Pediatr Infect Dis J, 2000;19:62530 Copyright 2000 by Lippincott Williams & Wilkins, Inc.

Vol. 19, No. 7 Printed in U.S.A.

Mycobacterium tuberculosis infection in pediatric liver transplant recipients

ANITA VERMA, MD, ANIL DHAWAN, MD, JIM J. WADE, MD, WAH H. LIM, MD, GARY RUIZ, MRCP, JOHN F. PRICE, FRCP, NEDIM HADZIC, MD, ALASTAIR J. BAKER, FRCP, MOHEMMED RELA, FRCS, NIGEL D. HEATON, FRCS AND GIORGINA MIELI-VERGANI, MD

Objectives. To study the incidence, clinical presentation, management, complications and outcome of tuberculosis in pediatric liver transplant recipients. Methods. A retrospective review of the medical records of children who underwent liver transplantation between 1991 and 1998. Results. Mycobacterium tuberculosis infection occurred in 6 of 254 (2.4%) children undergoing liver transplantation between 1991 and 1998. Cough, pyrexia and poor appetite were common presentations; one-half had normal chest radiographs. The median time to confirmation of diagnosis was 8 months (range, 1 to 17 months). Tests contributing to diagnosis included: ZiehlNeelsen (ZN) stain (2 patients), M. tuberculosis polymerase chain reaction (1 patient), Mantoux test (1 patient) and histopathology (4 patients). Family health screening was productive for 4 patients. Duration of treatment varied from 9 to 18 months. Isoniazid-induced hepatitis was observed in 2 patients but resolved with dose reduction. Two patients died while receiving treatment, one of Klebsiella spp. septicemia and the other of pulmonary hemorrhage. Conclusions. Tuberculosis after liver transplantation has a significant morbidity and mortality. Pretransplantation a personal and family history of tuberculosis must be sought, and screening of patients and their families should be considered. Standard regimens incorporating isoniazid and rifampin are effective, but regular

monitoring of liver function is essential to detect drug-induced hepatotoxicity.

INTRODUCTION

Infection continues to be a major cause of morbidity and mortality after liver transplantation. Although the overall incidence of tuberculosis in the West is falling, it is encountered increasingly in individuals immunocompromised by HIV infection, other diseases or immunosuppressive medication.15 The incidence of tuberculosis after liver transplantation in Europe and the United States is 0.9 to 2.3% with most reported cases in adults.2, 6 12 The published experience in pediatric liver transplant recipients is limited to a series of three children.13 Management of tuberculosis in such children is difficult and complicated by the interactions between immunosuppressive regimens and antituberculous drugs and the hepatotoxicity of the latter.14 16 At our center 6 of 254 (2.4%) children undergoing liver transplantation between 1991 and 1998 developed tuberculosis during follow-up. We report here the presentation, diagnosis, management, complications and outcome in these patients.
CASE REPORTS

Accepted for publication April 5, 2000. From Dulwich Public Health Laboratory and Medical Microbiology (AV, JW), Pediatric Liver Service (AD, NH, AJB, GMV), and the Pediatric Respiratory Service (WHL, GR, JFP) and Liver Transplant Surgery (MR, NDH), Kings College Hospital, London, UK. Key words: Mycobacterium tuberculosis, liver transplant, children, tuberculosis, antituberculous drugs, Mantoux test. Address for reprints: Dr. Anita Verma, Dulwich Public Health Laboratory and Medical Microbiology, Guys, Kings and St. Thomas Medical School, Bessemer Road, London SE5 9RS, UK. Fax 0171 346 3404; E-mail anil.dhawan@kcl.ac.uk.

Case reports are summarized in Table 1. Mycobacterium tuberculosis infection was seen in six children ages 7 months to 16 years at transplantation. Most (five of six) were born in the United Kingdom, and only one had a history of tuberculosis before liver transplantation. The median time from transplantation to confirmation of tuberculosis was 8 months (range, 1 to 17 months). Four patients had a household contact with a positive chest radiograph (CXR) or Mantoux test. A regimen containing isoniazid and rifampin was used for five patients. Isoniazid-induced hepatitis was observed in two patients and resolved after dose reduction. Two patients died during treatment, one of Klebsiella spp. septicemia and the other of pulmonary hemorrhage. Case 1 who had living-related liver transplantation presented with cervical lymphadenopathy which, on biopsy, revealed necrosis and granulomas, although culture was negative after 8 weeks. After 1 week of 625

626

THE PEDIATRIC INFECTIOUS DISEASE JOURNAL

Vol. 19, No. 7, July, 2000

TABLE 1. Details of history, presentation, diagnosis and outcome for six children developing
Pre- and PeriTransplantation Case Sex 1 2 F M Age 14 mo 9 mo TB history None None Diagnosis Biliary atresia Biliary atresia Immunosuppression Cyclo, Aza, and Pred Tacro and Pred Complications pre-TB therapy None Acute rejection Presentation of TB CXR

Pyrexia and cervical lymphadenopathy Cough and decreased appetite

Normal Consolidation

7 mo

None

Idiopathic ALF

Tacro and Pred

Severe acute rejection Rejection and sepsis GVHD and sepsis

Cough, pyrexia and decreased appetite Deafness, headache and abnormal gait Pyrexia

Pleural effusion

16 yr

None

Biliary atresia

Tacro and Pred

Normal

16 yr

Yes

GVHD post-BMT

Tacro and Pred

Effusion, consolidation

4 yr

None

Biliary atresia

Cyclo, Aza and Pred

None

Pyrexia and abnormal liver function

Normal

TB, tuberculosis; ALF, acute liver failure; GVHD, graft vs. host disease; BMT, bone marrow transplant; Tacro, tacrolimus; Pred, prednisolone; Cyclo, cyclosporin; BAL, pyrazinamide; Clar, clarithromycin.

antituberculous treatment the serum aspartate aminotransferase increased, and liver biopsy suggested druginduced hepatitis. The isoniazid dose was halved to 5 mg/kg/day, azathioprine was withdrawn and within 2 weeks the aspartate aminotransferase returned to normal. She received 9 months of treatment and remains asymptomatic 6 years later. Case 2 initially received a regimen of cyclosporin, azathioprine and prednisolone which was changed to tacrolimus plus prednisolone after an episode of rejection. Ten weeks posttransplant he was seen with signs and symptoms compatible with tuberculosis. Isoniazid (10 mg/kg/day), rifampin (10 mg/kg/day) and pyrazinamide (20 mg/kg/day) were commenced, and immunosuppression was withdrawn. Streptomycin (15 mg/kg/ day) was added after 2 weeks because of persisting fever. The patient was apyrexic after 26 days of treatment, and immunosuppression with cyclosporin was reinstated after 1 month. Streptomycin was stopped after 8 weeks of therapy. After 3 months of therapy aspiration of an enlarged cervical lymph node revealed acid-fast bacilli. This abscess resolved after pyrazinamide was increased to 30 mg/kg/day and intramuscular streptomycin (15 mg/kg/day) was restarted. Streptomycin was continued for 5 months, pyrazinamide for 10 months and rifampin and isoniazid both for 18 months. However, 1 month after treatment was stopped graft function deteriorated, and biopsy showed a single granuloma. Biopsy culture was negative but CXR revealed new changes. Treatment was restarted with rifampin, isoniazid and pyrazinamide. At 3 months liver biopsy for abnormal liver enzymes re-

vealed drug-induced hepatitis, and antituberculosis treatment was stopped. Graft function returned to normal after 2 weeks and the patient remains asymptomatic 2 years after stopping antituberculous therapy. Case 3 developed cough and poor appetite 2 weeks after transplantation. The CXR revealed pleural effusion, microscopy and culture of pleural fluid were negative and M. tuberculosis PCR was positive. Because of poor graft function due to rejection he was given streptomycin (15 mg/kg/day), ciprofloxacin (10 mg/kg/day) and ethambutol (15 mg/kg/day) and immunosuppression was reduced. Fever and effusion had not resolved by 2 weeks, and treatment was changed to rifampin (10 mg/kg/day), pyrazinamide (20 mg/kg/day) and streptomycin (15 mg/kg/day) with clinical improvement. After retransplantation at Day 45 for early chronic rejection (biopsy revealing that more than one-half of the portal tracts were devoid of bile ducts) he developed Klebsiella spp. septicemia and died. Case 4 presented 16 months after transplantation with impaired hearing, bilateral ear discharge, headache, confusion, unsteady gait and signs of meningitis. A magnetic resonance imaging scan revealed multiple spherical enhancing cerebellar and thalamic lesions. The cerebrospinal fluid white blood cell count was 81 cells/mm3 (60% lymphocytes), with glucose and protein concentrations of 0.1 mmol/l and 2296 mg/l, respectively; ZN stain was negative. Quadruple antituberculous therapy was commenced, the tacrolimus dose was halved and steroids were stopped. After gradual neurologic improvement ethambutol was stopped at 3 months. Graft function remained normal without any

Vol. 19, No. 7, July, 2000

THE PEDIATRIC INFECTIOUS DISEASE JOURNAL

627

Mycobacterium tuberculosis infection after liver transplantation

Investigation and Diagnosis of TB Mantoux test ZN stain CSF findings M. tuberculosis culture PCR Positive Time from Family histo- transplant to health pathology confirmation screening LN Bx Liver Bx 8 mo 10 wk Anti-TB therapy Therapy and Outcome Duration ImmunoGraft Outcome/ of suppression dysfunction cause of death therapy 9 mo 18 mo Reduced INH Well hepatitis

1:10 000, Negative LP not Negative Negative 14 indicated LN Bx mm Negative BAL, Normal Positive Not done bronchial BAL Bx and and LN Bx bronchial BX Negative Negative LP not Negative Positive indicated pleural fluid Negative Middle ear Bx and fluid Negative Negative Elevated Positive WBC middle and ear Bx protein LP not Positive indicated BAL and pleural fluid LP not Negative indicated Not done

Negative INH, Rif Positive INF, Rif, Pyr SM SM, Cip Eth INH, Rif, Pyr Eth

Well Stopped for INH 2 wk hepatitis

Not done

1 mo

Positive

2 mo

Reduced

Rejection

Middle ear Bx Not done

17 mo

Positive

18 mo

Reduced

None

Died: sepsis and multiorgan failure Well

Negative pleural fluid and BAL Not done

2 mo

Not done INH, Rif, Pyr Char Positive INH, Rif Cip

2 wk

Reduced

GVHD and Died: sepsis pulmonary hemorrhage None Well

Negative Negative

Liver Bx

2 wk

12 mo

Reduced

bronchoalveolar lavage; Bx, biopsy; LN, lymph node; LP, lumbar puncture; CXR, chest radiograph; Sm, streptomycin; Cip, ciprofloxacin; Eth, ethambutol; INH, isoniazid; Pyr,

rejection while the patient was receiving antituberculosis therapy. He continues to receive the remaining three drugs with the intention of completing 18 months of treatment. Case 5 underwent allogeneic bone marrow transplantation at 16 years of age for CD24 ligand deficiency at another hospital. He had a history of tuberculous osteomyelitis at 3 years of age but did not receive isoniazid prophylaxis. He was transferred for liver transplantation for acute liver failure due to severe graft vs. host disease. Nine days later CXR revealed bilateral effusions, but ZN stain, culture and M. tuberculosis PCR of pleural fluid were negative; PCR for Cryptosporidium was positive. Pleural biopsy was not attempted because of persistent thrombocytopenia. Twenty days later a CT scan showed a focal lesion in the right lung with persisting effusions and antituberculous therapy plus clarithromycin was begun to cover atypical mycobacteria. Bronchoalveolar lavage was ZN and M. tuberculosis PCR-negative. Antituberculous treatment was stopped because of deranged graft function and persistently negative results for M. tuberculosis. He subsequently developed disseminated cryptosporidiosis with failing liver and marrow grafts and died of pulmonary hemorrhage. Culture of bronchoalveolar lavage yielded M. tuberculosis after 8 weeks, the result becoming available postmortem. Case 6 presented 48 days posttransplantation. Liver biopsy revealed epithelioid granuloma and other causes of granulomatous hepatitis were excluded. After 1 month of antituberculous therapy clinical and biochemical improvement was noted. Ciprofloxacin (10

mg/kg/day) was stopped after 3 months and she continues on rifampin (10 mg/kg/day) and isoniazid (10 mg/ kg/day) to complete a 9-month course; she is asymptomatic 6 months after start of treatment.
DISCUSSION

The incidence of tuberculosis in the United Kingdom increased in the late 1980s but the number of new cases diagnosed has since stabilized.13, 14 However, the increasing number of solid organ transplant recipients represent an expanding population of patients predisposed to opportunistic infections. In a concise worldwide review of tuberculosis in solid organ transplant recipients mainly in adults, Singh and Paterson2 found the frequency of tuberculosis in liver grafts recipients from the United States and Europe to be between 0.9 and 2.3%. At our institution tuberculosis occurred in 6 of 254 children transplanted between 1991 and 1998. This represents the largest cohort of pediatric patients from a single center. Most of our patients (five of six) were born in the United Kingdom, one of whom was of Vietnamese extraction, an observation pertinent in the context of reported experience.7, 8 In countries with a higher incidence of tuberculosis than the United Kingdom, especially developing countries, the risk of M. tuberculosis infection after transplantation is significantly higher. Clinicians in transplant centers serving low incidence populations must maintain a high index of suspicion of tuberculosis in those patients originating from high incidence areas, or having family or other close contacts from such areas. Family health screen-

628

THE PEDIATRIC INFECTIOUS DISEASE JOURNAL

Vol. 19, No. 7, July, 2000

24

ing was positive in four of our patients. McDiarmid et al.13 also reported positive skin tests in close family members of three children who developed tuberculosis after liver transplantation. These observations underline the importance of seeking a positive family history of tuberculosis and performing health screening for the families of children listed for liver transplantation. The donor organ has been implicated as a source of M. tuberculosis infection in renal transplant recipients,17 but there is no evidence to suggest that this was the source in any of our patients. Deafness is an unusual presentation of tuberculosis (Case 4), and the middle ear is an occult site of infection to be added to other rare presentations of tuberculosis in transplant recipients.18 21 None of our patients developed disseminated tuberculosis, which has been reported in onethird of patients receiving OKT3 or higher cumulative doses of immunosuppression,2 probably because our patients were receiving relatively low doses of tacrolimus or cyclosporin and were not treated with monoclonal antibodies. The median time between transplantation and confirmation of a diagnosis of tuberculosis was 8 months (range, 1 to 16 months), contrasting with 4 months reported by Singh and Paterson.2 Although tuberculosis in adult transplant recipients may result from reactivation rather than primary infection,10, 16, 22 only one of our patients had a history of tuberculosis, suggesting recurrence. Most of our patients were younger than 12 years of age, the age at which bacillus of Calmette-Guerin vaccine is administered in the United Kingdom. Making a diagnosis of tuberculosis can be difficult in the immunocompromised. Four of our patients had culture-proved disease, but the diagnosis was made only after invasive investigations such as bronchoalveolar lavage or biopsy of the middle ear, lymph node or bronchi. Although M. tuberculosis PCR on pleural fluid assisted the diagnosis in one patient, it is important to acknowledge that a negative PCR test result does not exclude TB. Similarly, although biopsies showing acid-fast bacilli and caseating granulomas help to establish the diagnosis, histopathologic changes may be nonspecific; histology showed epithelioid cell granuloma in only one of our patients and caseating granuloma in three. A negative Mantoux test in an immunocompromised patient is unreliable23 but a positive test is significant. The test is simple and should be considered when pursuing a diagnosis of tuberculosis in liver transplant patients. Interpretation of skin tests in immunocompromised children is critical; induration of 5 mm should be considered positive in high risk patients (close contact with an infected case, HIV positivity, other immunocompromised) whereas in the immunocompetent, with no risk factors, 15 mm of

induration is significant. Induration of 5 mm in HIV-infected individuals is accepted by CDC as positive and sufficient to warrant prophylaxis.25 The optimum treatment of tuberculosis in liver transplant recipients has not been established. Recent guidelines of the British Thoracic Society do not address organ transplant recipients.26 It is important to resist initiating treatment empirically in immunocompromised children if possible: aggressive investigation to establish a diagnosis must be undertaken. The closest model for the pediatric solid organ transplant recipient appears to be the HIV-infected child, but there is a paucity of reports of tuberculosis and its management in this population.4, 5 The treatment regimens used for our patients were variable. This reflects both the lack of published experience and the need to tailor therapy to the individuals condition and complications. We monitored closely the blood level of cyclosporin and tacrolimus and adjusted doses accordingly. Although we reduced immunosuppression this precipitated rejection in only one patient. Isoniazid, rifampin and pyrazinamide were avoided in one patient with severe graft dysfunction because of rejection. Goncalves et al.27 attributed 27% of lost grafts in renal transplant recipients to chronic rejection, probably due to suboptimal immunosuppression after the interactions between cyclosporin, rifampin and isoniazid.27 Aguado et al.28 reported increased mortality in patients developing acute rejection as a result of suboptimal immunosuppression resulting from drug interaction. Rifampin and isoniazid are microsomal enzyme inducers with particular affinity for liver cytochrome p450 IIIa, the major cyclosporinmetabolizing enzyme system29; concurrent administration of rifampin and cyclosporin increases metabolism and decreases bioavailability of the latter. Rifampin, isoniazid and pyrazinamide are potentially hepatotoxic.30 Hepatotoxicity is reported in 6.8% of the general population receiving isoniazid, although the incidence of transient elevation of liver enzymes can be as high as 18%.3133 Two of our patients experienced liver biopsy-proved hepatotoxicity, and in both reducing the dose of isoniazid resulted in resolution of abnormal liver function tests. The optimal duration of treatment for tuberculosis in the immunosuppressed child is debatable, but typically 9- to 18-month courses are used, given that low cure rates have been reported with shorter courses.2 Khouri et al.5 suggest a minimum of 9 months of therapy for culture-positive and at least 6 month for culturenegative HIV patients. We treated two of our patients, one with cerebral tuberculosis (Case 4) and one with severe pulmonary tuberculosis (Case 3), for 18 months (Case 3 did not respond to quadruple therapy for the

Vol. 19, No. 7, July, 2000

THE PEDIATRIC INFECTIOUS DISEASE JOURNAL

629

first 2 months of therapy) and one patient was treated for 9 months. The mortality in our patients was comparable with other studies7, 13 but it is difficult to implicate the mortality attributable specifically to M. tuberculosis per se. Two of the six children died, one of Gramnegative septic shock and the other of pulmonary hemorrhage. The nature of the underlying disease and infection early after transplantation appear to be risk factors for an adverse outcome. Whereas both of the children who died presented with acute liver failure and appear to have acquired tuberculosis in the first month after transplantation, the remainder had primary diseases such as biliary atresia. Patients with a history of tuberculosis have a higher risk of reactivation during the first month after transplantation, independent of the type of immunosuppression.9 Tuberculosis after liver transplantation in children is not rare and can contribute to increased mortality. Bacillus of Calmette-Guerin vaccination of stable chronic liver disease children who might require transplantation should be considered. A Mantoux test should be performed for children undergoing liver transplantation and for their family members. Interpretation of the skin test should be according to guidelines given for immunosuppressed children. Isoniazid prophylaxis should be offered to liver transplant recipients with a past history of tuberculosis and to those with a positive skin test. Duration of prophylaxis is controversial though 6 to 9 months is usually recommended. This seems logical for liver transplant patients in whom immunosuppression is maximal up to 3 months posttransplant and reduced thereafter. All posttransplant patients with pyrexia of unknown origin should be investigated aggressively to exclude the diagnosis of tuberculosis, and to this end the advice of a medical microbiologist or infectious diseases physician should be sought. Increasingly the use of molecular diagnostic tests, such as PCR, will broaden experience and delineate their role. Concurrent administration of conventional antituberculous drugs and cyclosporin or tacrolimus decreases bioavailability of the latter. Close monitoring of the levels of these agents is essential. Standard antituberculous drugs can be used in these patients with careful monitoring of liver function. Withdrawing immunosuppression during therapy of tuberculosis does not inevitably precipitate rejection.
REFERENCES
1. Sinott JT, Emmanuel PJ. Mycobacterial infections in the transplant patient. Semin Respir Infect 1990;5:6573. 2. Singh N, Paterson LD. Mycobacterium tuberculosis in solid organ transplant recipients: impact and implications for management. Clin Infect Dis 1998;27:1266 77. 3. Skoberg K, Runtu P, Kiainen TU, Voltonen V. Effect of immunosuppressive therapy on clinical presentation and outcome of tuberculosis. Clin Infect Dis 1993;17:10127. 4. Moss WJ, Dedyo T, Suarez M, Nicholas SW, Abrams E.

5.

6. 7. 8. 9.

10. 11. 12. 13.

14. 15. 16. 17. 18. 19. 20. 21.

22. 23. 24. 25.

26. 27.

Tuberculosis in children infected with human immunodeficiency virus: a report of five cases. Pediatr Infect Dis J 1992;11:114 6. Khouri YF, Mastrucci MT, Hutto C, Mitchell CD, Scott GB. Mycobacterium tuberculosis in children with human immunodeficiency virus type 1 infection. Pediatr Infect Dis J 1992;11:950 5. Nishizaki T, Yanaga K, Soejima Y, et al. Tuberculosis following liver transplantation: report of a case and review of the literature. Trans Int 1996;9:589 92. Meyers Br, Halpern M, Sheiner P, Mendelson MH, Neibart E, Miller C. Tuberculosis in liver transplant recipients. Transplantation 1994;58:301 6. Higgens R, Kusne S, Reys J, et al. Mycobacterium infection after liver transplantation. Transplant Proc 1990;22:454. Aguado JM, Herrero JA, Gavalda J, et al., and the Spanish Transplantation Infection Study Group, GESITRA. Clinical presentation and outcome of tuberculosis in kidney, liver and heart transplant recipients in Spain. Transplantation 1999; 63:1278 86. Stoblen F, Neuhas R, Neuman K, et al. Tuberculosis in liver transplant recipients: recurrence after transplantation? Transplant Proc 1994;26:3604 5. Grauhan O, Lahmann R, Lemmens P, et al. Mycobacterial infections after liver transplantation. Langenbecks Arch Chir 1995;38:1715. Salizzoni JL, Tiruviluamala P, Reichman LB. Liver transplantation: an unheralded probable risk for tuberculosis. Tuber Lung Dis 1992;73:232 8. McDiarmid SV, Blumberg DA, Remotti H, et al. Mycobacterial infections after pediatric liver transplantation: a report of three cases and review of the literature. J Pediatr Gastroenterol Nutr 1995;20:42531. Watson JM, Fern KJ, Porter JDH, et al. Notification of tuberculosis in England and Wales 1:9829. Comm Dis Report 1990;1:R135. Hayward AC, Watson JM. Tuberculosis in England and Wales 198293: notification exceeded predictions. Comm Dis Report 1995;5:R29 33. Hall CM, Willcox PA, Swanepoel CR, Kahn D, Van Zyl-Smit R. Mycobacterial infection in renal transplant recipients. Chest 1994;106:4359. Peters TG, Reiter CG, Boswell RL. Transmission of tuberculosis by kidney transplantation. Transplantation 1984;38: 514 6. Ahsan N, Blanchard RL, Mai ML. Case report: gastrointestinal tuberculosis in renal transplantation: a case report and review. Clin Transplant 1995;9:349 52. Forslund T, Laasonen L, Hockerstedt K, Stenman S, Edgren J. Tuberculosis of the colon in a kidney transplant patient. Acta Med Scand 1984;215:181 4. ODonoghue DJ, Short CD, Gokal R, Johnson RWG. Pancreatic tuberculosis following renal transplantation. Nephron 1992;60:367. Papagianni A, Alexopoulos E, Vakianis P, Papadimitriou M. Tuberculous meningoencephalitis after cytomegalovirus infection in renal transplant recipients. Nephrol Dial Transplant 1994;9:438 42. John GT, Vincent L, Jeyaseelan L, Jacob CK, Shastry JCM. Cyclosporine immunosuppression and mycobacterial infections. Transplantation 1994;58:2479. Lichtenstein IH, Macgregar RR. Mycobacterial infection in renal transplant recipients: report of five cases and review of the literature. Rev Infect Dis 1983;5:216 20. Starke JR, Jacobs RF, Jereb J. Resurgence of tuberculosis in children. J Pediatr 1992;120:839 55. Centers for Disease Control and Prevention. Purified protein derivative PPD-tuberculin anergy and HIV infection: guidelines for anergy testing and management of anergic persons at risk of tuberculosis. MMWR 1991;40(RR-5):2733. Chemotherapy and management of tuberculosis in the United Kingdom: BTS recommendation 1998. Thorax 1998; 153:536 48. Goncalves AR, Caetano MA, Paula FJ, Ianhez LE, Saldanha

630

THE PEDIATRIC INFECTIOUS DISEASE JOURNAL

Vol. 19, No. 7, July, 2000

LB, Sabbaga E. Tuberculosis interstitial granulomatous nephritis in renal transplants: report of three cases. Transplant Proc 1992;24:1911. 28. Aguado JM, Herrero JA, Gavalda J, et al. Clinical presentation and outcome of tuberculosis in kidney, liver and heart transplant recipients in Spain. Transplantation 1997;63:1278 86. 29. Obrien RJ, Long MW, Cross FS, Lyle MA, Suides DE. Hepatotoxicity from isoniazid and rimpicin among children treated for tuberculosis. Pediatrics 1983;72:4919. 30. Kronbach T, Fisher V, Meyers UA. Cyclosporin metabolism in human liver; identification in human liver: identification of a cytochrome P-450 gene family as the major cyclosporine

metabolising enzyme explains interaction of cyclosporin with other drugs. Clin Pharmacol Ther 1988;43:630 5. 31. Scharer L, Smith J. Serum transaminase elevations and other hepatic abnormalities in patients receiving isoniazid. Ann Intern Med 1969;71:111320. 32. Bailey WC, Taylor SL, Dascomb HE, Greenberg HB, Ziskind M. Disturbed hepatic function during isoniazid chemoprophylaxis: monitoring the hepatic function of 427 hospital employees receiving isoniazid chemoprophylaxis for tuberculosis. Am Rev Respir Dis 1973;107:5239. 33. Byrd RB, Horn BR, Grigg GA, Solomon DA. Isoniazid chemoprophylaxis. Arch Intern Med 1977;137:1130 3.

Pediatr Infect Dis J, 2000;19:6304 Copyright 2000 by Lippincott Williams & Wilkins, Inc.

Vol. 19, No. 7 Printed in U.S.A.

Bacteremic urinary tract infection in children

OLLI HONKINEN, MD, TIMO JAHNUKAINEN, MD, JUSSI MERTSOLA, MD, JUHANI ESKOLA, MD AND OLLI RUUSKANEN, MD

Objective. To assess the clinical characteristics of bacteremic urinary tract infection (UTI) in children. Design. Clinical data of Finnish children with bacteremic UTI (n 134) from 1985 to 1994 were analyzed. Their symptoms, laboratory and imaging findings were compared with those of ageand sex-matched patients hospitalized for blood culture negative UTI. Results. Generally, no major differences were seen in clinical findings between bacteremic and nonbacteremic patients. Bacteremic children had more frequent feeding problems (P 0.02), and children >12 months of age tended more often to have abdominal pain and vomiting than did nonbacteremic patients. Fever was the major initial symptom in both study groups, but no significant difference occurred in the mean highest temperature or in the mean of duration of fever before admission to the hospital. The mean concentration of serum C-reactive protein on admission was significantly higher in bacteremic patients (116 vs. 76 mg/l; P < 0.01). After onset of

antimicrobial treatment fever lasted significantly longer in bacteremic patients than in control patients (means, 2.3 vs. 1.1 days; P < 0.01). Anatomic or functional abnormalities in the urinary tract were detected in 51% vs. 46%, respectively. Obstruction of the urinary tract (9% vs. 1%, P < 0.01) and Grade 3 to 5 vesicoureteral reflux (30% vs. 16%, P 0.02) were significantly more frequent in bacteremic patients with UTI. Obstruction or vesicoureteral reflux was found in 46% of children with bacteremic UTI caused by Escherichia coli vs. 89% of children with non-E. coli infection (P < 0.01). Conclusions. Clinical symptoms do not significantly distinguish bacteremic from nonbacteremic children with UTI. Outcome of bacteremic UTI was comparable with that of nonbacteremic UTI. Bacteremic children, especially those with non-E. coli UTI, more often had anatomical or functional abnormalities in the urinary tract.
INTRODUCTION

Accepted for publication March 28, 2000. From the Department of Pediatrics, University of Turku, Turku (OH, TJ, JM, OR), and the National Public Health Institute, Helsinki (JE), Finland. Key words: Urinary tract infection, bacteremia, Escherichia coli. Address for reprints: Olli Ruuskanen, M.D., Department of Pediatrics, Turku University Hospital, PL 52, 20521 Turku, Finland.

Urinary tract infection (UTI) is a common bacterial illness in children.1, 2 In a study of febrile infants the prevalence of UTI was 5.3%,3 and in another study of 3553 school entrants symptomatic UTI was found in 7.8% of girls and 1.6% of boys.4 In a Finnish study from the 1980s the annual incidence of symptomatic UTI requiring hospitalization was found to be 1.9/1000 girls and 0.6/1000 boys.5 In a retrospective analysis of 22 051 blood cultures in a pediatric hospital, the incidence of bacteremia was