j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 9 ( 2 0 1 5 ) 3 7 8 e3 8 5

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.JournalofSurgicalResearch.com

Differential impact of obesity and diabetes

mellitus on survival after liver resection for
colorectal cancer metastases

Sousana Amptoulach, MD, MSc, PhD,a Gillis Gross, MD,b

and Evangelos Kalaitzakis, MD, PhDb,*
a
Department of Oncology, Skåne University Hospital, University of Lund, Malmö, Sweden
b
Institution of Clinical Sciences, Faculty of Medicine, University of Lund, Lund, Sweden

article info abstract

Article history: Background: Data on the potential effect of obesity and diabetes mellitus on survival after
Received 16 March 2015 liver resection due to colorectal cancer (CRC) metastases are very limited.
Received in revised form Methods: Patients undergoing liver resection for CRC metastases in a European institution
27 May 2015 in 2004e2011 were retrospectively enrolled. Relevant data, such as body mass index, extent
Accepted 29 May 2015 of resection, chemotherapy, and perioperative outcome, were collected from medical re-
Available online 4 June 2015 cords. The relation of obesity and diabetes mellitus with overall and disease-free survival
was assessed using adjusted Cox models.
Keywords: Results: Thirty of 207 patients (14.4%) included in the study were obese (BMI
30 kg/m2) and
Colorectal cancer metastases 25 (12%) had diabetes mellitus. Major hepatectomy was performed in 46%. Although both
Liver resection obese patients and those with diabetes had higher American Society of Anesthesiologist
Obesity scores (P < 0.05 for both), neither obesity nor diabetes was significantly related to primary
Diabetes mellitus tumor characteristics, liver metastasis features, extent or radicality of resection, extrahe-
Outcome patic disease at hepatectomy, preoperative or postoperative oncologic therapy, or periop-
Survival erative outcome (P > 0.05 for all). Patients were followed up for a median of 39 mo
posthepatectomy (interquartile range, 13e56 mo). After adjustment for confounders,
obesity was an independent predictor of improved (hazard ratio, 0.305, 95% confidence
interval, 0.103e0.902) and diabetes of worse overall survival (hazard ratio, 3.298, 95%
confidence interval, 1.306e8.330). Obese patients with diabetes had also worse disease-free
survival compared with the rest of the cohort (P < 0.05).
Conclusions: After hepatectomy for CRC metastases, obesity does not seem to be associated
to poor outcome while diabetes mellitus has a negative impact on prognosis.
ª 2015 Elsevier Inc. All rights reserved.

1. Introduction a small fraction may be amenable to curative liver resec-

tion. Thus, up to 40% of patients with CRC metastases un-
Obesity is increasing worldwide and it is a risk factor for dergoing hepatectomy may be obese [2e4]. Although several
colorectal cancer (CRC) development [1]. A significant pro- reports have shown that obesity is associated with higher
portion of patients with CRC, develop liver metastases, and odds of perioperative complications after hepatectomy

* Corresponding author. Digestive Disease Center, Copenhagen University Hospital/Bispebjerg, University of Copenhagen, 2400
Copenhagen, Denmark. Tel.: þ45 35313531; fax: þ45 35313911.
E-mail addresses: evangelos.kalaitzakis@medicine.gu.se, kalvag@hotmail.com (E. Kalaitzakis).
0022-4804/$ e see front matter ª 2015 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jss.2015.05.059
 j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 9 ( 2 0 1 5 ) 3 7 8 e3 8 5
[2,5e7], published data are not unanimous [4,8e10]. Also, in
most studies, only a fraction of enrolled patients had un-
dergone hepatectomy due to CRC metastases, as patients
with other indications including benign conditions, and
with underlying end-stage liver disease were also included
[2,5e7,9e11].
Obesity may affect prognosis after cancer diagnosis [1]. In
advanced CRC, it has been reported to have a detrimental
effect on overall survival and disease recurrence [12,13], but
published reports are inconsistent [14]. Data on the impact of
obesity on survival after liver resection for CRC metastases
are very limited. Although overall survival has been shown to
be prolonged with increasing body mass index (BMI) after
hepatectomy for various types of cancer [2], this was not
confirmed in a recent study in patients undergoing liver
resection due to CRC metastases [3]. Finally, diabetes mellitus
is frequent in obesity and may have a negative impact on
oncologic outcomes in CRC [15], but data on its potential ef-
fect on survival after liver resection for CRC metastases are
scarce.
The primary aim of the present study was to evaluate the
potential role of obesity and diabetes mellitus in survival after
liver resection for CRC metastases. Secondarily, we also aimed
to assess the potential impact of obesity and diabetes on
perioperative morbidity in these patients.
2. Patients and methods
2.1. Patients
All consecutive patients who underwent liver resection due
to CRC metastases at the Skåne University Hospital in Lund,
Sweden in 2004e2011 were retrospectively identified through
a search of the computerized discharge register of the hos-
pital. Repeat resections were excluded. Medical records were
scrutinized, and patients without an available preoperative
height and/or weight were also excluded. Relevant data were
collected, such as demographics, site, and TNM staging of the
primary tumor, number and size of liver metastases, preop-
erative chemotherapy and its intent (adjuvant after resection
of primary CRC or neoadjuvant before hepatectomy), date
and extent of liver resection, American Society of Anesthe-
siologists (ASA) score, postoperative chemotherapy, and its
intent (adjuvant after hepatectomy or palliative). Pre-
operative locoregional treatment, such as portal vein embo-
lization, was also registered. History of diabetes and arterial
hypertension were noted. The Charlson comorbidity index
was calculated as a measure of the burden of comorbid
illness [16], excluding malignant tumors. A resection margin
of <1 mm was considered a positive margin, whereas a
margin of
1 mm was considered a tumor-free margin.
Resections of
3 liver segments were considered major
resections. The cohort was stratified into underweight (BMI
<18.5 kg/m2), normal (BMI 18.5e24.9 kg/m2), overweight (BMI
25e29.9 kg/m2), and obese patients (BMI
30 kg/m2) accord-
ing to the World Health Organization criteria. Analyses were
performed between obese (BMI
30 kg/m2) and nonobese
patients (BMI <30 kg/m2). The study protocol was approved
by the local ethics committee.
379
2.2. Outcome parameters
The primary outcome end point of the study was overall
survival (from hepatectomy until death from any cause). Alive
patients were censored at last follow-up. Death versus alive
status was ascertained through medical records and the
regional health care administrations system. Secondary end
points were disease-free survival, calculated from the date of
liver resection to tumor recurrence or death due to any cause,
postoperative mortality, defined as mortality from any cause
within 30 and 90 d after liver resection, and postoperative
complications. Length of intensive care unit (ICU) stay and
total length of stay (LOS), from liver resection until discharge,
were also analyzed. Complications were graded according to
the DindoeClavien classification [17], and those with a score
of
3 were considered severe. Complications were further
classified as noneliver related (cardiovascular, respiratory,
infection, bleeding, acute kidney injury, and other) and liver
related. Bleeding was defined as need for intraoperative or
postoperative blood transfusions. Acute kidney injury was
defined as previously described [18]. Liver-related complica-
tions included postoperative hepatic failure, defined as peak
postoperative total bilirubin >7 mg/dL [19] and/or ascites, and
bile leak.
2.3. Statistics
Data are presented as mean and standard deviation (SD) or n
(%) as appropriate. When comparing groups, the chi-square
test or Fisher exact test was used for categorical variables,
whereas analysis of variance or the ManneWhitney test was
used for continuous variables, as appropriate. Logistic
regression analyses were used to adjust for confounders in the
relation between, on the one hand, obesity and, on the other
hand, complications, prolonged length of ICU stay (
1 d), and
prolonged LOS (
10 d, i.e.,
the 75th percentile of the whole
cohort). Confounding factors included age, gender, the ASA
score, the Charlson comorbidity index, type of resection, and
diabetes, and any parameter in our cohort found to be related
to postoperative complications or prolonged ICU stay and LOS
in univariate analysis at P < 0.1. The relationships of obesity
and diabetes with overall and disease-free survival were
assessed with survival analysis (KaplaneMeier), and groups
were compared with the log-rank test. Cox regression analysis
was used to adjust for confounders, that is, node positive
primary, >3 liver metastases, size of the largest
5 cm, car-
cinoembryonic antigen >60 ng/mL, positive resection margin,
and extrahepatic disease [20], as well as any parameter found
to be related to overall survival (ASA score, the Charlson co-
morbidity index, the number of perioperative complications,
liver-related complications, T stage of the primary tumor, and
adjuvant chemotherapy after hepatectomy) or disease-free
survival (ASA score, multiple or severe perioperative compli-
cations, perioperative bleeding, locoregional treatment after
surgery, preoperative chemotherapy, and adjuvant chemo-
therapy after hepatectomy) in univariate analysis at P < 0.1 in
our cohort. For overall survival, our data set would have a
power of 0.78 to detect a hazard ratio of 0.4 for obesity, given a
type 1 error of 5%. Reported P values are two-tailed, and the
significance level was set at P < 0.05.
380 j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 9 ( 2 0 1 5 ) 3 7 8 e3 8 5

diabetes was more common among obese patients, who also

3. Results
In all, 215 patients underwent first-time hepatectomy for CRC
metastases during the study period but 8 (4%) were excluded
due to missing preoperative height and/or weight data. In all,
four patients (2%) were underweight (BMI <18.5 kg/m2), 80
(38.6%) normal (BMI 18.5e25 kg/m2), 93 (45%) overweight (BMI
25e30 kg/m2), and 30 (14.4%) obese (BMI
30 kg/m2). Only 3 of
30 obese patients had severe obesity (BMI
35 kg/m2). De-
mographics and primary tumor characteristics were similar in
obese and nonobese patients (Table 1). Although the burden of
comorbid illness, expressed as the Charlson comorbidity
index, was similar in the two groups, not unexpectedly,
had higher ASA scores. Furthermore, obese versus nonobese
patients tended to have liver metastasis size >5 cm and pos-
itive resection margins more often. However, the two groups
did not differ significantly regarding primary tumor charac-
teristics, other liver metastasis features, hepatectomy extent,
the presence of extrahepatic disease at the time of hepatec-
tomy, and preoperative or postoperative oncologic therapy
(Table 1). The proportion of patients with obesity undergoing
hepatectomy in or after 2008 did not differ significantly from
those operated before 2008 (17% versus 11%, P ¼ 0.236). Only
three patients in the cohort (all nonobese) had known chronic
liver disease (two hemochromatosis and one hepatitis C) but
none was known to have liver cirrhosis.

Table 1 e Characteristics of patients undergoing first-time liver resection for CRC liver metastases in 2004e2011 stratified
according to the presence of obesity.
Characteristics All patients Obese Nonobese P value
(n ¼ 207) (BMI
30 kg/m2) (n ¼ 30) (BMI <30 kg/m2)
(n ¼ 177)

Age, y 65 (10) 64 (11) 65 (10) 0.592

Female gender 79 (38) 11 (37) 68 (38) 0.855
Primary tumor characteristics
Rectal 67 (32) 8 (27) 59 (33) 0.470
T stage 3 or 4 159 (86) 20 (77) 139 (87) 0.182
Positive nodal status 128 (66) 19 (68) 109 (66) 0.821
Comorbid illness
Charlson comorbidity index* 2.5 (1.4) 2.7 (1.5) 2.3 (1.3) 0.368
Diabetes mellitus 25 (12) 8 (27) 17 (10) 0.014
Arterial hypertension 90 (44) 15 (50) 75 (42) 0.436
ASA score
I 27 (8) 2 (7) 15 (8) 0.004
II 161 (78) 18 (60) 143 (81)
III 29 (14) 10 (33) 19 (11)
Liver metastases
Metachronous 83 (40) 14 (47) 69 (39) 0.427
More than three metastases 28 (14) 4 (13) 24 (14) 0.973
Largest tumor diameter >5 cm 31 (17) 8 (29) 23 (15) 0.099
Extrahepatic disease at hepatectomy 29 (14) 3 (10) 26 (15) 0.494
Carcinoembryonic antigen >60 ng/mLy 22 (11) 4 (13) 18 (11) 0.650
Preoperative oncologic treatment
Preoperative chemotherapy 123 (59) 18 (60) 105 (59) 0.944
Neoadjuvant before liver resection 84 (41) 8 (27) 76 (43) 0.109
Number of cycles 6 (4) 4 (1) 6 (4) 0.838
Locoregional treatment 37 (18) 7 (23) 30 (17) 0.399
Major liver resection 95 (46) 14 (47) 81 (46) 0.927
Extrahepatic procedurez 7 (3) 1 (3) 6 (3) 0.987
Positive resection margin 26 (14) 7 (24) 19 (12) 0.084
Postoperative chemotherapy 129 (66) 17 (59) 112 (68) 0.353
Adjuvant after liver resection 113 (58) 13 (45) 100 (60) 0.121
Number of cycles 7 (3) 6 (4) 7 (3) 0.123

CEA ¼ carcinoembryonic antigen.

Data are presented as mean (SD) or n (%) as appropriate.
Significant results (P < 0.05) are indicated in bold.
Missing values were as follows: primary tumor T stage n ¼ 21; primary tumor nodal status n ¼ 13; carcinoembryonic antigen n ¼ 6; size of liver
metastases n ¼ 25; neoadjuvant chemotherapy prior to liver resection n ¼ 1; number of cycles of adjuvant chemotherapy after liver resection
n ¼ 16; resection margin n ¼ 17; postoperative chemotherapy n ¼ 12; adjuvant chemotherapy after liver resection n ¼ 12; no of cycles of adjuvant
chemotherapy after liver resection n ¼ 28.
*
The Charlson comorbidity index was calculated as a measure of the burden of comorbid illness [16], excluding malignant tumors as all patients
in the cohort had this parameter in common. Calculation of the Charlson comorbidity index, excluding diabetes mellitus as well, yields a value
of 2.3 (SD, 1.3) for the whole cohort and 2.4 (1.4) for obese versus 2.3 (1.3) for nonobese patients (P ¼ 0.797).
y
Four percent of nonobese patients had CEA >200 ng/mL versus 0% of obese patients, P ¼ 0.597.
z
Concomitant colon resection in all cases.
 j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 9 ( 2 0 1 5 ) 3 7 8 e3 8 5 381

Diabetic patients compared with those without diabetes
had more frequently ASA scores >II (32% versus 12%, P ¼ 0.01)
and a history of arterial hypertension (76% versus 39%
P < 0.001) and they also had a higher Charlson comorbidity
index (3.4 [0.9] versus 2.3 [1.4], P < 0.001). However, they did not
differ significantly in any other parameter of Table 1 (P > 0.05
for all; data not shown).
3.1. Perioperative outcome
The occurrence of complications was similar in obese and
nonobese patients (Table 2). Although severe complications
(DindoeClavien score
3) were more common in the former,
the difference was not statistically significant, which was also
true for postoperative mortality. Only in the case of liver-related
complications there was a trend (P ¼ 0.077) in favor of the
nonobese group, due to less frequent postoperative liver failure
and bile leak (Table 2). Patients with versus without diabetes had
acute kidney injury more frequently (24% versus 9%, P ¼ 0.033).
However, the two groups did not differ significantly in any
parameter of Table 2 (data not shown). After adjustment for
confounders with logistic regression analysis, neither obesity
nor diabetes was independently related to postoperative com-
plications, prolonged ICU stay, or LOS (Table 2).
3.2. Overall survival
Patients included in the study were followed up for a mean of
37 mo (median 39, SD 28, interquartile range 13e56). Mean
survival was 37 mo (median 38, SD 25, interquartile range
18e72) and 5-y survival was 20%. Overall, 72 patients died
during follow-up (cause of death: 49 metastatic CRC, 3 car-
diovascular, 1 liver failure perioperatively, 20 other or un-
known). In univariate analysis, diabetes but not obesity was
related to significantly worse overall survival (Figs. 1 and 2).
However, after adjustment for confounders with Cox regres-
sion analysis, obesity was associated with improved and dia-
betes with worse overall survival posthepatectomy (Table 3).
Inclusion of an interaction term between diabetes and obesity
did not alter these findings (Table 3). Seventy-one percent of
diabetic patients (versus 53% nondiabetic) and 59% of obese
patients (versus 61% nonobese) died due to metastatic CRC.
3.3. Disease-free survival
Recurrence status could not be ascertained in 1 patient,
whereas in 11 patients, with recurrent disease during follow-
up, recurrence dates were unclear (all referred back to refer-
ring institutions). These patients (n ¼ 12; two obese and 10

Table 2 e Perioperative outcome after liver resection for CRC liver metastases.
All patients Univariate analyses Multivariate analyses (risk for event)
(n ¼ 207)
Obese Nonobese P Obesity Diabetes
(BMI
30 kg/m2) (BMI <30 kg/m2) (n ¼ 177) value (BMI
30 kg/m2) mellitus
(n ¼ 30)
OR (95% CI) OR (95% CI)

Complications 93 (45) 13 (43) 80 (45) 0.849 0.703 (0.292e1.691) 1.467 (0.562e3.830)

Multiple (>1) complications 33 (16) 6 (20) 27 (15) 0.511 0.899 (0.281e2.880) 0.766 (0.221e2.656)
Severe complications* 18 (9) 4 (13) 14 (8) 0.330 1.285 (0.294e5.615) 0.768 (0.167e3.538)
Multiple and/or 38 (18) 7 (23) 31 (18) 0.446 1.053 (0.357e3.111) 0.804 (0.251e2.577)
severe complications
Noneliver related 92 (44) 13 (43) 79 (45) 0.895 0.822 (0.350e1.930) 1.265 (0.497e3.224)
complications
Cardiovascular 12 (6) 3 (10) 9 (5) 0.287
Respiratory 7 (3) 1 (3) 6 (3) 0.987
Infection 21 (10) 3 (10) 18 (10) 0.977
Acute kidney injury 22 (11) 3 (10) 19 (11) 0.904
Need of RBC transfusion 54 (26) 5 (17) 49 (28) 0.204
Other 8 (4) 2 (7) 6 (3) 0.389
Liver-related complicationsy 12 (6) 4 (13) 8 (5) 0.077 4.835 (0.846e27.621) 0.287 (0.024e3.425)
Liver failurez 8 (4) 3 (10) 5 (3) 0.093
Bile leak 9 (3) 3 (10) 3 (1.7) 0.041
Length of ICU stay, d 1.5 (2.9) 3.4 (6.8) 1.2 (1.2) 0.451
>1 21 (10) 6 (20) 15 (9) 0.093 0.822 (0.692e7.460) 1.265 (0.497e3.224)
LOS, dx 10 (7.6) 13.4 (14.9) 9.5 (5.3) 0.384
>10 45 (22) 9 (30) 36 (20) 0.238 1.554 (0.584e4.137) 0.407 (0.111e1.488)
30-d mortalityk 2 (1) 1 (3) 1 (0.6) 0.269 Not performed Not performed
90-d mortalityk 4 (2) 2 (7) 2 (1) 0.101 Not performed Not performed

CI ¼ confidence interval; OR ¼ odds ratio; RBC ¼ red blood cell.

Data are expressed as mean (SD) or n (%) as appropriate.
Significant results (P < 0.05) are indicated in bold.
*
DindoeClavien score
3.
y
Postoperative liver failure and/or bile leak.
z
Peak postoperative total bilirubin >7 mg/dl [19] and/or ascites.
x
Total hospital stay including ICU stay.
k
No multivariate analyses performed due to low numbers.
382 j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 9 ( 2 0 1 5 ) 3 7 8 e3 8 5

Fig. 1 e KaplaneMeier curves for overall survival (A) and disease-free survival (B) of obese (dashed line) and nonobese
(continuous line) patients after liver resection for CRC liver metastases.

nonobese) were excluded from disease-free survival analyses.
In univariate analysis, both obesity and diabetes mellitus were
related to worse postoperative disease-free survival (Figs. 1
and 2). After adjustment for confounders, only diabetes was
found to have a significant impact on disease-free survival, but
when an interaction term between diabetes and obesity was
entered into the model, only the interaction term was found to
be significant (Table 3). This suggests that the effect of obesity
on disease-free survival is different depending on diabetes
status. Survival analysis showed that obese patients with dia-
betes had worse disease-free survival than patients with any
other combination of obesity and diabetes status (Fig. 3).
4. Discussion
Because of the rising prevalence of obesity worldwide and its
relation with CRC [1], an increasing number of patients un-
dergoing liver resection for CRC metastases are likely to be
obese. In the present study, obesity in patients undergoing
liver resection due to CRC metastases was associated to
improved overall survival, after adjustment for confounders.
Not unexpectedly, diabetes mellitus, a common comorbidity
in obese patients, had a negative impact on overall and
disease-free survival. To our knowledge, ours is the first study
investigating the potential role of obesity in relation to that of
diabetes in survival after liver resection with a specific focus
on patients with CRC liver metastases.
The global obesity epidemic and its relation with cancer [1]
has led to an increased awareness of the potential association
of obesity with oncologic outcomes after surgery in different
types of cancer, with somewhat conflicting findings [21e23]. A
recent report on the impact of obesity on posthepatectomy
oncologic outcomes found that increasing BMI was related to
better overall and recurrence-free survival after liver resection
due to malignancy [2]. However, patients with both primary
and metastatic liver cancer were included in this study and
obesity per se was not significantly associated with survival [2].
In patients with CRC liver metastases undergoing liver
resection, patient survival has been shown to be unrelated to
obesity [3], overweight, or central adiposity [24] in two previ-
ous studies. However, no adjustment for confounders was
performed in one of these reports [3], whereas the potential
role only of overweight (BMI
25 kg/m2) and central adiposity,
but not specifically obesity, was explored in the other [24].
Our finding of improved posthepatectomy survival in obese
patients with CRC liver metastases, might be part of the “obesity
paradox” reported in other cancer types [22,25]. It has been

Fig. 2 e KaplaneMeier curves for overall survival (A) and disease-free survival (B) of patients with (dashed line) versus those
without diabetes mellitus (continuous line) after liver resection for CRC liver metastases.
 j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 9 ( 2 0 1 5 ) 3 7 8 e3 8 5 383

survival [26]. However, in a recent study, neither sarcopenia nor

Table 3 e The impact of diabetes mellitus and obesity on
overall and disease-free survival following hepatectomy sarcopenic obesity was found to have an impact on survival
for CRC liver metastases in adjusted Cox models. after liver resection for CRC liver metastases [4]. In the same
report, excessive body fat, estimated from measurements of
Model Overall survival Disease-free survival
skeletal mass areas on available computed tomography scans
HR (95% CI) HR (95% CI) and body weight, correlated with improved overall and disease-
Model A free survival [4], which is in accordance with our findings. On the
Obesity other hand, among surgical patients those at the extremes of
No 1 (Reference) 1 (Reference) BMI (underweight and morbidly obese) have the highest post-
Yes 0.356 (0.156e0.814) 0.937 (0.505e1.740) operative mortality, which persists at long term [27]. As very few
Diabetes mellitus
patients were underweight (n ¼ 4) or severely obese (n ¼ 3) in our
No 1 (Reference) 1 (Reference)
cohort, no firm conclusions can be drawn on these patient
Yes 3.682 (1.672e8.109) 2.073 (1.109e3.875)
Model B groups in our study. It is conceivable that obesity in our study
Obesity 0.305 (0.103e0.902) 0.595 (0.277e1.274) may not have been severe enough to have a negative impact on
Diabetes mellitus 3.298 (1.306e8.330) 1.323 (0.601e2.910) postoperative survival, but it may have contributed to an
Obesity x 1.510 (0.271e8.402) 6.106 (1.572e23.712) increased physiological reserve and, in turn, to improved overall
diabetes mellitus survival. In particular, it is conceivable that the observed rela-
CI ¼ confidence interval; HR ¼ hazard ratio. tion between obesity and overall survival in our study may
Significant results (P < 0.05) are indicated in bold. possibly represent the effect of reduced frailty and improved
In Model A, obesity and diabetes mellitus were entered separately ability to tolerate adjuvant chemotherapy [12,28]. Although it
with no interaction term. In model B, obesity and diabetes mellitus seems that obesity does not worsen posthepatectomy overall or
entered along with an interaction term. Please refer to the statistics
disease-free survival in patients with CRC liver metastases,
section for parameters adjusted for in the Cox models. Inclusion of
further studies are clearly warranted to fully delineate the effect
the year of liver resection (in 2008 or later versus before 2008) or the
Charlson comorbidity index calculated after exclusion of the of obesity and body composition on prognosis in these patients.
diabetes-related dimensions in the Cox models did not change the Our finding that diabetes is a predictor of poor survival
statistical significance of the results. after liver resection for CRC metastases is in line with previous
studies showing worse long-term outcomes in diabetic pa-
tients with various CRC stages [29] and with its negative effect
suggested that the “obesity paradox” is present in patients with on life expectancy in the general population. The potential
cancer only when obesity is defined by BMI and that sarcopenia, effect of diabetes on long-term outcomes after hepatectomy
that is reduced muscle mass, even in the presence of excessive for CRC metastases has been assessed in two previous studies
fat mass (i.e., sarcopenic obesity), is a major predictor of poor yielding nonsignificant results [24,30], whereas a trend to
worse overall survival (P ¼ 0.08) has been shown in diabetic
patients undergoing hepatectomy due to various types of
cancer [2]. Our data further indicate that diabetes might
modulate the positive relation of obesity with survival after
hepatectomy for CRC metastases.
The effect of diabetes on patient outcome in CRC has been
attributed to an association with cardiovascular mortality,
underlying diabetes-associated comorbid illness limiting the
administration of chemotherapy [31], or reduced effect of
chemoradiotherapy (in diabetic patients with rectal cancer)
[32]. Hyperinsulinemina, in the context of insulin resistance,
and increased levels of insulin-like growth factor have also
been proposed as potential factors affecting tumor aggres-
siveness in diabetic patients with CRC [33]. In our cohort, most
diabetic patients died because of metastatic cancer. Also,
diabetic compared with nondiabetic patients did not differ
significantly in primary tumor T or node stage nor in the
administration of preoperative or postoperative chemo-
Fig. 3 e KaplaneMeier curves for disease-free survival of therapy or the number of cycles they received. Further studies
obese patients with diabetes mellitus (dashed line) versus are necessary to elucidate the mechanisms behind the rela-
obese patients without diabetes (dotted line) versus tion between diabetes and mortality in these patients.
nonobese patients with diabetes (continuous line) versus With patient populations becoming increasingly obese,
nonobese patients without diabetes (dash-dotted line) there has been an increased interest in the relationship be-
after liver resection for CRC liver metastases. Obese tween obesity and surgical outcomes [8]. Despite some
patients with diabetes had significantly worse survival inconsistency in the literature [4,9,10], several articles have
compared to any other group (P < 0.05 compared to all). shown an increased risk for perioperative complications in
The other groups did not differ significantly from one obese patients undergoing hepatectomy [2,5e7]. In our cohort,
another (P > 0.05 for all). neither obesity nor diabetes was identified as a major
384 j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 9 ( 2 0 1 5 ) 3 7 8 e3 8 5
predictor of perioperative morbidity or mortality. Although
this could be attributed to the modest size of our cohort, it
may also be related to the low proportion of severely obese
patients included (n ¼ 3/207), as extreme obesity has been
reported to be an independent factor for increased perioper-
ative mortality in one study [9]. Furthermore, hepatic steatosis
and steatohepatitis, both occurring frequently in obese and
diabetic patients, are considered predictors of poor perioper-
ative outcome after hepatectomy [34,35]. Histologic features of
the nonetumor-bearing liver have not been assessed in the
present study, which may have limited detection of a sub-
group of obese patients prone to worse perioperative outcome.
Certain limitations should be taken into consideration
when interpreting the results of the present study. First, it is
retrospective in nature. Thus, there may have been unmea-
sured confounding variables, as, for example, no reliable data
on smoking habits or physical activity could be collected
through medical record scrutinization. Both factors may have
modified the relationship between obesity and survival in our
cohort. Finally, our findings would need to be reproduced in
larger samples of patients undergoing hepatectomy due to
CRC metastases with a larger proportion of patients with
obesity, or, ideally, in larger samples of patients with obesity
matched with nonobese patients, and with histologic assess-
ment of the tumor-surrounding liver for the presence of
steatosis or steatohepatitis.
5. Conclusions
In summary, obesity in the absence of diabetes is not associ-
ated with worse prognosis, whereas diabetic obese patients
appear to have a shorter overall and disease-free survival after
liver resection for CRC metastases. Further translational
studies are warranted to investigate the potential role of
obesity and adiposity as well as diabetes mellitus and insulin
resistance in the pathophysiology of CRC recurrence.
Acknowledgment
This work was supported by a grant from the Inger Persson
fund for cancer research. The fund had no involvement in the
study design, the collection, analysis and interpretation of
data, in the writing of the report, or the decision to submit the
article for publication.
Authors’ contributions: S.A. and G.G. contributed to the
data acquisition and critical revision of the article for impor-
tant intellectual content. S.A., G.G., and E.K. contributed to the
data interpretation. E.K. contributed to the study conception,
design, and drafting of the article. All authors approved the
version of the article to be submitted.
Disclosure
The authors reported no proprietary or commercial interest in
any product mentioned or concept discussed in this article.
None of the authors has any conflict of interest to declare.
references
[1] Calle EE, Rodriguez C, Walker-Thurmond K, Thun MJ.
Overweight, obesity, and mortality from cancer in a
prospectively studied cohort of U.S. adults. N Engl J Med 2003;
348:1625.
[2] Mathur AK, Ghaferi AA, Sell K, Sonnenday CJ, Englesbe MJ,
Welling TH. Influence of body mass index on complications
and oncologic outcomes following hepatectomy for
malignancy. J Gastrointest Surg 2010;14:849.
[3] Pathak S, Tang JM, Terlizzo M, Poston GJ, Malik HZ. Hepatic
steatosis, body mass index and long term outcome in
patients undergoing hepatectomy for colorectal liver
metastases. Eur J Surg Oncol 2010;36:52.
[4] Lodewick TM, van Nijnatten TJ, van Dam RM, et al. Are
sarcopenia, obesity and sarcopenic obesity predictive of
outcome in patients with colorectal liver metastases? HPB
(Oxford) 2015;17:438.
[5] Balzan S, Nagarajan G, Farges O, et al. Safety of liver
resections in obese and overweight patients. World J Surg
2010;34:2960.
[6] Vigano L, Kluger MD, Laurent A, et al. Liver resection in obese
patients: results of a case-control study. HPB (Oxford) 2011;
13:103.
[7] Mathur AK, Ghaferi AA, Osborne NH, et al. Body mass index
and adverse perioperative outcomes following hepatic
resection. J Gastrointest Surg 2010;14:1285.
[8] Mullen JT, Davenport DL, Hutter MM, et al. Impact of body
mass index on perioperative outcomes in patients
undergoing major intra-abdominal cancer surgery. Ann Surg
Oncol 2008;15:2164.
[9] Saunders JK, Rosman AS, Neihaus D, Gouge TH, Melis M.
Safety of hepatic resections in obese veterans. Arch Surg
2012;147:331.
[10] Wiggans MG, Lordan JT, Shahtahmassebi G, Aroori S,
Bowles MJ, Stell DA. The interaction between diabetes, body
mass Index, hepatic steatosis, and risk of liver Resection:
insulin dependent diabetes is the greatest risk for major
complications. HPB Surg 2014;2014:586159.
[11] Cescon M, Vetrone G, Grazi GL, et al. Trends in
perioperative outcome after hepatic resection: analysis of
1500 consecutive unselected cases over 20 years. Ann
Surg 2009;249:995.
[12] Meyerhardt JA, Catalano PJ, Haller DG, et al. Influence of
body mass index on outcomes and treatment-related
toxicity in patients with colon carcinoma. Cancer 2003;
98:484.
[13] Dignam JJ, Polite BN, Yothers G, et al. Body mass index and
outcomes in patients who receive adjuvant chemotherapy
for colon cancer. J Natl Cancer Inst 2006;98:1647.
[14] Simkens LH, Koopman M, Mol L, et al. Influence of body mass
index on outcome in advanced colorectal cancer patients
receiving chemotherapy with or without targeted therapy.
Eur J Cancer 2011;47:2560.
[15] Meyerhardt JA, Catalano PJ, Haller DG, et al. Impact of
diabetes mellitus on outcomes in patients with colon cancer.
J Clin Oncol 2003;21:433.
[16] Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new
method of classifying prognostic comorbidity in longitudinal
studies: development and validation. J Chronic Dis 1987;
40:373.
[17] Dindo D, Demartines N, Clavien PA. Classification of surgical
complications: a new proposal with evaluation in a cohort of
6336 patients and results of a survey. Ann Surg 2004;240:205.
[18] Mehta RL, Kellum JA, Shah SV, et al. Acute kidney injury
network: report of an initiative to improve outcomes in acute
kidney injury. Crit Care 2007;11:R31.
 j o u r n a l o f s u r g i c a l r e s e a r c h 1 9 9 ( 2 0 1 5 ) 3 7 8 e3 8 5
[19] Mullen JT, Ribero D, Reddy SK, et al. Hepatic insufficiency
and mortality in 1,059 noncirrhotic patients undergoing
major hepatectomy. J Am Coll Surg 2007;204:854.
discussion 62e4.
[20] Rees M, Tekkis PP, Welsh FK, O’Rourke T, John TG. Evaluation
of long-term survival after hepatic resection for metastatic
colorectal cancer: a multifactorial model of 929 patients. Ann
Surg 2008;247:125.
[21] Vitolins MZ, Kimmick GG, Case LD. BMI influences prognosis
following surgery and adjuvant chemotherapy for lymph
node positive breast cancer. Breast J 2008;14:357.
[22] Schrader AJ, Rustemeier J, Rustemeier JC, et al. Overweight is
associated with improved cancer-specific survival in patients
with organ-confined renal cell carcinoma. J Cancer Res Clin
Oncol 2009;135:1693.
[23] Fleming JB, Gonzalez RJ, Petzel MQ, et al. Influence of obesity
on cancer-related outcomes after pancreatectomy to treat
pancreatic adenocarcinoma. Arch Surg 2009;144:216.
[24] van Vledder MG, Levolger S, Ayez N, Verhoef C, Tran TC,
Ijzermans JN. Body composition and outcome in patients
undergoing resection of colorectal liver metastases. Br J Surg
2012;99:550.
[25] Park SM, Lim MK, Shin SA, Yun YH. Impact of prediagnosis
smoking, alcohol, obesity, and insulin resistance on survival
in male cancer patients: National Health Insurance
Corporation Study. J Clin Oncol 2006;24:5017.
[26] Gonzalez MC, Pastore CA, Orlandi SP, Heymsfield SB. Obesity
paradox in cancer: new insights provided by body
composition. Am J Clin Nutr 2014;99:999.
385
[27] Valentijn TM, Galal W, Tjeertes EK, Hoeks SE, Verhagen HJ,
Stolker RJ. The obesity paradox in the surgical population.
Surgeon 2013;11:169.
[28] Meyerhardt JA, Tepper JE, Niedzwiecki D, et al. Impact of
body mass index on outcomes and treatment-related toxicity
in patients with stage II and III rectal cancer: findings from
Intergroup Trial 0114. J Clin Oncol 2004;22:648.
[29] Mills KT, Bellows CF, Hoffman AE, Kelly TN, Gagliardi G.
Diabetes mellitus and colorectal cancer prognosis: a meta-
analysis. Dis Colon Rectum 2013;56:1304.
[30] Little SA, Jarnagin WR, DeMatteo RP, Blumgart LH, Fong Y.
Diabetes is associated with increased perioperative mortality
but equivalent long-term outcome after hepatic resection for
colorectal cancer. J Gastrointest Surg 2002;6:88.
[31] Gross CP, McAvay GJ, Guo Z, Tinetti ME. The impact of
chronic illnesses on the use and effectiveness of adjuvant
chemotherapy for colon cancer. Cancer 2007;109:2410.
[32] Caudle AS, Kim HJ, Tepper JE, et al. Diabetes mellitus affects
response to neoadjuvant chemoradiotherapy in the
management of rectal cancer. Ann Surg Oncol 2008;15:1931.
[33] Giovannucci E. Metabolic syndrome, hyperinsulinemia, and
colon cancer: a review. Am J Clin Nutr 2007;86:s836.
[34] Gomez D, Malik HZ, Bonney GK, et al. Steatosis predicts
postoperative morbidity following hepatic resection for
colorectal metastasis. Br J Surg 2007;94:1395.
[35] Reddy SK, Marsh JW, Varley PR, et al. Underlying
steatohepatitis, but not simple hepatic steatosis, increases
morbidity after liver resection: a case-control study.
Hepatology 2012;56:2221.