkins L, Whorwell PJ. Suicidal ideation in patients with irritable bowel syndrome. Clin Gastroenterol Hepatol 2004;3: 1064 1068. 7. Rushton JL, Forcier M, Schectman RM. Epidemiology of depressive symptoms in the National Longitudinal Study of Adolescent Health. J Am Acad Child Adolesc Psychiatry 2002;41:199 205. 8. Bearman PS, Jones J, Udry JR. The National Longitudinal Study of Adolescent Health: research design 1997. Chapel Hill, NC: Carolina Population Center, 1997. Available at: http://www. cpc.unc.edu/addhealth.
9. Radloff LS. The use of the Center for Epidemiologic Studies Depression Scale in adolescent and young adults. J Youth Adolesc 1991;20:149 166. 10. Roberts RE, Lewinsohn PM, Seeley JR. Screening for adolescent depression: a comparison of depression scales. J Am Acad Child Adolesc Psychiatry 1990;30:58 66. 11. Campo JV, et al. Recurrent abdominal pain, anxiety, and depression in primary care. Pediatrics 2004;113:817 824. 12. Ghandour RM, et al. Headache, stomachache, backache, and morning fatigue among adolescent girls in the United States. Arch Pediatr Adolesc Med 2004;158:797 803. 13. Burke P, Elliott M, Fleissner R. Irritable bowel syndrome and recurrent abdominal pain: a comparative review. Psychosomatics 1999;40:277281. 14. Little CA, Williams SE, Puzanovova M, et al. Multiple somatic symptoms linked to positive screen for depression in pediatric patients with chronic abdominal pain. J Pediatr Gastroenterol Nutr 2007;44:58 62. 15. Egger HL, Costello EJ, Erkanli A, et al. Somatic complaints and pyschopathology in children and adolescents: stomach aches, musculoskeletal pains and headaches. J Am Acad Child Adolesc Psychiatry 1999;38:852 860. 16. Goodman E, Hinden B, Khandelwal S. Accuracy of teen and parental reports of obesity and body mass index. Pediatrics 2000;106:5258. 17. Strauss RS, Pollack HA. Social marginalization of overweight children. Arch Pediatr Adolesc Med 2003;157:746 752. 18. Eaton DK, et al. Youth risk behavior surveillanceUnited States 2005. MMWR Surveillance Summaries 2006;55:1108. 19. Spiegel B, Schoenfeld P, Nabiloff B. Systematic review: the prevalence of suicidal behaviour in patients with chronic abdominal pain and irritable bowel syndrome. Aliment Pharmacol Ther 2007;26:183193.
Address correspondence to: Nader N. Youssef, MD, Center for Pediatric Irritable & Motility Disorders, Goryeb Childrens Hospital at Atlantic Health, 100 Madison Avenue, Internal Box 82, Morristown, New Jersey. The Marguerite and Joseph P. Goryeb Endowment to the Center for Pediatric Irritable Bowel & Motility Disorders provided partial funding for this study. This research used data from Add Health, a program project designed by J. Richard Udry, Peter S. Bearman, and Kathleen Mullan Harris, and funded by a grant P01-HD31921 from the National Institute of Child Health and Human Development, with cooperative funding from 17 other agencies. Persons interested in obtaining data les from Add Health should contact Add Health, Carolina Population Center, 123 West Franklin Street, Chapel Hill, NC 27516-2524 (addhealth@unc.edu). Special acknowledgment is owed to Ronald R. Rindfuss and Barbara Entwisle for assistance in the original design.
Sharing Genetic Test Results in Lynch Syndrome: Communication With Close and Distant Relatives
ELENA M. STOFFEL,*,, BETH FORD, ROWENA C. MERCADO, DARASHANA PUNGLIA, WENDY KOHLMANN, PEGGY CONRAD, AMIE BLANCO, KRISTEN M. SHANNON,# MARK POWELL,** STEPHEN B. GRUBER, JONATHAN TERDIMAN, DANIEL C. CHUNG,,#, and SAPNA SYNGAL*,,
*Division of Gastroenterology, Brigham and Womens Hospital, Boston, Massachusetts; Division of Population Sciences, **Department of Biostatistics and Computational Biology, Dana-Farber Cancer Institute, Boston, Massachusetts; Harvard Medical School, Boston, Massachusetts; Cancer Genetics Clinic, University of Michigan Medical Center, Ann Arbor, Michigan; University of California San Francisco Cancer Center, San Francisco, California; and the #Center for Cancer Risk Analysis, Gastroenterology Unit, Massachusetts General Hospital, Boston, Massachusetts
Background & Aims: Clinical genetic testing can help

direct cancer screening for members of Lynch syndrome families; however, there is limited information about family communication of genetic test results. Methods: A total of 174 probands who had genetic testing for Lynch syndrome were enrolled through 4 US cancer genetics clinics. Subjects were asked whether they had disclosed their genetic test results to rst-, second-, and third-degree relatives. Univariate and multivariate analyses were used to identify clinical and demographic factors associated with informing immediate and extended family of genetic test results. Results: One hundred seventy-one of 174 probands (98%; 95% condence interval, 95%100%) reported that they had disclosed their genetic test result to a rstdegree relative. Communication of test results to other relatives occurred signicantly less often, with only 109 of 162 (67%; 95% condence interval, 59%74%) subjects with second- or third-degree relatives sharing their results. Individuals with a pathogenic mutation were signicantly more likely to inform distant relatives than were subjects with a negative or indeterminate test result (odds ratio, 2.49; 95% condence interval, 1.14 5.40). Probands age, sex, and cancer status did not inuence communication of genetic test results. Lack of closeness and concerns that relatives would worry or not understand the implications of test results were the primary reasons for not sharing genetic test results. Conclusions: Most individuals who undergo genetic testing for Lynch syndrome share their test results with rst-degree family members; however, these results reach more distant relatives signicantly less often. Interventions to improve communication of genetic test results to members of the extended family are necessary to provide optimal cancer prevention care to at-risk families. enetic testing plays an increasing role in the care of patients at risk for cancer as a result of hereditary cancer syndromes. Lynch syndrome, also known as hereditary nonpolyposis colorectal cancer, is the most common hereditary colorectal cancer syndrome, and accounts for approximately 3% to 5% of all diagnosed colorectal cancer cases.1 Genetic testing is clinically available for mutations in the DNA mismatch repair genes MLH1, MSH2, and MSH6, which are the most common causes of Lynch syndrome. Because of the increased risk for colorectal and extracolonic cancers, individuals at risk for Lynch syn-
drome require a colonoscopy every 1 to 2 years starting at age 20 to 25,2 and women should have screening for endometrial and ovarian cancers or consider prophylactic hysterectomy.2 Identication of a pathogenic mutation through genetic testing conrms the clinical diagnosis of Lynch syndrome and provides an opportunity to stratify cancer risk for other family members. Individuals who undergo genetic testing for Lynch syndrome appear to be more likely to adhere to recommended cancer screening guidelines.3 A recent survey of cancer centers in the United States indicated that the demand for genetic evaluation services for familial cancer syndromes has increased rapidly over the past decade.4 Although genetic testing is expensive, economic analyses have supported the clinical utility of genetic testing for Lynch syndrome and have shown that cost effectiveness increases substantially when the benets of testing are extended to probands family members.5 Studies have suggested that most relatives of patients with colorectal cancer would be interested in genetic testing for cancer predisposition6; however, there are limited data regarding how information about genetic testing actually is communicated in families undergoing molecular evaluation for Lynch syndrome. The objectives of our study were to examine how genetic testing information is communicated in families at risk for Lynch syndrome, and to identify factors associated with disclosure of genetic test results to close and distant family members.
Methods
We conducted a cross-sectional questionnaire study among individuals with a personal or family history fullling clinical criteria for Lynch syndrome. Subjects were recruited through 4 cancer genetics clinics in the United States: DanaFarber Cancer Institute (Boston, MA), Massachusetts General Hospital (Boston, MA), University of Michigan (Ann Arbor, MI), and University of California San Francisco (San Francisco, CA). Eligible subjects included individuals whose personal or family history fullled Bethesda Guidelines for Lynch syndrome.7 All participants were age 18 years or older, and were required to read and write English.
Abbreviations used in this paper: FDR, rst-degree relative; SDR, second-degree relative; TDR, third-degree relative. 2008 by the AGA Institute 1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.014
334
STOFFEL ET AL
Eligible individuals were identied through visits at 1 of 4 cancer genetics clinics or through referral by a family member. Potential subjects were invited to enroll in the study either at a clinical visit or by mail. Individuals approached by mail received an initial study packet with an introduction letter and questionnaire, as well as a decline to participate form. Individuals who did not return study materials after 2 follow-up telephone calls and 2 mailings were considered nonresponders. Subjects who had undergone genetic testing were enrolled at least 3 months after their genetic test result had been disclosed to them. Questionnaire data were scanned and entered into a computerized database. The study was approved by the institutional review board of each participating study site. A total of 466 eligible individuals were approached for enrollment. Of these, 270 (58%) completed the study questionnaire, 34 (7%) declined to participate, and 158 (34%) were nonresponders. Women and college graduates were more likely to complete the study questionnaires. There were no signicant differences between other demographic characteristics (such as age and cancer status) of study responders and nonresponders. Of the 270 subjects who completed study questionnaires, 174 (64%) reported that they had undergone genetic testing for Lynch syndrome; only these individuals were included in this analysis.
The potential effects of clinical and demographic factors on the decision to disclose genetic test results were explored using univariate tests of association (Fisher Exact test and Student t test). Factors that were found to be signicant on univariate analysis or that were believed to have empiric clinical relevance were included in multivariable logistic regression models to identify variables associated with sharing genetic test results with FDRs and SDRs/TDRs. Generalized estimating equations were used to account for potential clustering of results among members of the same family. Analyses were performed using SAS software (SAS Institute, Cary, NC). All P values are 2-sided and a P value of less than .05 was considered signicant.
Results Subject Characteristics

Most of the 174 subjects who reported having undergone genetic testing for Lynch syndrome were women (70%), white (91%), college graduates (69%), and married (76%). The mean age of the participants was 46.7 years (range, 18 79 y). More than half of study participants (61%) had a cancer diagnosis, and 104 (60%) individuals had a conrmed positive genetic mutation associated with Lynch syndrome (Table 1).
Measures
The study questionnaires collected standard demographic data including age, sex, race, ethnicity, marital status, household income, level of education, and type of health insurance. Subjects provided details about personal cancer history and were asked to estimate their own risk for developing cancer and whether they had ever undergone genetic testing for Lynch syndrome. In a detailed family history section, subjects provided information about number of siblings; children; prevalence of cancers among rst-, second-, and third-degree relatives; as well as history of genetic testing and specic genetic test results for individuals in their immediate and extended family who had undergone genetic testing. A family pedigree was constructed for each subject. Subjects were asked, Have you shared your genetic test result with any of the following people: mother, father, sisters, brothers, spouse/partner, daughters, sons, aunts/uncles, or cousins? Subjects were asked to choose among reasons why they had or had not shared genetic test results with each of those family members and were permitted to select more than one response.
Disclosure of Genetic Test Results to First-Degree Family Members

Overall, 171 of 174 (98%; 95% condence interval, 95 100) subjects reported that they had shared their genetic test result with at least one FDR (Figure 1). Only 5 subjects with a living parent reported that they had not disclosed their genetic test results to their mother or father, 4 had not disclosed their results to a sister, and 2 had not informed a brother. Nearly 90% of subjects with children informed their sons or daughters of their test result, and most of the others indicated that they would wait until their children were older before discussing the testing. There were no observed differences in rates of disclosure to FDRs by probands genetic test results or sex.
Disclosure of Genetic Test Results to More Distant Relatives

Overall, 109 of 162 (67%) subjects with living second- or third-degree family members reported that they had shared their genetic test result with one or more of these SDRs/TDRs (Figure 1). Of the 97 individuals whose genetic test result showed a pathogenic mutation (positive test), 73 (75%) disclosed their test result to a relative beyond their nuclear family. Rates of disclosure to SDRs/TDRs were signicantly lower among subjects with indeterminate or true negative results, with only 24 of 42 (57%) and 12 of 23 (52%), respectively, indicating they had shared test results with relatives beyond rst degree (P .03) (Table 1). In univariate analysis, subjects who had more relatives diagnosed with cancers associated with Lynch syndrome appeared more likely to share their genetic test result with family members beyond FDRs (P .05) (Table 1). Aside from the genetic test result, there were no other signicant associations between the disclosure of genetic test result to an SDR/TDR and probands sociodemographic characteristics such as sex, age, level of education, race, marital status, having children, and personal history of cancer. Similarly, individuals with a higher level of cancer worry, previous history of genetic testing in the family, or prior evaluation at a high-risk/genetics
For subjects who reported having undergone genetic testing, each participants family history and pedigree was reviewed to determine that they had at least one living rst-degree relative (FDR) and at least 1 living second- or third-degree relative (SDR/TDR). Subjects who indicated they had shared their genetic test result with their mother, father, brothers, sisters, or children were classied as having disclosed the result to an FDR. Subjects who indicated they had shared their result with uncles, aunts, or cousins were classied as having disclosed results to an SDR/TDR. Subjects without at least one living FDR or SDR/TDR were not included in the corresponding analysis.
March 2008
SHARING GENETIC TEST RESULTS IN LYNCH SYNDROME
335
Table 1. Characteristics of the Study Population (N 174) and Univariate Analysis of Factors Predicting Disclosure of Genetic Test Results to Any Family Members Beyond First Degree
Subject disclosure to family members All subjects who had genetic testing (N 174) Characteristics Mean age (range) Sex Male Female Race White Non-white Unknown/missing Education Less than college graduate At least college graduate Unknown/missing Marital status Married Not married Unknown/missing Cancer diagnosis Yes No Test results Positive Indeterminate True negative Cancer worry Low Moderate High Unknown/missing Prior genetic testing in family Yes No Do not know Mutation previously identied in family Yes No History of Lynch syndrome cancer in 1 or more relatives Yes No Mean number of relatives with Lynch syndrome cancers Subjects with children Yes No Ever evaluated in a genetics/high-risk clinic Yes No Unknown/missing
aThere
Told beyond FDR (N 109)a Frequency 45.96 28 81 96 13 59.6 70.4 65.3 86.7 %
Did not tell beyond FDR (N 53)a Frequency 45.26 19 34 51 2 40.4 29.6 .15 34.7 13.3 .72 % P valueb .73 .20
Frequency 46.7 52 122 157 16 1 53 118 3 132 38 4 106 68 104 47 23 58 65 50 1 128 39 7 113 61
% 1879 29.9 70.1 90.8 9.3 31.0 69.0 75.9 21.8 2.3 60.9 39.1 59.8 27.0 13.2 33.5 37.6 28.9 73.6 22.4 4.0 64.9 35.1
34 73 2 81 26 2 66 43 73 24 12 35 44 30
69.4 65.8 66.9 68.4 69.5 64.2 75.3 57.1 52.2 66.0 72.1 62.5
15 38 40 12 1 29 24 24 18 11 18 17 18
30.6 34.2 1.00 (33.1) (31.6) .50 30.5 35.8 .03 24.7 42.9 47.8 .55 34.0 27.9 37.5 1.00
82 24 3 75 34
67.2 68.6 69.4 63.0
40 11 2 33 20
32.8 31.4 .48 30.6 37.0 1.00
155 89.1 19 10.9 3.71 (2.0)
98 67.1 11 68.8 4.04 (2.1)
48 32.9 5 31.2 3.38 (1.8)
.05 .72
122 52 154 19 1
70.1 29.9 89.0 11.0
76 33 98 10 1
68.5 64.7 67.6 62.5
35 18 47 6
31.5 35.3 .78 32.4 37.5
NOTE. Bolded entries have a signicance of P .05. were 162 subjects who indicated they have at least 1 SDR or TDR to tell. bComparison between subjects who did and did not tell beyond FDRs, the Fisher exact test was used for all categoric variables; the t test was used for all continuous variables.
336
STOFFEL ET AL
Figure 1. Percentage of participants disclosing test results to family members, by test result. Error bars represent 95% condence intervals.
clinic were not more likely to disclose their result to SDRs/ TDRs (Table 1). There were no differences in rates of disclosure of genetic test results among subjects enrolled from any of the 4 study sites. In multivariate analysis controlling for subjects sex and personal and family history of cancer, having a genetic test result that revealed a pathogenic mutation was the only significant predictor of disclosing the test result to one or more SDRs/TDRs (odds ratio, 2.49; 95% condence interval, 1.14 5.40) (Table 2).
Reasons for Disclosing or Not Disclosing Genetic Test Results

Subjects were asked to choose from a list of reasons why they had or had not disclosed genetic test results to one or more family members (Figure 2). The most frequently cited reasons for sharing a genetic test result with family members were as follows: (1) to inform them of their risk, (2) to encourage testing, and (3) to obtain emotional support. Only 1 in 10 respondents reported that they informed their family members of their genetic test result because their physician told them to do so. The most frequently cited reasons for not sharing genetic test results were as follows: (1) they were not close to their family members, (2) concern that family members would not understand the test result, and (3) they did not want relatives to worry. Only 3 individuals listed concerns about condentiality as a reason why they did not share their test result. None listed guilt about their result as a reason for not sharing information about their testing. Among other reasons were wanting to wait until young children were older before informing them about results of genetic tests, inability to contact specic family members, and concern that the information would be too distressing. Although 53 of 162 (32.7%) individuals with living SDRs/ TDRs said they did not share genetic test results beyond their immediate family, only 26 (15%) reported that they had deliberately withheld their genetic test results from any family member.
nearly all (98%) had disclosed their test results to FDRs; and two thirds had shared these results with more distant family members such as cousins, aunts, or uncles. Seventy-three of 97 (75%) subjects whose testing identied a gene mutation communicated this result to one or more SDRs/TDRs. Having a true positive genetic test result was the only clinical or demographic factor signicantly associated with disclosure of a genetic test result to relatives beyond the immediate family. To date, most of the research examining family communication about genetic testing has focused on individuals undergoing evaluation for hereditary breast and ovarian cancer syndrome. These studies have shown that most patients that are tested for BRCA 1 and 2 mutations inform their FDRs of their genetic test result; however, communication to more distant relatives occurs less frequently.8 In the predominantly female BRCA 1 and 2 cohorts, patients were more likely to disclose genetic test results to other female family members rather than to male relatives also at risk,8,9 and positive results were more likely to be shared, as compared with negative or uninformative test results.8,9 Previous studies on disclosure of genetic test results in Lynch syndrome families, each of which reported on fewer than 40 subjects, also found that genetic test results are disclosed less frequently to at-risk relatives outside the nuclear family.10 12 One study suggested that male probands were less likely than female probands to communicate genetic test results to family members11 and proposed that male patients might require additional counseling resources to ensure appropriate communication of results. Our ndings, from a much larger Lynch syndrome cohort, show that probands sex, race, cancer history, and concerns about privacy and condentiality do not appear to inuence most patients decisions about sharing genetic test results with family; however, the genetic test result is still a signicant factor. Most subjects who did not disclose test results said they did not intentionally withhold this information. The fact that disclosure was nearly universal among FDRs, but occurred less frequently among more distant relatives, suggests that subjects may not be fully aware of the potential impact of their genetic test result on health care of SDRs/TDRs and/or may encounter
Table 2. Multivariate Analysis of Factors Predicting Disclosure to any Family Members Beyond First Degree
Characteristic Personal history of cancer Yes No Sex Female Male Positive mutation carrier Yes No Mean number of relatives with Lynch syndrome cancers Odds ratio (95% CI) P value
1.23 (0.612.51) 1.68 (0.823.44) 2.49 (1.145.40) 0.13a
.57
.17
.02 .21
Discussion
In a multicenter study of 174 individuals who had undergone genetic testing for Lynch syndrome, we found that
NOTE. Analysis used generalized estimating equation to control for family effects, with exchangeable working correlation 0.02. Bold entry has a signicance of P .05. aParameter estimate.
March 2008
SHARING GENETIC TEST RESULTS IN LYNCH SYNDROME
337
Figure 2. (A) Reasons for disclosing results to family members. (B) Reasons for not disclosing results to family members. *Subjects were permitted to select more than 1 option.
other barriers to communicating results to more distant relatives. Why is sharing genetic test results important for other relatives in a family with Lynch syndrome? In the case of a true positive test, identifying a clearly pathogenic gene mutation in a proband conrms the diagnosis and allows unaffected family members to ascertain their cancer risk through informative (and less expensive) mutation-specic testing. A true positive genetic test result affects clinical management for FDRs, SDRs, and TDRs, who, through testing, would learn whether they require high-risk cancer screening. By comparison, a true negative genetic test result (which occurs when an individual tests negative for a mutation previously identied in the family) changes clinical management for only the tested individual and his/her progeny (children/grandchildren), providing reassurance that they did not inherit the increased cancer risk. Because a true negative test result does not change management for other relatives who are not direct descendants, lower rates of disclosure of true negative test results to SDRs/TDRs might be expected. In the case of an indeterminate/uninformative genetic test result (which occurs when an individuals genetic test result does not reveal a mutation and there has not been a mutation previously identied in the family), the diagnosis of Lynch syndrome can be neither conrmed nor rejected and the importance of disclosing this test result is less obvious. However, an indeterminate/uninformative result still may have clinical
relevance. Discussion of the genetic testing provides an opportunity to share information about Lynch syndrome with other family members who may be at risk and may benet from specialized cancer surveillance. In our cohort more than half of the subjects with an indeterminate/uninformative genetic test result met Amsterdam Criteria and still would be considered at risk for Lynch syndrome. Overall, our results show that in most cases information about a genetic test result does reach members of the immediate and extended family. In 75% of cases in which a pathogenic mutation was identied, this information was shared with one or more SDRs/TDRs. If this information led other family members to get tested, this would support models of cost effectiveness for genetic testing, suggesting benets are likely to extend to family members beyond the proband and his/her immediate family. With regard to individuals with an indeterminate genetic test result, our nding that rates of disclosure to SDRs/ TDRs are signicantly lower (57%) may reect the difculty in interpreting the clinical signicance of an uninformative result and conveying this information in a way patients and their families can understand. Some prior studies have suggested that individuals with an indeterminate genetic test result may be falsely reassured that their cancer risk is now lower because there was no mutation identied, and may not feel that information about the genetic test result is important to disseminate.8
338
STOFFEL ET AL
One of the major challenges in improving the effectiveness of genetic testing for cancer prevention is to develop ways to distribute this information to family members and health care providers. At present, most genetic testing is conducted in specialized centers where patients meet with genetic counselors before and after disclosure of their genetic test result. It generally is accepted that health care providers have an obligation to inform probands of implications of the genetic diagnosis for other family members and encourage them to share their test result with relatives.13 In the United States, privacy laws prevent physicians from disseminating this information without a patients consent. However, simply telling patients that they need to inform family members of their genetic test result may not be enough. Our ndings suggest that even under ideal conditions, some patients do not inform relatives of their genetic test result, in many cases because of concerns their family members will worry, or will not understand, or because they are not close. Providing patients with a detailed letter describing the implications of the test result and giving them an annotated copy of the family tree indicating which family members should receive genetic testing information may help ensure that this information is shared with others who may benet. There are several limitations to consider in this study. Subjects who were willing to spend 30 minutes completing study questionnaires may have been more motivated and thus more likely to communicate their test result to family members, or more likely to report that they had. We asked subjects whether they had shared their genetic test result with specic types of relatives (sisters, aunts, cousins, and so forth); however, we did not have total counts of each probands living relatives, so it was not possible to calculate the proportion of at-risk relatives informed of genetic test results. We did not have any way to conrm that subjects actually shared genetic testing information with relatives who were at risk, nor could we assess the quality of the communication or the accuracy of the information transmitted to family members. Consequently, our ndings may overestimate the true rates of genetic test disclosure. Finally, our study did not collect information about outcomes of disclosure of genetic testing information to relatives; therefore we could not ascertain whether clinical care of family members changed as a result of genetic testing. Despite these limitations, our report provides useful data about patterns of communication about genetic testing in Lynch syndrome families. Our study of patients from 4 US cancer centers shows that genetic test results often are shared with members of the immediate family, but less often are communicated to more distant at-risk relatives. Because it is expected that a greater share of genetic testing will move from specialized cancer centers to physicians private ofces, it is important to develop and implement strategies that will help patients communicate with family members about genetic testing. Specic interventions, such as providing patients with documentation of their genetic test result and screening recommendations and providing them with strategies for disseminating their test result to at-risk family members, may help to
remove barriers to family communication and to improve effectiveness of genetic testing for cancer prevention. References
1. Lynch HT, Watson P, Shaw TG, et al. Clinical impact of molecular genetic diagnosis, genetic counseling, and management of hereditary cancer. Part II: hereditary nonpolyposis colorectal carcinoma as a model. Cancer 1999;86(Suppl):24572463. 2. Giardiello FM, Brensinger JD, Petersen GM. AGA technical review on hereditary colorectal cancer and genetic testing. Gastroenterology 2001;121:198 213. 3. Wagner A, van Kessel I, Kriege MG, et al. Long term follow-up of HNPCC gene mutation carriers: compliance with screening and satisfaction with counseling and screening procedures. Fam Cancer 2005;4:295300. 4. Epplein M, Koon KP, Ramsey SD, et al. Genetic services for familial cancer patients: a follow-up survey of National Cancer Institute Cancer Centers. J Clin Oncol 2005;23:4713 4718. 5. Ramsey SD, Clarke L, Etzioni R, et al. Cost-effectiveness of microsatellite instability screening as a method for detecting hereditary nonpolyposis colorectal cancer. Ann Intern Med 2001; 135:577588. 6. Kinney AY, Choi YA, DeVellis B, et al. Interest in genetic testing among rst-degree relatives of colorectal cancer patients. Am J Prev Med 2000;18:249 252. 7. Rodriguez-Bigas MA, Boland CR, Hamilton SR, et al. A National Cancer Institute Workshop on Hereditary Nonpolyposis Colorectal Cancer Syndrome: meeting highlights and Bethesda guidelines. J Natl Cancer Inst 1997;89:1758 1762. 8. Claes E, Evers-Kiebooms G, Boogaerts A, et al. Communication with close and distant relatives in the context of genetic testing for hereditary breast and ovarian cancer in cancer patients. Am J Med Genet A 2003;116:1119. 9. Patenaude AF, Dorval M, DiGianni LS, et al. Sharing BRCA1/2 test results with rst-degree relatives: factors predicting who women tell. J Clin Oncol 2006;24:700 706. 10. Peterson SK, Watts BG, Koehly LM, et al. How families communicate about HNPCC genetic testing: ndings from a qualitative study. Am J Med Genet C Semin Med Genet 2003;119:78 86. 11. Gaff CL, Collins V, Symes T, et al. Facilitating family communication about predictive genetic testing: probands perceptions. J Genet Couns 2005;14:133140. 12. Mesters I, Ausems M, Eichhorn S, et al. Informing ones family about genetic testing for hereditary non-polyposis colorectal cancer (HNPCC): a retrospective exploratory study. Fam Cancer 2005;4:163167. 13. Oft K, Groeger E, Turner S, et al. The duty to warn a patients family members about hereditary disease risks. JAMA 2004;292: 1469 1473.
Address requests for reprints to: Elena M. Stoffel, MD, MPH, Division of Gastroenterology, Brigham and Womens Hospital, 75 Francis Street, Boston, Massachusetts 02115. e-mail: estoffel@partners.org; fax: (617) 632-4088. Supported by an American College of Gastroenterology Junior Faculty Award (2004), a GlaxoSmithKline Institute for Digestive Health Clinical Research Award (2004), and K07 NCI CA 120448-01 (all to E.M.S.), and K24 NCI CA 113433 (S.S.). Drs Syngal, Chung, and Terdiman have consultant/advisory relationships with Myriad Genetic Laboratories, Salt Lake City, Utah.
ORIGINAL ARTICLESLIVER, PANCREAS, AND BILIARY TRACT

Prognostic Implications of Lactate, Bilirubin, and Etiology in German Patients With Acute Liver Failure
JOHANNES HADEM,* PENELOPE STIEFEL,* MATTHIAS J. BAHR,* HANS L. TILLMANN, KINAN RIFAI,* RGEN KLEMPNAUER, HEINER WEDEMEYER,* MICHAEL P. MANNS,* and ANDREA S. SCHNEIDER* JU
*Department of Gastroenterology, Hepatology and Endocrinology, and Department of Visceral and Transplant Surgery, Hannover Medical School, Hannover; and Medical Clinic and Policlinic II, University of Leipzig, Leipzig, Germany
Background & Aims: Among the potentially helpful

indicators of poor prognosis in acute liver failure (ALF) are etiology, encephalopathy grade, blood lactate, and Kings College Criteria (KCC). The accuracy of these parameters in predicting transplantation or death shows signicant variation in different countries. Methods: We retrospectively analyzed 102 patients with ALF treated at our institution between 1996 and 2005. Baseline parameters, simplied acute physiology score III (SAPS-III), KCC, Model for EndStage Liver Disease (MELD) score, and a novel score of bilirubin, lactate, and etiology (BiLE score) were compared between transplant-free survivors and patients who required liver transplantation or died, by using multivariate linear regression analysis and receiver operating characteristics (ROC). Results: The most common causes of ALF were indeterminate liver failure (21%), acute hepatitis B (18%), acetaminophen ingestion (16%), and Budd-Chiari syndrome (9%). Transplantation-free survival was 38%, 44% of patients underwent liver transplantation, and 18% died without transplantation. Eight-week survival was 77%. The BiLE score was the best predictor of death or need of transplantation, with 79% sensitivity and 84% specicity. ROC analysis revealed a better performance of BiLE score when compared with bilirubin, lactate, MELD score, and SAPS-III (area under the curve: 0.87 0.04, 0.73 0.51, 0.73 0.52, 0.71 0.05, and 0.68 0.59, respectively). Conclusions: The simple, combined BiLE score emerged as the best predictor of poor outcome in our patient cohort and should be prospectively evaluated in other populations. n 1970, Trey and Davidson1 dened acute liver failure (ALF) as an acute deterioration of liver function resulting in the development of encephalopathy within 8 weeks of the onset of symptoms in a patient with a previously healthy liver. According to the recent position paper of the American Association for the Study of Liver Diseases, ALF is dened as decline in liver function of less than 26 weeks duration, absence of cirrhosis, evidence of coagulation abnormality (usually indicated by an international normalized ratio [INR] 1.5), and any degree of encephalopathy. Patients with Wilsons disease, vertically acquired HBV, or autoimmune hepatitis might be included in
spite of possible cirrhosis if the disease has only been recognized for less than 26 weeks.2 The composition of etiologies varies among geographic regions, an issue that has been addressed by a number of studies.3 Kings College Hospital London reported 57% acetaminophen toxicity and 9% viral hepatitis among 1014 patients between 1973 and 1991.4 A large prospective study involving 17 sites in the United States (Acute Liver Failure Group) collected data of 308 patients regarding etiology, clinical and laboratory features, and outcome of patients presenting with ALF. The most common etiologies were acetaminophen overdose (39%), indeterminate (17%), idiosyncratic drug reactions (13%), and viral hepatitis A and B (12%).5 Early reports from France indicated viral hepatitis to be the most important cause of ALF in central Europe,6 with acute hepatitis B accounting for up to 45% of cases.7 The diversity of etiologies was just recently assessed by an international survey of 6 liver transplantation centers summarizing 284 ALF cases. Although acetaminophen was the dominant etiology in London and Copenhagen with 55% and 74% of cases, respectively, hepatitis B (including acute exacerbations of chronic hepatitis B) was the most prominent etiology in Hong Kong (74%) and Karachi (38%). In addition, a high proportion of cases in Karachi (35%) were due to hepatitis E.8 A major problem in the management of ALF is the prediction of spontaneous recovery. Of numerous proposed prognostic criteria for ALF, the Kings College criteria (KCC) have been the most commonly used and most frequently evaluated.9 KCC are probably the best validated tool currently available.10 However, the predictive value of KCC in the U.S. has been questioned11 and has not been sufciently evaluated in central Europe, where acetaminophen is not the leading cause of ALF.12 Depending on etiology, KCC were shown to have a positive predictive value of 79% 88% and negative predictive value of
Abbreviations used in this paper: ALF, acute liver failure; AUC, area under the curve; BiLE, bilirubin-lactate etiology; ICU, intensive care unit; INR, international normalized ratio; KCC, Kings College Criteria; MELD, Model for End-Stage Liver Disease; OLT, orthotopic liver transplantation; ROC, receiver operating characteristics; SAPS-III, simplied acute physiology score III. 2008 by the AGA Institute 1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.039
340
HADEM ET AL
50% 65%.13 In France, Clichy criteria are widely used to select ALF patients in need of orthotopic liver transplantation (OLT). Besides the presence of hepatic encephalopathy, they include patients age and factor V levels.14 Only a few studies have attempted to directly compare the Clichy criteria and KCC, suggesting superiority of the KCC.15,16 In acetaminophen-induced ALF, blood lactate levels above 3.0 mmol/L had a sensitivity of 76% and specicity of 97% in predicting death,17 but results diverged in nonacetaminophen-related ALF.18 22 According to the large prospective U.S. multi-center study, encephalopathy grade at admission and etiology were the only predictors of outcome.5 Recently, Yantorno et al23 have suggested a cut-off value above 30 of the Model for End-Stage Liver Disease (MELD) score as predictor of poor outcome. The objectives of our retrospective analysis were to evaluate these prognostic parameters in a central European cohort.
Table 1. BiLE Score

BiLE score Bilirubin (mol/L)/100 lactate (mmol/L) 4 (in case of indeterminate ALF, Budd-Chiari syndrome, or phenprocoumon toxicity) 2 (in case of acetaminophen toxicity) 0 (in case of any other ALF etiology) NOTE. Sensitivity and specicity in predicting liver transplantation or death were calculated by using a BiLE score cut-off value 6.9.
Methods Patients and Clinical Assessment

We retrospectively identied 210 patients with acute hepatic dysfunction treated at the intensive care unit of our institution between 1996 and 2005. Of those, 102 patients (Figure 1) fullled the diagnostic criteria of ALF as initially published by Trey and Davidson1 as well as those recently dened by the American Association for the Study of Liver Diseases (ie, hepatic encephalopathy, acute-onset increase of INR 1.5, and absence of signs of chronic liver disease in clinical and ultrasound examination).2 The study was approved by the human research committee of the Medical School Hannover, thereby conrming that the protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki. Informed consent for retrospective data acquisition was obtained from all patients whose current postal address was available or from patients next of kin in case of a fatal outcome. All parameters except the maximum encephalopathy grade were documented at admission to intensive care unit (ICU). Outcome end points included liver transplantation, death, and
survival without transplantation for at least 8 weeks after admission at ICU. Hepatic encephalopathy was graded on a standard scale of IIV as described previously.24 Simplied acute physiology score (SAPS) III, which is currently being evaluated as a predictive model in critical care patients, was calculated by using an EXCEL (Microsoft Corp, Redmond, WA) le provided by the SAPS-III Outcomes Research Group (www.saps3.org). KCC were determined as described.9 The calculation of MELD score was performed according to the equation: MELD score [9.57 loge creatinine (mg/dL) 3.78 loge bilirubin (mg/dL) 11.20 loge INR 6.43].25 In addition to the established predictive tools, we introduced a simple combined bilirubin-lactate-etiology score (or BiLE score) that can be calculated right at the beginning of the ICU stay. The calculation of BiLE score, which was developed empirically, is BiLE score (baseline bilirubin [mol/L]/100) baseline lactate [mmol/L] 4 [in case of indeterminate ALF, Budd-Chiari syndrome, or phenprocoumon toxicity] 2 [in case of acetaminophen toxicity] 0 [in case of any other ALF etiology] (Table 1). These parameters were then compared between the cohort of transplant-free survivors and patients who needed liver transplantation or died. Diagnoses were based on accepted diagnostic criteria that involved history, laboratory values, ultrasound imaging, and in individual cases histopathologic examination. Indeterminate ALF was assumed when the above mentioned diagnostic procedures including toxicology screens and serology for hepatitis A, B, and C, herpes simplex virus, varicella zoster virus, cytomegalovirus, and Epstein-Barr virus as well as autoantibodies were inconclusive. Decision to enroll a patient in the high urgency liver transplantation program was made according to the guidelines of the Eurotransplant International Foundation (see www.eurotransplant.nl for details).
All calculations were performed with SPSS software (version 13.0; SPSS Inc, Chicago, IL). Results are expressed as medians and ranges unless otherwise stated. Differences in discrete variables were tested with the 2 test or the Fisher exact test, wherever appropriate. Continuous variables were compared with the t test for normally distributed variables and with the Mann-Whitney test for non-normally distributed variables. Multivariate linear regression analysis was performed to evaluate prognostic value. Receiver operating characteristic (ROC) curves were calculated for the most relevant parameters.
Results Demographic Characteristics and Clinical Data

Figure 1. Outcome of 102 patients included in the retrospective analysis.
Of the 102 patients with acute liver failure, 72 (71%) were women. The median age of the group was 38 years (range,
March 2008
PROGNOSTIC IMPLICATIONS OF LACTATE, BILIRUBIN, AND ETIOLOGY IN ALF
341
Table 2. Characteristics of Patients at Admission According to Outcome

Variable Age (y) Women, n (%) Maximum encephalopathy grade I (n) II (n) III (n) IV (n) Prothrombin time (% of normal) INR WBC (per L) Hemoglobin (g/dL) Factor V (% of normal) Sodium (mmol/L) Creatinine (mol/L) ALT (U/L) AST (U/L) Alkaline phosphatase (U/L) Bilirubin (mol/L) Ammonia (mol/L) Lactate (mmol/L) MELD score KCC fullled, n (%) SAPS-III BiLE score BiLE score 6.9, n (%) Survival without transplantation (n 39) 37 (1674) 23 (59) II (IIV) 15 11 10 3 24 (875) 3.5 (1.27.4) 9.4 (4.623.5) 13.6 (8.019.9) 29 (4165) 136 (121147) 69 (33416) 2633 (5516,726) 2154 (4060,620) 151 (19333) 103 (10624) 41 (6185) 2.9 (0.613.1) 28 (1349) 7 (18) 54 (3998) 4.3 (014.7) 6 (16) Liver transplantation or death (n 63) 39 (1774) 49 (78) III (IIV) 8 11 25 19 20 (448) 4.0 (1.510.0) 11.2 (0.242.1) 12.6 (5.217.2) 20 (463) 137 (124153) 82 (281746) 1062 (1210,011) 769 (238122) 197 (8597) 263 (20848) 72 (16634) 4.7 (1.226.7) 34 (1462) 35 (58) 61 (42103) 9.7 (1.430.0) 50 (79) P value NS NS .001 NS .05 .01 .005 NS NS NS .01 .05 .04 .001 .005 .001 .001 .001 .004 .001 .001 95% Condence interval of differences 0.76.4 0.52.6 2492681 195 75220 1690 1.295.17 310 4.48.4
NOTE. All variables are expressed as medians (and ranges), unless otherwise stated. Abbreviation: WBC, white blood count.
16 74 years). Exact data on the duration of jaundice before onset of encephalopathy were available for 54 patients. The median interval between jaundice and onset of encephalopathy was 6 days (range, 0 40 days) in the overall cohort, 1 day (range, 0 39 days) in patients surviving without OLT (referred to as OLT-free survival), and 9 days (0 40 days) in those who underwent transplantation or died (referred to as OLT or death) (P .01). Median value for individual maximum hepatic encephalopathy grade was III (range, IIV) for the whole group, II (range, IIV) in case of OLT-free survival, and III (range, IIV) in case of OLT or death (P .001). A serum creatinine of 2 mg/dL (177 mol/L) or greater was present in 6 of 39 (15%) patients with OLT-free survival versus 17 of 63 (27%) patients with OLT or death (not signicant). Details are presented in Table 2.
lymphoblastic leukemia and another one with myelodysplastic syndrome who developed veno-occlusive disease after stem cell transplantation. An illustration of ALF etiologies is given in Figure 2.
Etiology of Acute Liver Failure

The distribution of etiologies of ALF was heterogeneous without predominance of one particular cause. Indeterminate ALF accounted for 21 patients (21%). Acute hepatitis B infection led to 18 cases (18%, 1 patient with hepatitis B/D coinfection), and 16 acetaminophen ingestions (16%) were observed. Another 9 patients (9%) had acute Budd-Chiari syndrome. Phenprocoumon toxicity and idiosyncratic drug reactions were deemed responsible for 7 (7%) and 5 (5%) cases of ALF, respectively. Diagnoses of Amanita phalloides ingestion, Wilsons disease, hepatitis A, ischemic hepatitis (shock liver), and halothane reaction accounted for 5 (5%), 5 (5%), 4 (4%), 4 (4%), and 3 (3%), respectively. Other causes were seen in 5 (5%) of ALF patients. Among them were one patient with acute
Figure 2. Causes and outcomes of ALF. Percentage of each etiology with regard to the whole study cohort is given on top of each bar.
342
HADEM ET AL
Table 3. Characteristics of Patients at Admission According to the 3 Most Common Etiologies

Variable Age (y) Women, n (%) Maximum encephalopathy grade Prothrombin time (% of normal) WBC (per L) Hemoglobin (g/dL) Coagulation factor V (% of normal) Sodium (mmol/L) Creatinine (mol/L) ALT (U/L) AST (U/L) Alkaline phosphatase (U/L) Bilirubin (mol/L) Ammonia (mol/L) Lactate (mmol/L) C-reactive protein (mg/L) MELD score KCC fullled, n (%) BiLE score BiLE score 6.9, n (%) Received transplant, n (%) Died, n (%) Indeterminate (n 21) 43 (2873) 18 (86) III (IIV) 22 (843) 11.0 (7.229.5) 12.9 (7.317.0) 21 (754) 138 (128153) 72 (28485) 1062 (757766) 803 (235389) 212 (114597) 310 (73660) 106 (24634) 4.1 (1.324.2) 9 (440) 34 (2350) 18 (86) 11.3 (7.129.8) 21 (100) 13 (62) 6 (29) Fulminant hepatitis B (n 18) 34 (2057) 11 (61) II (IIV) 19 (439) 8.6 (5.417.2) 13.3 (11.615.5) 28 (4165) 137 (123142) 65 (3482) 3345 (6669792) 1435 (2208108) 193 (19333) 214 (67624) 94 (6218) 4.3 (0.621.1) 22 (4148) 33 (2044) 2 (11) 6.9 (3.923.7) 8 (44) 9 (50) 4 (22) Acetaminophen (n 16) 37 (2052) 9 (56) II (IIV) 24 (848) 11.6 (5.635.3) 13.0 (6.319.9) 19 (742) 138 (128147) 87 (331746) 3497 (898471) 2474 (2229826) 179 (45331) 58 (120193) 83 (17496) 3.2 (0.926.7) 38 (2208) 27 (1362) 4 (25) 1.9 (025.2) 3 (19) 2 (13) 3 (19)
NOTE. All variables are expressed as medians (and ranges), unless otherwise stated. Abbreviation: WBC, white blood count.
Comparison of Clinical and Laboratory Features by Etiology

A comparison of the 3 most common etiologies of ALF in our cohort, ie, indeterminate ALF, hepatitis B, and acetaminophen ingestion, is given in Table 3. There is a notably high proportion of women (18 of 21, 86%) in indeterminate ALF. Patients with indeterminate ALF also had a higher maximum encephalopathy grade (median III) and higher bilirubin level, but ALT and C-reactive protein were markedly lower than in the total cohort, respectively. In contrast, acetaminophen toxicity affected men and women equally and was associated with higher ALT values. Although information on jaundice duration was available for 54 patients only, time interval between jaundice and onset of encephalopathy appeared to vary between groups.
death were multiorgan failure, septicemia, malignancies, and respiratory failure, among others. Eight-week outcome was strongly inuenced by the cause of ALF. Survival without transplantation was high in hepatitis A (4 of 4 patients, 100%), A phalloides toxicity (4 of 5, 80%), idiosyncratic drug reaction other than phenprocoumon (4 of 5, 80%), and acetaminophen ingestion (12 of 16, 75%). Low rates of survival without OLT were seen in Wilsons disease (0 of 5 patients, 0%), indeterminate ALF (1 of 21, 5%), Budd-Chiari syndrome (1 of 9, 11%), and phenprocoumon toxicity (1 of 7, 14%). Transplantation was performed in only 12% of the acetaminophen group as compared with 62% of the indeterminate ALF group. Women were more likely to require liver transplantation or die than men (Figure 1).
Outcome of Acute Liver Failure

As shown in Figure 1, overall survival 8 weeks after admission at ICU was 77% (79 of 102 patients). Thirty-nine patients (38%) survived without transplantation, and 45 (44%) received a liver graft. Five patients (5%) died after transplantation, 5 (5%) died awaiting a graft, and 13 (13%) died without being listed for transplantation. In the latter group, causes of
Predictive Factors for a Fatal Outcome

Several clinical and laboratory parameters as well as prognostic models emerged as discriminators between survival without OLT on the one hand and OLT or death on the other (Table 2). Lactate (cutoff 3.5 mmol/L), MELD score (cutoff 32), and KCC achieved a sensitivity and specicity of 59% and 66%, 65% and 69%, and 58% and 82%, respectively (Table 4). No
Table 4. Predicting Death or Need of Transplantation

Parameter Lactate Bilirubin MELD score KCC BiLE score Cut-off value 3.5 mmol/L 140 mol/L 32 6.9 Sensitivity (%) 59 67 65 58 79 Specicity (%) 66 64 69 82 84 Positive predictive value (%) 74 75 77 83 89 Negative predictive value (%) 49 54 55 55 71
March 2008
343
Figure 3. Lactate, MELD score, and BiLE score discriminate between survivors without liver transplantation and liver transplantation or death. Survivors without liver transplantation (white boxes) are compared with patients who required transplantation or died (grey boxes). Box plots show medians as well as interquartile ranges. Error bars show ranges.
statistical difference was observed for sodium, creatinine, prothrombin time, and factor V. Multivariate linear regression analysis revealed that bilirubin and lactate were the most predictive individual laboratory parameters in our patients. We attempted to create a prognostic score simple enough for bedside use that could be calculated right after ICU admission. Therefore, baseline bilirubin (mol/L) was divided by 100, thereby transferring the values of this parameter to the value range of baseline lactate (mmol/L). To take into account the predictive importance of ALF etiology, the sum of bilirubin/100 and lactate was then modied by adding positive or negative summands for certain etiologic entities. The novel combined BiLE score (calculation formula given in Table 1) achieved highest signicance in predicting liver transplantation or death (Figure 3; Table 4). With a cut-off value of 6.9, its sensitivity, specicity, and positive and negative predicitive values were 79%, 84%, 89%, and 71%, respectively. Sensitivity of BiLE score was particularly high (100%) in patients with indeterminate ALF. In addition, the performance of the BiLE score in comparison to other parameters was measured by ROC analysis. Area under the curve (AUC) values revealed a better performance of the BiLE score in comparison with bilirubin, lactate, KCC, MELD score, and the SAPS III score (Figure 4, Table 5).
Figure 4. ROC curves for prediction of death or need of liver transplantation. (A) Comparison of lactate and bilirubin with BiLE score. (B) Comparison of MELD score and SAPS-III with BiLE score. Abbreviations are shown in Table 2.
Discussion
ALF is estimated to cause approximately 3.5 deaths per million people per year in industrialized countries.26 The currently largest prospective study on ALF, which enrolled 308 patients in the United States, demonstrated a 43% chance of survival without OLT and an overall survival of 67%.5 Our study represented a large cohort of patients with acetaminophen and nonacetaminophen-induced ALF in central Europe. Of 102 patients, 38% survived without transplantation. The overall survival 8 weeks after ICU admission in our patient group was 77%, slightly higher than in the recently published U.S. study. This might be due to differences between the European (Eurotransplant) and U.S. (United Network for Organ Sharing) liver allocation systems (mortality on the waiting list was 5% in our
cohort [data not shown] versus 10% in the U.S. study).5 Obviously, our data might be subject to referral bias. Analyzing patients from a single center might on the other hand imply a more homogeneous patient cohort as a result of low interindividual variability regarding diagnostic criteria and standard of care. Finally, we did not include patients with acute exacerbations of chronic hepatitis B, although arguably this scenario might be difcult to distinguish from and overlap with fulminant hepatitis Binduced ALF.8 Recent studies have shown a remarkable geographic diversity of ALF etiologies.35,79,14,17 Our data found indeterminate ALF
Table 5. Predicting Death or Need of Transplantation: ROC

Parameter Lactate Bilirubin MELD score BiLE score SAPS-III n 96 102 101 101 87 AUC SE (95% CI) 0.73 0.52 (0.630.83) 0.73 0.51 (0.640.83) 0.71 0.05 (0.610.82) 0.87 0.04 (0.800.95) 0.68 0.59 (0.570.79) P value .001 .001 .001 .001 .005
Abbreviations: SE, standard error; 95% CI, 95% condence interval.
344
HADEM ET AL
(21%) and acute hepatitis B (18%) as the most common etiologic ALF entities, whereas acetaminophen ingestion and BuddChiari syndrome accounted for only 16% and 9%, respectively. In that sense, overall survival rates of greater than 75% are even more surprising, considering the low survival rate without OLT (17% in the U.S. study, only 5% in our study) in patients with ALF of indeterminate origin.5 The high incidence of indeterminate ALF in Germany is one of the most interesting results of this study. As in the U.S. ALF study, 71% of patients were women. But interestingly, women accounted for 86% of patients with indeterminate ALF. Whether women are innately more susceptible to ALF or are more commonly exposed to different kinds of drugs remains to be determined.27 Other statistically signicant features of indeterminate ALF were the low levels of liver enzymes, C-reactive protein, the relatively high bilirubin levels, and a higher encephalopathy grade. One can speculate that a high proportion of cases of indeterminate ALF take a subacute course with late ICU admission, which has been shown to be associated with poor prognosis in other reports7 as well as in this study. Furthermore, it would be interesting to know whether infectious complications contribute to encephalopathy in these patients.28 The survival rate without OLT of 38% in acute hepatitis B might be an underestimation of todays prognosis of these patients. Although several case reports have suggested that early treatment with lamivudine in patients with subfulminant or fulminant hepatitis B might reduce the need for life-saving liver transplantation,29 there is no agreement whether antiviral therapy should be instituted in this setting. In fact, a randomized placebo-controlled trial from India did not support lamivudine treatment in acute hepatitis B.30 In our center, patients with ALF have been treated with lamivudine since 2000 (ie, 12 of 18 patients with acute hepatitis B in the present cohort). Patients with ALF caused by phenprocoumon toxicity represented a unique feature of this German patient cohort. Phenprocoumon, the most widely used coumarin anticoagulant in Germany, has been associated with hepatitis and subacute liver failure (2% and 0.2%, respectively, of phenprocoumon-associated adverse events in Germany).31 The poor outcome in this group was probably inuenced by contraindications to transplantation imposed by the multimorbidity of these patients. The major aim of our study was to evaluate commonly used prognostic models in ALF, comparing transplant-free survivors with patients who underwent transplantation or died. Limitations come from the retrospective data analysis. A number of patients had to be excluded from the study because of incomplete documentation of hepatic encephalopathy at ICU admission. Our study was also limited by the fact that treatment and outcome might have changed during the 9-year study period; one example is the treatment of fulminant hepatitis B with lamivudine as discussed above.29 Finally, dening poor prognosis as death or liver transplantation might have implicated a bias related to the decision whether to perform transplantation in an individual patient. We found that in German ALF patients with a high proportion of indeterminate ALF and acute viral hepatitis maximum encephalopathy grade, bilirubin, lactate, KCC, and MELD score could all discriminate between transplant-free survivors and patients who needed liver transplantation or died. Our observation is in line with a recent study on 144 patients with ALF caused by acute viral hepatitis from India. The authors found
that the presence of 3 of 6 clinical parameters (among them, encephalopathy grade and jaundice-encephalopathy interval) and to a lesser extent also MELD and KCC discriminated survivors from nonsurvivors.32 The prognostic value of MELD in acetaminophen-induced liver injury has recently been investigated prospectively in 460 patients. MELD was not found to be superior over KCC or INR in predicting death.33 Two other studies in patients with non-acetaminophen ALF and seronegative hepatitis suggested that KCC might be of similar predictive strength as MELD, and that its application results in high patient survival rates.34,35 Multivariate linear regression analysis of our data revealed that bilirubin and lactate were the most predictive of all parameters evaluated. Although the prognostic value of lactate is well-recognized in acetaminophen-induced ALF,17,36 its clinical utility in nonacetaminophen-related ALF is less well-established because of the small patient number examined and etiologic heterogeneity of this group of ALF patients. Dabos et al20 studied 59 patients with nonacetaminophen-related ALF, of whom 34 underwent liver transplantation or died and had signicantly higher lactate levels. MacQuillan et al22 presented probability curves of nonsurvival on the basis of blood lactate levels in 27 patients with nonacetaminophen-related ALF. Baseline and 8-hour lactate levels did not differ signicantly between 7 spontaneous survivors and 20 patients who needed liver transplantation or died. Finally, hyperlactatemia was not considered to be of prognostic value in 63 patients with non acetaminophen-related ALF. However, the interpretation of this study by Taura et al21 is complicated by the fact that only liver transplant candidates were included. Because 86 ALF cases (84%) of our patient cohort were nonacetaminophen-related, our study added evidence to the assumption that lactate is among the best predictors of a fatal outcome even in non acetaminophen-related ALF. Bilirubin and lactate in combination with etiology were subsequently integrated into a novel, empirically developed prognostic score that can be calculated at bedside right after ICU admission. This combined BiLE score achieved the highest sensitivity and specicity in predicting poor outcome and was revealed to be superior over other parameters in the ROC analysis. The particular advantage of BiLE score might be its very high sensitivity in predicting death or transplantation in indeterminate ALF. It would therefore be interesting to learn more about the implications of this parameter in areas of the world in which non-acetaminophen ALF is predominant. In our opinion, BiLE score deserves to be validated in other centers with other patient cohorts. References
1. Trey C, Davidson CS. The management of fulminant hepatic failure. In: Popper H, Schaffner F, eds. Progress in liver diseases. New York: Grune and Stratton, 1970:282298. 2. Polson J, Lee WM. AASLD position paper: the management of acute liver failure. Hepatology 2005;41:1179 1197. 3. Goldberg E, Chopra S. Fulminant hepatic failure: denition, etiology, and prognostic indicators. In: Uptodate 2004, version 12.2. Available at: http://www.uptodateonline.com. Accessed on May 15, 2004. 4. OGrady JG, Portmann B, Williams R. Fulminant hepatic failure. In: Schiff L, Schiff R, eds. Diseases of the liver. Philadelphia: Lippincott, 1993:852 876. 5. Ostapowicz G, Fontana RJ, Schidt FV, et al. Results of a pro-
March 2008
345
6.
7.
8.
9.
10.
11. 12. 13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
spective study of acute liver failure at 17 tertiary care centers in the United States. Ann Intern Med 2002;137:947954. Bernuau J, Rueff B, Benhamou JP. Fulminant and subfulminant liver failure: denitions and causes. Semin Liver Dis 1986;6:97 106. OGrady JG. Acute liver failure. In: OGrady JG, Lake JR, Howdle PD, eds. Comprehensive clinical hepatology. London: Mosby, 2000:30.130.20. Bernal W, Schidt FV, Hamid S, et al. Etiologies and outcomes for acute liver failure at 6 sites around the world. Dig Dis Sci 2005; 24:A318. OGrady JG, Alexander GJM, Hayllar KM, et al. Early indicators of prognosis in fulminant hepatic failure. Gastroenterology 1989; 97:439 455. Renner EL. How to decide when to list a patient with acute liver failure for liver transplantation? Clichy or Kings College criteria, or something else? J Hepatol 2007;46:554 557. Lee WM. Acute liver failure in the United States. Semin Liver Dis 2003;23:217226. Escorsell A, Mas A, de la Mata M. Acute liver failure in Spain: analysis of 267 cases. Liver Transplant 2007;13:1389 1395. Anand AC, Nightingale P, Neuberger JM. Early indicators of prognosis in fulminant hepatic failure: an assessment of the Kings criteria. J Hepatol 1997;26:62 68. Bismuth H, Samuel D, Castaing D, et al. Orthotopic liver transplantation in fulminant and subfulminant hepatitis: the Paul Brousse experience. Ann Surg 1995;222:109 119. Izumi S, Langley PG, Wendon J, et al. Coagulation factor V levels as a prognostic indicator of fulminant hepatic failure. Hepatology 1996;23:15071511. Pauwels A, Mostefa-Kara N, Florent C, et al. Emergency liver transplantation for acute liver failure. J Hepatol 1993;17:124 127. Bernal W, Donaldson N, Wyncoll D, et al. Blood lactate as an early predictor of outcome in paracetamol-induced acute liver failure: a cohort study. Lancet 2002;359:558 563. Bihari D, Gimson AE, Lindridge J, et al. Lactic acidosis in fulminant hepatic failure. Some aspects of pathogenesis and prognosis. J Hepatol 1985;1:405 416. Murphy ND, Kodakat SK, Wendon JA, et al. Liver and intestinal lactate metabolism in patients with acute hepatic failure undergoing liver transplantation. Crit Care Med 2001;29:21112118. Dabos KJ, Newsome PN, Parkinson JA, et al. Biochemical prognostic markers of outcome in non-paracetamol-induced fulminant hepatic failure. Transplantation 2004;27:200 205. Taura P, Martinez-Palli G, Martinez-Ocon J, et al. Hyperlactatemia in patients with non-acetaminophen-related acute liver failure. World J Gastroenterol 2006;12:1949 1953. MacQuillan GC, Seyam MS, Nightingale P, et al. Blood lactate, but not serum phosphate levels can predict patient outcome in fulminant hepatic failure. Liver Transpl 2005;9:10731079.
23. Yantorno SET, Trentadue JJ, Ruf AE. The model for end stage liver disease (MELD): a useful tool to assess prognosis in fulminant hepatic failure. Liver Transl 2004;10:C36. 24. Lee WM. Acute liver failure. N Engl J Med 1993;329:18621872. 25. Wiesner R, Edwards E, Freeman R, et al. Model for end-stage liver disease (MELD) and allocation of donor livers. Gastroenterology 2003;124:9196. 26. Hoofnagle JH, Carithers RL, Shapiro C, et al. Fulminant hepatic failure: summary of a workshop. Hepatology 1995;21:240. 27. Third National Health and Nutrition Examination Survey 19881994, NHANES III Second Laboratory Data File (CD-ROM, series 11, no 2A), US Department of Health and Human Services, National Center for Health Statistics. Hyattsville, MD. Centers for Disease Control and Prevention, 1998. 28. Vaquero J, Polson J, Chung C, et al. Infection and the progression of encephalopathy in ALF. Gastroenterology 2003;125:755 764. 29. Tillmann HL, Hadem J, Leifeld L, et al. Safety and efcacy of lamivudine in patients with severe acute or fulminant hepatitis B, a multicenter experience. J Viral Hepat 2006;13:256 263. 30. Kumar M, Satapathy S, Monga R, et al. A randomized controlled trial of lamivudine to treat acute hepatitis B. Hepatology 2007; 45:97101. 31. Schimanski CC, Burg J, Mhler M, et al. Phenprocoumon-induced liver disease ranges from mild acute hepatitis to (sub-) acute liver failure. J Hepatol 2004;41:6774. 32. Dhiman RK, Jain S, Maheshwari U, et al. Early indicators of prognosis in fulminant hepatic failure: an assessment of the Model for End-Stage Liver Disease (MELD) and Kings College Hospital Criteria. Liver Transpl 2007; Mar 16 [epub ahead of print]. 33. Schmidt LE, Larsen FS. MELD score as a predictor of liver failure and death in patients with acetaminophen-induced liver injury. Hepatology 2007;45:789 796. 34. Katoonizadeh A, Decaestecker J, Wilmer A, et al. MELD score to predict outcome in adult patients with non-acetaminophen-induced acute liver failure. Liver Int 2007;27:329 334. 35. Wigg AJ, Gunson BK, Mutimer DJ. Outcomes following liver transplantation for seronegative acute liver failure: experience during a 12-year period with more than 100 patients. Liver Transpl 2005;1:2734. 36. Schmidt LE, Larsen FS. Prognostic implications of hyperlactatemia, muliple organ failure, and systemic inammatory response syndrome in patients with acetaminophen-induced acute liver failure. Crit Care Med 2006;34:337.
Address requests for reprints to: Johannes Hadem, MD, Department of Gastroenterology, Hepatology, and Endocrinology, Medical School Hannover, Carl-Neuberg-Strasse 1, 30625 Hannover, Germany. e-mail: hadem.johannes@mh.hannover.de; fax: 49-511-532-4896.
Bleeding Stomal Varices: Case Series and Systematic Review of the Literature
BRET J. SPIER,* ABDULLAH A. FAYYAD,* MICHAEL R. LUCEY,* ERIC A. JOHNSON,* MYRON WOJTOWYCZ, LAYTON RIKKERS, BRUCE A. HARMS, and MARK REICHELDERFER*
*Section of Gastroenterology and Hepatology, Department of Medicine, Department of Radiology, and the Department of Surgery, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin
Background & Aims: Bleeding stomal varices are a

common problem in patients with surgical stomas and portal hypertension, and remain difcult to diagnose and manage. Methods: We identied all patients at our institution with bleeding stomal varices from 1989 to 2004. We surveyed all patients undergoing ileal pouchanal anastomosis from 1997 to 2007 for bleeding anastomotic varices. Finally, we performed a systematic review of the literature focusing on diagnosis and treatment of bleeding stomal varices that included 74 English language studies of 234 patients. Results: We identied 8 patients with bleeding stomal varices. Recognition of stomal varices typically was delayed, particularly when failing to examine the ostomy without the appliance. Stomal variceal bleeding was conrmed by Doppler ultrasound or angiographic imaging. Simple local therapy usually stopped bleeding, albeit temporarily. Sclerotherapy was effective, but at the expense of unacceptable stomal damage. Decompressive therapy was required for secondary prophylaxis, including transjugular intravascular transhepatic shunts (2 patients), surgical portosystemic shunts (2 patients), and liver transplantation (1 patient). No patient with an ileal pouchanal anastomosis developed anastomotic bleeding from varices. Conclusions: Primary prevention of bleeding stomal varices requires avoidance of creating enterocutaneous stomas in patients with portal hypertension. Careful inspection of the uncovered ostomy is essential for bleeding stomal varices diagnosis. Once identied, conservative measures will stop bleeding temporarily with denitive therapy required, including transjugular intravascular transhepatic shunts, surgical shunts, or liver transplantation. lthough varices arising at sites other than the gastroesophageal junction account for only 5% of all variceal bleeding, bleeding stomal varices (BSV) at surgically created mucocutaneous anastomoses, are a common cause of recurrent hemorrhage in patients with portal hypertension and surgical stomas.1 Previous reports of the incidence, natural history, and management of BSV have been anecdotal or small series with only short-term follow-up periods.2 68 In this article, we combine a report of our case series of BSV patients and a survey of our patients who received an ileal pouchanal anastomosis (IPAA) with a systematic review of the literature on stomal varices.
liam S. Middleton Memorial Veterans Hospital (Madison, WI) from 1989 to 2004. Eight patients were identied (Table 1), with records retrieved from the respective electronic charts after institutional review board approval. If patients were lost to follow-up evaluation, we used the social security database to determine health status (alive vs deceased) as of June 18, 2007.69 To examine the possibility that IPAA reconstruction (as opposed to permanent ileostomy) reduces the risk of bleeding from portal hypertension, we reviewed all IPAA procedures performed from 1997 to 2007. Charts were reviewed for evidence of gastrointestinal hemorrhage in the presence of portal hypertension. Surrogate markers for portal hypertension included radiographic imaging revealing cirrhotic-appearing liver, varices, splenomegaly, recanalized umbilical vein, portal venous thrombosis, or clinical history of chronic liver disease because direct pressure measurements were not obtained.
Search Strategy for Systematic Review

Relevant studies were identied by searching the following databases on July 3, 2007: Medline, Cochrane Database of Clinical Trails, and Cumulative Index to Nursing and Allied Health Literature (Figure 1). The databases were searched using the following key words or their combinations: stoma, stomal, ectopic, ileostomy, colostomy, bleed, bleeding, hemorrhage, portal hypertension, cirrhosis, end stage liver disease. In addition, we reviewed reference lists from all included studies and consulted experts. The review included all English language abstracts and manuscripts, and was performed by a single researcher, with previous less formal searches performed on multiple occasions. We included all study designs. Two of the authors (BJS and MR) independently reviewed the titles and abstracts of available literature for possible inclusion in the review. A data collection form was used to extract study time, intervention, treatment, and various outcome measures.
Results Case Series

Demographics. There were 4 men and 4 women with a mean age at onset of bleeding of 54 years (range, 2174 y); 7 were from the University of Wisconsin Hospital and Clinics and 1 was from the William S. Middleton Memorial Veterans HosAbbreviations used in this paper: BSV, bleeding stomal varices; CT, computerized tomography; IPAA, ileal pouchanal anastomosis; TIPS, transjugular intrahepatic portosystemic shunting. 2008 by the AGA Institute 1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.047
Methods
We retrieved the records of all patients with BSV seen at the University of Wisconsin Hospital and Clinics and the Wil-
March 2008
STOMAL VARICES
347
Table 1. Patient Characteristics

Patient (sex) 1 (F) 2 (F) 3 (F) 4 (M) 5 (M) 6 (M) 7 (F) Cause of PVH AIH PSC HCV HBC, HCV NASH HCC with PVT PVT caused by pylephlebitis Type of ostomy Ileostomy Ileostomy Ileostomy End colostomy End sigmoid colostomy Loop ileostomy and colostomy Transverse loop colostomy Age when ostomy was placed 32 38 32 17 73 47 54 Reason for ostomy Total colectomy for UC Total colectomy for UC Total colectomy for UC HIV lymphoma with visceral perforation APR for rectal cancer Bladder cancer invading rectum Streptococcus viridans septicemia with multiple abdominal abscesses Total colectomy for diverticular bleeding Time with ostomy before stomal variceal bleed 23 y 10 y 6y 4y 1 mo 11 y 9y
8 (M)
Cardiac cirrhosis Ileostomy
64
5y
AIH, Autoimmune hepatitis; APR, abdominoperitoneal resection; F, female; HBC, hepatitis B cirrhosis; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; HIV, human immunodeciency virus; M, male; NASH, nonalcoholic steatohepatitis; PSC, primary sclerosing cholangitis; PVH, portal venous hypertension; PVT, portal venous thrombosis; UC, ulcerative colitis.
pital. Cirrhosis caused portal hypertension in 6 of 8 patients, whereas the other 2 had portal vein thrombosis. The rst 3 patients in the series had ileostomal varices arising from permanent ileostomies placed after colectomy for ulcerative colitis. The indications for ileostomy at our institutions then changed after 1984 when IPAA replaced Brooke ileostomy for reconstruction after total colectomy (Table 1). Overall, 9 stomas were placed in our 8 patients: 4 patients had ileostomies, 3 had colostomies, and 1 patient had both a colostomy and an ileal conduit. The mean interval from stomal surgery to rst bleeding was 8.5 years (range, 1 mo to 23 y). Clinical presentation and physical examination. All patients presented with gastrointestinal bleeding, which often was recurrent and massive. There was a history of stomal variceal bleeding (made retrospectively) before diagnosis in 6 of 8 patients with a mean delay in diagnosis of 24 months (range, 1 mo to 13 y; Table 2). The diagnosis of portal hypertension was known in 5 patients but not in the other 3 patients. Removal of the appliance, a maneuver often omitted during previous episodes of bleeding, was necessary to see the full extent of the typical surface changes of hypertensive stomopathy (see later). Before diagnosis, all patients had signicant bleeding that often
was rapid and in one case lead to shock requiring rapid resuscitation: all patients required a blood transfusion (217 units per patient). Spurting or gushing of blood was reported in 6 patients. Indeed, the clinical record reported that blood spurted for a distance of 2 feet in 1 patient after appliance removal. Before the diagnosis of BSV, the treating physicians often attributed bleeding to stomal irritation or granulation tissue. The appearance of the stomas and surrounding skin was abnormal at the time of diagnosis in all patients; changes that were obvious once the appliance was removed. Typically, the enteric side was congested with a raspberry hue with the skin side showing a purple discoloration and dilated veins giving the appearance of protruding blue ridges. A large varix was never seen on the mucosal surface of the stoma. Bleeding could be provoked by stomal manipulation and was observed to be of high pressure in all instances. Imaging. Doppler ultrasound ndings included multiple, enlarged (2 mm) stomal veins with abnormal ow (Ta-
Table 2. Duration of Prior Bleeding and Techniques Used to Diagnose BSV

Bleeding before Patient (sex) denitive diagnosis 1 (F) 2 (F) 3 (F) 4 (M) 5 (M) 6 (M) Yes, 1 mo No Yes, 13 y Yes, 4 mo Yes, 7 mo Yes, 1 mo Modality used to diagnose stomal varices Doppler ultrasound Spurting 612 in; raspberry stoma; Doppler ultrasound Magnetic resonance angiogram; Doppler ultrasound Doppler ultrasound; spurting with needle puncture Tagged red blood cell scan; traditional angiogram CT angiogram showed PVT; spurting at home with ostomy change CT angiogram and traditional angiogram Vigorous venous oozing; Doppler ultrasound
7 (F) 8 (M)
Yes, 2 y No
Figure 1. Selection of the articles for the systematic review.
F, female; M, male; PVT, portal venous thrombosis.
348
SPIER ET AL
Figure 2. (A) CT angiogram performed 2 years after the rst episodes of stomal bleeding in patient 7. The study was performed during selective injection of iodinated contrast medium into her superior mesenteric artery with scans obtained during the portal venous phase. Tortuous dilated veins are present at the left-sided ileostomy. (B) Reprojected CT image from the same study during the portal venous phase (anteroposterior view of the abdomen). The patient has chronic occlusion of the main portal vein, immediately superior to the superior mesenteric vein (open arrow). The veins of the stoma (solid arrow) are enlarged, congested, and in communication with multiple prominent body wall collateral vessels (curved arrows).
ble 2). Computerized tomography (CT) scanning was diagnostic in 2 patients, showing portal hypertension and stomal varices in both (Figure 2). A tagged red blood cell scan localized active bleeding to the stoma in 1 patient with subsequent percutaneous angiography conrming stomal varices. Magnetic resonance angiography conrmed the diagnosis as well as dened the vascular anatomy in 1 patient being considered for liver transplantation. Upper- and lower-gastrointestinal endoscopic procedures were performed in all patients (often on multiple occasions), with internal ileal varices seen in none of our patients; gastroesophageal varices were found in 2 patients, leading to the diagnosis of portal hypertension. Treatment. At least temporary hemostasis was established by pressure with sandbags, cautery with silver nitrate, having the patient assume a recumbent position, and local suturing. One patient repeatedly cauterized his stoma with silver nitrate at home.
Pharmacotherapy including octreotide for acute bleeding and propranolol for prevention was tried in 2 patients (Table 3). Freehand sclerotherapy of the stoma was performed in 3 patients, using 5% sodium morrhuate in 0.5- to 2-mL amounts per injection (up to 5 mL total).66 Needle injection often provoked high-pressure venous bleeding. In 1 patient, injection sites were selected and marked by ultrasound.66 Sclerotherapy invariably produced signicant stomal damage, including stricturing of the stoma, requiring dilation and eventually causing retraction of the stoma, poor sealing of the appliance, and fecal leakage. In 1 patient, pain after injection beyond the squamocolumnar junction required hospitalization. Subsequent injections on the mucosal side caused no pain, but produced signicant stomal damage. Percutaneous embolization of a bleeding varix was performed in 2 patients in whom local therapies failed. Decompressive therapy was required for secondary prophylaxis in 4 patients, including transjugular intrahepatic portosystemic shunting (TIPS) (2 patients), surgical shunts (2 patients), and liver transplantation (1 patient previously undergoing TIPS). Survival and outcomes. The mean follow-up period after diagnosis was 5 years (range, 1 mo to 25 y). None of our patients died from bleeding, although 7 died from underlying liver disease (Table 3).70 Overall survival after the rst bleeding episode ranged from 5 months to 25 years. Two patients received local therapy only. One patient, who had inoperable metastatic hepatocellular carcinoma and portal vein thrombosis, lived for 5 months with repeated surgical suturing and self-administration of silver nitrate cautery. The second patient refused further therapy and died of undetermined causes after 2 years. Among those treated with sclerotherapy, 1 patient survived almost 2 years while undergoing sclerotherapy every 3 months. Another patient survived 4.5 years with the combination of maintenance propranolol plus sclerotherapy and subcutaneous octreotide for acute bleeding. Scheduled sclerotherapy every 1 to 2 months controlled bleeding for 3 years in the third patient until recurrent BSV and sclerotherapy-induced stomal damage required TIPS. Denitive therapy was used in 4 patients. The rst patient (patient 3) underwent TIPS without further bleeding until liver failure necessitated liver transplantation. This patient survived 5 years after transplant and 25 years after the initial stomal variceal bleed. The second patient (patient 5) underwent transhepatic coil embolization of the stomal varices along with TIPS, and survived 1.5 years after the initial bleed. Two patients underwent shunt surgery for portal hypertension, including a successful end-to-side portocaval shunt (patient 1), which was patent at 4 years, and a mesocaval shunt (patient 7) that failed as a result of shunt thrombosis.
Anastomotic Bleeding in Ileal PouchAnal Anastomosis Recipients

We identied 489 patients with IPAAs who were observed for a mean of 6.5 years (range, 19.5 y), including 10 patients with surrogate markers of portal hypertension (4 had primary sclerosing cholangitis). Pouch variceal bleeding did not occur in any patient.
March 2008
STOMAL VARICES
349
Table 3. Acute and Chronic Therapy With Complication and Survival Data
Patient (sex) 1 (F) 2 (F) Treatment at rst presentation Pressure with cautery Octreotide drip 50 mcg/h Therapy (in chronologic order) End-to-side portacaval shunt Propranolol 60 mg orally twice a day; sclerotherapy with ultrasound guidance; pressure; self-administered octreotide 50 mg as necessary Surgical revision of ostomy 2; propranolol 80 mg orally twice a day; TIPS; OLTx Sclerotherapy (monthly) TIPS Pressure and self-applied cautery with silver nitrate Mesocaval shunt Posttreatment complications None Sclerotherapy caused signicant scarring/stricturing requiring dilation, leading to retraction of the stoma and the bag not tting Appliance did not t because of sclerotherapy Pain with sclerotherapy requiring PCA None None Thrombosis of mesocaval shunt with subsequent rebleeding from stomal varices None Alive vs dead Dead Dead Survival after rst stomal variceal bleed 4y 4 y 7 mo
3 (F)
Sclerotherapy with cautery 2 Sclerotherapy Embolization with coiling Pressure and cautery; surgical suturing Embolization with coiling Pressure
Dead
25 y 7 mo
4 (M) 5 (M) 6 (M) 7 (F)
Dead Dead Dead Alive
1 y 10 mo 1 y 7 mo 5 mo 5 y 3 mo
8 (M)
Pressure
Dead
2 y 4 mo
OLTx, orthotopic liver transplant; PCA, patient-controlled analgesia.
Systematic Review
There were 74 case reports and case series consisting of 234 patients (Table 4). Imaging. Doppler ultrasound, CT, and magnetic resonance angiography may identify large variceal collections in the region of the stoma, aid the diagnosis of cirrhosis and/or portal hypertension, and show patency of the portal venous system.11,26,27,30,37,58 There were 14 reports of percutaneous angiography showing stomal varices as well as localization of active bleeding.2,8,12,13,15,18,21,22,25,31,46,50,53,56 Treatment and outcome. The systematic review conrmed that local therapy is effective as initial treatment of acute hemorrhage. Local pressure and recumbent body positioning provide rapid control of BSV.43,46 Surgical suturing, ligation, and cautery are useful approaches if the bleeding site can be visualized and have been documented in 27 patients in 16 reports.2,4,8,11,14,17,18,26,29,43,45,54,57,62,64,66 Data regarding the use of -blockers for secondary prophylaxis were limited to 19 patients in 8 reports; -blockers were used as monotherapy with subsequent bleeding occurring in most instances.16,30,32,42,43,58,65,67 Octreotide was not reported in the systematic review. There were 9 reports of sclerotherapy used in 14 patients.4,17,28,38,40,57,58,64,66 One report described visualization and marking of varices using Doppler ultrasound before sclerotherapy. There were no reports of injury to the stoma by repeated injection of sclerosant. There were 20 reports of local surgical approaches, including stomal revision or relocation, often without long-term follow-up evaluation.2,4,6,14 16,19,24,27,29,38,43,45,47,52,53,55,57,66,67 The more extensive local procedure of mucocutaneous disconnection was documented in 3 reports comprising at least 13 patients.6,15,24 Recurrent bleeding was common after these procedures. Placement of surgical shunts (including mesocaval, portocaval, and splenorenal shunts) for control of bleeding and portal
decompression was documented in 13 reports comprising 21 patients.2,8,12,13,21,22,46,49,50,71 In patients with well-compensated cirrhosis (ChildPugh A or low Model for End-Stage Liver Disease), shunting reduced the need for repeated intervention.72 Therapy with coil embolization after direct angiography allowed for acute control of bleeding, although vessel recanalization or new stomal variceal formation eventually mandated further intervention.3,31,34,35,49,56,61 The use of coil embolization at the time of TIPS placement was reported in 3 patients, but there was no evidence to suggest this technique is better than TIPS alone.3,26,61 TIPS has controlled bleeding successfully from stomal varices in 19 reports comprising 40 patients.3,4,7,9,10,26,27,30,33,39,41,48,51,59 61,64,67,68 For example, Vidal et al60 reported successful therapy in stomal varices in 7 of 8 patients, 1 of whom required shunt dilation for recurrent bleeding and 1 of whom eventually underwent surgical portacaval shunt; the overall cumulative rate of rebleeding was 23% and 31% at 1 and 2 years, respectively. Inter-
Table 4. Our Study Data Totaled With the Data Compiled From the Systematic Review
Our study Patients, n Conservative therapy Sclerotherapy Surgical suturing, ligation, or cautery -blocker Recumbent positioning Endovascular therapy TIPS Coil embolization Surgical shunt Liver transplant Deaths caused by bleeding 8 3 1 2 1 2 2 2 1 0 Systematic review 226 14 27 19 2 40 18 21 13 6 Total 234 17 28 21 3 42 20 23 14 6
350
SPIER ET AL
estingly, there were several reports of rebleeding by ectopic varices after TIPS despite lowering the portosystemic gradient to less than 12 mm Hg.59,60,64 Finally, our experience, along with 8 reports of 13 patients, conrmed that liver transplantation controls BSV successfully.4,6,20,35,64 66
Discussion
Stomal varices, which caused signicant morbidity, multiple hospitalizations, and blood transfusions, were rst described in 1968 in a study of the use of surgical colon exclusion to treat hepatic encephalopathy.45 Subsequently, there were case reports and small series leading to a misperception that BSV is a rare phenomenon.21,56,73 This orthodoxy was challenged by several reports from the Mayo Clinic; with 1 study reporting a 27% incidence (17 of 62 patients) of BSV.20,43,65 The Mayo Clinic investigators described the following risk factors for stomal varices in primary sclerosing cholangitis patients with ostomies who developed BSV65: at least stage 3 brosis on liver biopsy and splenomegaly, esophageal varices, and ascites. Internal anastomoses did not lead to bleeding varices in the 4 patients from the rst Mayo series who underwent IPAA reconstruction after colectomy in the presence of signicant portal hypertension (2 had splenomegaly and 1 had esophageal varices). Additional reports from the same institution of 40 patients with primary sclerosing cholangitis and IPAA followed up for 21 months revealed no ileal varices or bleeding from the pouch.74,75 Our study provides further evidence that primary prevention of stomal variceal formation is achievable when stomas are avoided in patients with known portal hypertension, and suggests that the unique venous anatomy of stomas (owing to the mucocutaneous connection) is what leads to the major risk of variceal bleeding. BSV may be missed. Visual inspection of the stoma after removing the appliance is critical and often is overlooked. Doppler ultrasound should be the rst imaging study ordered and usually will conrm the diagnosis of stomal varices, with evaluation of the liver and portal venous system allowing diagnosis of underlying cirrhosis or portal venous thrombosis. CT or magnetic resonance also will show stomal variceal collections and provide important additional anatomic information. Local therapies (including local pressure, positioning of the patient, suturing, or cautery) are effective at acute control of BSV and are always the rst step. Our study shows that certain types of secondary prophylaxis have unacceptable morbidity. First, sclerotherapy should be avoided on account of stomal damage and/or recurrent BSV in nearly all patients. Second, based on our systematic review, we suggest that mucocutaneous disconnection and surgical relocation of the stoma should not be used on account of recurrent bleeding and the perioperative surgical risk. Finally, data from our series, as well as the systematic review, suggest that pharmacotherapy is ineffective as monotherapy. Although the use of subcutaneous octreotide was not described in the systematic review, our experience suggests that it may reduce acute hemorrhage and should be considered for intermittent use in patients who are not candidates for denitive therapy. TIPS has become the preferred treatment modality in most patients in need of portal decompression, with surgical shunting reserved for TIPS failures.72 TIPS is also an effective treat-
Figure 3. Suggested algorithm for management of bleeding stomal varices. Epi, epinephrine; MELD, Model for End-Stage Liver Disease.
ment for BSV. However, there are reports of bleeding even when the trans-sinusoidal pressure gradient has been reduced to less than 12 mm Hg.59,60,64 These data suggest that the threshold portal pressure gradient leading to BSV may be lower than for esophageal or gastric varices. Liver transplantation is appropriate for patients with decompensated liver failure and BSV. Based on our data, and our systematic review of the literature, we have proposed a treatment algorithm for management of patients with bleeding from stomal varices (Figure 3). References
1. Kinkhabwala M, Mousavi A, Iyer S, et al. Bleeding ileal varicosity demonstrated by transhepatic portography. AJR Am J Roentgenol 1977;129:514 516. 2. Adson MA, Fulton RE. Ileal stoma and portal-hypertension uncommon site of variceal bleeding. Arch Surg 1977;112:501 504. 3. Ahari HK, Feldman L, Kaufman JA, et al. Vascular and interventional case of the day case 1: peristomal varices. AJR Am J Roentgenol 1999;173:829 832. 4. Alkari B, Shaath NM, El-Dhuwaib Y, et al. Transjugular intrahepatic porto-systemic shunt and variceal embolisation in the management of bleeding stomal varices. Int J Colorectal Dis 2005; 20:457 462. 5. Arcidiaco M, Dicorato G, Pasqualin G, et al. Portal hypertension and enterostomy: a dangerous combination. Colostomy-induced varices as a rare cause of gastrointestinal hemorrhage. Report of a case and review of the literature. Ann Ital Chir 1999;70:61 65. 6. Beck DE, Fazio VW, Grundfestbroniatowski S. Surgical-management of bleeding stomal varices. Dis Colon Rectum 1988;31: 343346. 7. Bernstein D, Yrizarry J, Reddy KR, et al. Transjugular intrahepatic portosystemic shunt in the treatment of intermittently bleeding stomal varices. Am J Gastroenterol 1996;91:22372238. 8. Cameron AD, Fone DJ. Portal hypertension and bleeding ileal varices after colectomy and ileostomy for chronic ulcerative colitis. Gut 1970;11:755759. 9. Carraello G, Lagana D, Giorgianni A, et al. Bleeding from peristomal varices in a cirrhotic patient with ileal conduit: treatment with transjugular intrahepatic portocaval shunt (TIPS). Emerg Radiol 2007;13:341343. 10. Chavez DR, Snyder PM, Juravsky LI, et al. Recurrent ileal conduit hemorrhage in an elderly cirrhotic man. J Urol 1994;152:951 953.
March 2008
STOMAL VARICES
351
11. Choi JW, Lee CH, Kim KA, et al. Ectopic varices in colonic stoma: MDCT ndings. Korean J Radiol 2006;7:297299. 12. Cooper MJ, Mackie CR, Dhorajiwala J, et al. Hemorrhage from ileal varices after total proctocolectomy. Am J Surg 1981;141: 178 179. 13. Crooks KK, Hensle TW, Heney NM, et al. Ileal conduit hemorrhage secondary to portal hypertension. Urology 1978;12:689 693. 14. Eade MN, Williams JA, Cooke WT. Bleeding from an ileostomy caput medusae. Lancet 1969;2:1166 118. 15. Eckhauser FE, Sonda LP, Strodel WE, et al. Parastomal ileal conduit hemorrhage and portal-hypertension. Ann Surg 1980; 192:620 624. 16. Farquharson AL, Bannister JJ, Yates SP. Peristomal variceslife threatening or luminal? Ann R Coll Surg Engl 2006;88:W6 W8. 17. Finemore RG. Repeated haemorrhage from a terminal colostomy due to mucocutaneous varices with coexisting hepatic metastatic rectal adenocarcinoma: a case report. Br J Surg 1979;66:806. 18. Firlit RS, Firlit CF, Canning J. Exsanguinating hemorrhage from urinary ileal conduit in patient with portal hypertension. Urology 1978;12:710 711. 19. Foulkes J, Wallace DM. Hemorrhage from stomal varices in an ileal conduit. Br J Urol 1975;47:630. 20. Fucini C, Wolff BG, Dozois RR. Bleeding from peristomal varices: perspectives on prevention and treatment. Dis Colon Rectum 1991;34:10731078. 21. Goldstein MB, Brandt LJ, Bernstein LH, et al. Hemorrhage from ileal varices: a delayed complication after total proctocolectomy in a patient with ulcerative colitis and cirrhosis. Am J Gastroenterol 1983;78:351354. 22. Goldstein WZ, Edoga J, Crystal R. Management of colostomal hemorrhage resulting from portal-hypertension. Dis Colon Rectum 1980;23:86 90. 23. Graeber GM, Ratner MH, Ackerman NB. Massive hemorrhage from ileostomy and colostomy stomas due to mucocutaneous varices in patients with coexisting cirrhosis. Surgery 1976;79: 107110. 24. Grundfestbroniatowski S, Fazio V. Conservative treatment of bleeding stomal varices. Arch Surg 1983;118:981985. 25. Hamlyn AN, Morris JS, Lunzer MR, et al. Portal-hypertension with varices in unusual sites. Lancet 1974;2:15311534. 26. Han SG, Han KJ, Cho HG, et al. A case of successful treatment of stomal variceal bleeding with transjugular intrahepatic portosystemic shunt and coil embolization. J Korean Med Sci 2007; 22:583587. 27. Handschin AE, Weber M, Weishaupt D, et al. Contrast-enhanced three-dimensional magnetic resonance angiography for visualization of ectopic varices. Dis Colon Rectum 2002;45:15411544. 28. Hesterberg R, Stahlknecht CD, Roher HD. Sclerotherapy for massive enterostomy bleeding resulting from portal hypertension. Dis Colon Rectum 1986;29:275277. 29. Hollands MJ. Parastomal hemorrhage from an ileal conduit secondary to portal-hypertension. Br J Surg 1982;69:675. 30. Johnson PA, Laurin J. Transjugular portosystemic shunt for treatment of bleeding stomal varices. Dig Dis Sci 1997;42:440 442. 31. Kishimoto K, Hara A, Arita T, et al. Stomal varices: treatment by percutaneous transhepatic coil embolization. Cardiovasc Intervent Radiol 1999;22:523525. 32. Konate A, Oberti F, Aube C, et al. Stomal varices treated by glue embolization. Gastroenterol Clin Biol 2007;31:300 302. 33. Lagier E, Rousseau H, Maquin P, et al. Treatment of bleeding stomal varices using transjugular intrahepatic portosystemic shunt. J Pediatr Gastroenterol Nutr 1994;18:501503. 34. Lashley DB, Saxon RR, Fuchs E, et al. Bleeding ileal conduit stomal varices: diagnosis and management using transjugular transhepatic angiography and embolization. Urology 1997;50: 612 614.
35. Macedo TA, Andrews JC, Kamath PS. Ectopic varices in the gastrointestinal tract: short- and long-term outcomes of percutaneous therapy. Cardiovasc Intervent Radiol 2005;28:178 184. 36. Mallick IH, Pearson HJ. Quiz HQ 36. An exceptional cause of bleeding from stoma. Varices around mucus stula as a result of portal hypertension complicating primary biliary cirrhosis. J Gastrointest Liver Dis 2007;16:83, 115. 37. Minami S, Okada K, Matsuo M, et al. Treatment of bleeding stomal varices by balloon-occluded retrograde transvenous obliteration. J Gastroenterol 2007;42:9195. 38. Morgan TR, Feldshon SD, Tripp MR. Recurrent stomal variceal bleeding. Successful treatment using injection sclerotherapy. Dis Colon Rectum 1986;29:269 270. 39. Morris CS, Najarian KE. Transjugular intrahepatic portosystemic shunt for bleeding stomal varices associated with chronic portal vein occlusion: long-term angiographic, hemodynamic, and clinical follow-up. Am J Gastroenterol 2000;95:2966 2968. 40. Mosquera DA, Walker SJ, McFarland JB. Bleeding stomal varices treated by sclerotherapy. J R Coll Surg Edinb 1988;33:337. 41. Nayar M, Saravanan R, Rowlands PC, et al. TIPSS in the treatment of ectopic variceal bleeding. Hepatogastroenterology 2006; 53:584 587. 42. Noubibou M, Douala HC, Druez PM, et al. Chronic stomal variceal bleeding after colonic surgery in patients with portal hypertension: efcacy of beta-blocking agents? Eur J Gastroenterol Hepatol 2006;18:807 808. 43. Peck JJ, Boyden AM. Exigent ileostomy hemorrhage. A complication of proctocolectomy in patients with chronic ulcerative colitis and primary sclerosing cholangitis. Am J Surg 1985;150:153 158. 44. Poilleux J, Damamme B. Colostomy-induced varices in portal hypertension. Chirurgie 1992;118:252258. 45. Resnick RH, Ishihara A, Chalmers TC, et al. A controlled trial of colon bypass in chronic hepatic encephalopathy. Gastroenterology 1968;54:10571069. 46. Ricci RL, Lee KR, Greenberger NJ. Chronic gastrointestinal-bleeding from ileal varices after total proctocolectomy for ulcerativecolitis correction by mesocaval shunt. Gastroenterology 1980; 78:10531058. 47. Roberts PL, Martin FM, Schoetz DJ, et al. Bleeding stomal varicesthe role of local treatment. Dis Colon Rectum 1990;33: 547549. 48. Ryu RK, Nemcek AA, Chrisman HB, et al. Treatment of stomal variceal hemorrhage with TIPS: case report and review of the literature. Cardiovasc Intervent Radiol 2000;23:301303. 49. Samaraweera RN, Feldman L, Widrich WC, et al. Stomal varicespercutaneous trans-hepatic embolization. Radiology 1989; 170:779 782. 50. Scaletscky R, Wright JK, Shaw J, et al. Ileal conduit venous varices from portal-hypertension as a cause of recurrent, massive hemorrhage case-report and review of the literature. J Urol 1994;151:417 419. 51. Shibata D, Brophy DP, Gordon FD, et al. Transjugular intrahepatic portosystemic shunt for treatment of bleeding ectopic varices with portal hypertension. Dis Colon Rectum 1999;42:15811585. 52. Smith S, Wiener ES, Starzl TE, et al. Stoma-related variceal bleeding: an under-recognized complication of biliary atresia. J Pediatr Surg 1988;23:243245. 53. Stansby G, Meyrick-Thomas J, Lewis AA. Pericolostomy varices. J R Coll Surg Edinb 1990;35:109 110. 54. Thomas DJ, Abercrombie GF. Peri-stomal varicesan unusual cause of bleeding from an ileal conduit. Br J Urol 1993;71:355 356. 55. Thompson J. Caput medusae: peristomal varices. J ET Nurs 1993;20:216 219.
352
SPIER ET AL
56. Toumeh KK, Girardot JD, Choo IW, et al. Percutaneous transhepatic embolization as treatment for bleeding ileostomy varices. Cardiovasc Intervent Radiol 1995;18:179 182. 57. Tricas JML, Celaya GZ, Lizoain JLH. Colostomal variceal hemorrhage in a patient with portal-hypertension. Rev Esp Enferm Dig 1991;79:437 438. 58. Van der Wouden EJ, D Westerveld B. Photo quiza patient with diabetes mellitus and recurrent peristomal bleedingwhat is your diagnosis? Portal hypertension. Neth J Med 2006;64:314. 59. Vangeli M, Patch D, Terreni N, et al. Bleeding ectopic varices treatment with transjugular intrahepatic porto-systemic shunt (TIPS) and embolisation. J Hepatol 2004;41:560 566. 60. Vidal V, Joly L, Perreault P, et al. Usefulness of transjugular intrahepatic portosystemic shunt in the management of bleeding ectopic varices in cirrhotic patients. Cardiovasc Intervent Radiol 2006;29:216 219. 61. Vuillemin E, Aube C, Rifet H, et al. Recurrent bleeding stomal varices treated with transjugular intrahepatic portosystemic shunt and stomal varices embolization. J Radiol 1998;79:690 692. 62. Wang MM, McGrew W, Dunn GD. Variceal bleeding from an ileostomy stoma. South Med J 1985;78:733737. 63. Watkins RM. Variceal hemorrhage from a colostomy due to portalhypertension secondary to intra-hepatic metastases from rectalcarcinoma. BMJ 1981;282:189 190. 64. Weinberg GD, Matalon TAS, Brunner MC, et al. Bleeding stomal varicestreatment with a transjugular intrahepatic portosystemic shunt in 2 pediatric-patients. J Vasc Interv Radiol 1995;6: 233236. 65. Wiesner RH, Larusso NF, Dozois RR, et al. Peristomal varices after proctocolectomy in patients with primary sclerosing cholangitis. Gastroenterology 1986;90:316 322. 66. Wolfsen HC, Kozarek RA, Bredfeldt JE, et al. The role of endoscopic injection sclerotherapy in the management of bleeding peristomal varices. Gastrointest Endosc 1990;36:472 474. 67. Wong RCK, Berg CL. Portal hypertensive stomapathy: a newly described entity and its successful treatment by placement
68.
69.
70. 71.
72.
73.
74.
75.
of a transjugular intrahepatic portosystemic shunt. Am J Gastroenterol 1997;92:1056 1057. Zimmerman G, Smith DC, Taylor FC, et al. Recurrent urinary conduit bleeding in a patient with portal-hypertensionmanagement with a transjugular intrahepatic portosystemic shunt. Urology 1994;43:748 751. Social Security Death Index Interactive Search. Available at http://ssdi.rootsweb.com/cgi-bin/ssdi.cgi. Accessed: June 18, 2007. Graham DY, Smith JL. The course of patients after variceal hemorrhage. Gastroenterology 1981;80:800 809. Warren WD, Millikan WJ, Henderson JM, et al. 10 years portal hypertensive surgery at Emoryresults and new perspectives. Ann Surg 1982;195:530 542. Henderson JM, Boyer TD, Kutner MH, et al. Distal splenorenal shunt versus transjugular intrahepatic portal systematic shunt for variceal bleeding: a randomized trial. Gastroenterology 2006; 130:16431651. Conte JV, Arcomano TA, Nacy MA, et al. Treatment of bleeding stomal varicesreport of a case and review of the literature. Dis Colon Rectum 1990;33:308 314. Kartheuser AH, Dozois RR, LaRusso NF, et al. Comparison of surgical treatment of ulcerative colitis associated with primary sclerosing cholangitis: ileal pouch-anal anastomosis versus Brooke ileostomy. Mayo Clin Proc 1996;71:748 756. Kartheuser AH, Dozois RR, Wiesner RH, et al. Complications and risk-factors after ileal pouch-anal anastomosis for ulcerative-colitis associated with primary sclerosing cholangitis. Ann Surg 1993;217:314 320.
Address requests for reprints to: Mark Reichelderfer, MD, Section of Gastroenterology and Hepatology, University of Wisconsin Hospital and Clinics, 600 Highland Avenue, H6/516-5124, Madison, Wisconsin 53792. e-mail: mxr@medicine.wisc.edu; fax: (608) 265-5677. The authors would like to thank Dr David Feldstein for his contributions to this article.
Dietary Counseling Versus Dietary Supplements for Malnutrition in Chronic Pancreatitis: A Randomized Controlled Trial
SIDDHARTH SINGH, SHALLU MIDHA, NAMRATA SINGH, YOGENDRA KUMAR JOSHI, and PRAMOD KUMAR GARG
Department of Gastroenterology and Human Nutrition, All India Institute of Medical Sciences, New Delhi, India
Background & Aims: Up to 50% of patients with chronic

pancreatitis (CP) are malnourished. There are limited data on the role of dietary intervention in improving the nutritional status of such patients. The aim was to compare the efcacy of medium chain triglyceride (MCT) enriched commercial dietary supplements with dietary counseling for homemade food in the management of malnutrition in patients with CP. Methods: In a randomized controlled trial, consecutive undernourished patients with CP (body mass index [BMI] <18.5 kg/m2) at a tertiary care hospital were randomized to receive either dietary counseling for regular homemade food or commercial MCT-enriched dietary supplements for a period of 3 months to compensate for the dietary calorie decit. All patients received standard management for CP including pancreatic enzyme supplements. Primary outcome measure was improvement in BMI. Results: Sixty malnourished patients with CP were randomized to counseling group (n 29; mean age, 32 10 years; male, 83%) and supplementation group (n 31; mean age, 28 10 years; male, 84%). BMI increased in both the counseling group and supplementation group (17.2 1.7 vs 18.1 1.8 kg/m2, P .001; 16.7 1.6 vs 18.2 1.6 kg/m2, P .001). There were similar improvements in triceps skinfold thickness, dietary intake, fecal fat, and pain score during a period of 3 months in both groups. There was, however, no signicant difference between the counseling and supplementation groups with regard to any of the outcome measures. Conclusions: Dietary counseling for a balanced homemade diet is as good as commercial food supplements in improving malnutrition in patients with CP. hronic pancreatitis (CP) is characterized by pancreatic inammation and brosis, eventually leading to destruction of pancreatic parenchyma and loss of exocrine and endocrine function. The most common cause of CP is alcohol in 40% 80% of cases, but it is idiopathic in 20% 60% of cases.1 Malnutrition is a common feature of CP, particularly in patients with alcoholic and idiopathic tropical CP. Although malnutrition has been thought of as a cause of or contributory factor in the pathogenesis of CP, others and we have shown that malnutrition is an effect and not a cause of CP.2,3 The etiology of malnutrition in these patients is multifactorial. Maldigestion caused by decreased pancreatic exocrine secretion and inadequate bicarbonate delivery to the duodenum leading to secondary inactivation of enzymes and bile acids by gastric acid is an important cause for malnutrition. Abdominal pain, nausea, vomiting, and postprandial satiety contribute by limiting di-
etary intake.4 Self-imposed dietary restriction caused by the fear of inducing pain also contributes to poor nutrition.5 Physicians also, as a general habit, advise patients with CP to reduce their fat intake to a minimum.6 In a study, we found that patients with CP had a selective dietary fat restriction due to food fads in spite of being adequately supplemented with pancreatic lipase and other enzymes and analgesics.2 Patients with alcoholic CP are malnourished as a result of continued alcohol intake as an important cause of undernutrition. Development of CP-related complications like diabetes, pancreatic pseudocyst, and pancreatic cancer also lead to nutritional decline.4 Hypermetabolic state with increased resting energy expenditure is another cause of malnutrition in 30%50% of patients with CP.7 The degree of undernutrition has a negative impact on the outcome of these patients.8,9 Micronutrient deciency might also contribute to decreased antioxidant capacity and increased oxidative stress in these patients.10 Although a lot of emphasis has been laid on treating abdominal pain by way of analgesics, pancreatic enzyme supplements, endoscopic therapy, and surgery, not many studies have looked at strategies for improving nutrition in these patients beyond pancreatic enzyme supplementation. Recently, commercially available dietary supplements containing hydrolyzed oligopeptides and medium chain triglycerides (MCT) have been regarded as useful for improving nutrition in patients with CP. These are readily digestible, well-tolerated preparations. Moreover, patients might be more compliant, thinking it is of medicinal value. Recent studies have shown that the presence of MCT in the commercially available food supplements results in only a minimal stimulation of postprandial CCK release and exocrine pancreatic secretion, which might decrease abdominal pain in patients with CP.11 No study has explored the efcacy of these supplements as compared with simple dietary counseling with homemade food in improving the nutritional status of malnourished patients with CP. We conducted a randomized controlled trial (RCT) to compare the efcacy of MCT-enriched commercially available food supplements with dietary counseling for regular homemade food for the management of malnutrition in patients with CP.
Abbreviations used in this paper: BMI, body mass index; CCK-RF, CCK releasing factor; CP, chronic pancreatitis; CHI, creatinine height index; CT, computed tomography; ERCP, endoscopic retrograde cholangiopancreatography; MRCP, magnetic resonance cholangiopancreatography; MCT, medium chain triglycerides; MUAC, mid upper arm circumference; RCT, randomized controlled trial; TSF, triceps skinfold thickness. 2008 by the AGA Institute 1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.040
354
SINGH ET AL
Methods
We conducted an RCT in our tertiary care academic center. Consecutive patients with CP attending the pancreas clinic at our center were included in the study during the period starting August 2000 July 2003. The diagnosis of CP was suspected on the basis of suggestive clinical features, ie, recurrent or chronic abdominal pain and/or presence of diabetes and/or steatorrhea. The diagnosis was conrmed if there was evidence of pancreatic calcication and/or ductal changes in the form of irregularity, dilation, and/or stricture of pancreatic duct on imaging studies that included ultrasonography and/or computed tomography (CT) of the abdomen and/or endoscopic retrograde cholangiopancreatography (ERCP)/magnetic resonance cholangiopancreatography (MRCP).12 Patients were assessed for their nutritional status. Patients with malnutrition formed the study group. Patients were considered malnourished if their body mass index (BMI) was less than 18.5 kg/m2, or if they had lost signicant weight (dened as recent loss of 10% of their usual body weight within the last 6 months) as a result of the primary disease.13 The patients with the following associated conditions were excluded from the study: (1) clinically apparent steatorrhea in the form of large, bulky, oily stools because any improvement in the nutritional status of patients with steatorrhea would have been attributed to pancreatic enzyme supplementation and not to dietary intervention; (2) cancer of the pancreas; (3) biliary obstruction in the form of deranged liver function test results and dilated bile duct on ultrasound; (4) patients currently undergoing endoscopic or surgical therapy; (5) patients with uncontrolled diabetes; (6) patients with acute exacerbation of pancreatitis; (7) patients with large pseudocyst (6 cm in size); (8) patients currently consuming alcohol 40 g/day; (9) opioid analgesic addicts; and (10) patients with comorbid conditions like chronic liver disease. All the patients underwent a battery of tests for the diagnosis of CP and its complications. These tests included the following: hematology and serum biochemistry and imaging, including ultrasound abdomen and CT. If required, MRCP was done.
the standard Table of Food Composition in India.15 The amounts of proteins, fats, carbohydrates, and the calories for these were computed. Protein metabolism was assessed by nitrogen balance and creatinine height index (CHI). Nitrogen balance was calculated by using the formula: [Nitrogen intake/day Nitrogen output/ day]. Nitrogen intake was calculated from 24-hour dietary protein intake (total protein intake/6.25), and nitrogen output was estimated from 24-hour urinary nitrogen excretion by using Kjeldahls method and an additional 5 mg/kg of nitrogen for integumental and other losses.16,17 CHI is a ratio of the patients 24-hour creatinine excretion and the expected normal creatinine excretion. CHI is calculated with the following formula: [(Measured urinary creatinine 100)/Ideal urinary creatinine for a given height]. Urinary creatinine is an estimate of body muscle mass calculated as urinary creatinine in grams per 24 hours. Exocrine and endocrine pancreatic functions were also assessed. Endocrine function was done by using blood sugar measurement (fasting and postprandial). Diabetes was diagnosed on the basis of World Health Organization criteria.18 Exocrine function was assessed by measuring fecal chymotrypsin concentration by spectrophotometric method.19 Stools were collected for 24 hours from patients receiving a normal diet; stools were subsequently homogenized, weighed, and stored at 20C. Fecal fat was measured to quantify fat loss according to van de Kamer et al.20 Patients were given 1 g/kg fat supplement per day for 3 days before stool collection. A pain score was devised to assess the severity of abdominal pain in patients with CP. It was calculated on the basis of frequency of pain (No episode of pain in last 12 months 0/one episode per 312 months 1/one episode per 3 months 2/one episode per month 3/one episode per week 4/two episodes per week or continuous 6) of CP, and treatment/severity (no treatment 0/oral analgesics 2/parenteral analgesics 4/hospitalization 6) of the pain.
Randomization
All the study patients were randomly assigned, by using computer-generated random number list, to either of the 2 groups, dietary counseling or dietary supplementation. Random allocation sequence, enrollment, and assigning participants to the 2 groups were done by separate individuals. The participants knew what intervention they were getting, those administering the intervention knew what was being administered, but the person assessing the outcome was blinded to the treatment the patient was receiving. The daily nutrient requirement of the patients in both groups was calculated on the basis of Harris Benedict equation, which takes into account the present weight, age, sex, and height of the patient, and this value was then multiplied by 1.9 to compensate for hypermetabolic state in chronic disease.21 Patients in both the groups received pancreatic enzyme supplementation (4 capsules 3 times a day to be taken at the start of, during, and at the end of meals). Each enteric-coated microsphere capsule contained lipase 8000 USP, amylase 30,000 USP, and protease 30,000 USP (Digestomen-P; Menarini Raunaq Pharma Limited, India). In the dietary counseling group, the patients usual dietary intake was assessed by 24-hour recall and food frequency questionnaire, and the calorie decit in the diet was calculated as the
Nutritional and Dietary Assessment

The patients underwent a detailed nutritional and dietary assessment. Nutritional assessment was assessed by the anthropometric prole of patients. Anthropometry included BMI, which was calculated by using the formula, Weight (in kg)/Height2 (in m). The triceps skinfold thickness (TSF) was measured midway between acromion process of scapula and olecranon process by using skinfold calipers (Harpenden). The mid upper arm circumference (MUAC) was measured in the left upper arm, with a nonstretch tape, at the mid-point between the tip of the shoulder and the tip of the elbow (olecranon process and the acromion). Three readings were recorded for each parameter, and the mean was calculated. For dietary assessment, a detailed dietary history was obtained from each subject at the time of entry into the study by a trained dietitian through an interview by using a food frequency questionnaire. The frequency of consumption of raw foodstuff before the onset of disease was elicited. A record of all the food consumed during the past 24 hours was made with the recall method. Nutrient intake was calculated with the 24-hour recall method.14 The result was estimated in accordance with
March 2008
NUTRITIONAL INTERVENTION IN CP
355
Figure 1. CONSORT chart.
difference between recommended calories and actual intake. An expert dietitian counseled and encouraged the patient to increase the dietary intake by eating small frequent servings of normal homemade diet including all food groups, that is, cereals, pulses, milk, vegetables, fruit, sugar, and oil. No emphasis was placed on using any particular type of oil. A diet chart was prescribed for the required amount of calories. The approximate distribution of calories into the nutrients was carbohydrates 60%, proteins 10%15%, and fat 25%30% of total energy intake. In the dietary supplementation group, the average current daily intake of calories by the patients was calculated, and the decit in the calorie intake (required minus actual) was supplemented by commercially available polymeric formula feeds (Nutren 1.0; Nestle India Ltd, India). One 250-mL serving of such enteral formula feed provided 250 kcal and 9.5 g proteins along with vitamins, minerals, choline, taurine, and carnitine. The distribution of calories into the nutrients was carbohydrates 51%, proteins 16%, and fat 33% of total energy intake. The feed was MCT-enriched with 25% of fat as MCT. It contained 50% casein and 50% whey as the protein source, had a high level of natural antioxidants, and was lactose-, cholesterol-, and glutenfree. To check for compliance of the patients, they were asked to bring back empty tins of the supplement consumed at every follow-up visit. The calculated dietary intake target was similar in both groups.
dietary intake (the total calorie intake of patient in each group), nitrogen balance, pancreatic exocrine function (assessed by change in fecal fat), and pain score. The study was approved by the Indian Council of Medical Research. All the patients were included in the study after an informed written consent. The study was conducted in accordance with the humane and ethical principles of research set forth in the Helsinki guidelines. The study followed CONSORT guidelines.22
Sample Size
In the absence of any previous data on this subject and on the basis of an estimate of around 30 50 new patients of CP
Table 1. Clinical and Imaging Features at Baseline: Counseling Versus Supplementation Groups
Variables Age (y) Male sex (n) Duration of disease (y) Etiology Alcoholic (n) Idiopathic (n) Diabetes (n) Pseudocyst Pancreatic calcication
aMean
Counseling (n 29) 32 (10)a 24 3.3 (2.7)a 14 15 7 6 (21%) 16 (55%)
Supplementation (n 31) 28 (10)a 26 4.4 (3.3)a 11 20 6 5 (16%) 20 (64%)
Outcome Measures
To evaluate improvement in anthropometry, improvement in BMI at 3 months was taken as the primary outcome measure. Secondary outcome measures included TSF, MUAC,
(standard deviation). P NS for all variables.
356
SINGH ET AL
Table 2. Hematology and Biochemical Prole at Baseline in Counseling Versus Supplementation Groups
Variable Hemoglobin (g/dL) Fasting sugar (mg/dL) Serum bilirubin (mg/dL) Serum total protein (g/dL) Serum albumin (g/dL) Serum AST (IU) Serum ALT (IU) Serum alkaline phosphate (IU) Serum calcium (mg/dl) Stool chymotrypsin (unit/g stool) Counseling 13.21 0.42 112.00 36.00 0.59 0.10 7.55 0.64 4.21 0.44 43.30 24.93 42.95 29.87 253.21 220.54 10.14 0.42 3.41 2.75 Supplementation P value 13.52 1.23 114.77 72 0.88 0.65 8.14 0.41 4.57 0.47 33.22 18.97 29.28 17.21 140.07 79.29 9.50 2.10 4.42 3.04 NS NS NS NS NS NS NS NS NS NS
ison at periods of follow-up was studied by using paired sample t test. All parameters were compared with intention-to-treat analysis. No ancillary analysis was performed. The data are presented as mean (standard deviation) or median (range) as appropriate. A P value of .05 was taken as signicant. All analyses were performed with the SPSS software (SPSS 12.0; SPSS Inc, Chicago, IL).
Results Baseline Clinical Prole

The clinical characteristics of patients randomized to dietary counseling and to dietary supplementation have been described in Table 1. The mean age of the patients was 30 years (10); 83% were men. The etiology of CP was alcoholic in 40% and idiopathic in 59%. The baseline hematology and biochemical parameters of patients in both the groups are given in Table 2. Of the 60 patients studied, 29 were randomized to receive dietary counseling alone, whereas 31 received dietary supplements. There was no evidence of any adverse events in either intervention group. The ow of patients is given in Figure 1. Effect of intervention on nutritional status. The effect of intervention on nutritional status is shown in Figure 2. The 2 groups were not different in their anthropometric prole at baseline (Table 3). At the end of 3 months of intervention, there was signicant improvement in all the anthropometric parameters including BMI, MUAC, and TSF in both groups. BMI increased from 17.2 1.7 to 18.1 1.8 kg/m2 (P .001) in the dietary counseling group and from 16.7 1.6 to 18.2 1.6 kg/m2 (P .001) in the dietary supplementation group. There was, however, no signicant difference between the counseling and supplementation groups at 3 months (18.1 1.8 vs 18.2 1.6 kg/m2; P NS). There was no signicant difference in the parameters of protein metabolism at baseline in the 2 groups (Table 4). At the end of 3 months, the dietary supplementation group showed signicant improvement in protein metabolism as assessed by a positive nitrogen balance and a high CHI. This trend was not seen in those receiving dietary counseling. The 2 groups were, however, still not different in terms of protein metabolism at 3 months.
per year in our hospital, we took a sample size of 60 patients to be recruited during a 3-year period. A total of 201 patients with CP were seen in our hospital during the 3-year period. Of these, 84 were not malnourished and hence did not meet the inclusion criteria. Of the remaining, 35 patients were excluded because of presence of biliary obstruction (n 9), clinically apparent steatorrhea (n 7), comorbidities like chronic liver disease, gastric outlet obstruction, intestinal or pulmonary tuberculosis (n 4), continued heavy alcohol intake (n 4), presence of large pseudocyst (n 3), uncontrolled diabetes mellitus (n 3), opioid addiction (n 2), carcinoma of the pancreas (n 1), acute exacerbation of pancreatitis (n 1), and patients undergoing endoscopic therapy (n 1). Twenty-two patients refused to participate. Hence, 60 patients were included in the study (Figure 1, CONSORT chart).
The distributions of fecal fat and fecal chymotrypsin were normalized by log transformation. Intergroup comparison was done with independent sample t test. Intragroup compar-
Figure 2. Effect of intervention on nutritional status.
March 2008
357
Table 3. Anthropometric Variables at Baseline and 3 Months: Counseling Versus Supplementation Groups
Counseling Parameter Body weight (kg) BMI (kg/m2) MUAC (cm) TSF (mm)
aPaired
Supplementation P value .001 .001 .001 .001 Baseline (n 31) 46.0 6.4 16.7 1.6 21.6 2.3 7.1 3.6 3 Mo (n 24) 50.9 6.9 18.2 1.6 23.4 2.0 9.7 3.9 P valuea .001 .001 .001 .001 P valueb .718 .793 .979 .524
Baseline (n 29) 47.1 6.3 17.2 1.7 22.2 2.4 7.7 3.6
3 Mo (n 25) 50.1 7.0 18.1 1.8 23.4 2.5 8.8 5.5
NOTE. Values shown as mean standard deviation. t test within the group. bt test between supplementation and counseling groups at 3 months.
Effect of intervention on dietary intake. There was a signicant improvement in the dietary intake of fat, carbohydrates, and proteins as well as total energy in both groups at 3 months (Figure 3, Table 5). In the dietary supplementation group, there were 41.25% and 66.66% increases in the mean calorie and fat intakes, respectively, of which 42.14% of calories and 37.5% of fat were provided by the dietary supplements. There was no difference between the 2 groups in terms of dietary intake at baseline and during subsequent follow up. Effect of intervention on pancreatic exocrine function. There was a signicant decrease in the fecal fat excretion, suggesting an improved fat absorption in both groups at 3 months. Fecal fat decreased from 14.8 to 8.0 g/d (P .007) in the dietary counseling group and from 12.8 to 6.9 g/d (P .001) in the dietary supplementation group. There was, however, no signicant difference in the fecal fat excretion between the 2 groups at baseline or during subsequent follow-up. Effect of intervention on pain score. The 2 groups showed signicant improvement in the pain score during follow-up. It decreased from 5.6 to 3.3 (P .001) in the dietary counseling group and from 5.0 to 3.4 (P .001) in the dietary supplementation group. There was, however, no signicant difference between the 2 groups at baseline and at the end of 3 months.
Discussion
Patients with CP might experience maldigestion and malnutrition.4 Chronic inammation and brosis in the gland can destroy exocrine tissue, leading to inadequate delivery of digestive enzymes to the duodenum in the prandial and postprandial periods and subsequent maldigestion. Maldigestion is augmented by inadequate bicarbonate delivery to the duodenum, with secondary inactivation of enzymes and bile acids by
gastric acid.23 Abdominal pain, sitophobia, nausea, vomiting, postprandial satiety, and ongoing alcohol abuse might contribute to poor oral intake. Gastric dysmotility and mechanical gastric outlet obstruction from brosis in the pancreatic head might contribute to malnutrition and clinical decline. Patients with CP might at times experience profound steatorrhea and weight loss.4 The standard therapy for CP is centered on management of pain. Little emphasis has been paid on adequate nutrition of these patients. Aggressive nutrition therapy in CP might have the ability to change the course of the disease and ameliorate the clinical outcome.24 Nutrition in CP patients has twin objectives; rst and foremost is improvement of the malnutrition, and second is to decrease pain by decreasing pancreatic gland stimulation. Recently, commercially available food supplements have been actively propagated for their easy digestibility and better tolerance for patients with pancreatic insufciency. The homogenized form of these foods might cause pancreatic stimulation for a shorter period of time as compared with complex solids.25 They also contain hydrolyzed oligopeptides and MCT, which have been shown to cause minimal pancreatic stimulation by way of decreased secretion of CCK.11 The efcacy of these food supplements vis--vis homemade diet, with regard to their ability to improve the nutritional status and symptomatology of patients with CP, has not been studied. We found in the present RCT that although both dietary supplementation with commercial food preparations and dietary counseling improved the nutritional status of the patients, commercially available food supplements were no more efcacious than routine dietary counseling for a balanced homemade diet in improving the overall nutritional status of patients. The signicant improvement in anthro-
Table 4. Parameters of Protein Metabolism at Baseline and 3 Months: Counseling Versus Supplementation Groups
Counseling Parameter Nitrogen balance (g/day) CHI Urinary creatinine (g/day)
aPaired
Supplementation P valuea .765 .210 .124 Baseline (n 31) 5.4 3.9 43.3 13.9 0.6 0.3 3 Mo (n 24) 8.8 5.5 56.3 18.9 0.8 0.3 P valuea .003 .007 .001 P valueb .205 .727 .925
Baseline (n 28) 6.5 4.8 48.7 15.4 0.7 0.2
3 Mo (n 24) 7.1 4.0 54.1 24.1 0.8 0.3
NOTE. Values shown as mean standard deviation. t test within the group. bt test between supplementation and counseling groups at 3 months.
358
SINGH ET AL
Figure 3. Effect of intervention on dietary intake.
pometric parameters, dietary intake, pancreatic exocrine test, and pain scores in both groups was perhaps related to overall improvement in their dietary intake coupled with the standard management for CP in a systematic manner and regular follow-up. One of the important reasons for malnutrition in CP is the misconstrued belief of patients, sometimes reinforced by physicians, that fat intake should be restricted to a minimum for fear of inducing pain. With simple dietary counseling for a wholesome food intake, we were able to alleviate undernutrition in these patients. Restriction of fat leads to impalatability of food besides decreasing the overall caloric value of the food. Fat intake need not be altered in quantity or modied in quality (use of MCTs) because these patients are receiving adequate pancreatic lipase supplementation by way of exogenous pancreatic enzymes. Moreover, it has been shown that gastric lipase secretion is also increased in CP, and this enzyme can achieve about 30% of the lipolysis observed in healthy volunteers.26,27 The use of MCT in place of long chain triglycerides has been shown to reduce postprandial CCK release.28 CCK, in turn, has been implicated as one of the factors responsible for pancreatic stimulation and subsequent pain in patients with CP.11 However, only a single small study of 10 patients with CP has shown a reduction in pain in patients receiving MCT-enriched food supplements.29 CCK release is dependent on CCK releasing factor (CCK-RF). In the presence of trypsin and other proteases, this releasing factor is easily degraded. However, CP leads to decreased secretion of proteases, leading to loss of the negative feedback on CCK-RF and elevated CCK levels.30 Supplementation with pancreatic enzymes, including proteases, could pos-
sibly result in destruction of CCK-RF, so that CCK levels are not high enough to result in pain. Moreover, MCTs also have a low energy density, are not very palatable, and hence have a poor compliance and might induce side effects such as nausea, abdominal pain, and diarrhea.24 Hence, it is usually recommended that MCTs be advised only to those patients who do not gain weight adequately in spite of standard management for CP and those with persistent steatorrhea. It is generally believed that 80% of the patients can be managed by dietary counseling, analgesics, and pancreatic enzyme supplements, whereas only 10%15% might require oral nutritional supplements.31 The improvement in protein anabolism, as assessed by nitrogen balance, was better in patients on supplements as compared with dietary counseling. However, the difference was not signicant at the end of 3 months between the 2 groups, which could be due to a small sample size. The reasons for better nitrogen balance in the intervention group could be related to (1) consumption of adequate proteins through nutritional supplements, (2) low dietary protein intake due to predominantly vegetarian diet in the counseling group, and (3) better assimilation or absorption of oligopeptides present in the supplements. The weight gain observed in both groups could be attributed to (1) increased calorie, fat, and protein intake; (2) use of exogenous pancreatic enzyme supplements resulting in decreased fecal fat; and (3) decreased pain experienced by patients in both groups, with resultant increase in calorie consumption, attributable to use of pancreatic enzyme supplements and regular patient follow-up. Although the possibility of Hawthorne effect cannot be excluded in our study, it is less likely because there was denite objective improvement in weight gain and nutritional status after the intervention in the groups under study.32 The Hawthorne/protocol effect is a component of the nonspecic benet from improved routine care within a trial.33 The effect, however, is minimal at best.34 There are several clinical implications of this study. (1) Homemade diet, which is cheaper, more palatable, and more physiologic, improves the nutritional status of patients with CP. (2) Patients with CP should be advised to take small, frequent, homemade balanced meals. Dietary counseling by a trained dietitian should be made an integral part of the management of these patients. (3) Nutritional supplements are not generally required; they might be used only in special situations. Thus, we concluded that dietary counseling for a balanced homemade diet is as good as commercial food supplements in improving malnutrition in patients with CP.
Table 5. Dietary Intake at Baseline and 3 Months: Counseling Versus Supplementation Groups
Counseling Parameter Energy (kcal/day) Carbohydrates (g/day) Proteins (g/day) Fats (g/day)
aPaired
Supplementation P valuea .069 .031 .044 .102 Baseline (n 31) 2053 689 324 126 69 22 48 18 3 Mo (n 24) 2900 195 407 147 97 30 80 39 P valuea .004 .004 .001 .829 P valueb .182 .664 .148 .195
Baseline (n 29) 2188 672 346 140 73 24 52 24
3 Mo (n 25) 2575 141 377 147 85 26 67 29
NOTE. Values are shown as mean standard deviation. t test within the group. bt test between supplementation and counseling groups at 3 months.
March 2008
359
References
1. Garg PK, Tandon RK. Survey on chronic pancreatitis in the AsiaPacic region. J Gastroenterol Hepatol 2004;19:998 1004. 2. Joshi YK, Midha S, Singh N, et al. Muscle and visceral proteins are preserved with a positive nitrogen balance in patients with chronic pancreatitis. Gastroenterology 2005;128(Suppl 2):A471. 3. Abu-Bakare A, Taylor R, Gill GV, et al. Tropical or malnutritionrelated diabetes: a real syndrome? Lancet 1986;1:11351138. 4. Petersen JM, Forsmark CE. Chronic pancreatitis and maldigestion. Semin Gastrointest Dis 2002;13:191199. 5. Goebell H, Hotz J, Hoffmeister H. Hypercaloric nutrition as aetiological factor in chronic pancreatitis. Z Gastroenterol 1980;18: 94 97. 6. Vaona B, Armellini F, Bovo P, et al. Food intake of patients with chronic pancreatitis after onset of the disease. Am J Clin Nutr 1997;65:851 854. 7. Hebuterne X, Hastier P, Peroux JL, et al. Resting energy expenditure in patients with alcoholic chronic pancreatitis. Dig Dis Sci 1996;41:533539. 8. Thorsgaard Pedersen N, Nyboe Andersen B, Pedersen G, et al. Chronic pancreatitis in Copenhagen: a retrospective study of 64 consecutive patients. Scand J Gastroenterol 1982;17:925931. 9. Worning H. Chronic pancreatitis: pathogenesis, natural history and conservative treatment. Clin Gastroenterol 1984;13:871 894. 10. Braganza JM, Schoeld D, Snehalatha C, et al. Micronutrient antioxidant status in tropical compared with temperate-zone chronic pancreatitis. Scand J Gastroenterol 1993;28:1098 1104. 11. Shea JC, Bishop MD, Parker EM, et al. An enteral therapy containing medium-chain triglycerides and hydrolyzed peptides reduces postprandial pain associated with chronic pancreatitis. Pancreatology 2003;3:36 40. 12. Tandon RK, Sato N, Garg PK, et al. Chronic pancreatitis: AsiaPacic consensus report. J Gastroenterol Hepatol 2002;17: 508 518. 13. WHO physical status: the use and interpretation of anthropometryreport of a WHO Expert Consultation. WHO Technical Report series number 854, 1995. 14. Thimmayamma BVS. A hand book of schedules and guidelines in socio economic and diet survey. Hyderabad, India: National Institute of Nutrition, ICMR, 1987. 15. Deosthale YG. Nutritive value of Indian foods: some recent studies. Indian J Med Res 1978;68(Suppl):116. 16. Hiller A, Plazin J, Van Slyke DD. A study of conditions for Kjeldahl determination of nitrogen in proteins: description of methods with mercury as catalyst, and titrimetric and gasometric measurements of the ammonia formed. J Biol Chem 1948;176:1401 1420. 17. Food and Agriculture Organization/World Health Organization. Report of a joint FAO/WHO ad hoc expert committee. WHO Technical Report series 1974 no. 52, WHO, Geneva.
18. Jarrett RJ. Impaired glucose tolerance and diabetes: WHO criteria. Br Med J (Clin Res Ed) 1981;282:990. 19. Kasper P, Moller G, Wahlefeld A. New spectrophotometric estimation of chymotrypsin in stool. Clin Chem 1984;30:1753 1757. 20. Van de Kamer JH, Bokkel-Huinik HB, Weyers HA. Rapid method for determination of fat in feces. J Biol Chem 1949;177:347 355. 21. Dickerson RN, Vehe KL, Mullen JL, et al. Resting energy expenditure in patients with pancreatitis. Crit Care Med 1991;19:484 490. 22. Altman DG, Schulz KF, Moher D, et al. The revised CONSORT statement for reporting randomized trials: explanation and elaboration. Ann Intern Med 2001;134:663 694. 23. Thiruvengadam R, DiMagno EP. Inactivation of human lipase by proteases. Am J Physiol 1988;255:G476 G481. 24. Giger U, Stanga Z, DeLegge MH. Management of chronic pancreatitis. Nutr Clin Pract 2004;19:37 49. 25. Miller LJ, Clain JE, Malagelada JR, et al. Control of human postprandial pancreatic exocrine secretion: a function of the gastroduodenal region. Dig Dis Sci 1979;24:150 154. 26. Moreau J, Bouisson M, Balas D, et al. Gastric lipase in alcoholic pancreatitis: comparison of secretive proles following pentagastrin stimulation in normal adults and patients with pancreatic insufciency. Gastroenterology 1990;99:175180. 27. Carriere F, Grandval P, Renou C, et al. Quantitative study of digestive enzyme secretion and gastrointestinal lipolysis in chronic pancreatitis. Clin Gastroenterol Hepatol 2005;3:28 38. 28. Symersky T, Vu MK, Frolich M, et al. The effect of equicaloric medium- chain and long-chain triglycerides on pancreas enzyme secretion. Clin Physiol Funct Imaging 2002;22:307311. 29. Freedman SD. Peptamen: a novel therapy in patients with chronic pancreatitis (abstract). Gastroenterology 1997;112:A267. 30. Spannagel AW, Green GM, Guan D, et al. Purication and characterization of a luminal cholecystokinin-releasing factor from rat intestinal secretion. Proc Natl Acad Sci U S A 1996;93:4415 4420. 31. Meier R, Ockenga J, Pertkiewicz M, et al. ESPEN guidelines on enteral nutrition: pancreas. Clin Nutr 2006;25:275284. 32. Franke RH, Kaul JD. The Hawthorne experiments: rst statistical interpretation. Am Sociol Rev 1978;43:623 643. 33. Braunholtz DA, Edwards SJ, Lilford RJ. Are randomized clinical trials good for us (in the short term)? evidence for a trial effect. J Clin Epidemiol 2001;54:217224. 34. Jones S. Was there a Hawthorne effect? Am J Sociol 1992;98: 451 468.
Address requests for reprints to: Pramod Kumar Garg, MD, DM, Associate Professor, Department of Gastroenterology, All India Institute of Medical Sciences, New Delhi 110029, India. e-mail: pkgarg@ aiims.ac.in; fax: 91-11-26588663. The study was supported by a research grant from the Indian Council of Medical Research, New Delhi.
BRIEF COMMUNICATIONS
Anti-Aging Therapy With Human Growth Hormone Associated With Metastatic Colon Cancer in a Patient With Crohns Colitis
GIL Y. MELMED,* SHANE M. DEVLIN,* GEORGE VLOTIDES,* DEEPTI DHALL, SORAYA ROSS,* RUN YU,* and SHLOMO MELMED*
*Department of Medicine and Department of Anatomic Pathology, Cedars-Sinai Medical Center, David Geffen School of Medicine, University of California, Los Angeles, California
Background & Aims: The nonapproved use of human

growth hormone (HGH) for anti-aging has been increasing. Theoretical concerns for neoplastic potentiation by HGH have been raised, but not proven clinically. Methods: We report the case of a 68-year-old man with colonic Crohns disease who was found to have aggressive metastatic colon cancer. The patient had been receiving HGH therapy for anti-aging purposes for 7 years before presentation. Normal and malignant colonic tissue was examined for qualitative and quantitative molecular proles of growth hormone (GH) and its signaling molecules, using immunohistochemistry and RNA extraction with polymerase chain reaction amplication. Results: Immunoreactivity was more robust in tumor tissue than in normal colon for insulin-like growth factor-1 receptor (IGF-1R) but not for IGF, GH, or GH receptor. RNA extraction with quantitative polymerase chain reaction showed that IGF-1R and vascular endothelial growth factor expression, but not IGF-1, GH receptor, or suppressor of cytokine signaling-2, were higher in tumor than in normal colonic tissue. Conclusions: Colorectal cancer development concurrent with administration of HGH for anti-aging purposes occurred in an individual already at increased risk for colon cancer. This underscores the need for further investigation of the proneoplastic potential of GH supplementation for anti-aging. he use of growth hormone (GH) in adults is approved for patients with proven pituitary deciency, or those with acquired immune deciency syndromeassociated wasting.1 Approximately 100,000 individuals obtain GH for anti-aging purposes annually in the United States without a prescription, and about 30% of GH prescriptions in the United States are for off-label use such as anti-aging.2,3 Because of the potential benets of GH administration on bone density, muscle mass, and cardiovascular health in GH-decient individuals, the practice of administering GH to healthy adults has become increasingly common as a result of claims of anti-aging properties. However, a recent meta-analysis of GH administration in healthy adults showed no important benet to body composition and that side effects were common.4 Furthermore, concerns have been raised regarding the carcinogenic potential of human growth hormone (HGH).5 Several epidemiologic studies have reported an association between increased insulin-like growth factor-1 (IGF-1) serum levels and the risk of colorectal cancer.6 8 Acromegaly, a condition of endogenous GH excess,
has been reported to be associated with increased risk of neoplasia including colorectal adenomas and carcinoma,9,10 and guidelines for colorectal cancer screening in these individuals have been published.11 Despite the growing off-label use of HGH for anti-aging and its potential carcinogenic risk, there is a paucity of published reports attributing its use to the occurrence of cancer in this setting. We describe a case of metastatic colon cancer in a patient with long-standing Crohns disease who was receiving daily anti-aging treatment with HGH for years before the diagnosis of cancer. We performed a molecular assessment for the presence of IGF-1, IGF-1 receptor (IGF-1R), GH, and growth hormone receptor (GHR) in malignant and nonmalignant colonic tissue. We also performed molecular analysis for suppressor of cytokine signaling (SOCS)-2, a GH-dependent negative signaling molecule implicated in the regulation of intestinal epithelial cell growth,12 and vascular endothelial growth factor (VEGF), an IGF-1 dependent angiogenic gene that promotes colon cancer growth.13,14
Case Report
A 68-year-old man with long-standing colonic Crohns disease was admitted with hematochezia and abdominal pain. He was receiving prednisone 6 mg/day and mesalamine, and had never taken immunomodulator or biologic therapy. For 7 years before admission, the patient sought therapy at anti-aging clinics, and was maintained with testosterone 200 mg weekly injection, dehydroepiandrosterone 100 mg/day, and recombinant HGH (0.6 mg subcutaneously twice daily). Colonoscopy identied a 7-cm cecal mass, and laparotomy revealed widely metastatic colon cancer. The patient had undergone 2 colonoscopies within 2 years of presentation, both of which showed changes consistent with chronic, active colitis in the right colon, but no denite dysplasia or malignancy; it is not known whether the patient had ever had a colonoscopy for dysplasia screening. At the time of admission, his carcinoembryonic antigen level was 38 ng/mL (normal, 3 ng/mL) and his IGF-1 level was 245 ng/mL (normal, 36 237 ng/mL). The
Abbreviations used in this paper: GH, growth hormone; GHR, growth hormone receptor; HGH, human growth hormone; IGF-1, insulin-like growth factor-1; IGF-1R, insulin-like growth factor-1 receptor; SOCS, suppressor of cytokine signaling; VEGF, vascular endothelial growth factor. 2008 by the AGA Institute 1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.017
March 2008
HGH AND COLON CANCER
361
patient died from his illness within 5 months of diagnosis. Posthumously, normal and malignant colonic tissue were examined for qualitative and quantitative molecular proles of GH, GHR, IGF-1, IGF-1R, SOCS-2, and VEGF-A by using immunohistochemistry and RNA extraction with polymerase chain reaction amplication.
Immunohistochemistry
Parafn slices (5 m thick) of resected malignant and normal tissue were deparafnated and rehydrated, and antigens were retrieved by microwave heating. Processed slices were blocked with 5% fetal bovine serum and 0.6% Tween 20 in buffered saline, and incubated with primary antibody at 1:100 for 2 hours followed by rhodamine-labeled goat anti-mouse immunoglobulin G or donkey anti-rabbit immunoglobulin G at 1:200 for 1 hour. Slices were counterstained with Hoechst 33342 at 1:5000. Stained slices were preserved in Prolong Gold antifade reagent (Invitrogen, Carlsbad, CA). Antibodies used included rabbit polyclonal anti-GH (Dako, Carpinteria, CA), mouse monoclonal anti-GHR (MAB263; Abcam, Cambridge, MA), rabbit polyclonal antiIGF-1 (H-70; Santa Cruz Biotechnology, Santa Cruz, CA), and rabbit polyclonal antiIGF-1R
Methods Human Subjects Approval

Permission to review medical records and to obtain normal and malignant colonic tissue from archived, formalinxed, parafn-embedded sections from laparotomy for molecular investigation was obtained from the institutional review board at Cedars-Sinai Medical Center (Los Angeles, CA) after the patients death.
Figure 1. Immunouorescent staining of GH, GHR, IGF-1, and IGF-1R in normal colon mucosa and colon carcinoma. Parafn sections of normal colon mucosa and colon carcinoma from the patient were stained for GH, GHR, IGF-1, or IGF-1R (red), and counterstained with Hoechst 33342 for nuclei (blue). (A) Increased cytoplasmic signal for IGF-1R, but not for GH, GHR, or IGF-1, in tumor relative to normal colon tissue. (B and C) Merged images for IGF-1R staining in normal and malignant colonic tissue, respectively.
362
MELMED ET AL
Figure 2. Increased expression of IGF-1R and VEGF, but not GHR or IGF-1, in tumor relative to normal colon tissue as determined by quantitative reverse-transcription polymerase chain reaction. **P .01.
(C-20; Santa Cruz). Sections were observed with a Nikon Eclipse TE200 uorescence microscope (Nikon Corporation, Tokyo, Japan), and pictures were acquired with a SPOT digital camera (Diagnostic Instruments, Inc Sterling Heights, MI).
RNA Extraction and Quantitative-Polymerase Chain Reaction

RNA extraction was performed from 10-mthick sections of freshly cut normal and tumor formalin-xed, parafnembedded tissue, respectively, using the RNeasy formalin-xed, parafn-embedded kit (Qiagen). Reverse transcription of 1.0 g total RNA was performed with M-MLV-Reverse Transcriptase (Invitrogen). Q-PCR was performed with SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA) in a 20-L reaction system, containing 10 L Master Mix, 500 nmol/L primer (each) and appropriate DNA aliquots, and analyzed with the Bio-Rad iQ5 Optical System Software (Bio-Rad Laboratories, Hercules, CA). To allow the use of fragmented RNA, the following primers were designed specically to amplify small amplicons: GH (Genebank Accession NM_000515): sense 5-CAGGAGTGTCTTCGCCAACA3, anti-sense 5-TCCCCATCAGCGTTTGGAT-3 (amplicon 101 nt); GHR (Genebank Accession NM_000163): sense 5-CAACCAGATCCACCCATTGC-3, anti-sense 5-GAATCCCAGTTAAACTGACGTTCAG-3 (amplicon 61 nt); IGF-I (Genebank Accession NM_000618): sense 5-GCTGGTGGATGCTCTTCAGTT-3, antisense 5-CCCTGTGGGCTTGTTGAAAT-3 (amplicon 64 nt); IGF-IR (Genebank Accession NM_000875): sense 5-ACGTGAAGATCCGCCATTCT-3, anti-sense 5-CCTAGGATGAGGCGAAGGTTT-3 (amplicon 70 nt); SOCS-2 (Genebank Accession NM_003877): sense 5-GCAAGGATAAGCGGACAGGTC-3, antisense 5-AGAGCGGTTTGGTCAGATAAAGG-3 (amplicon 72 nt); VEGF-A (Genebank Accession NM_001025366): sense 5-GCTTGCCATTCCCCACTTG-3, anti-sense 5-TTTCCTCTTTCTGCTGGTTTCC-3 (amplicon 97 nt); and -actin (Genebank Accession BC002409): sense 5-TTGAATGATGAGCCTTCGTG-3, anti-sense 5-GCCTTCATACATCTCAAGTTGG-3 (amplicon 67 nt).
noreactivity was more robust in tumor tissue than in normal colon for IGF-1R, but not for IGF, GH, or GHR (Figure 1). When RNA was subjected to Q-PCR, IGF-1R and VEGF expression were shown to be higher in the tumor than in normal tissue. However, IGF-1, GHR, and SOCS-2 expression were not increased (Figure 2).
Discussion
The attributable risk of cancer in patients receiving GH therapy has not been established denitively. In a meta-analysis evaluating the reported incidence of neoplasia in individuals with GH deciency treated with exogenous GH, there was a modest but signicant increased incidence of overall cancer mortality, colorectal cancer, and Hodgkins disease.15 However, surveillance studies of GH replacement for GH-decient adults have not shown enhanced cancer incidence or mortality. The potential risk of adenomatous polyps and cancer death in the setting of excess GH may be owing to increased serum IGF-1, which can stimulate cell growth and inhibit apoptosis.16 IGF1Rs are expressed on human colorectal cancer cells and IGF-1 is known to be a stimulator of colorectal cancer cell growth in vitro.17 IGF-1R also has been shown to play an important role in angiogenesis and growth of human colon cancer cells.18 Our ndings of increased IGF-1R and VEGF expression supports the notion that VEGF acts to promote tumor angiogenesis via IGF-1 signaling.13 Despite our inability to detect SOCS-2 in either normal or malignant tissue, SOCS-2 may yet be implicated in this patients cancer because haplotype insufciency for SOCS-2 promotes colonic polyposis in the setting of excess GH in mice.19 In this case report, the patient had long-standing Crohns colitis, a known risk factor for colorectal cancer. The degree to which exogenous HGH anti-aging treatment may have contributed to the malignant transformation of colonic epithelium or accelerated growth of a pre-existing neoplasm cannot be determined on the basis of our ndings. However, 2 colonoscopic evaluations within 2 years of presentation did not detect malignancy or dysplasia, suggestive of an accelerated process of malignant transformation and/or growth. IGF-1, which was increased in this patients serum at the time of admission, has been shown both to inhibit apoptosis and to allow progression through the cell cycle, which could potentiate both malignant transformation as well as accelerated growth of an existing neoplasm.20 Conversely, inammatory bowel disease has been characterized by GH resistance and relative IGF-1 deciency,21 a
Results are expressed as the mean SD and differences were evaluated using the Student t test (statistical signicance was set at P .05).
Results
Light histology showed distinct areas of normal colon and poorly differentiated adenocarcinoma (not shown). Immu-
March 2008
HGH AND COLON CANCER
363
nding that has prompted clinical investigations for the use of HGH and other growth factors for the treatment of both Crohns disease and ulcerative colitis. A randomized, placebocontrolled study showed signicant short-term improvement in adults with Crohns disease treated with HGH.22 However, concerns about malignant transformation of chronically inamed, rapidly dividing colonic epithelial tissue in response to growth factors have been raised.23 Our ndings of increased tumor tissue IGF-1R expression as compared with normal colon lends support to an etiologic role for the GH/IGF-1 axis in the development and progression of colon cancer, as has been described previously.24,25 However, our ndings do not discriminate between accelerated growth inuenced by increased local GH signaling or increased circulating IGF-1. Furthermore, the molecular basis of colitis-related colorectal cancer is thought to involve multiple genetic alterations, including mutations in p53, APC, and k-ras.26 The assessment of these markers and description of the complex interplay between these molecular alterations and endocrine and paracrine signals is important to understanding the molecular pathogenesis of cancer in this patient, but beyond the scope of this report. Colorectal cancer development concurrently with the administration of HGH for anti-aging purposes occurred in an individual already at increased risk for colon cancer with inammatory bowel disease. This underscores the need for further investigation of the proneoplastic potential of GH supplementation in individuals for anti-aging as its use for this nonapproved indication becomes more widespread. References
1. Gharib H, Cook DM, Saenger PH, et al. American Association of Clinical Endocrinologists medical guidelines for clinical practice for growth hormone use in adults and children2003 update. Endocr Pract 2003;9:64 76. 2. Perls TT, Reisman NR, Olshansky SJ. Provision or distribution of growth hormone for antiaging: clinical and legal issues. JAMA 2005;294:2086 2090. 3. Vance ML. Can growth hormone prevent aging? N Engl J Med 2003;348:779 780. 4. Liu H, Bravata DM, Olkin I, et al. Systematic review: the safety and efcacy of growth hormone in the healthy elderly. Ann Intern Med 2007;146:104 115. 5. Melmed S. Supplemental growth hormone in healthy adults: the endocrinologists responsibility. Nat Clin Pract Endocrinol Metab 2006;2:119. 6. Ma J, Pollak MN, Giovannucci E, et al. Prospective study of colorectal cancer risk in men and plasma levels of insulin-like growth factor (IGF)-I and IGF-binding protein-3. J Natl Cancer Inst 1999;91:620 625. 7. Giovannucci E, Pollak MN, Platz EA, et al. A prospective study of plasma insulin-like growth factor-1 and binding protein-3 and risk of colorectal neoplasia in women. Cancer Epidemiol Biomarkers Prev 2000;9:345349. 8. Kaaks R, Toniolo P, Akhmedkhanov A, et al. Serum C-peptide, insulin-like growth factor (IGF)-I, IGF-binding proteins, and colorectal cancer risk in women. J Natl Cancer Inst 2000;92:1592 1600. 9. Jenkins PJ. Acromegaly and cancer. Horm Res 2004;62(Suppl 1): 108 115.
10. Domenech-Santasusana M, Carles J, Goday A, et al. Association of acromegaly and two malignancies: colorectal cancer and nonHodgkins lymphoma. Ann Oncol 1994;5:659. 11. Jenkins PJ, Fairclough PD. Screening guidelines for colorectal cancer and polyps in patients with acromegaly. Gut 2002; 51(Suppl 5):V13V14. 12. Miller ME, Michaylira CZ, Simmons JG, et al. Suppressor of cytokine signaling-2: a growth hormone-inducible inhibitor of intestinal epithelial cell proliferation. Gastroenterology 2004;127: 570 581. 13. Wu Y, Yakar S, Zhao L, et al. Circulating insulin-like growth factor-I levels regulate colon cancer growth and metastasis. Cancer Res 2002;62:1030 1035. 14. Warren RS, Yuan H, Matli MR, et al. Induction of vascular endothelial growth factor by insulin-like growth factor 1 in colorectal carcinoma. J Biol Chem 1996;271:2948329488. 15. Swerdlow AJ, Higgins CD, Adlard P, et al. Risk of cancer in patients treated with human pituitary growth hormone in the UK, 1959-85: a cohort study. Lancet 2002;360:273277. 16. Sperling MA, Saenger PH, Hintz R, et al. Special editorial: growth hormone treatment and neoplasia-coincidence or consequence? J Clin Endocrinol Metab 2002;87:53515352. 17. Lahm H, Amstad P, Wyniger J, et al. Blockade of the insulin-like growth-factor-I receptor inhibits growth of human colorectal cancer cells: evidence of a functional IGF-II-mediated autocrine loop. Int J Cancer 1994;58:452 459. 18. Reinmuth N, Fan F, Liu W, et al. Impact of insulin-like growth factor receptor-I function on angiogenesis, growth, and metastasis of colon cancer. Lab Invest 2002;82:13771389. 19. Michaylira CZ, Ramocki NM, Simmons JG, et al. Haplotype insufciency for suppressor of cytokine signaling-2 enhances intestinal growth and promotes polyp formation in growth hormonetransgenic mice. Endocrinology 2006;147:16321641. 20. Giovannucci E. Insulin, insulin-like growth factors and colon cancer: a review of the evidence. J Nutr 2001;131:3109S3120S. 21. Theiss AL, Fruchtman S, Lund PK. Growth factors in inammatory bowel disease: the actions and interactions of growth hormone and insulin-like growth factor-I. Inamm Bowel Dis 2004;10:871 880. 22. Slonim AE, Bulone L, Damore MB, et al. A preliminary study of growth hormone therapy for Crohns disease. N Engl J Med 2000;342:16331637. 23. Sinha A, Nightingale J, West KP, et al. Epidermal growth factor enemas with oral mesalamine for mild-to-moderate left-sided ulcerative colitis or proctitis. N Engl J Med 2003;349:350 357. 24. Hakam A, Yeatman TJ, Lu L, et al. Expression of insulin-like growth factor-1 receptor in human colorectal cancer. Hum Pathol 1999;30:1128 1133. 25. Weber MM, Fottner C, Liu SB, et al. Overexpression of the insulin-like growth factor I receptor in human colon carcinomas. Cancer 2002;95:2086 2095. 26. Itzkowitz SH. Molecular biology of dysplasia and cancer in inammatory bowel disease. Gastroenterol Clin North Am 2006;35: 553571.
Address requests for reprints to: Gil Y. Melmed, MD, Cedars-Sinai Medical Center, 8635 West Third Street, #960-W, Los Angeles, California 90048. e-mail: melmedg@cshs.org; fax: (310) 967-0131. Supported by the National Institutes of Health (CA 75979 to S.M.), the Doris Factor Molecular Endocrinology Laboratory, and by a National Institutes of Healthsponsored Gastroenterology training grant (T32 DK 07180-31 to G.Y.M.). S.M. is an ad hoc scientic advisor to Eli Lilly. G.Y.M. and S.M.D. contributed equally to this article.
Rituximab Therapy for Refractory Biliary Strictures in Immunoglobulin G4 Associated Cholangitis

MARK TOPAZIAN,* THOMAS E. WITZIG, THOMAS C. SMYRK, JOSE S. PULIDO, MICHAEL J. LEVY,* PATRICK S. KAMATH,* and SURESH T. CHARI*
*Miles and Shirley Fiterman Center for Digestive Diseases, Department of Medicine, Division of Gastroenterology and Hepatology, Department of Internal Medicine, Division of Hematology and Oncology, Department of Pathology, and Department of Ophthalmology, Mayo Clinic, Rochester, Minnesota
See Ghazale AH et al on page 706 for companion article in the March 2008 issue of Gastroenterology. Background & Aims: Biliary strictures occur in a third
of patients with autoimmune pancreatitis and have been termed immunoglobulin G subclass 4 (IgG4) associated cholangitis (IAC). IAC often responds to steroid therapy. Methods: A patient with autoimmune pancreatitis and (IAC) refractory to steroids and 6-mercaptopurine was treated with rituximab, a monoclonal antibody directed against the CD20 antigen on B lymphocytes. Results: The patients biliary strictures improved after rituximab therapy, permitting removal of his biliary stents. Systemic manifestations of IgG4-associated disease also improved. Conclusions: Rituximab may be a treatment option for patients with refractory or recurrent autoimmune pancreatitis or IAC. teroid-responsive biliary strictures occur in autoimmune pancreatitis (AIP), and have been termed immunoglobulin G subclass 4 (IgG4) associated cholangitis (IAC).1 Patients with IAC who relapse after withdrawal of steroids often improve with azathioprine.2 This report describes a patient with IAC and refractory hilar biliary strictures who responded to treatment with rituximab, a monoclonal antibody directed against the CD20 antigen on B cells.
Case Report
A 64-year-old man developed pancreatitis in the fall of 2004. Serum liver tests were normal except for an alkaline phosphatase level of 130 U/L (normal, 80 120 U/L). History, laboratory evaluation, abdominal ultrasound, and computerized tomography scan did not reveal an etiology. Magnetic resonance imaging showed an irregular main pancreatic duct with a stenosis in the neck. His serum IgG4 level was 568 mg/dL (normal, 8 140 mg/dL). Endoscopic ultrasound showed a hypoechoic, enlarged pancreas. Endoscopic retrograde cholangiopancreatography (ERCP) showed multifocal strictures of the pancreatic duct. Cholangiography was normal. Pancreatic duct brushings were negative for malignancy. Based on the increased serum IgG4 level and compatible pancreatography, a diagnosis of AIP was made and he was treated with an 8-week tapering course of oral prednisone starting with 50 mg/day. His postprandial pain recurred a
month after completing steroid therapy, and he was re-treated with a 12-week course of prednisone with immediate symptomatic improvement. Steroid therapy was complicated by ophthalmic herpes zoster. In the fall of 2005 he became jaundiced. His serum alkaline phosphatase level was 378 U/L, his aspartate aminotransferase level was 215 U/L, his total bilirubin level was 12.7 mg/dL, his albumin level was 3.1 g/dL, and his IgG4 level was 993 mg/dL. ERCP showed a stenosis of the biliary conuence and common hepatic duct (Figure 1A). Biliary brushings for cytology, digital image analysis, and uorescent in situ hybridization (FISH) were negative for malignancy. Intraductal biliary biopsy specimens showed a lymphoplasmacytic inammatory inltrate with many IgG4-positive cells, compatible with IAC (Figure 2A and B). Biliary stents were placed and his jaundice resolved. He was treated with a 12-week tapering course of oral prednisone, but repeat ERCP showed a persistent hilar stenosis extending further into the intrahepatic ducts than previously. Intraductal brushings and biopsy specimens again were negative for malignancy. He received another 12-week course of oral prednisone as well as oral 6-mercaptopurine at a dose of 1.5 mg/kg/day. ERCP was repeated 3 months later; the hilar strictures were somewhat improved, and the biliary stents were removed. The patient was hospitalized 2 weeks later with jaundice and cholangitis. Cholangiography again showed a tight stenosis of the biliary conuence, and 3 plastic biliary stents were placed and subsequently exchanged every 3 months. In the fall of 2006, after 8 months of 6-mercaptopurine therapy, the patient lost weight and developed steatorrhea with decreased vision in the right eye. His serum bilirubin level increased to 4.4 mg/dL and remained increased despite exchange of his biliary stents. A positron emission tomography scan showed mild to moderate [18F]-2-uoro-deoxy-D-glucose uptake consistent with inammatory disease in the hepatic hilum, the pancreas, and the right eye. Ophthalmologic examination revealed deep choroidal inltrates. Magnetic resonance imaging showed thickening of the pancreatic body and tail with an irregular pancreatic duct, and hilar bile duct thickening. Cholangiography showed worsening of his intrahepatic biliary strictures, and it was difcult to ll his intrahepatic ducts with contrast despite occlusion cholangiography (Figure 1B). AddiAbbreviations used in this paper: AIP, autoimmune pancreatitis; ERCP, endoscopic retrograde cholangiopancreatography; IAC, IgG4associated cholangitis; IgG4, immunoglobulin G subclass 4. 2008 by the AGA Institute 1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.020
March 2008
RITUXIMAB FOR IgG4 ASSOCIATED CHOLANGITIS
365
Figure 1. (A) Cholangiogram showing hilar biliary strictures before treatment. (B) Progression of strictures after steroid and 6-mercaptopurine therapy. Intrahepatic ducts ll poorly despite occlusion cholangiography. (C) Improvement 4 months after the initiation of rituximab therapy.
tional biliary biopsy specimens again showed ndings of IAC, and biliary brushings were negative for malignancy. EUS was repeated and showed a hypoechoic pancreas. EUSguided Tru-cut biopsy of the pancreatic body showed chronic pancreatitis with a lymphoplasmacytic inltrate and storiform brosis (Figure 2C). Immunostains showed many IgG4-positive cells, supporting the diagnosis of AIP (Figure 2D). The lymphocytic inltrate was composed of both CD3- and CD20-positive lymphocytes. Stains for and light chains did not show light chain restriction. Peripheral blood ow cytometric immunophenotyping showed no monotypic B-cell population. T- and B-cell quantitative markers showed normal CD4 and natural killer cell numbers. The serum IgG4 level was 1480 mg/dL. A diagnosis was made of AIP and IAC with ocular involvement, refractory to steroids and 6-mercaptopurine, which were discontinued. Rituximab immunotherapy was initiated at 375 mg/m2 weekly for 4 doses in November of 2006. Within a month of initiating rituximab therapy the patients sense of well being improved steadily; he regained weight and resumed normal activities. Vision in his right eye returned to normal. His jaundice and symptoms of steatorrhea resolved. An uppergastrointestinal hemorrhage occurred and was attributed to gastric varices; a transjugular liver biopsy in January of 2007 showed cirrhosis with rare, scattered IgG4-positive cells. Oral propranolol was begun. Repeat ophthalmologic examination of the right eye showed the choroidal lesions were now atrophic and inactive. The patient was placed on maintenance rituximab 375 mg/m2 every 3 months. ERCP performed 4 months after initiation of rituximab showed improvement in his hilar biliary strictures (Figure 1C) and his biliary stents were removed. The serum IgG4 level was 1020 mg/dL at the time of stent removal. He remained anicteric and without evidence of cholangitis. Nine months after stent removal his total serum bilirubin level was 1.1 mg/dL, his alkaline phosphatase level was 267 IU/L (normal, 45115 IU/L), his aspartate aminotransferase level was 58 IU/L, his alanine aminotransferase level was 40 IU/L, and his IgG4 level was 299 mg/dL. The patient consented to review of his medical record for research purposes.
obliterative phlebitis. Bile duct strictures are reported in a third of patients with AIP and have been termed IAC.3 IAC also may occur in patients without pancreatitis. Stenoses may occur in the intrapancreatic bile duct, biliary conuence, and/or peripheral intrahepatic ducts. Histology of resected IAC shows IgG4positive lymphoplasmacytic inltrates in the bile duct wall.4 Similar inltrates have been reported in some cases of thyroiditis, Mikuliczs disease, chronic sialadenitis, retroperitoneal brosis, tubulointerstitial nephritis, and autoimmune hypophysitis,57 and these all may be manifestations of an IgG4-related systemic disease.1,8 IAC typically responds to oral steroid therapy.1,3,4,9 Relapse after withdrawal of steroids is common, particularly with hilar or intrahepatic duct strictures, and additional steroid or immunomodulator therapy often is successful.4 The patient described in this report did not respond to steroid or 6-mercaptopurine therapy. Rituximab therapy resulted in cholangiographic improvement and allowed his biliary stents to be removed. Rituximab is a chimeric IgG1 monoclonal antibody directed against CD20, a phosphoprotein expressed on the surface of B lymphocytes. Initially used to treat lymphoma, rituximab has efcacy in some immune-mediated diseases, including rheumatoid arthritis,10 thrombotic thrombocytopenic purpura,11 benign orbital pseudolymphoma,12 and pemphigus vulgaris13 (a disease characterized by the presence of pathogenic IgG4 antibodies). The mechanism of action of rituximab in these diseases is presumed to be B-cell depletion, resulting in decreased production of pathogenic autoantibodies. Because the patient was treated solely with rituximab, which depletes only B cells, this case suggests that B lymphocytes play a central role in the pathogenesis of IAC and AIP. The lymphoplasmacytic inltrates that characterize AIP include polyclonal B cells, plasma cells, and T cells.14 In addition to inltrating affected tissues, B lymphocytes affect the inammatory response by interaction with regulatory T cells.15 The patients serum IgG4 level initially remained increased, sharply declining toward normal 7 months after rituximab therapy was begun.
Discussion
AIP is an inammatory disorder characterized by pancreatic enlargement, multifocal pancreatic duct strictures, increased serum IgG4 levels, and a pancreatic IgG4-positive lymphoplasmacytic inltrate with narrowing of small ducts and
Figure 2. (A) Biliary biopsy specimen showing a lymphoplasmacytic inltrate. (B) IgG4 stain of the biliary biopsy specimen shows abundant positive plasma cells, consistent with IAC. (C) Pancreas Tru-cut biopsy shows features of chronic pancreatitis with a lymphoplasmacytic inltration and storiform brosis, consistent with AIP. (D) Pancreatic IgG4 stain shows abundant positive plasma cells.
366
TOPAZIAN ET AL
The initially persistent increase of IgG4 levels is not surprising because immunoglobulins are secreted by mature plasma cells that are not depleted by rituximab. This phenomenon also has been reported in patients with rheumatoid arthritis10 and Waldenstroms macroglobulinemia16 who receive rituximab. The patients clinical and cholangiographic response occurred while his serum IgG4 level was increased persistently, suggesting that IgG4 was not directly pathogenic. Cirrhosis in this case most likely was caused by chronic biliary obstruction caused by IAC. A liver biopsy performed 2 months after the initiation of rituximab therapy did show rare, scattered IgG4-positive cells. By improving biliary strictures, rituximab therapy might lead to some reversal of hepatic brosis.17,18 Could our patient have a biliary malignancy rather than IAC? Pseudotumors of the biliary tree have been described in association with AIP.19 The diagnosis of IAC in this case was based on the marked increase of the serum IgG4 levels, the presence of co-existent AIP, and compatible biliary histology. A positron emission tomography scan showed mild to moderate activity suggestive of inammation rather than malignancy. Histologic and hematologic evaluations for lymphoma were unrevealing. It is unlikely that this patients biliary stricture was malignant. The required duration of rituximab therapy is unclear. The short-term efcacy of rituximab therapy has been documented in rheumatoid arthritis10 and other inammatory diseases, but the duration of response and the benet of maintenance therapy is uncertain. Further data are needed in this regard. In conclusion, rituximab therapy was effective in this case of refractory IgG4-associated cholangitis. Rituximab may be a treatment option for patients with refractory or recurrent IgG4associated disease. Rituximab has few side effects, and may be an attractive therapeutic option in patients who are intolerant of steroids or suffer relapses of IgG4-associated disease, as well as in patients such as this one with AIP and refractory IgG4associated cholangitis. References
1. Bjrnsson E, Chari S, Smyrk T, et al. IgG4 associated cholangitis: description of an emerging clinical entity based on review of the literature. Hepatology 2007;45:15471554. 2. Ghazale A, Chari S, Smyrk T, et al. Value of serum IgG4 in the diagnosis of autoimmune pancreatitis and in distinguishing it from pancreatic cancer. Am J Gastroenterol 2007;102:1646 1653. 3. Nishino T, Toki F, Oyama H, et al. Biliary tract involvement in autoimmune pancreatitis. Pancreas 2005;30:76 82. 4. Ghazale A, Chari S, Zhang L, et al. IgG4-associated cholangitis (IAC): clinical prole and response to therapy. Gastroenterology 2008;134:706 715. 5. Ohara H, Nakazawa T, Sano H, et al. Systemic extrapancreatic lesions associated with autoimmune pancreatitis. Pancreas 2005;31:232237.
6. Yamamoto M, Takahashi H, Ohara M, et al. A new conceptualization for Mikuliczs disease as an IgG4-related plasmacytic disease. Mod Rheumatol 2006;16:335340. 7. Chari S, Smyrk T, Levy M, et al. Diagnosis of autoimmune pancreatitis: the Mayo Clinic experience. Clin Gastroenterol Hepatol 2006;4:1010 1016. 8. Hamed G, Tsushima K, Yasuo M, et al. Inammatory lesions of the lung, submandibular gland, bile duct and prostate in a patient with IgG4-associated multifocal systemic brosclerosis. Respirology 2007;12:455 457. 9. Matsushita M, Yamashina M, Ikeura T, et al. Effective steroid pulse therapy for the biliary stenosis caused by autoimmune pancreatitis. Am J Gastroenterol 2007;102:220 221. 10. Cohen S, Emery P, Greenwald M, et al. Rituximab for rheumatoid arthritis refractory to anti-tumor necrosis factor therapy: results of a multicenter, randomized, double-blind, placebo-controlled, phase III trial evaluating primary efcacy and safety at twenty-four weeks. Arthritis Rheum 2006;54:27932806. 11. Scully M, Cohen H, Cavenagh J, et al. Remission in acute refractory and relapsing thrombotic thrombocytopenic purpura following rituximab is associated with a reduction in IgG antibodies to ADAMTS-13. Br J Haematol 2007;136:451 461. 12. Witzig T, Inwards D, Habermann T, et al. Treatment of benign orbital pseudolymphomas with the monoclonal anti-CD20 antibody rituximab. Mayo Clin Proc 2007;82:692 699. 13. Ahmed A, Spigelman Z, Cavacini L, et al. Treatment of pemphigus vulgaris with rituximab and intravenous immune globulin. N Engl J Med 2006;355:17721779. 14. Kojima M, Sipos B, Klapper W, et al. Autoimmune pancreatitis: frequency, IgG4 expression, and clonality of T and B cells. Am J Surg Pathol 2007;31:521528. 15. Yang Z, Novak A, Ziesmer S, et al. Attenuation of CD8() T-cell function by CD4()CD25() regulatory T cells in B-cell nonHodgkins lymphoma. Cancer Res 2006;15:1014510152. 16. Ghobrial I, Fonseca R, Greipp P, et al. Initial immunoglobulin M are after rituximab therapy in patients diagnosed with Waldenstrom macroglobulinemia: an Eastern Cooperative Oncology Group Study. Cancer 2004;101:25932598. 17. Bonis P, Friedman S, Kaplan M. Is liver brosis reversible? N Engl J Med 2001;344:452 454. 18. Hammel P, Couvelard A, OToole D, et al. Regression of liver brosis after biliary drainage in patients with chronic pancreatitis and stenosis of the common bile duct. N Engl J Med 2001;344: 418 423. 19. Kurihara T, Itoi T, Sofuni A, et al. Pseudotumor of the bile duct associated with autoimmune pancreatitis. Endoscopy 2007; epub ahead of print.
Address requests for reprints to: Dr Mark Topazian, Miles and Shirley Fiterman Center for Digestive Diseases, Department of Medicine, Division of Gastroenterology and Hepatology, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55906. e-mail: topazian.mark@ mayo.edu; fax: (507) 266-3939. Dr Witzig serves on the Genentech and BiogenIDEC advisory boards and receives research support from these companies for clinical trials. This information was disclosed to participants.
LETTER TO THE EDITOR

Readers are encouraged to write letters to the editor concerning articles that have been published in CLINICAL GASTROENTEROLOGY AND HEPATOLOGY. Short, general comments are also considered, but use of the Letters to the Editor section for publication of original data in preliminary form is not encouraged. Letters should be typewritten and submitted electronically to http://www.editorialmanager.com/cgh. Please be sure to send 2 hard copies of any gures to the editorial ofce.
Ribavirin Posology, Observance, and Sustained Virologic Response

Dear Editor: The article by Reddy et al1 seems very encouraging for clinicians and patients; minor ribavirin dose reductions do not appear to signicantly affect sustained virologic response (SVR) of genotype 1 hepatitis C with 48 weeks of treatment if cumulative exposure is more than 60% of the prescribed dose. However, it is also confusing. The seminal article used for Food and Drug Administration and European Agency enregistration pointed out that 1000 1200 mg ribavirin is better than 800 mg ribavirin for 48 weeks of treatment of genotype 1 hepatitis C.2 This dose difference is between 60% and 100%! It will be problematic if clinicians conclude that the efcacy difference between ribavirin posology of 100% and, for example, 70% is very small, in opposition to the rule 80-80-80.2 Importance of compliance and observance will be weakened and also the interest of prescribing erythropoietin in case of ribavirin-induced anemia. When looking more carefully at pooled data in the article by Reddy et al,1 this apparent discordance could be explained; in that article there is no analysis taking into account the results by low or high pretreatment viremia. In the article by Hadziyannis et al,3 SVR difference between 24-week treatment with 1000 1200 mg ribavirin (42%) and 48-week treatment with 800 mg ribavirin (41%) is not signicant, especially in case of low pretreatment viral load (52% and 55%, respectively). Recently the possibility to treat only for 24 weeks patients with low pretreatment viral load and rapid virologic response (RVR) at week 4 has been shown.4,5 So it is not surprising to illustrate that 48 weeks with standard dose of ribavirin or only greater than 60% dose of ribavirin is equivalent if the data analysis does not differentiate low and high pretreatment viral load. In genotype 1 patients, low pretreatment viral load patients especially with RVR are easier to treat than high viral load patients, and a decrease of ribavirin dose could be effective with less consequences in SVR than for high viral load patients if treatment duration is 48 weeks; 24 weeks with full dose 1000 1200 mg ribavirin in case of low pretreatment viral load is the standard of care if RVR is obtained. The real clinical questions from the article by Reddy et al1 are in case of high pretreatment viral load, is 60% of ribavirin dose
equivalent to standard dose (1000 1200 mg) ribavirin during 48 weeks? In case of low pretreatment viral load and RVR, is 60% of ribavirin dose equivalent to standard dose (800 mg) ribavirin during 24 weeks? Therefore, we ask Reddy et al for data reanalysis, taking into account the pretreatment viral load, if possible with different cutoff (the most discriminant cutoff could be around 400,000 IU/mL6) and also the RVR. P. COUZIGOU J. FOUCHER L. CASTERA V. DE LEDINGHEN HepatoGastroEnterology Department University Victor Segalen Bordeaux Haut Levque Hospital Pessac, France
1. Reddy KR, Shiffman ML, Morgan TR, et al. Impact of ribavirin dose reductions in hepatitis C virus genotype 1 patients completing peginterferon alfa-2a/ribavirin treatment. Clin Gastroenterol Hepatol 2007;5:124 129. 2. Mc Hutchison JG, Manns M, Patel K, et al. Adherence to combinaison therapy enhances sustained response in genotype 1-infected patients with chronic hepatitis C. Gastroenterology 2002; 123:10611069. 3. Hadziyannis SJ, Sette H Jr, Morgan TR, et al. Peginterferon alpha 2a (40 KD) and ribavirin combination therapy in chronic hepatitis C: randomized study of the effect of duration and ribavirin dose. Ann Intern Med 2004;140:346 355. 4. Zeuzem S, Buti M, Ferenci P, et al. Efcacy of 24 weeks treatment with peginterferon alfa-2b plus ribavirin in patients with chronic hepatitis C infected with genotype 1 and low pretreatment viremia. J Hepatol 2006;44:97103. 5. Jensen D, Morgan TR, Marcellin P, et al. Early identication of HCV genotype 1 patients responding to 24 weeks peginterferon alpha-2a (40 kd)/ribavirin therapy. Hepatology 2006;43:954 956. 6. Zeuzem S, Fried M, Reddy KR, et al. Improving the clinical relevance of pre-treatment viral load as a predictor of sustained virological response SVR in patients infected with hepatitis C genotype 1 treated with peginterferon alpha2a (40KD) (pegasys) plus ribavirin (copegus). Hepatology 2006;44:267A. doi:10.1016/j.cgh.2007.09.008
The Th e Official fficial Clinical linical Practice Journal urnal of t the e AGA Institute
Information for Readers

SUBSCRIPTIONS STATEMENT OF RESPONSIBILITY
Correspondence regarding subscriptions or change of address should be directed to CLINICAL GASTROENTEROLOGY AND HEPATOLOGY, Elsevier, Periodicals Department, 6277 Sea Harbor Dr, Orlando, FL 32887-4800. Change of address notices, including both the old and new addresses of the subscriber and the mailing label, should be sent at least one month in advance. Customer Service: (800) 654-2452; outside the United States and Canada, (407) 563-6020. Yearly subscription rates: United States and possessions: individuals, $224; institutions, $396; students and residents, $114; single issues, $38. All other countries: individuals, $279; institutions, $455; students and residents, $153; single issues, $38. For all areas outside the United States and possessions, there is no additional charge for surface delivery. To receive student/resident rate, orders must be accompanied by name of afliated institution, date of term, and the signature of program/residency coordinator on institution letterhead. Orders will be billed at individual rate until proof of status is received. Checks should be made payable to Elsevier and sent to Clinical Gastroenterology and Hepatology, Elsevier, Periodicals Department, PO Box 628239, Orlando, FL 32862-8239. Prices are subject to change without notice. Current prices are in effect for back volumes and back issues. Single issues, both current and back, exist in limited qualities and are offered for sale subject to availability. Back issues sold in conjunction with a subscription are on a prorated basis.
CLAIMS FOR MISSING ISSUES
The ideas and opinions expressed in Clinical Gastroenterology and Hepatology do not necessarily reect those of the American Gastroenterological Association (AGA) Institute, the Editor, or the Publisher. Publication of an advertisement or other product mention in Clinical Gastroenterology and Hepatology should not be construed as an endorsement of the product or the manufacturers claims. Readers are encouraged to contact the manufacturer with any questions about the features or limitations of the products mentioned. Neither the American Gastroenterological Association (AGA) Institute nor the Publisher assumes any responsibility for any injury and/or damage to persons or property arising out of or related to any use of the material contained in this periodical. The reader is advised to check the appropriate medical literature and the product information currently provided by the manufacturer of each drug to be administered to verify the dosage, the method and duration of administration, or contraindications. It is the responsibility of the treating physician or other health care professional, relying on independent experience and knowledge of the patient, to determine drug dosages and the best treatment for the patient.
PERMISSIONS
Claims for missing issues can be honored only up to 3 months for domestic addresses and 6 months for foreign addresses. Duplicate copies will not be sent to replace ones undelivered through failure to notify Elsevier of change of address.
REPRINTS
For 100 or more copies of an article in this publication, please contact the Commercial Reprints Department, Elsevier Science Inc., 360 Park Avenue South, New York, New York 10010-1710. Tel: (212) 633-3813; fax: (212) 633-3820; email: reprints@elsevier.com.
ADVERTISING REPRESENTATIVE
For display advertising, contact: Cunningham Associates, 180 Old Tappan Rd, Old Tappan, NJ 07675; telephone (201) 767-4170; fax (201) 767-8065. For classied advertising, contact: Jay Hong, 360 Park Avenue South, New York, NY 10010-1710; telephone (212) 633-3713; fax (212) 6333820.
Copyright 2008 by The AGA Institute. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means now or hereafter known, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the Publisher. Printed in the United States of America. Correspondence regarding permission to reprint all or part of any article published in this journal may be sought directly from Elseviers Global Rights Department in Oxford, UK: phone 215-239-3804 or 44 (0)1865 843830, fax 44 (0)1865 853333, e-mail healthpermissions@elsevier.com. Requests may also be completed on-line via the Elsevier homepage (www. elsevier.com/permissions). The appearance of the code at the bottom of the rst page of an article in this journal indicates the copyright owners consent that copies of the article may be made for personal or internal use, or for the personal or internal use of specic clients, for those registered with the Copyright Clearance Center, Inc. (222 Rosewood Drive, Danvers, MA 01923; (978) 750-8400; www. copyright.com). This consent is given on the condition that the copier pay the stated per-copy fee for that article through the Copyright Clearance Center, Inc. for copying beyond that permitted by Sections 107 or 108 of the US Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale. Absence of the code indicates that the material may not be processed through the Copyright Clearance Center, Inc.
EDITOR
EDITORIAL STAFF
C. Mel Wilcox
SENIOR ASSOCIATE EDITOR
Michael B. Fallon
ASSOCIATE EDITORS
Miguel R. Arguedas Mohamad A. Eloubeidi Charles O. Elson
Thoba Khumalo, Managing Editor Brook Simpson, Assistant Managing Editor Sarah Williamson, Medical Illustrator Lindsey Guerin, Editorial Assistant Shannon Dean, Editorial Assistant Erin Dubnansky, Senior Director of Scholarly Publishing Christine Charlip, Division Director of Publications
EDITORIAL BOARD
Neena Abraham, Houston, TX Julio Bai, Buenos Aires, Argentina Alan Barkun, Montreal, Canada Deepak Bhasin, Chandigarh, India Laurence Blendis, Tel Aviv, Israel Steven R. Brandt, Baltimore, MD Alan Buchman, Chicago, IL Peter Bytzer, Copenhagen, Denmark Michael Camilleri, Rochester, MN Marcia Canto, Baltimore, MD Brooks Cash, Bethesda, MD Naga Chalasani, Indianapolis, IN Henry L. Chan, Hong Kong, China William D. Chey, Ann Arbor, MI Marcia Cruz-Correa, San Juan, PR Byron Cryer, Dallas, TX Robert Fraser, Daw Park, South Australia Michael Goggins, Baltimore, MD Benjamin Gold, Atlanta, GA Takuji Gotoda, Tokyo, Japan Ian Gralnek, Haifa, Israel Steven-Huy B. Han, Los Angeles, CA Stephen A. Harrison, San Antonio, TX Jeremy Jass, Montreal, Quebec Sunanda Kane, Chicago, IL Meredith Kilgore, Birmingham, AL Richard Kozarek, Seattle, WA Uri Ladabaum, San Francisco, CA Angel Lanas, Zaragoza, Spain Michael Levy, Rochester, MN Stephen McClave, Louisville, KY Klaus Monkemuller, Magdeburg, Germany Koenraad Mortele, Boston, MA John Pandolno, Chicago, IL Eamonn Quigley, Cork, Ireland Jose Remes-Troche, Mexico City, Mexico David Rubin, Chicago, IL Bruce Sands, Boston, MA Mark Schattner, New York, NY Robert Schoen, Pittsburgh, PA Nick Shaheen, Chapel Hill, NC Prateek Sharma, Kansas City, MO Kirti Shetty, Washington, DC Rhonda F. Souza, Dallas, TX Brennan Spiegel, Los Angeles, CA Richard Sterling, Richmond, VA Jonathan P. Terdiman, San Francisco, CA Pier A. Testoni, Milan, Italy Radu Tutuian, Zurich, Switzerland Shyam Varadarajulu, Birmingham, AL William Whitehead, Chapel Hill, NC Khay-Guan Yeoh, Singapore, Republic of Singapore
EDITOR EMERITUS
Michael Camilleri
OFFICERS OF THE AGA INSTITUTE
Nicholas F. LaRusso, President Rochester, Minnesota Gail Hecht, Vice President Chicago, Illinois
Robert S. Sandler, President-Elect Chapel Hill, North Carolina Damian H. Augustyn, Secretary/Treasurer San Francisco, CA
AGA Institute: AGA Education, Practice and Research

Published by Elsevier, 1600 John F. Kennedy Boulevard, Philadelphia, Pennsylvania 19103-2899. Clinical Gastroenterology and Hepatology (ISSN 1542-3565), 2008 by the AGA Institute, is published monthly by Elsevier, 360 Park Avenue South, New York, NY 10010. Business and Editorial Ofces: 1600 John F. Kennedy Boulevard, Suite 1800, Philadelphia, PA 19103-2899. Accounting and Circulation Ofces: 6277 Sea Harbor Dr, Orlando, FL 32887-4800. Periodicals postage paid at New York, NY and additional mailing ofces. Postmaster: Send change of address to Clinical Gastroenterology and Hepatology, Elsevier, Periodicals Department, 6277 Sea Harbor Dr, Orlando, FL 32887-4800.
Contents
IMAGES OF THE MONTH
xxiv Metastatic Malignant Melanoma of the Gastrointestinal Tract K. R. Parisian, J. E. McFarland, and A. N. Shah xxvi Transmural Perforation of the Stomach by a Fishbone I. J. Nastaskin, K. Ransibrahmanakul, and W. Trudeau xxviii Hairball in the Stomach: A Case of Gastric Trichobezoar K. Hisamuddin and C. P. Brandt
Vol. 6, No. 3, March 2008
ELECTRONIC IMAGES OF THE MONTH

Available only online at www.cghjournal.org
e11 Gastric Outlet Obstruction Caused by a Large Gallstone in the Duodenum (Bouverets
Syndrome) K. Kishi, K. Yamada, and T. Sugiyama

e12 Mirizzi With Pre-Bouverets Syndrome A. Shakouri and S.-J. Tang e15 Bouverets Syndrome: Diagnosis and Endoscopic Treatment L. E. V. V. C. Ferreira, M. D. Topazian, and T. H. Baron
CME ACTIVITIES
Take the CME exams online at http://www.cghjournal.org/content/cme
261 Exam 1: Pancreatic Cancer: A Review of the Evidence on Causation M. R. Arguedas 263 Exam 2: What Is the Role of Serological Testing in Celiac Disease? A Prospective, Biopsy-
Conrmed Study With Economic Analysis M. R. Arguedas
ON THE COVER
(Top Figure) Reprojected CT image during the portal venous phase of study (anteroposterior view of the abdomen). The patient has chronic occlusion of the main portal vein, immediately superior to the superior mesenteric vein (open arrow). The veins of the stoma (solid arrow) are enlarged, congested, and in communication with multiple prominent body wall collateral vessel (curved arrows) See Speir et al on page 346. (Bottom Figure) Esophagogastroduodenoscopy displaying a sh bone penetrating the stomach wall (arrow). See Nastaskin et al on page xxvi.

264 Visit CGH online at www.cghjournal.org to link to these and additional papers of interest C. M. Wilcox
ELECTRONIC ABSTRACTS FROM AROUND THE WORLD

e17 Available only at www.cghjournal.org C. M. Wilcox
EDITORIAL
266 Photodynamic Therapy: Standard of Care for Palliation of Cholangiocarcinoma? T. H. Baron
REVIEWS
link
268 Chronic Hepatitis B: Preventing, Detecting, and Managing Viral Resistance E. B. Keeffe, D. T. Dieterich, J.M. Pawlotsky, and Y. Benhamou
See companion article by Deng G et al on page 716 in the March issue of Gastroenterology.
E 275 Pancreatic Cancer: A Review of the Evidence on Causation

CME
A. R. Hart, H. Kennedy, and I. Harvey
CLINICAL IMAGING
283 Computerized Tomography Enterography and Its Role in Small-Bowel Imaging J. G. Fletcher, J. Huprich, E. V. Loftus, Jr, D. H. Bruining, and J. L. Fidler
ENDOSCOPY CORNER
290 Unresectable Cholangiocarcinoma: Comparison of Survival in Biliary Stenting Alone
Versus Stenting With Photodynamic Therapy M. Kahaleh, R. Mishra, V. M. Shami, P. G. Northup, C. L. Berg, P. Bashlor, P. Jones, K. Ellen, G. R. Weiss, C. M. Brenin, B. E. Kurth, T. A. Rich, R. B. Adams, and P. Yeaton
298 Performance Characteristics of the Suspected Blood Indicator Feature in Capsule
Endoscopy According to Indication for Study J. M. Buscaglia, S. A. Giday, S. V. Kantsevoy, J. O. Clarke, P. Magno, E. Yong, and G. E. Mullin
ORIGINAL ARTICLESALIMENTARY TRACT

link
302 Patient Predictors of Esophageal Stricture Development After Photodynamic Therapy P. Yachimski, W. P. Puricelli, and N. S. Nishioka
See companion article by Curvers W et al on page 670 in the March issue of Gastroenterology.
309 Peptic Ulcerations Are Related to Systemic Rather Than Local Effects of Low-Dose
Aspirin M. G. H. Van Oijen, J. P. Dieleman, R. J. F. Laheij, M. C. J. M. Sturkenboom, J. B. M. J. Jansen, and F. W. A. Verheugt
E 314 What Is the Role of Serologic Testing in Celiac Disease? A Prospective, Biopsy-Conrmed
CME
Study With Economic Analysis A. D. Hopper, M. Hadjivassiliou, D. P. Hurlstone, A. J. Lobo, M. E. McAlindon, W. Egner, G. Wild, and D. S. Sanders
321 Symptom Severity but Not Psychopathology Predicts Visceral Hypersensitivity in
Irritable Bowel Syndrome P. P. J. Van der Veek, Y. R. Van Rood, and A. A. M. Masclee
329 Chronic Abdominal Pain and Depressive Symptoms: Analysis of the National
Longitudinal Study of Adolescent Health N. N. Youssef, K. Atienza, A. L. Langseder, and R. S. Strauss

333 Sharing Genetic Test Results in Lynch Syndrome: Communication With Close and
Distant Relatives E. M. Stoffel, B. Ford, R. C. Mercado, D. Punglia, W. Kohlmann, P. Conrad, A. Blanco, K. M. Shannon, M. Powell, S. B. Gruber, J. Terdiman, D. C. Chung, and S. Syngal
ORIGINAL ARTICLESLIVER, PANCREAS, AND BILIARY TRACT

339 Prognostic Implications of Lactate, Bilirubin, and Etiology in German Patients With
Acute Liver Failure J. Hadem, P. Stiefel, M. J. Bahr, H. L. Tillmann, K. Rifai, J. Klempnauer, H. Wedemeyer, M. P. Manns, and A. S. Schneider
346 Bleeding Stomal Varices: Case Series and Systematic Review of the Literature B. J. Spier, A. A. Fayyad, M. R. Lucey, E. A. Johnson, M. Wojtowycz, L. Rikkers, B. A. Harms, and M. Reichelderfer 353 Dietary Counseling Versus Dietary Supplements for Malnutrition in Chronic
Pancreatitis: A Randomized Controlled Trial S. Singh, S. Midha, N. Singh, Y. K. Joshi, and P. K. Garg
BRIEF COMMUNICATIONS
360 Anti-Aging Therapy With Human Growth Hormone Associated With Metastatic Colon
Cancer in a Patient With Crohns Colitis G. Y. Melmed, S. M. Devlin, G. Vlotides, D. Dhall, S. Ross, R. Yu, and S. Melmed
link
364 Rituximab Therapy for Refractory Biliary Strictures in Immunoglobulin G4 Associated
Cholangitis M. Topazian, T. E. Witzig, T. C. Smyrk, J. S. Pulido, M. J. Levy, P. S. Kamath, and S. T. Chari

See companion article by Ghazale AH et al on page 706 in the March issue of Gastroenterology.
LETTER TO THE EDITOR

367 Ribavirin Posology, Observance, and Sustained Virologic Response P. Couzigou, J. Foucher, L. Castera, and V. de Ledinghen
Access to the full content of Clinical Gastroenterology and Hepatology Online is available to all subscribers! New! AGA members have seamless access to full Clinical Gastroenterology and Hepatology content from the AGA Web site. Simply sign in to http://www.cghjournal.org, visit the Publications section of the Website, and click on Clinical Gastroenterology and Hepatology. You will be directed to the Journal Website and will have access to all content without having to supply a different username and password. Members can also visit http:// www.cghjournal.org directly and click on AGA Memebers Log-in. Non-member subscribers must create an online user account and activate their subscription to access the full text of articles on Clinical Gastroenterology and Hepatology Online. To activate your individual online subscription, please visit http://www.cghjournal.org and click on Activate Online Access. To activate your account, you will your subscriber acccount or member number, which you can nd on your mailing label (AGA members must use their 10-digit member number, including leading zeros). If you have any questions regarding your online subscription, please call Elsevier Customer Service at (800) 654-2452. Sample mailing label This is your American Gastroenterological Association member number
0000101864 JANE DOE 531 MAIN ST CENTER CITY, NY 10001
1GAST V91-4 1234567-8 J.H. DOE, MD 531 MAIN ST CENTER CITY, NY 10001
This is your subscription account number
Institutional access to Clinical Gastroenterology and Hepatology Online will be allowed only by limited site licensing. Further instructions will be available online. Personal subscriptions to Clinical Gastroenterology and Hepatology Online are for individual use only and may not be transferred. Use of Clinical Gastroenterology and Hepatology Online is subject to agreement to the terms and conditions of use as indicated online.
INFORMATION FOR AUTHORS
Content CLINICAL GASTROENTEROLOGY AND HEPATOLOGY publishes original papers, review articles, brief communications, and special category articles on all aspects of the digestive tract and the liver. Manuscripts must be prepared in accordance with the Uniform Requirements for Manuscripts Submitted to Biomedical Journals developed by the International Committee of Medical Journal Editors (http://www.icmje.org). Ethics CLINICAL GASTROENTEROLOGY AND HEPATOLOGY strongly discourages the submission of more than one article dealing with related aspects of the same study. In almost all cases, a single study is best reported in a single paper. The Journal editors consider research/publication misconduct to be a serious breach of ethics and will take action as necessary to address such misconduct, which includes submission or publication of information that: Is intentionally erroneous,

manuscript be automatically posted to PubMed Central (PMC). If the author wishes to make this request, he or she should check the appropriate box located on the Copyright Assignment Form under the NIH Funding portion of the form. In addition, the author must provide the NIH grant number. If an author makes this request, the manuscript will automatically be entered in the PMC system and will be available to the public, free of charge, twelve months from the manuscripts publication date. Disclosure Policy I. Authors (Original Articles and Review Articles) The following information must be included on the title page of submitted manuscripts: A list for all authors that discloses any nancial arrangement (e.g., consultancies, stock ownership, equity interests, patent-licensing arrangements, research support, major honoraria, etc.) they may have with a company whose product gures prominently in the submitted manuscript or with a company making a competing product. Interactions that occur from the start of the research activity in the specic program until the time when the paper is anticipated to be published or one year from submission date, whichever is longer, are pertinent. In the absence of any conict of interest, authors must make the statement that there is no conict to disclose. A statement of all funding sources supporting the work and all institutional or corporate afliations. Individuals who provided writing assistance for the manuscript and the funding source for this assistance must be disclosed. Investigators must disclose potential conicts to study participants and must state whether they have done so. All phase II or III trials pertaining to a commercial product (pharmaceutical or device) require: a. A statement that the statistical analysis of the entire data sets pertaining to efcacy (specically primary and major secondary efcacy endpoints) and safety (specically, serious adverse events as dened in federal guidelines) have been independently conrmed by a biostatistician who is not employed by the corporate entity (the name of the biostatistician must be provided); and b. A statement from the corresponding author that he or she had full access to all of the data and takes full responsibility for the veracity of the data and analysis. All authors for each manuscript will also need to complete and submit to the journals editorial ofce the Authorship Responsibility portion of the Copyright Assignment, Authorship Responsibility, NIH Funding, Financial Disclosure, Institutional Review Board/ Animal Care Committee Approval, and Sponsorship form upon submission of a manuscript.
Has been published elsewhere by a different author without acknowledgment (plagiarism), Has been published elsewhere by the same author without acknowledgment (duplicate publication), or
Is subsequently published elsewhere by the same author without acknowledgment, attribution, or permission from the AGA Institute, as holder of the copyright, to reprint or adapt the material. Breaches in these standards may result in proscribed submission for all authors of the concerned manuscript and, when appropriate, notication of the authors institutions. All authors are fully responsible for the content of the manuscript. The publication of abstracts is not considered duplicate publication but should be disclosed in the cover letter accompanying the manuscript submission. Authorship Each author must have participated sufciently in the work to take public responsibility for the content of the paper and must approve of the nal version of the manuscript. Authorship should be based on substantive contributions to each of the following: Conception and design of the study;

Generation, collection, assembly, analysis and/or interpretation of data; Drafting or revision of the manuscript; Approval of the nal version of the manuscript.
National Institutes of Health (NIH) Funding If a manuscript is accepted for publication and was supported in part, or in whole, by the NIH, the author may request that the
INFORMATION FOR AUTHORS (Continued)
Authors must describe the role of the study sponsor(s), if any, in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the report for publication. II. Reviewers and Editorialists Reviewers and editorialists will be disqualied from reviewing or from writing an editorial if they: a. have had an ongoing collaboration, original publications or grants with the authors within the previous two years, except in the case of being a part of a multi-center group from a different site; or b. are from the same institution as the authors. Reviewers and editorialists should disclose at the time of submission of their review or editorial any nancial arrangement (e.g., consultancies, stock ownership, equity interests, patentlicensing arrangements, research support, major honoraria, etc.) they may have with a company whose product gures prominently in the submitted manuscript or with a company making a competing product. Interactions that occur from the start of the research activity in the specic program until the time when the paper is anticipated to be published or one year from submission date, whichever is longer, are pertinent.
Continuing Medical Education (CME)
site, http://www.editorialmanager.com/cgh. Complete instructions for online submission are located on the website. Note: If you have submitted your manuscript electronically, you do not need to submit a hard copy. If your manuscript is accepted you will be asked to provide the editorial ofce with hard copies of the missing materials.
Other Enquiries. Visit http://authors.elsevier.com/trackpaper.
html for the facility to track accepted articles and set up e-mail alerts to inform you of when an articles status has changed. The Authors Homepage (http://www.elsevier.com/authors) also provides detailed artwork guidelines, copyright information, frequently asked questions, and more. Contact details for questions arising after acceptance of an article, especially those relating to proofs, are provided after registration of an article for publication.
Manual Submission. Manual submissions may be mailed to: AGA Institute, 4930 Del Ray Avenue, Bethesda, MD 20814, Attention: CGH Editorial Ofce. Authors of manual submissions will incur a nonrefundable charge of $75. Manual submissions must be accompanied by a check (payable to the AGA Institute), purchase order, or credit card information (card type, card number, cardholders name, and expiration date). Title Page
Each issue of CLINICAL GASTROENTEROLOGY AND HEPATOLOGY will contain two to three continuing medical education exams associated with articles that appear in the issue. AGA members can take the exams online free of charge. Non-AGA members are required to pay a $15 processing fee. For CME exams that accompany original articles, readers can claim 1.0 AMA PRA Category 1 Credits. Reviewers of manuscripts can also claim CME credit. After reviewing a manuscript on the Journals manuscript tracking system, Editorial Manager, the reviewer will be prompted to claim up to 3.0 AMA PRA Category 1 Credits.
Cover Letter
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY strongly encourages authors to suggest 3 to 4 referees (include the mailing address, electronic address, phone, and fax numbers) and the Associate Editor they believe best qualied to review their paper. Authors may also list a non-preferred Associate Editor and non-preferred referees, but the ultimate selection of an Associate Editor and referees is at the sole discretion of the Editor and Associate Editor, respectively. State reasons for deviations, if any, from standard format and clarify any potential conict related to the exclusive nature of the publication. The cover letter must also categorize the manuscript into one of two groups: Alimentary Tract or Liver/Pancreas/Biliary. Manuscript Preparation
Submission. Submit 1 complete manuscript typed in 12-point
TitleUse no abbreviations. Limit: 120 characters with spaces. Short TitleLimit: 45 characters with spaces. AuthorsInclude rst names of all authors and name and full location of department and institution where work was performed. Grant SupportList grant support and other assistance. AbbreviationsList abbreviations alphabetically. (Note: In general, the use of abbreviations is discouraged.) CorrespondenceProvide name, complete address, e-mail address, telephone number, and fax number of corresponding author. Financial DisclosuresAll authors must disclose any nancial arrangement(s) they may have with a company whose product gures prominently in the submitted manuscript or with a company making a competing product. Writing AssistanceThe names and funding source for individuals who provided writing assistance must be listed. Abstract Limit: 250 words. Do not use abbreviations, footnotes, or references. Authors should submit a structured abstract of no more than 250 words organized into the following categories as applicable: Background & Aims: Describe the importance of the study and the precise research objective(s) or study question(s). Methods: Methods should include information on the following aspects of study design when applicable. The methods section may employ subheadings at the discretion of the author.
font size and double-spaced with 1-inch margins. Limit of 15 pages (approximately 4000 words) for original articles or about 20 pages (5000 6000) words for review articles, including tables, gures, and references (2 gures 1 typed page). All manuscripts submitted to CLINICAL GASTROENTEROLOGY AND HEPATOLOGY are made available for online review. Authors may submit their manuscripts, with gures and tables, electronically via our web-
Designdescribe the basic study design, e.g., randomized controlled trial, cross-sectional study, cohort study, case series, survey, etc. Settingspecify whether the study was conducted in a primary or tertiary care setting, in an ambulatory care clinic or hospital, in the general community, etc. Participantsindicate the number of study subjects and how they were selected, recruited, and assigned to the intervention.
Interventionreport the method of administration and duration of the intervention.
Results: Provide the main outcomes of the study, including condence intervals or P values. Report the absolute values and risk differences so that readers can determine the absolute, as well as the relative, impact of the results. Conclusions: State only conclusions that are directly supported by the evidence and the implications of the ndings. Body of Paper Describe ethical guidelines followed; cite approval of institutional human research review committee or animal welfare committee; describe in detail hazardous procedures or chemicals involved, including precautions observed. Outline statistical methods used. Identify drugs and chemicals used by generic name (if trademarks are mentioned, give manufacturer name and city). References Cite references in order of appearance in text using superscripted Arabic numerals. Cite personal communications and unpublished data directly in text without being numbered. Conform abbreviations to those used in Index Medicus. Conform style and punctuation to CLINICAL GASTROENTEROLOGY AND HEPATOLOGY requirements: Article (list 3 authors followed by et al): 13. Meltzer SJ, Ahnen DJ, Battifour H, et al. Protooncogene abnormalities in colon cancers and adenomatous polyps. Gastroenterology 1987;92:11741180. Book: 18. Day RA. How to write and publish a scientic paper. Philadelphia: Institute for Scientic Information, 1979. Article in Book: 22. Costa M, Furness JB, Llewellyn-Smith IF. Histochemistry of the enteric nervous system. In: Johnson LR, ed. Physiology of the gastrointestinal tract. Volume 1. 2nd ed. New York: Raven, 1987:1 40.
Tables
Tables should be prepared without the use of tabs; most table editor programs can be uploaded successfully. If your table contains decimal fractions, please round your numbers to two places after the decimal point.
Figures
authors upload grayscale or black-and-white les only to allow review of the data as they will ultimately be published. Figure Legends: Please do not embed or atten the text into the image les. Figure legends should be typed and submitted in .rtf (rich text format). This text will be reformatted in the style of the AGA Institute journals. Accepted Figure File Formats: We support the following among dozens of le formats: .bmp, .gif, .jpg, .pbm, .pcx, .png, .tif, .eps, .xbm, .psd, and .tga les. When sending image les, please do not embed them in Word. You may submit mixed le formats (image1.jpg, image2.tif, image3.eps, etc.). Preferred Figure File Formats: .tiff, .psd, and .jpg. If you have created Photoshop image les containing separate layers with arrows or text, please send us the layered les (unattened). Image File Formats not Supported at this Time: ChemDraw, CorelDraw, Canvas, FreeHand, Excel, SigmaPlot, Quark Xpress, and Equation Editor. You may export image les from these programs as PDF, JPEG, or other acceptable le formats. File Naming Convention: Figures should be named consecutively such as gure 1.tif, gure 2.jpg, etc., with the le extension appended (.tif, .jpg, .eps, etc). Each gure should be saved as a separate electronic le. Color Figures: Files should be submitted in the CMYK color space. Authors are encouraged to present color gures in a manner that will allow the data to be interpreted by colorblind readers. CLINICAL GASTROENTEROLOGY AND HEPATOLOGY suggests that authors present dual labeled images in green and magenta rather than in green and red. See the website if the Jy data depository for Drosophila researchers (http://jy.iam.u-tokyo.ac.jp/color/) for more information on how to make gures and presentations intelligible for a colorblind audience. Font: If your gures include text, an 8 to 10 point font should be used. Acceptable fonts are sans serif fonts such as Arial, Helvetica, and Myriad. Examples of unacceptable fonts (serif fonts) are Times, Palatino, and Georgia. Lettering should begin with an upper case letter, followed by lower case lettering. Multiple Panel Figures: Do not mount multiple part gures. Please submit each panel (image) separately. However, you may submit a multiple panel version to suggest the order in which you would like the panels arranged. You may also include a written, suggested layout. Each individual panel should be of the highest possible quality (300 dpi or higher) at actual print size.
Reporting Clinical Trials
Images: Images can be clinical, pathologic (gross or microscopic), endoscopic, or radiographic. They should be of high quality (300 dpi or greater, clear, and in good focus) and illustrate the diagnosis well. Photographs: Photographs of identiable patients must be accompanied by written permission to publish from the patient. Line Art and Graphs: Graphs, charts, and other line art may be reformatted and/or redrawn by our Medical Illustration Department (if needed) for consistency with the overall style of the AGA Institute journals. Please be sure that any graphs or line art you submit are at a resolution of at least 150 dpi so that they are readable to reviewers. Cost: Authors will be required to pay for the printing of color gures ($650 for the rst color gure and $100 each for additional gures). If the manuscript is reviewed with color gures, it must be published with color gures with printing fees paid for by the author. If the author does not wish to pay for printing color gures, then the
As of January 1, 2008, it is mandatory for authors to provide full registration of their clinical trial(s). A clinical trial is dened as any research project that prospectively assigns human subjects to intervention and comparison groups of study the cause-and-effect relationship between a medical intervention and a health outcome. The trial must have at least one prospectively assigned concurrent control or comparison group in order to trigger the requirement for registration. CLINICAL GASTROENTEROLOGY AND HEPATOLOGY has adopted the recommendations put forth by the International Committee of Medical Journal Editors (ICMJE). For more information on the ICMJE recommendations, please go to http://www.icmje.org/. Appropriate online trial registries include: www.clinicaltrials. gov, www.isrctn.org, www.umin.ac.jp/ctr/index.htm, www.actr. org.au, www.trialregister.nl, or any primary registers that participate in the World Health Organizations International Clinical Trial Registry Platform. The clinical trial registry URL and the clinical trial number must be included in the body of the manuscript and
must be provided in the manuscript management system upon submission.

Reporting Meta-Analyses of Genetic Studies
editorialmanager.com/cgh. Authors are required to follow the Image of the Month guidelines provided above.
Letters to the Editor
As of January 1, 2008 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY will require all metaanalyses of genetic studies to be registered and to follow the Human Genome Epidemiology Network (HuGENet) guidelines. To review the guidelines, go to http://www.cdc.gov/genomics/ hugenet/reviews/guidelines.htm/.
Image of the Month
Image of the Month presents a striking clinical image that is meant to challenge and inform readers. The Image of the Month is presented as a short case report including up to three images. The section is intended to illustrate and teach important medical points. If you would like to submit an image for publication in the Journal, please follow the instructions below.
Letters to the Editor offer opinions on manuscripts submitted to CLINICAL GASTROENTEROLOGY AND HEPATOLOGY. Text should not exceed 250 words with a limit of 3 references; no more than 3 authors are allowed on each submission. All letters become the property of CLINICAL GASTROENTEROLOGY AND HEPATOLOGY and are subject to editing by the editors. Letters commenting on manuscripts are sent to the authors of those manuscripts for a response. Letters are selected for their importance and not all letters submitted can be published. Letters should be submitted within a month from the publication of the manuscripts.
Brief Communications
Images can be either clinical, pathologic (gross or microscopic), endoscopic, or radiographic. They should be of high quality and illustrate the diagnosis well. Images for hardcopy submission should be original color or black and white photographic or digital prints 5 7 inches in size and should be submitted in duplicate. Arrows or other symbols, or label identiers should be marked only on the second print. The back of each print and slide must be labeled with the last names of the contributors and an arrow indicating the top of image. The case should be described in one typed, double-spaced page or less. Format should be as follows: Short pertinent history, physical examination and laboratory ndings, and initial clinical course. The image(s) (1 to 2, no more than 2) should then be described with all labeled structures explained in the text. The answer should discuss the image ndings and the diagnosis in no more than one double-spaced, typed page. The diagnosis and discussion should make an important medical teaching point and include from 1 to 3 pertinent references. Information regarding the specic patient in terms of follow-up and response to therapy should be given as appropriate. No more than three authors are allowed on each submission. Contributors must provide their names, addresses, phone, and e-mail addresses. Contributors must sign and return the copyright form which assigns copyright to the AGA Institute and attest that the gure has not been submitted or published elsewhere. Image of the Month submissions must be submitted online at http://www.editorialmanager.com/cgh.
Brief Communications are short reports of preliminary or limited results of original research, observations, or case series on the causes, mechanisms, diagnosis, course, treatment, and prevention of digestive diseases. The formatting guidelines for brief communication are as follows: Abstract must not exceed 175 words, manuscript body must not exceed 1500 words, 12 gures or tables, and 1520 references. Case Reports The journal will publish only a select number of high quality and critical case reports. Authors have the option of reformatting their case report as a Brief Communication or Image of the Month. Publication Accepted manuscripts are sent to the publisher, Elsevier, soon after acceptance. Once authors have had the opportunity to review galleys of their manuscripts, these author-corrected proofs will be uploaded to Articles in Press on www.cghjournal.org and will be indexed on PubMed, several weeks ahead of print. Manuscripts are copyedited to make them consistent with Journal style; if a particular section in the manuscript is not clear or requires additional information, the copy editor will direct questions to the author. These questions, or author queries, will appear in the margins of the proofs that are sent to the author. The time between acceptance and publication is currently about 3 months. The corresponding author can expect proofs of the article approximately 2 months after acceptance. Authors receive proofs for the primary purpose of checking the accuracy of the typesetting; authors are not to revise or rewrite their articles at this stage. If after acceptance of their paper, authors become aware of important information they believe should be added to their manuscript, they should contact the editor of CLINICAL GASTROENTEROLOGY AND HEPATOLOGY. Authors are required to return proofs to the publisher within 48 hours. Reprints. Forms for ordering article reprints will be sent with proofs to authors and should be returned with the corrected proofs. Authors do not receive free reprints, and therefore are responsible for ordering their own reprints (minimum order, 100) from the publisher.
Editorial Ofce
Electronic Image of the Month
Due to a high number of Image of the Month submissions and the increasing use of video in reports on clinical cases, CLINICAL GASTROENTEROLOGY AND HEPATOLOGY publishes Electronic Image of the Month, where some accepted images or those that include video clips can be published online only. When an image is accepted for the Electronic Image of the Month section, authors will receive a decision letter requesting approval to publish their article online only. If an author does not agree to these terms the article will not be considered further for publication. If authors choose to have their accepted image published in the Electronic Image of the Month section, their article will be posted on our website, www.cghjournal.org, within 2 to 3 months. To submit your Image of the Month to CLINICAL GASTROENTEROLOGY AND HEPATOLOGY, log-on to http://www.
The address for the submission of manuscripts or correspondence is: C. Mel Wilcox, MD, Editor, CLINICAL GASTROENTEROLOGY AND HEPATOLOGY, AGA Institute, 4930 Del Ray Avenue, Bethesda, Maryland 20814. E-mail: cgh@gastro.org; tel: (301) 654-2055 ext. 683. 2008 by the AGA Institute
Copyright Assignment, Authorship Responsibility, NIH Funding, Financial Disclosure, Institutional Review Board/Animal Care Committee Approval, and Sponsorship
Copyright Assignment. In consideration of the American Gastroenterological Association (AGA) Institute (the AGA Institute) taking action to review and credit the below-identied submission (the Manuscript), and for other valuable consideration, the receipt and sufciency of which is hereby acknowledged, the undersigned authors and/or creators (the Authors), jointly and severally, hereby transfer, convey, and assign to the AGA Institute, free and clear of any liens, licenses or encumbrances, the entire right, title, and interest in and to the Manuscript throughout the world, including without limitation in and to any and all copyrights for the Manuscript (including but not limited to rights to copy, publish, excerpt, collect royalties and make derivative works) in print, electronic, Internet, broadcast, and all other forms and media now or hereafter known, and for any and all causes of action heretofore accrued in Authors favor for infringement of the aforesaid copyrights, to have and to hold the same unto the AGA Institute, its successors and assigns, for and during the existence of the aforesaid copyrights, and all renewals and extensions thereof. At any time and from time to time hereafter, the Authors shall upon the AGA Institutes written request take any and all steps and execute, acknowledge and deliver to the AGA Institute any and all further instruments and assurances necessary or expedient in order to vest the aforesaid copyrights and causes of action more effectively in the AGA Institute. The Authors retain the nonexclusive permission to use all or part of the Manuscript in future works of their own in a noncompeting way, provided proper copyright credit is given to the AGA Institute. Should the AGA Institute nally determine that it will not publish the Manuscript, the AGA Institute agrees to assign its rights therein back to the Authors. (Note: material prepared by employees of the federal government in the course of their ofcial duties may not be copyrightable.) Authorship Responsibility. I, the undersigned Author, certify that I have participated sufciently in the intellectual content, the analysis of data, if applicable, and the writing of the Manuscript, to take public responsibility for it. I have reviewed the nal version of the Manuscript, believe it represents valid work, and approve it for publication. As an Author of this Manuscript, I certify that, except to the extent expressly credited to others in the text of the Manuscript: the entire Manuscript is an original creation of the Authors; and none of the material in the Manuscript has been published previously, is included in another manuscript, or is currently under consideration for publication elsewhere. I also certify that this Manuscript has not been accepted for publication elsewhere, and that I have not assigned, licensed, or otherwise transferred any right or interest in the Manuscript to anyone. Moreover, should the AGA Institute or the editors of CLINICAL GASTROENTEROLOGY AND HEPATOLOGY request the data upon which the Manuscript is based, I shall produce it. Authors are responsible for applying for permission for both print and electronic rights for all borrowed materials and are responsible for paying any fees related to the applications of these permissions. Institutional Review Board/Animal Care Committee Approval. I, the undersigned Author, certify that my institution has approved the protocol for any investigation involving humans or animals and that all experimentation was conducted in conformity with ethical and humane principles of research. Sponsorship. I, the undersigned author, certify that I had full access to all of the data in this study and I take responsibility for the integrity of the data and the accuracy of the data analysis.
Manuscript title: Signatures: Each Author must sign and date this statement and assignment. In the case of a work made for hire, the employer(s) must also sign. For example, for any Manuscript including any portion created in the course of employment for another (whether as a regular employee or as an independent contractor) requires the signature of the relevant employer(s).
Print Name:
Date:
Print Name:
Date:
Print Name:
Date:
Print Name:
Date:
Employer signature(s) as Author (required for works made for hire): (Employer) By: Print Name: Date:
Employer signature(s) as Author (required for works made for hire): (Employer) By: Print Name: Date:
If this Manuscript exists in the public domain because it was written as part of the ofcial duties of the Authors as employees of the U.S. government, check this box.
National Institutes of Health (NIH) Funding. My manuscript was supported in part, or in whole, by the NIH. In accordance with the NIH
Public Access Policy, I would like to request that my manuscript, should it be accepted for publication, be submitted to PubMed Central (PMC). I understand that my manuscript will therefore be freely accessible by the public via PMC twelve months from the date of its publication. I would like my manuscript automatically submitted to PMC, should it ultimately be accepted. My NIH grant number is
Financial Disclosure. Check the appropriate box and sign where indicated. All Authors must sign one of the statements below.
I, the undersigned Author, certify that I have no nancial arrangements (e.g., consultancies, stock ownership, equity interests, patent-licensing arrangements, research support, major honoraria, etc.) with a company whose product gures prominently in the submitted manuscript or with a company making a competing product except as disclosed on a separate attachment. All funding sources supporting the work are acknowledged on the title page.
I, the undersigned Author, certify that I have included on the title page of the manuscript any nancial arrangements (e.g., consultancies, stock ownership, equity interests, patent-licensing arrangements, research support, major honoraria, etc.) that I have with a company whose product gures prominently in the submitted manuscript or with a company making a competing product. All funding sources supporting the work are acknowledged on the title page.

Clinical Gastroenterology2008!06!03

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Clinical Gastroenterology2008!06!03

Uploaded by

Copyright:

Available Formats

www.cghjournal.

March 2008 Volume 6

Ofcial Clinical Practice Journal of the

Image of the Month

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 6, No. 3

Image of the Month

2008 by the AGA Institute

1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.024 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2008;6:xxvi

Image of the Month

2008 by the AGA Institute

1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.013 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2008;6:xxviii

IMAGE OF THE MONTH

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 6, No. 3

Image of the Month

2008 by the AGA Institute

1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.027 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2008;6:e11

Image of the Month

1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.006 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2008;6:e12 e13

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY

Image of the Month

2008 by the AGA Institute

1542-3565/08/$34.00 doi:10.1016/j.cgh.2007.12.055 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2008;6:e15

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2008;6:e17

ABSTRACTS FROM AROUND THE WORLD

Additional Papers of Interest

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2008;6:261262

CME ActivitiesExams 1 and 2

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 6, No. 3

Exam 1: Pancreatic Cancer: A Review of the Evidence on Causation

Contact hours: 1.0

Expiration Date: March 31, 2009

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2008;6:263

Contact hours: 1.0

Expiration Date: March 31, 2009

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2008;6:264 265

ABSTRACTS FROM AROUND THE WORLD

Your Mom Was RightEat Your Fruits and Vegetables

Take Your Probiotic Before Your Antibiotic

Capsule Endoscopy Is Not the Answer for Surveillance in FAP

ABSTRACTS FROM AROUND THE WORLD

Our Expanding Arsenal Against Hepatitis B Virus Infection

Additional Papers of Interest

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2008;6:268 274

Viral Replication and Selection of Resistant Strains

CHRONIC HEPATITIS B AND ANTIVIRAL RESISTANCE

Mutations in the Polymerase Gene and Antiviral Resistance

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 6, No. 3

Detecting, Preventing, and Managing Hepatitis B Virus Resistance

Assays for Detection of Resistance-Conferring Mutations

CHRONIC HEPATITIS B AND ANTIVIRAL RESISTANCE

Management of Patients With Resistant Virus

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 6, No. 3

Avoiding Resistance Development: Selection of Initial Therapy

CHRONIC HEPATITIS B AND ANTIVIRAL RESISTANCE

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 6, No. 3

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2008;6:275282

Pancreatic Cancer: A Review of the Evidence on Causation

See CME exam on page 262.

Family History and Genetic Disorders

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 6, No. 3

CAUSE OF PANCREATIC CANCER

Total Energy Intake

Meat H pylori Infection