European Journal of Clinical Nutrition (2014) 68, 964–969 © 2014 Macmillan Publishers Limited All rights reserved 0954-3007/14 www.nature.com/ejcn
ORIGINAL ARTICLE
Preoperative immunonutrition in liver resection—a propensity score matched case-control analysis T Zacharias, N Ferreira and A-J Carin BACKGROUND/OBJECTIVES: The value of preoperative nutritional support in liver resection remains questionable. The aim of the present study was to compare the incidence of postoperative complications after liver resection between those patients that received preoperative immunonutrition versus those patients without preoperative nutritional support. SUBJECTS/METHODS: Patients undergoing elective liver resection between 9 November 2007 and 14 May 2013 were considered for the study: 84 with preoperative immunonutrition (Oral-Impact, Nestle, 3 × 237 ml per day for seven days at home) and 63 control patients without preoperative nutritional support. To reduce selection bias, propensity score matching was performed. Primary endpoint was the overall complication rate. Secondary endpoints were infectious and major complications. RESULTS: Ninety-eight patients could be matched (49 in each group). Seventy-seven patients (78.6%) had a minor and 21 patients (21.4%) a major liver resection. The two groups were balanced for age, gender distribution, American Society of Anesthesiology score ⩾ 3, NRS 2002 score, weight loss410%, cirrhosis, steathosis, preoperative chemotherapy, pathology, major liver resection and extrahepatic procedures. No significant differences were observed in the overall complication rate (53.0% versus 51.0%), infectious complications (38.7% versus 28.5%) and major complications (12.2% versus 10.2%) for the immunonutrition and control group, respectively. CONCLUSIONS: The present study did not permit to demonstrate an impact of preoperative immunonutrition with Oral-Impact on postoperative complications after minor liver resection. European Journal of Clinical Nutrition (2014) 68, 964–969; doi:10.1038/ejcn.2014.113; published online 25 June 2014
INTRODUCTION Liver resection remains an intervention with considerable morbidity.1,2 Infectious complications were registered in 22 to 41% of patients after liver resection.1,3 Surgical site infection was observed in 9–18% of patients.4 Infectious complications after liver resection were associated with increased mortality, longer hospital stay and poorer long-term survival.5,6 Therefore, strategies to reduce the risk of infectious and overall complications are needed. Immunonutrition is a therapeutic approach aimed at restoring immune function by correcting arginine deficiency and modulating the systemic inflammatory response.7 Immunonutrition has been studied in gastrointestinal surgery since 1992,8 and has shown to reduce the overall complication rate from 48% to 25%9 and the infectious complication rate from 36% to 18%.10–12 Perioperative immunonutrition in major abdominal surgery was recommended by the American Society for Parenteral and Enteral Nutrition and the European Society for Parenteral and Enteral Nutrition (ESPEN) and implemented in national guidelines.13,14 According to recent reports the value of nutritional support in liver resections remains questionable.15,16 To the best of the authors' knowledge, only one randomized trial of preoperative immunonutrition in liver resection has been reported.17 In the authors' institution preoperative immunonutrition (Oral-Impact, Nestle) was used in major abdominal surgery including hepatobiliary surgery since 2009. The aim of the present study was to compare the incidence of postoperative complications after liver resection between those patients that received preoperative immunonutrition versus those patients without preoperative nutritional support.
SUBJECTS AND METHODS The present study was performed in a single institution. One-hundred and sixty patients had 198 liver resections in the authors' institution between 9 November 2007 and 14 May 2013. The data were retrieved from a surgery department based liver resection database and retrospectively analyzed. In the database the following variables were prospectively recorded by the authors: age, gender, pathology, TNM-stage, histology, preoperative nutritional support and chemotherapy, preoperative laboratory values (hemoglobin, albumin, bilirubin, prothrombin-level and so on), bodyweight, weight loss, body-height, co-morbidities, NRS 2002 score, American Society of Anesthesiology (ASA) score, type of liver resection and associated procedures, length of operation and Pringle clamping time, estimated blood loss, perioperative transfusion, drainage, any postoperative complications with treatment, mortality, length of hospital stay and latest news. Only the first liver resections (n = 160) were considered in the present study, in order to avoid paired data which are statistically not independent in patients with more than one liver resection. Nine patients with emergency liver resection and four patients with other preoperative nutritional support were excluded from the analysis. Therefore 147 patients were available for analysis (63 without nutritional support and 84 with Oral-Impact). In order to reduce selection bias because of the nonrandomized study design, propensity score matching was performed.18 Immunonutrition was given at home before hospital admission with Oral-Impact, Nestle, 3 × 237 ml per day for 7 consecutive days. Oral-Impact is a ready to drink hypercaloric (1.41 kcal/ml) and hyperproteinic (18 g per unit = 76 g/l) solution, enriched with arginine 4.2 g per unit of 237 ml, omega-3 fatty acids (eicosapentaenoic acid and docosahexaenoic acid) 1.1 g per unit and nucleic acid (RNA) 0.43 g per unit. Immunonutrition was added to normal diet. The control group had normal diet without nutritional support. Preoperative immunonutrition was introduced in the authors' institution for major abdominal surgery including hepatobiliary
Service de chirurgie générale, endocrinienne et digestive, Pôle de Pathologies Digestives et Urologiques, Hôpital Emile Muller, 20 rue du Dr Laennec, BP 1370, 68070 Mulhouse Cedex, France. Correspondence: Dr T Zacharias, Service de chirurgie générale, endocrinienne et digestive, Pôle de Pathologies Digestives et Urologiques, Hôpital Emile Muller, 20 rue du Dr Laennec, BP 1370, 68070 Mulhouse Cedex, France. E-mail: [email protected] Received 9 August 2013; revised 29 March 2014; accepted 9 May 2014; published online 25 June 2014
Immunonutrition in liver resection T Zacharias et al
965 surgery in 2009.14 Patient’s compliance with immunonutrition was verified the day before surgery by asking the patient if he had really taken OralImpact. However, the number of units was not registered. Patients who required extrahepatic procedures like simultaneous colorectal resection or others were included in the study. All patients gave informed written consent for surgery and data collection. Ethics approvals were not required because of the retrospective monocentric study design and local data analysis without data transmission.19 Patients were followed-up for at least 3 months.
score, cirrhosis, hepatocellular carcinoma, colorectal metastasis, cholangiocarcinoma, extrahepatic procedure and major liver resection, to predict the probability to receive immunonutrition. One to one greedy, nearestneighbor matching without replacement on the logit of the propensity score and using a caliper distance equal to 0.25 of the standard deviation of the logit of the propensity score18 was performed to form pairs of treated and untreated subjects. Balance diagnostics for comparing the distribution of baseline covariates between immunonutrition and control groups in the unmatched and propensity score matched samples were performed using standardized differences. 29
Surgical Technique The surgical technique was described elsewhere.20 After a bilateral subcostal incision, liver parenchyma transection was performed by the clamp crushing technique under intermittent portal triad clamping. Ultrasonography and a bile leakage test were performed routinely. Red blood cells transfusion was given when hemoglobin level dropped below 8 g/dl and according to hemodynamic tolerance. Albumin was occasionally given postoperatively for the treatment of ascites. Postoperatively patients were monitored in the intensive care unit for at least 24 h. Antibiotics were given perioperatively according to guidelines.21 During the study period—with the exception of the introduction of immunonutrition in 2009—there was no significant modification in the perioperative management: an enhanced recovery protocol was not yet introduced, carbohydrate loading was not used, transfusion practice and fluid management were not changed during the study period.
Definitions Major abdominal surgery was defined as any gastric, hepatic, pancreatic, intestinal and colorectal resection. Major hepatectomy was defined as the resection of three or more liver segments. The Brisbane terminology was used for classification of the type of liver resection.22 Extrahepatic procedures were defined as partial or total resection of other organs (colon, rectum, stomach, pancreas, diaphragm, adrenal gland) and biliodigestive anastomosis. Cholecystectomy, liver biopsy and lymph node sampling or dissection were not considered as extrahepatic procedures. Overall complications were defined as any deviation from an uneventful postoperative course within 90 days after surgery. The diagnosis of an infectious complication was based on clinical, biological and radiological data and included: catheter, surgical site, pulmonary and urinary infections. The definition of infectious complications were given according to the criteria established by the Centers for Disease Control and Prevention and the Consensus Conference in 1991.23 Infectious complications were treated adequately with either antibiotics, percutaneous drainage or surgical revision. Wound infection were classified as surgical site infection. Severity of complication were defined according to the Clavien–Dindo classification.24 Hospital stay was defined as postoperative hospital stay. Nutritional risk was evaluated at admission according to the Nutritional Risk Screening 2002 (NRS 2002), as recommended by ESPEN.25 Nutritional status was evaluated according to the Nutritional Risk Index (NRI), defined as: NRI = 1.519xalbuminemia (g/l)+41.7xactual weight/usual weight, as recommended by the French Society of Digestive Surgery.14 According to the NRI patients were classified as: no malnutrition NRI ⩾ 97.5, moderate malnutrition NRIo97.5, and severe malnutrition NRIo83.5.26
Endpoints The primary endpoint was the overall complication rate. Secondary endpoints were infectious and major complications. The endpoints were chosen according to studies of immunonutrition in abdominal surgery.9–12,27 A power analysis was not performed owing to the retrospective study design.28
Statistical analysis Continuous variables were reported as mean with standard deviation or as median with range and compared with a non-parametric Mann–Whitney test. Dichotomous variables were compared with the Pearson χ2 or Fisher’s exact test, as appropriate. For paired variables (after matching), a McNemar test was used for dichotomous and a Wilcoxon Signed-Ranks test for continuous variables. All statistical tests were two-sided, and Po 0.050 was considered significant. To correct for baseline selection bias, propensity score matching was performed.18 A propensity score was constructed with logistic regression analysis for each of the 147 patients using 11 preoperatively known variables: age in years, female gender, cancer, ASA score ⩾ 3, NRS 2002 © 2014 Macmillan Publishers Limited
RESULTS Un-matched sample Baseline characteristics of all 147 patients were shown in Table 1. There were significant covariate imbalances between both groups: patients with immunonutrition were older and had more often cancer and hepatocellular carcinoma. Cirrhosis, ASA score ⩾ 3, comorbidity and malnutrition defined by weight loss410%, or NRIo 97.5 were more frequently recorded in the immunonutrition group. However, the differences were not significant as shown in Table 1. Matched sample Ninety-eight patients (66.7%) could be matched. Baseline data are shown in Table 2. The immunonutrition and control groups were balanced for age, gender distribution, ASA score ⩾ 3, NRS 2002 score, weight loss410%, cirrhosis, steathosis, preoperative chemotherapy, pathology, major liver resection and extrahepatic procedures (standardized differences below or near 0.1). For Charlson Comorbidity Index, body mass index, NRI o 97.5, and mean weight loss standardized differences remained between 0.1 and 0.15 after propensity score matching. The baseline covariate imbalances for the unmatched and matched samples were illustrated in Figure 1. All patients with cirrhosis (n = 8) were classified as Child-Pugh A. The cause of cirrhosis was viral hepatitis in five patients (2 B and 3 C). Cirrhotic patients had a median age of 58.5 years (range 49–81), all had cancer: hepatocellular carcinoma (seven patients) and colorectal liver metastasis (one patient), and four had nutritional risk: NRS 2002 score ⩾ 3. The type of liver resection, simultaneous extrahepatic procedures and intraoperative data were shown in Table 3. Postoperative course was uneventful in 47 patients. Median hospital stay after the operation was 10 days (range 4–62). Complications in matched patients Complications were registered in 51 patients (52.0%). No patient required reoperation. Five patients (5.1%) had a percutaneous radiological drainage for infected intraabdominal collections. Anastomotic leakage after simultaneous colorectal resection (n = 20) were not registered in the study patients. Major complications (Clavien–Dindo ⩾ 3) were recorded in 11 patients (11.2%) and infectious complications in 33 patients (33.6%). Pulmonary infections were registered in 13 patients (13.3%), surgical site infection (including wound infection) in 11 patients (11.2%), urinary tract infection in five patients (5.1%), and central venous catheter infection in eight patients (8.2%). Some patients had more than one infection. Data were shown in Table 4. There were no significant differences between the immunonutrition and control group for any of the studied endpoints. Mortality in matched patients There were no intraoperative deaths. Two patients (2.0%) died within 90 days; causes of death were: septic choc and multiorgan failure after extended right hepatectomy with bile duct resection after preoperative biliary drainage for hilar European Journal of Clinical Nutrition (2014) 964 – 969
Immunonutrition in liver resection T Zacharias et al
966 Table 1.
Baseline characteristics of patients in the immunonutrition group and in the control group before matching
Female gender* Age* in years Cancer* Charlson comorbidity index Body mass index NRS 2002*
Immunonutrition
Control
N = 84 (%)
N = 63 (%)
28 66.9 79 5.7 26.1 3.5
(33.3%) (±12.2) (94.0%) (±2.2) (±4.8) (±1.3)
17 62.8 50 4.8 26.4 3.2
P
Standardized difference
(26.9%) (±14.0) (79.3%) (±2.5) (±4.7) (±1.1)
0.471 0.047 0.010 0.101 0.960 0.167
0.13867 0.31037 0.44294 0.38406 0.07862 0.27616
Nutritional risk index o97.5 o83.5
26 (30.9%) 6 (7.1%)
17 (26.9%) 3 (4.7%)
0.714 0.732
0.08756 0.10075
Weight loss in % 410% ASA score ⩾ 3* Cirrhosis* Steathosis
6.5 26 30 14 36
(±7.2) (30.9%) (35.7%) (16.6%) (42.8%)
5.4 14 14 5 28
(±5.7) (22.2%) (22.2%) (7.9%) (44.4%)
0.674 0.265 0.101 0.141 1
0.16611 0.19857 0.30077 0.26817 0.03200
Diagnosis Colorectal metastases* Hepatocellular carcinoma* Cholangiocarcinoma* Other diagnosis Preoperative chemotherapy Major hepatectomy* Extrahepatic procedure*
39 24 11 10 19 29 28
(46.4%) (28.5%) (13.1%) (11.9%) (22.6%) (34.5%) (33.3%)
27 9 9 18 13 14 17
(42.8%) (14.3%) (14.3%) (28.5%) (20.6%) (22.2%) (26.9%)
0.738 0.046 1 0.018 0.841 0.142 0.471
0.07188 0.35355 0.03463 0.42404 0.04820 0.27545 0.13867
Abbreviations: ASA, American society of anesthesiology; NRS 2002, nutritional risk screening 2002. Continuous variables are reported as mean and standard deviation. Dichotomous variables are reported as N (%). Continuous variables were compared using a Mann–Whitney test. Dichotomous variables were compared using Fisher’s exact test. Eleven variables (*) were used to construct the propensity score. Absolute standardized differences were calculated according to Austin.18
cholangiocarcinoma, and septic cerebral embolism as a result of endocarditis in a patient with hepatocellular carcinoma and liver cirrhosis. DISCUSSION The aim of the present study was to compare the incidence of postoperative complications after liver resections between those patients who received preoperative immunonutrition with OralImpact versus those patients who didn't receive preoperative nutritional support. The present study did not permit to demonstrate an impact of preoperative immunonutrition with Oral-Impact on overall, major and infectious complications after minor liver resection. To the best of the authors' knowledge, only Mikagi et al.17 have analyzed in a randomized study the effect of preoperative immunonutrition versus no nutritional support in liver resection. Twenty-six patients were included in their study. Mikagi et al.17 have shown a decreased postoperative inflammatory response, measured by significantly lower white blood cell counts and interleukin 6-levels for patients with immunonutrition versus control patients. In their study the incidence of postoperative complications was not significantly different between the immunonutrition group versus the control group. Immunonutrition was studied in liver transplantation. In 2005 a pilot study by Plank et al.30 compared 15 patients with preoperative immunonutrition with 17 historical controls. Infectious complications were registered in five patients (33.3%) with immunonutrition versus 12 patients (70.5%) in the control group (P = 0.074). However, the results were biased by the nonrandomized study design and a high reoperation rate (47.0%) in the control group. In 2010 Plank et al.31 reported a randomized European Journal of Clinical Nutrition (2014) 964 – 969
study including 101 patients with liver transplantation. Infectious complication were recorded in 60% patients with perioperative immunonutrition versus 57% control patients with isonitrogenous, isocaloric nutrition (P = 0.841).31 It has been shown that preoperative immunonutrition may reduce the overall and infectious complication rate in colorectal, pancreatic and gastric surgery.9–11 However, patients with liver resection were not included in these studies. Recently, Hübner et al.27 included 62 patients undergoing hepatobiliary surgery in a randomized trial with a total of 145 patients (73 with Oral-Impact and 72 with isocaloric isonitrogenous nutrition) at nutritional risk (NRS 2002 score ⩾ 3) and undergoing major gastrointestinal surgery. Complications were registered in 39 immunonutrition patients (53.4%) versus 33 control patients (45.8%) (P = 0.408). Subgroup analysis for patients with liver resection was not performed.27 A Cochrane review by Koretz et al.15 analyzed nutritional support with oral nutritional supplements for surgical disorders in liver disease. Besides the study of Mikagi et al.17 three other trials of liver resection, and analyzing oral nutritional supplements versus a control group without nutritional support, were included in the review.32–34 No differences were found with regard to postoperative complications in these trials.15 In summary, trials for liver resection reported no benefit for oral nutritional supplements (immunonutrition or others) versus no nutritional support;17,32–34 and trials for liver transplantation,31 and gastrointestinal surgery including hepatobiliary surgery27 found no benefit for immunonutrition versus an isonitrogenous, isocaloric nutrition with regard to postoperative complications. The present study, which did not permit to demonstrate an impact of immunonutrition on postoperative complications after liver resection, seems to confirm these trials.17,31–34 © 2014 Macmillan Publishers Limited
Immunonutrition in liver resection T Zacharias et al
967 Table 2.
Baseline characteristics of patients in the immunonutrition group and in the control group after propensity score matching
Female gender* Age* in years Cancer* Charlson comorbidity index Body Mass Index NRS 2002*
Immunonutrition
Control
N = 49 (%)
N = 49 (%)
11 65.9 44 5.6 26.5 3.3
(22.4%) (±13.1) (89.7%) (±2.5) (±5.1) (±1.2)
12 65.6 43 5.3 25.9 3.3
P
Standardized difference
(24.4%) (±10.8) (87.7%) (±2.3) (±4.5) (±1.1)
1 0.548 1 0.624 0.459 0.555
0.04816 0.02840 0.0468 0.12452 0.10644 0.06808
Nutritional risk index o97.5 o83.5
10 (20.4%) 3 (6.1%)
13 (26.5%) 3 (6.1%)
0.303 0.500
0.14485 0
Weight loss in % 410% ASA score ⩾ 3* Cirrhosis* Steathosis
5.0 12 11 4 20
(±6.6) (24.4%) (22.4%) (8.1%) (40.8%)
5.7 11 12 4 21
(±5.8) (22.4%) (24.4%) (8.1%) (42.8%)
0.490 0.500 1 1 1
0.11374 0.04816 0.04816 0 0.04138
Diagnosis Colorectal metastases* Hepatocellular carcinoma* Cholangiocarcinoma* Other diagnosis Preoperative chemotherapy Major hepatectomy* Extrahepatic procedure*
24 11 5 9 13 10 16
(49.0%) (22.4%) (10.2%) (18.3%) (26.5%) (20.4%) (32.6%)
25 9 5 10 12 11 16
(51.0%) (18.3%) (10.2%) (20.4%) (24.4%) (22.4%) (32.6%)
1 0.753 1 1 1 1 1
0.04082 0.10140 0 0.05163 0.04682 0.04975 0
Abbreviations: ASA, American society of anesthesiology; NRS 2002, nutritional risk screening 2002. Continuous variables are reported as mean and standard deviation. Dichotomous variables are reported as N (%). Continuous variables were compared using a Wilcoxon Signed-Ranks test. Dichotomous variables were compared using a McNemar test. Eleven variables (*) were used to construct the propensity score. Absolute standardized differences were calculated according to Austin.18
Absolute standardized differences for baseline covariates comparing treated to control subjects in the original and in the matched sample 0.5
Absolute standardized difference
0.45 0.4 0.35 0.3 original sample
0.25
matched sample
0.2 0.15 0.1 0.05
Ch
ar
lso
Fe
m al
eg en de r n Co A ge m o C N ut W rbid anc rit eig it er y io h na t l Ind l R os e isk s > x Sc 10% re A enin SA g C Sc H olo ep re o at ct Cirr re oc al h o el m s Ch lula eta is ol r c stas an ar Pr gi cin is o eo o pe O carc ma ra th tiv er inom di e a M che agn m os a Ex jo o i tra r h the s he epa rap pa te y tic cto pr my oc ed ur e
0
Baseline covariates
Figure 1. Absolute standardized differences for baseline covariates comparing treated (preoperative immunonutrition) to control subjects in the original (147 patients) and in the matched sample (98 patients) for liver resection.
© 2014 Macmillan Publishers Limited
European Journal of Clinical Nutrition (2014) 964 – 969
Immunonutrition in liver resection T Zacharias et al
968 Table 3.
Perioperative data for 49 matched pairs of immunonutrition and control patients.
Major hepatectomy Right hepatectomy Left hepatectomy Trisegmentectomy Othera
Immunonutrition
Control
N = 49 (%)
N = 49 (%)
P
10 (20.4%)
11 (22.4%)
1
3 4 2 1
(6.1%) (8.1%) (4%) (2%)
5 4 1 1
(10.2%) (8.1%) (2%) (2%)
0.726 1 1 1
Minor hepatectomy Bisegmentectomy 2,3 Bisegmentectomy (other) Segmentectomy and limited resection
39 9 17 13
(79.5%) (18.3%) (34.6%) (26.5%)
38 5 15 18
(77.5%) (10.2%) (30.6%) (36.7%)
1 0.387 0.814 0.404
Extrahepatic procedures Total Colorectal resection Otherb Median operating time in min (range)
16 9 7 210
Transfusion Yes 18 Intra-operatively 15 Postoperatively 4 Median estimated blood loss in ml (range) 300
(32.6%) 16 (18.3%) 11 (14.2%) 5 (110–360) 247
(32.6%) 1 (22.4%) 0.774 (10.2%) 0.753 (120–505) 0.013
(36.7%) (30.6%) (8.1%) (50–2000)
(22.4%) 0.189 (16.3%) 0.167 (8.1%) 1 (30–1300) 0.936
11 8 4 300
Continuous variables are reported as median and range. Dichotomous variables are reported as N (%). Continuous variables were compared using a Wilcoxon Signed-Ranks test. Dichotomous variables were compared using a McNemar test. aOne patient had bisegmentectomy+unisegmentectomy (three resected segments) and one patient had two bisegmentectomies classifying those resections as major hepatectomy. bOther extrahepatic procedures included: total gastrectomy (n = 2), left pancreatectomy (n = 2) partial resection of the right diaphragm (n = 2), right adrenalectomy (n = 2) and biliodigestive anastomosis (n = 4).
Limitations In the present study in the unmatched sample, there were significant covariate imbalances between both groups (Table 1). Therefore, propensity score matching was performed to reduce baseline covariate imbalances between groups.18 The immunonutrition and control group in the matched sample were balanced for the majority of measured baseline covariates (standardized differences below or near 0.1). For some baseline covariates standardized differences remained between 0.1 and 0.15 after matching: NRI o 97.5 (moderate malnutrition) was more frequently observed in control patients, and mean weight loss was 5.7% in control patients versus 5.0% in immunonutrition. On the other hand mean Charlson Comorbidity Index was 5.6 in immunonutrition versus 5.3 in control patients. (Table 2). These differences were not significant. However, the authors cannot rule out that more sophisticated methods for determining the nutritional status would show a significant difference between groups in the matched sample. The remaining covariates imbalance after matching were partially explained by the rather small number of patients (n = 147), which limited the possibilities of matching. Furthermore, propensity score matching will, on an average, result in measured baseline covariates being balanced between treatment groups.18 Non-measured covariates can only be balanced between treatment groups by randomization. The absence of randomization remains the major limitation of the present study. Furthermore, the rather small number of patients available for analysis after propensity score matching limited the statistical value of the present study. Type II error may be an explanation for the absence of benefit of immunonutrition observed in the present study. A further limitation of the present study was the European Journal of Clinical Nutrition (2014) 964 – 969
Table 4. Postoperative outcome data: number of patients with complications (infectious and others), severity of complications according to the Clavien–Dindo classification,24 and hospital stay for 49 matched pairs of immunonutrition and control patients. Immunonutrition
Control
N = 49 (%)
N = 49 (%)
P
Number of patients with complications
26 (53%)
25 (51%)
1
Number of patients with infectious complications Pulmonary infection Venous catheter infection Urinary tract infection Surgical site infection
19 8 4 2 6
(38.7%) (16.3%) (8.1%) (4%) (12.2%)
14 5 4 3 5
7 (14.2%) 0 1 (2%) 1 (2%) 1 (2%) 1 (2%) 1 (2%) 2 (4%)
11 2 1 4 2 1
Number of patients with other complications Ascites Bile leakage Ileus Thrombosis Arrhythmia Confusion Pleural effusions Number of patients with complications grade Clavien–Dindo 1 Clavien–Dindo 2 Clavien–Dindo 3 Clavien–Dindo 4 Clavien–Dindo 5 (90 day mortality) Major complications (Clavien–Dindo 3, 4 and 5) Median hospital stay in days (range)
3 17 2 3 1 6 10
(6.1%) (34.6%) (4%) (6.1%) (2%) (12.2%) (5–62)
(28.5%) 0.383 (10.2%) 0.548 (8.1%) 1 (6.1%) 1 (10.2%) 1
(22.4%) 0.480 (4%) 0.500 (2%) 1 (8.1%) 0.375 (4%) 1 (2%) 1 0 1 1 (2%) 1
6 14 3 1 1 5 10
(12.2%) (28.5%) (6.1%) (2%) (2%) (10.2%) (4–32)
0.5078 0.6072 1 0.6250 1 1 0.9522
Continuous variables are reported as median and range. Dichotomous variables are reported as N (%). Continuous variables were compared using a Wilcoxon Signed-Ranks test. Dichotomous variables were compared using a McNemar test.
small number of major liver resection 21.4% (21/98) in the matched sample. Therefore, the findings of the present study are limited to minor hepatectomy. A possible bias may be a shorter median operation time and a higher transfusion rate for the immunonutrition group. It has been shown, that a shorter operation time was associated with a reduced complication rate and inversely.35 On the other hand, a higher transfusion rate has been reported to increase the complication rate.36 In consequence, the effect of both variables on the complication rate in the immunonutrition group may be ‘neutralized’. The duration of the operation and the transfusion rate were not included in the propensity score model, as only variables that are measured at baseline (before the administration of treatment) should be included in the model.18 A further bias may be the compliance with immunonutrition. In the present study, patients started immunonutrition 7 days before surgery at home. The day before surgery the patients were asked if they have really taken Oral-Impact and all patients in the immunonutrition group have indeed done so. However, the number of units of immunonutrition taken by patients were not recorded. This was a shortcoming of the present study, as it is assumed that the effect of immunonutrition depends on dose. A limited compliance with preoperative immunonutrition has been reported in the outpatient setting and may influence postoperative outcome.27 Concerning the timing of the immunonutrition recent metaanalyses have underlined the importance of peri and postoperative administration of immunonutrition.37,38 On the other hand, preoperative administration was recommended in 2005 by French guidelines14 for major abdominal surgery based on studies that have demonstrated a benefit of pre and perioperative © 2014 Macmillan Publishers Limited
Immunonutrition in liver resection T Zacharias et al
immunonutrition for reducing infectious10 and overall9 complications after gastrointestinal surgery. For logistical reasons OralImpact was not available inside the hospital during the study period. Therefore, the administration of immunonutrition was limited to the preoperative outpatient setting. The absence of postoperative immunonutrition may contribute to the ‘negative’ results of the present study.38 With these limitations, the present study may be regarded as a ‘real life’ picture of immunonutrition for liver resection in the authors' institution. The authors expected to observe a reduced complication rate after introducing immunonutrition for liver resection, based on reports showing a reduction of complications in gastrointestinal surgery.9–12 The authors continue to use immunonutrition for gastrointestinal surgery,12 and selectively in patients undergoing liver resections. However, further study of immunonutrition in liver resection is needed. CONCLUSION The present study did not permit to demonstrate an impact of preoperative immunonutrition with Oral-Impact on postoperative complications after minor liver resection. CONFLICT OF INTEREST The authors declare no conflict of interest.
REFERENCES 1 Jarnagin WR, Gonen M, Fong Y, DeMatteo RP, Ben-Porat L, Little S et al. Improvement in perioperative outcome after hepatic resection: analysis of 1803 consecutive cases over the past decade. Ann Surg 2002; 236: 397–407. 2 Aloia TA, Fahy BN, Fischer CP, Jones SL, Duchini A, Galati J et al. Predicting poor outcome following hepatectomy: analysis of 2313 hepatectomies in the NSQIP database. HPB (Oxford) 2009; 11: 510–515. 3 Koperna T, Kisser M, Schulz F. Infections after liver resection in the elderly. Langenbecks Arch Chir 1997; 382: 192–196. 4 Moreno Elola-Olaso A, Davenport DL, Hundley JC, Daily MF, Gedaly R. Predictors of surgical site infection after liver resection: a multicentric analysis using national Surgical Quality Improuvement Prgram data. HPB (Oxford) 2012; 14: 136–141. 5 Neal CP, Mann CD, Garcea G, Briggs CD, Dennison AR, Berry DP. Preoperative systemic inflammation and infectious complications after resection of colorectal liver metastases. Arch Surg 2011; 146: 471–478. 6 Farid SG, Aldouri A, Morris-Stiff G, Khan AZ, Toogood GJ, Lodge JP et al. Correlation between postoperative infective complications and long-term outcome after hepatic resection for colorectal liver metastasis. Ann Surg 2010; 251: 91–100. 7 Mizock BA, Sriram K. Perioperative immunonutrition. Expert Rev Clin Immunol 2011; 7: 1–3. 8 Daly JM, Lieberman MD, Goldfine J, Shou J, Weintraub F, Rosato EF et al. Enteral nutrition with supplemental arginine, RNA, and omega-3 fatty acids in patients after operation: immunologic, metabolic, and clinical outcome. Surgery 1992; 112: 56–67. 9 Braga M, Gianotti L, Nespoli L, Radaelli G, Di Carlo V. Nutritional approach in malnourished surgical patients. A prospective randomized study. Arch Surg 2002; 137: 174–180. 10 Gianotti L, Braga M, Nespoli L, Radaelli G, Beneduce A, Di Carlo V. A randomized controlled trial on preoperative oral supplementation with a specialized diet in patients with gastrointestinal cancer. Gastroenterology 2002; 122: 1763–1770. 11 Senkal M, Zumtobel V, Bauer KH, Marpe B, Wolfram G, Frei A et al. Outcome and cost-effectiveness of perioperative enteral immunonutrition in patients undergoing elective upper gastrointestinal tract surgery: a prospective randomized study. Arch Surg 1999; 134: 1309–1316. 12 Cerantola Y, Hübner M, Grass F, Demartines N, Schäfer M. Immunonutrition in gastrointestinal Surgery. Br J Surg 2011; 98: 37–48. 13 Weimann A, Braga M, Harsanyi L, Laviano A, Ljungqvist O, Soeters P et al. ESPEN guidelines on enteral nutrition: surgery including organ transplantation. Clin Nutr 2006; 25: 224–244. 14 Mariette C, Alves A, Benoist S, Bretagnol F, Mabrut JY, Slim K. Perioperative care in digestive surgery. Guidelines for the French Society of Digestive Surgery. Ann Chir 2005; 130: 108–124.
© 2014 Macmillan Publishers Limited
969 15 Koretz RL, Avenell A, Lipman TO. Nutritional support for liver disease. Cochrane Database Syst Rev 2012; 5: CD008344. 16 Richter B, Schmandra TC, Golling M, Bechstein WO. Nutritional support after open liver resection: a systematic review. Dig Surg 2006; 23: 139–145. 17 Mikagi K, Kawahara R, Kinoshita H, Aoyagi S. Effect of preoperative immunonutrition in patients undergoing hepatectomy; a randomized controlled trial. Kurume Med J 2011; 58: 1–8. 18 Austin PC. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behav Res 2011; 46: 399–424. 19 Claudot F, Alla F, Fresson J, Calvez T, Coudane H, Bonaïti-Pellié C. Ethics and observational studie in medical research: various rules in a common framework. Int J Epidemiol 2009; 38: 1104–1108. 20 Zacharias T, Ferreira N. Carrier-bound fibrin sealant compared to oxidized cellulose application after liver resection. HPB (Oxford) 2012; 14: 839–847. 21 Société française d’anesthesie et de reanimation. Antibioprophylaxis in surgery and interventional medicine (adult patients). Actualization 2010. Ann Fr Anesth Reanim 2011; 30: 168–190. 22 Strasberg SM. Nomenclature of hepatic anatomy and resections: a review of the Brisbane 2000 system. J Hepatobiliary Pancreat Surg 2005; 12: 351–355. 23 Bone RC, Balk RA, Cerra FB, Dellinger RP, Fein AM, Knaus WA et al. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Chest 1992; 101: 1644–1655. 24 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: 205–213. 25 Kondrup J, Rasmussen HH, Hamberg O, Stanga Z, Ad Hoc ESPEN Working Group. Nutritional risk screening (NRS 2002): a new method based on an analysis of controlled clinical trials. Clin Nutr 2003; 22: 321–336. 26 The Veterans Affairs Total Parenteral Nutrition Cooperative Study Group. Perioperative total parenteral nutrition in surgical patients. N Engl J Med 1991; 325: 525–532. 27 Hübner M, Cerantola Y, Grass F, Bertrand PC, Schäfer M, Demartines N. Preoperative immunonutrition in patients at nutritional risk: results of a double-blinded randomized clinical trial. Eur J Clin Nutr 2012; 66: 850–855. 28 Schulz KF, Grimes DA. Sample size calculations in randomized trials: mandatory and mystical. Lancet 2005; 365: 1348–1353. 29 Austin PC. Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples. Statist Med 2009; 28: 3083–3107. 30 Plank LD, McCall JL, Gane EJ, Rafique M, Gillanders LK, McIlroy K et al. Pre- and postoperative immunonutrition in patients undergoing liver transplantation: a pilot study of safety and efficacy. Clin Nutr 2005; 24: 288–296. 31 Plank LD, Mathur S, Gane EJ, Gillanders LK, McIlroy K, McCall JL. Perioperative immunonutrition in liver transplantation: results of a double-blind randomized controlled trial. Clin Nutr 2010; 5 (Suppl2): 3. 32 Hendry PO, van Dam RM, Bukkems SF, McKeown DW, Parks RW, Preston T et al. Randomized clinical trial of laxatives and oral nutritional supplements within an enhanced recovery after surgery protocol following liver resection. Br J Surg 2010; 97: 1198–1206. 33 Ishikawa Y, Yoshida H, Mamada Y, Taniai N, Matsumoto S, Bando K et al. Prospective randomized controlled study of short-term perioperative oral nutrition with branched chain amino acids in patients undergoing liver surgery. Hepatogastroenterology 2010; 57: 583–590. 34 Meng WC, Leung KL, Ho RL, Leung TW, Lau WY. Prospective randomized control study on the effect of branched-chain amino acids in patients with liver resection for hepatocellular carcinoma. Aust N Z J Surg 1999; 69: 811–815. 35 Tzeng CW, Katz MH, Fleming JB, Pisters PW, Lee JE, Abdalla EK et al. Risk of venous thromboembolism outweighs post-hepatectomy bleeding complications: analysis of 5651 National Surgical Quality Improuvement Program patients. HPB (Oxford) 2012; 14: 506–513. 36 Kooby DA, Stockmann J, Ben-Porat L, Gonen M, Janargin WR, Dematteo R et al. Influence of transfusion on perioperative and long-term outcome in patients following hepatic resection for colorectal metastases. Ann Surg 2003; 237: 860–870. 37 Marimuthu K, Varadhan KK, Ljungqvist O, Lobo DN. A meta-analysis of the effect of combinations of immune modulating nutrients on outcome in patients undergoing major open gastrointestinal surgery. Ann Surg 2012; 255: 1060–1068. 38 Osland E, Hossain MB, Khan S, Memon MA. Effect of timing of pharmaconutrition (immunonutrition) administration on outcomes of elective surgery for gastrointestinal malignancies: a systematic review and meta-analysis. JPEN J Parenter Enteral Nutr 2014; 38: 53–69.
European Journal of Clinical Nutrition (2014) 964 – 969
ORIGINAL ARTICLE
Preoperative immunonutrition in liver resection—a propensity score matched case-control analysis T Zacharias, N Ferreira and A-J Carin BACKGROUND/OBJECTIVES: The value of preoperative nutritional support in liver resection remains questionable. The aim of the present study was to compare the incidence of postoperative complications after liver resection between those patients that received preoperative immunonutrition versus those patients without preoperative nutritional support. SUBJECTS/METHODS: Patients undergoing elective liver resection between 9 November 2007 and 14 May 2013 were considered for the study: 84 with preoperative immunonutrition (Oral-Impact, Nestle, 3 × 237 ml per day for seven days at home) and 63 control patients without preoperative nutritional support. To reduce selection bias, propensity score matching was performed. Primary endpoint was the overall complication rate. Secondary endpoints were infectious and major complications. RESULTS: Ninety-eight patients could be matched (49 in each group). Seventy-seven patients (78.6%) had a minor and 21 patients (21.4%) a major liver resection. The two groups were balanced for age, gender distribution, American Society of Anesthesiology score ⩾ 3, NRS 2002 score, weight loss410%, cirrhosis, steathosis, preoperative chemotherapy, pathology, major liver resection and extrahepatic procedures. No significant differences were observed in the overall complication rate (53.0% versus 51.0%), infectious complications (38.7% versus 28.5%) and major complications (12.2% versus 10.2%) for the immunonutrition and control group, respectively. CONCLUSIONS: The present study did not permit to demonstrate an impact of preoperative immunonutrition with Oral-Impact on postoperative complications after minor liver resection. European Journal of Clinical Nutrition (2014) 68, 964–969; doi:10.1038/ejcn.2014.113; published online 25 June 2014
INTRODUCTION Liver resection remains an intervention with considerable morbidity.1,2 Infectious complications were registered in 22 to 41% of patients after liver resection.1,3 Surgical site infection was observed in 9–18% of patients.4 Infectious complications after liver resection were associated with increased mortality, longer hospital stay and poorer long-term survival.5,6 Therefore, strategies to reduce the risk of infectious and overall complications are needed. Immunonutrition is a therapeutic approach aimed at restoring immune function by correcting arginine deficiency and modulating the systemic inflammatory response.7 Immunonutrition has been studied in gastrointestinal surgery since 1992,8 and has shown to reduce the overall complication rate from 48% to 25%9 and the infectious complication rate from 36% to 18%.10–12 Perioperative immunonutrition in major abdominal surgery was recommended by the American Society for Parenteral and Enteral Nutrition and the European Society for Parenteral and Enteral Nutrition (ESPEN) and implemented in national guidelines.13,14 According to recent reports the value of nutritional support in liver resections remains questionable.15,16 To the best of the authors' knowledge, only one randomized trial of preoperative immunonutrition in liver resection has been reported.17 In the authors' institution preoperative immunonutrition (Oral-Impact, Nestle) was used in major abdominal surgery including hepatobiliary surgery since 2009. The aim of the present study was to compare the incidence of postoperative complications after liver resection between those patients that received preoperative immunonutrition versus those patients without preoperative nutritional support.
SUBJECTS AND METHODS The present study was performed in a single institution. One-hundred and sixty patients had 198 liver resections in the authors' institution between 9 November 2007 and 14 May 2013. The data were retrieved from a surgery department based liver resection database and retrospectively analyzed. In the database the following variables were prospectively recorded by the authors: age, gender, pathology, TNM-stage, histology, preoperative nutritional support and chemotherapy, preoperative laboratory values (hemoglobin, albumin, bilirubin, prothrombin-level and so on), bodyweight, weight loss, body-height, co-morbidities, NRS 2002 score, American Society of Anesthesiology (ASA) score, type of liver resection and associated procedures, length of operation and Pringle clamping time, estimated blood loss, perioperative transfusion, drainage, any postoperative complications with treatment, mortality, length of hospital stay and latest news. Only the first liver resections (n = 160) were considered in the present study, in order to avoid paired data which are statistically not independent in patients with more than one liver resection. Nine patients with emergency liver resection and four patients with other preoperative nutritional support were excluded from the analysis. Therefore 147 patients were available for analysis (63 without nutritional support and 84 with Oral-Impact). In order to reduce selection bias because of the nonrandomized study design, propensity score matching was performed.18 Immunonutrition was given at home before hospital admission with Oral-Impact, Nestle, 3 × 237 ml per day for 7 consecutive days. Oral-Impact is a ready to drink hypercaloric (1.41 kcal/ml) and hyperproteinic (18 g per unit = 76 g/l) solution, enriched with arginine 4.2 g per unit of 237 ml, omega-3 fatty acids (eicosapentaenoic acid and docosahexaenoic acid) 1.1 g per unit and nucleic acid (RNA) 0.43 g per unit. Immunonutrition was added to normal diet. The control group had normal diet without nutritional support. Preoperative immunonutrition was introduced in the authors' institution for major abdominal surgery including hepatobiliary
Service de chirurgie générale, endocrinienne et digestive, Pôle de Pathologies Digestives et Urologiques, Hôpital Emile Muller, 20 rue du Dr Laennec, BP 1370, 68070 Mulhouse Cedex, France. Correspondence: Dr T Zacharias, Service de chirurgie générale, endocrinienne et digestive, Pôle de Pathologies Digestives et Urologiques, Hôpital Emile Muller, 20 rue du Dr Laennec, BP 1370, 68070 Mulhouse Cedex, France. E-mail: [email protected] Received 9 August 2013; revised 29 March 2014; accepted 9 May 2014; published online 25 June 2014
Immunonutrition in liver resection T Zacharias et al
965 surgery in 2009.14 Patient’s compliance with immunonutrition was verified the day before surgery by asking the patient if he had really taken OralImpact. However, the number of units was not registered. Patients who required extrahepatic procedures like simultaneous colorectal resection or others were included in the study. All patients gave informed written consent for surgery and data collection. Ethics approvals were not required because of the retrospective monocentric study design and local data analysis without data transmission.19 Patients were followed-up for at least 3 months.
score, cirrhosis, hepatocellular carcinoma, colorectal metastasis, cholangiocarcinoma, extrahepatic procedure and major liver resection, to predict the probability to receive immunonutrition. One to one greedy, nearestneighbor matching without replacement on the logit of the propensity score and using a caliper distance equal to 0.25 of the standard deviation of the logit of the propensity score18 was performed to form pairs of treated and untreated subjects. Balance diagnostics for comparing the distribution of baseline covariates between immunonutrition and control groups in the unmatched and propensity score matched samples were performed using standardized differences. 29
Surgical Technique The surgical technique was described elsewhere.20 After a bilateral subcostal incision, liver parenchyma transection was performed by the clamp crushing technique under intermittent portal triad clamping. Ultrasonography and a bile leakage test were performed routinely. Red blood cells transfusion was given when hemoglobin level dropped below 8 g/dl and according to hemodynamic tolerance. Albumin was occasionally given postoperatively for the treatment of ascites. Postoperatively patients were monitored in the intensive care unit for at least 24 h. Antibiotics were given perioperatively according to guidelines.21 During the study period—with the exception of the introduction of immunonutrition in 2009—there was no significant modification in the perioperative management: an enhanced recovery protocol was not yet introduced, carbohydrate loading was not used, transfusion practice and fluid management were not changed during the study period.
Definitions Major abdominal surgery was defined as any gastric, hepatic, pancreatic, intestinal and colorectal resection. Major hepatectomy was defined as the resection of three or more liver segments. The Brisbane terminology was used for classification of the type of liver resection.22 Extrahepatic procedures were defined as partial or total resection of other organs (colon, rectum, stomach, pancreas, diaphragm, adrenal gland) and biliodigestive anastomosis. Cholecystectomy, liver biopsy and lymph node sampling or dissection were not considered as extrahepatic procedures. Overall complications were defined as any deviation from an uneventful postoperative course within 90 days after surgery. The diagnosis of an infectious complication was based on clinical, biological and radiological data and included: catheter, surgical site, pulmonary and urinary infections. The definition of infectious complications were given according to the criteria established by the Centers for Disease Control and Prevention and the Consensus Conference in 1991.23 Infectious complications were treated adequately with either antibiotics, percutaneous drainage or surgical revision. Wound infection were classified as surgical site infection. Severity of complication were defined according to the Clavien–Dindo classification.24 Hospital stay was defined as postoperative hospital stay. Nutritional risk was evaluated at admission according to the Nutritional Risk Screening 2002 (NRS 2002), as recommended by ESPEN.25 Nutritional status was evaluated according to the Nutritional Risk Index (NRI), defined as: NRI = 1.519xalbuminemia (g/l)+41.7xactual weight/usual weight, as recommended by the French Society of Digestive Surgery.14 According to the NRI patients were classified as: no malnutrition NRI ⩾ 97.5, moderate malnutrition NRIo97.5, and severe malnutrition NRIo83.5.26
Endpoints The primary endpoint was the overall complication rate. Secondary endpoints were infectious and major complications. The endpoints were chosen according to studies of immunonutrition in abdominal surgery.9–12,27 A power analysis was not performed owing to the retrospective study design.28
Statistical analysis Continuous variables were reported as mean with standard deviation or as median with range and compared with a non-parametric Mann–Whitney test. Dichotomous variables were compared with the Pearson χ2 or Fisher’s exact test, as appropriate. For paired variables (after matching), a McNemar test was used for dichotomous and a Wilcoxon Signed-Ranks test for continuous variables. All statistical tests were two-sided, and Po 0.050 was considered significant. To correct for baseline selection bias, propensity score matching was performed.18 A propensity score was constructed with logistic regression analysis for each of the 147 patients using 11 preoperatively known variables: age in years, female gender, cancer, ASA score ⩾ 3, NRS 2002 © 2014 Macmillan Publishers Limited
RESULTS Un-matched sample Baseline characteristics of all 147 patients were shown in Table 1. There were significant covariate imbalances between both groups: patients with immunonutrition were older and had more often cancer and hepatocellular carcinoma. Cirrhosis, ASA score ⩾ 3, comorbidity and malnutrition defined by weight loss410%, or NRIo 97.5 were more frequently recorded in the immunonutrition group. However, the differences were not significant as shown in Table 1. Matched sample Ninety-eight patients (66.7%) could be matched. Baseline data are shown in Table 2. The immunonutrition and control groups were balanced for age, gender distribution, ASA score ⩾ 3, NRS 2002 score, weight loss410%, cirrhosis, steathosis, preoperative chemotherapy, pathology, major liver resection and extrahepatic procedures (standardized differences below or near 0.1). For Charlson Comorbidity Index, body mass index, NRI o 97.5, and mean weight loss standardized differences remained between 0.1 and 0.15 after propensity score matching. The baseline covariate imbalances for the unmatched and matched samples were illustrated in Figure 1. All patients with cirrhosis (n = 8) were classified as Child-Pugh A. The cause of cirrhosis was viral hepatitis in five patients (2 B and 3 C). Cirrhotic patients had a median age of 58.5 years (range 49–81), all had cancer: hepatocellular carcinoma (seven patients) and colorectal liver metastasis (one patient), and four had nutritional risk: NRS 2002 score ⩾ 3. The type of liver resection, simultaneous extrahepatic procedures and intraoperative data were shown in Table 3. Postoperative course was uneventful in 47 patients. Median hospital stay after the operation was 10 days (range 4–62). Complications in matched patients Complications were registered in 51 patients (52.0%). No patient required reoperation. Five patients (5.1%) had a percutaneous radiological drainage for infected intraabdominal collections. Anastomotic leakage after simultaneous colorectal resection (n = 20) were not registered in the study patients. Major complications (Clavien–Dindo ⩾ 3) were recorded in 11 patients (11.2%) and infectious complications in 33 patients (33.6%). Pulmonary infections were registered in 13 patients (13.3%), surgical site infection (including wound infection) in 11 patients (11.2%), urinary tract infection in five patients (5.1%), and central venous catheter infection in eight patients (8.2%). Some patients had more than one infection. Data were shown in Table 4. There were no significant differences between the immunonutrition and control group for any of the studied endpoints. Mortality in matched patients There were no intraoperative deaths. Two patients (2.0%) died within 90 days; causes of death were: septic choc and multiorgan failure after extended right hepatectomy with bile duct resection after preoperative biliary drainage for hilar European Journal of Clinical Nutrition (2014) 964 – 969
Immunonutrition in liver resection T Zacharias et al
966 Table 1.
Baseline characteristics of patients in the immunonutrition group and in the control group before matching
Female gender* Age* in years Cancer* Charlson comorbidity index Body mass index NRS 2002*
Immunonutrition
Control
N = 84 (%)
N = 63 (%)
28 66.9 79 5.7 26.1 3.5
(33.3%) (±12.2) (94.0%) (±2.2) (±4.8) (±1.3)
17 62.8 50 4.8 26.4 3.2
P
Standardized difference
(26.9%) (±14.0) (79.3%) (±2.5) (±4.7) (±1.1)
0.471 0.047 0.010 0.101 0.960 0.167
0.13867 0.31037 0.44294 0.38406 0.07862 0.27616
Nutritional risk index o97.5 o83.5
26 (30.9%) 6 (7.1%)
17 (26.9%) 3 (4.7%)
0.714 0.732
0.08756 0.10075
Weight loss in % 410% ASA score ⩾ 3* Cirrhosis* Steathosis
6.5 26 30 14 36
(±7.2) (30.9%) (35.7%) (16.6%) (42.8%)
5.4 14 14 5 28
(±5.7) (22.2%) (22.2%) (7.9%) (44.4%)
0.674 0.265 0.101 0.141 1
0.16611 0.19857 0.30077 0.26817 0.03200
Diagnosis Colorectal metastases* Hepatocellular carcinoma* Cholangiocarcinoma* Other diagnosis Preoperative chemotherapy Major hepatectomy* Extrahepatic procedure*
39 24 11 10 19 29 28
(46.4%) (28.5%) (13.1%) (11.9%) (22.6%) (34.5%) (33.3%)
27 9 9 18 13 14 17
(42.8%) (14.3%) (14.3%) (28.5%) (20.6%) (22.2%) (26.9%)
0.738 0.046 1 0.018 0.841 0.142 0.471
0.07188 0.35355 0.03463 0.42404 0.04820 0.27545 0.13867
Abbreviations: ASA, American society of anesthesiology; NRS 2002, nutritional risk screening 2002. Continuous variables are reported as mean and standard deviation. Dichotomous variables are reported as N (%). Continuous variables were compared using a Mann–Whitney test. Dichotomous variables were compared using Fisher’s exact test. Eleven variables (*) were used to construct the propensity score. Absolute standardized differences were calculated according to Austin.18
cholangiocarcinoma, and septic cerebral embolism as a result of endocarditis in a patient with hepatocellular carcinoma and liver cirrhosis. DISCUSSION The aim of the present study was to compare the incidence of postoperative complications after liver resections between those patients who received preoperative immunonutrition with OralImpact versus those patients who didn't receive preoperative nutritional support. The present study did not permit to demonstrate an impact of preoperative immunonutrition with Oral-Impact on overall, major and infectious complications after minor liver resection. To the best of the authors' knowledge, only Mikagi et al.17 have analyzed in a randomized study the effect of preoperative immunonutrition versus no nutritional support in liver resection. Twenty-six patients were included in their study. Mikagi et al.17 have shown a decreased postoperative inflammatory response, measured by significantly lower white blood cell counts and interleukin 6-levels for patients with immunonutrition versus control patients. In their study the incidence of postoperative complications was not significantly different between the immunonutrition group versus the control group. Immunonutrition was studied in liver transplantation. In 2005 a pilot study by Plank et al.30 compared 15 patients with preoperative immunonutrition with 17 historical controls. Infectious complications were registered in five patients (33.3%) with immunonutrition versus 12 patients (70.5%) in the control group (P = 0.074). However, the results were biased by the nonrandomized study design and a high reoperation rate (47.0%) in the control group. In 2010 Plank et al.31 reported a randomized European Journal of Clinical Nutrition (2014) 964 – 969
study including 101 patients with liver transplantation. Infectious complication were recorded in 60% patients with perioperative immunonutrition versus 57% control patients with isonitrogenous, isocaloric nutrition (P = 0.841).31 It has been shown that preoperative immunonutrition may reduce the overall and infectious complication rate in colorectal, pancreatic and gastric surgery.9–11 However, patients with liver resection were not included in these studies. Recently, Hübner et al.27 included 62 patients undergoing hepatobiliary surgery in a randomized trial with a total of 145 patients (73 with Oral-Impact and 72 with isocaloric isonitrogenous nutrition) at nutritional risk (NRS 2002 score ⩾ 3) and undergoing major gastrointestinal surgery. Complications were registered in 39 immunonutrition patients (53.4%) versus 33 control patients (45.8%) (P = 0.408). Subgroup analysis for patients with liver resection was not performed.27 A Cochrane review by Koretz et al.15 analyzed nutritional support with oral nutritional supplements for surgical disorders in liver disease. Besides the study of Mikagi et al.17 three other trials of liver resection, and analyzing oral nutritional supplements versus a control group without nutritional support, were included in the review.32–34 No differences were found with regard to postoperative complications in these trials.15 In summary, trials for liver resection reported no benefit for oral nutritional supplements (immunonutrition or others) versus no nutritional support;17,32–34 and trials for liver transplantation,31 and gastrointestinal surgery including hepatobiliary surgery27 found no benefit for immunonutrition versus an isonitrogenous, isocaloric nutrition with regard to postoperative complications. The present study, which did not permit to demonstrate an impact of immunonutrition on postoperative complications after liver resection, seems to confirm these trials.17,31–34 © 2014 Macmillan Publishers Limited
Immunonutrition in liver resection T Zacharias et al
967 Table 2.
Baseline characteristics of patients in the immunonutrition group and in the control group after propensity score matching
Female gender* Age* in years Cancer* Charlson comorbidity index Body Mass Index NRS 2002*
Immunonutrition
Control
N = 49 (%)
N = 49 (%)
11 65.9 44 5.6 26.5 3.3
(22.4%) (±13.1) (89.7%) (±2.5) (±5.1) (±1.2)
12 65.6 43 5.3 25.9 3.3
P
Standardized difference
(24.4%) (±10.8) (87.7%) (±2.3) (±4.5) (±1.1)
1 0.548 1 0.624 0.459 0.555
0.04816 0.02840 0.0468 0.12452 0.10644 0.06808
Nutritional risk index o97.5 o83.5
10 (20.4%) 3 (6.1%)
13 (26.5%) 3 (6.1%)
0.303 0.500
0.14485 0
Weight loss in % 410% ASA score ⩾ 3* Cirrhosis* Steathosis
5.0 12 11 4 20
(±6.6) (24.4%) (22.4%) (8.1%) (40.8%)
5.7 11 12 4 21
(±5.8) (22.4%) (24.4%) (8.1%) (42.8%)
0.490 0.500 1 1 1
0.11374 0.04816 0.04816 0 0.04138
Diagnosis Colorectal metastases* Hepatocellular carcinoma* Cholangiocarcinoma* Other diagnosis Preoperative chemotherapy Major hepatectomy* Extrahepatic procedure*
24 11 5 9 13 10 16
(49.0%) (22.4%) (10.2%) (18.3%) (26.5%) (20.4%) (32.6%)
25 9 5 10 12 11 16
(51.0%) (18.3%) (10.2%) (20.4%) (24.4%) (22.4%) (32.6%)
1 0.753 1 1 1 1 1
0.04082 0.10140 0 0.05163 0.04682 0.04975 0
Abbreviations: ASA, American society of anesthesiology; NRS 2002, nutritional risk screening 2002. Continuous variables are reported as mean and standard deviation. Dichotomous variables are reported as N (%). Continuous variables were compared using a Wilcoxon Signed-Ranks test. Dichotomous variables were compared using a McNemar test. Eleven variables (*) were used to construct the propensity score. Absolute standardized differences were calculated according to Austin.18
Absolute standardized differences for baseline covariates comparing treated to control subjects in the original and in the matched sample 0.5
Absolute standardized difference
0.45 0.4 0.35 0.3 original sample
0.25
matched sample
0.2 0.15 0.1 0.05
Ch
ar
lso
Fe
m al
eg en de r n Co A ge m o C N ut W rbid anc rit eig it er y io h na t l Ind l R os e isk s > x Sc 10% re A enin SA g C Sc H olo ep re o at ct Cirr re oc al h o el m s Ch lula eta is ol r c stas an ar Pr gi cin is o eo o pe O carc ma ra th tiv er inom di e a M che agn m os a Ex jo o i tra r h the s he epa rap pa te y tic cto pr my oc ed ur e
0
Baseline covariates
Figure 1. Absolute standardized differences for baseline covariates comparing treated (preoperative immunonutrition) to control subjects in the original (147 patients) and in the matched sample (98 patients) for liver resection.
© 2014 Macmillan Publishers Limited
European Journal of Clinical Nutrition (2014) 964 – 969
Immunonutrition in liver resection T Zacharias et al
968 Table 3.
Perioperative data for 49 matched pairs of immunonutrition and control patients.
Major hepatectomy Right hepatectomy Left hepatectomy Trisegmentectomy Othera
Immunonutrition
Control
N = 49 (%)
N = 49 (%)
P
10 (20.4%)
11 (22.4%)
1
3 4 2 1
(6.1%) (8.1%) (4%) (2%)
5 4 1 1
(10.2%) (8.1%) (2%) (2%)
0.726 1 1 1
Minor hepatectomy Bisegmentectomy 2,3 Bisegmentectomy (other) Segmentectomy and limited resection
39 9 17 13
(79.5%) (18.3%) (34.6%) (26.5%)
38 5 15 18
(77.5%) (10.2%) (30.6%) (36.7%)
1 0.387 0.814 0.404
Extrahepatic procedures Total Colorectal resection Otherb Median operating time in min (range)
16 9 7 210
Transfusion Yes 18 Intra-operatively 15 Postoperatively 4 Median estimated blood loss in ml (range) 300
(32.6%) 16 (18.3%) 11 (14.2%) 5 (110–360) 247
(32.6%) 1 (22.4%) 0.774 (10.2%) 0.753 (120–505) 0.013
(36.7%) (30.6%) (8.1%) (50–2000)
(22.4%) 0.189 (16.3%) 0.167 (8.1%) 1 (30–1300) 0.936
11 8 4 300
Continuous variables are reported as median and range. Dichotomous variables are reported as N (%). Continuous variables were compared using a Wilcoxon Signed-Ranks test. Dichotomous variables were compared using a McNemar test. aOne patient had bisegmentectomy+unisegmentectomy (three resected segments) and one patient had two bisegmentectomies classifying those resections as major hepatectomy. bOther extrahepatic procedures included: total gastrectomy (n = 2), left pancreatectomy (n = 2) partial resection of the right diaphragm (n = 2), right adrenalectomy (n = 2) and biliodigestive anastomosis (n = 4).
Limitations In the present study in the unmatched sample, there were significant covariate imbalances between both groups (Table 1). Therefore, propensity score matching was performed to reduce baseline covariate imbalances between groups.18 The immunonutrition and control group in the matched sample were balanced for the majority of measured baseline covariates (standardized differences below or near 0.1). For some baseline covariates standardized differences remained between 0.1 and 0.15 after matching: NRI o 97.5 (moderate malnutrition) was more frequently observed in control patients, and mean weight loss was 5.7% in control patients versus 5.0% in immunonutrition. On the other hand mean Charlson Comorbidity Index was 5.6 in immunonutrition versus 5.3 in control patients. (Table 2). These differences were not significant. However, the authors cannot rule out that more sophisticated methods for determining the nutritional status would show a significant difference between groups in the matched sample. The remaining covariates imbalance after matching were partially explained by the rather small number of patients (n = 147), which limited the possibilities of matching. Furthermore, propensity score matching will, on an average, result in measured baseline covariates being balanced between treatment groups.18 Non-measured covariates can only be balanced between treatment groups by randomization. The absence of randomization remains the major limitation of the present study. Furthermore, the rather small number of patients available for analysis after propensity score matching limited the statistical value of the present study. Type II error may be an explanation for the absence of benefit of immunonutrition observed in the present study. A further limitation of the present study was the European Journal of Clinical Nutrition (2014) 964 – 969
Table 4. Postoperative outcome data: number of patients with complications (infectious and others), severity of complications according to the Clavien–Dindo classification,24 and hospital stay for 49 matched pairs of immunonutrition and control patients. Immunonutrition
Control
N = 49 (%)
N = 49 (%)
P
Number of patients with complications
26 (53%)
25 (51%)
1
Number of patients with infectious complications Pulmonary infection Venous catheter infection Urinary tract infection Surgical site infection
19 8 4 2 6
(38.7%) (16.3%) (8.1%) (4%) (12.2%)
14 5 4 3 5
7 (14.2%) 0 1 (2%) 1 (2%) 1 (2%) 1 (2%) 1 (2%) 2 (4%)
11 2 1 4 2 1
Number of patients with other complications Ascites Bile leakage Ileus Thrombosis Arrhythmia Confusion Pleural effusions Number of patients with complications grade Clavien–Dindo 1 Clavien–Dindo 2 Clavien–Dindo 3 Clavien–Dindo 4 Clavien–Dindo 5 (90 day mortality) Major complications (Clavien–Dindo 3, 4 and 5) Median hospital stay in days (range)
3 17 2 3 1 6 10
(6.1%) (34.6%) (4%) (6.1%) (2%) (12.2%) (5–62)
(28.5%) 0.383 (10.2%) 0.548 (8.1%) 1 (6.1%) 1 (10.2%) 1
(22.4%) 0.480 (4%) 0.500 (2%) 1 (8.1%) 0.375 (4%) 1 (2%) 1 0 1 1 (2%) 1
6 14 3 1 1 5 10
(12.2%) (28.5%) (6.1%) (2%) (2%) (10.2%) (4–32)
0.5078 0.6072 1 0.6250 1 1 0.9522
Continuous variables are reported as median and range. Dichotomous variables are reported as N (%). Continuous variables were compared using a Wilcoxon Signed-Ranks test. Dichotomous variables were compared using a McNemar test.
small number of major liver resection 21.4% (21/98) in the matched sample. Therefore, the findings of the present study are limited to minor hepatectomy. A possible bias may be a shorter median operation time and a higher transfusion rate for the immunonutrition group. It has been shown, that a shorter operation time was associated with a reduced complication rate and inversely.35 On the other hand, a higher transfusion rate has been reported to increase the complication rate.36 In consequence, the effect of both variables on the complication rate in the immunonutrition group may be ‘neutralized’. The duration of the operation and the transfusion rate were not included in the propensity score model, as only variables that are measured at baseline (before the administration of treatment) should be included in the model.18 A further bias may be the compliance with immunonutrition. In the present study, patients started immunonutrition 7 days before surgery at home. The day before surgery the patients were asked if they have really taken Oral-Impact and all patients in the immunonutrition group have indeed done so. However, the number of units of immunonutrition taken by patients were not recorded. This was a shortcoming of the present study, as it is assumed that the effect of immunonutrition depends on dose. A limited compliance with preoperative immunonutrition has been reported in the outpatient setting and may influence postoperative outcome.27 Concerning the timing of the immunonutrition recent metaanalyses have underlined the importance of peri and postoperative administration of immunonutrition.37,38 On the other hand, preoperative administration was recommended in 2005 by French guidelines14 for major abdominal surgery based on studies that have demonstrated a benefit of pre and perioperative © 2014 Macmillan Publishers Limited
Immunonutrition in liver resection T Zacharias et al
immunonutrition for reducing infectious10 and overall9 complications after gastrointestinal surgery. For logistical reasons OralImpact was not available inside the hospital during the study period. Therefore, the administration of immunonutrition was limited to the preoperative outpatient setting. The absence of postoperative immunonutrition may contribute to the ‘negative’ results of the present study.38 With these limitations, the present study may be regarded as a ‘real life’ picture of immunonutrition for liver resection in the authors' institution. The authors expected to observe a reduced complication rate after introducing immunonutrition for liver resection, based on reports showing a reduction of complications in gastrointestinal surgery.9–12 The authors continue to use immunonutrition for gastrointestinal surgery,12 and selectively in patients undergoing liver resections. However, further study of immunonutrition in liver resection is needed. CONCLUSION The present study did not permit to demonstrate an impact of preoperative immunonutrition with Oral-Impact on postoperative complications after minor liver resection. CONFLICT OF INTEREST The authors declare no conflict of interest.
REFERENCES 1 Jarnagin WR, Gonen M, Fong Y, DeMatteo RP, Ben-Porat L, Little S et al. Improvement in perioperative outcome after hepatic resection: analysis of 1803 consecutive cases over the past decade. Ann Surg 2002; 236: 397–407. 2 Aloia TA, Fahy BN, Fischer CP, Jones SL, Duchini A, Galati J et al. Predicting poor outcome following hepatectomy: analysis of 2313 hepatectomies in the NSQIP database. HPB (Oxford) 2009; 11: 510–515. 3 Koperna T, Kisser M, Schulz F. Infections after liver resection in the elderly. Langenbecks Arch Chir 1997; 382: 192–196. 4 Moreno Elola-Olaso A, Davenport DL, Hundley JC, Daily MF, Gedaly R. Predictors of surgical site infection after liver resection: a multicentric analysis using national Surgical Quality Improuvement Prgram data. HPB (Oxford) 2012; 14: 136–141. 5 Neal CP, Mann CD, Garcea G, Briggs CD, Dennison AR, Berry DP. Preoperative systemic inflammation and infectious complications after resection of colorectal liver metastases. Arch Surg 2011; 146: 471–478. 6 Farid SG, Aldouri A, Morris-Stiff G, Khan AZ, Toogood GJ, Lodge JP et al. Correlation between postoperative infective complications and long-term outcome after hepatic resection for colorectal liver metastasis. Ann Surg 2010; 251: 91–100. 7 Mizock BA, Sriram K. Perioperative immunonutrition. Expert Rev Clin Immunol 2011; 7: 1–3. 8 Daly JM, Lieberman MD, Goldfine J, Shou J, Weintraub F, Rosato EF et al. Enteral nutrition with supplemental arginine, RNA, and omega-3 fatty acids in patients after operation: immunologic, metabolic, and clinical outcome. Surgery 1992; 112: 56–67. 9 Braga M, Gianotti L, Nespoli L, Radaelli G, Di Carlo V. Nutritional approach in malnourished surgical patients. A prospective randomized study. Arch Surg 2002; 137: 174–180. 10 Gianotti L, Braga M, Nespoli L, Radaelli G, Beneduce A, Di Carlo V. A randomized controlled trial on preoperative oral supplementation with a specialized diet in patients with gastrointestinal cancer. Gastroenterology 2002; 122: 1763–1770. 11 Senkal M, Zumtobel V, Bauer KH, Marpe B, Wolfram G, Frei A et al. Outcome and cost-effectiveness of perioperative enteral immunonutrition in patients undergoing elective upper gastrointestinal tract surgery: a prospective randomized study. Arch Surg 1999; 134: 1309–1316. 12 Cerantola Y, Hübner M, Grass F, Demartines N, Schäfer M. Immunonutrition in gastrointestinal Surgery. Br J Surg 2011; 98: 37–48. 13 Weimann A, Braga M, Harsanyi L, Laviano A, Ljungqvist O, Soeters P et al. ESPEN guidelines on enteral nutrition: surgery including organ transplantation. Clin Nutr 2006; 25: 224–244. 14 Mariette C, Alves A, Benoist S, Bretagnol F, Mabrut JY, Slim K. Perioperative care in digestive surgery. Guidelines for the French Society of Digestive Surgery. Ann Chir 2005; 130: 108–124.
© 2014 Macmillan Publishers Limited
969 15 Koretz RL, Avenell A, Lipman TO. Nutritional support for liver disease. Cochrane Database Syst Rev 2012; 5: CD008344. 16 Richter B, Schmandra TC, Golling M, Bechstein WO. Nutritional support after open liver resection: a systematic review. Dig Surg 2006; 23: 139–145. 17 Mikagi K, Kawahara R, Kinoshita H, Aoyagi S. Effect of preoperative immunonutrition in patients undergoing hepatectomy; a randomized controlled trial. Kurume Med J 2011; 58: 1–8. 18 Austin PC. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behav Res 2011; 46: 399–424. 19 Claudot F, Alla F, Fresson J, Calvez T, Coudane H, Bonaïti-Pellié C. Ethics and observational studie in medical research: various rules in a common framework. Int J Epidemiol 2009; 38: 1104–1108. 20 Zacharias T, Ferreira N. Carrier-bound fibrin sealant compared to oxidized cellulose application after liver resection. HPB (Oxford) 2012; 14: 839–847. 21 Société française d’anesthesie et de reanimation. Antibioprophylaxis in surgery and interventional medicine (adult patients). Actualization 2010. Ann Fr Anesth Reanim 2011; 30: 168–190. 22 Strasberg SM. Nomenclature of hepatic anatomy and resections: a review of the Brisbane 2000 system. J Hepatobiliary Pancreat Surg 2005; 12: 351–355. 23 Bone RC, Balk RA, Cerra FB, Dellinger RP, Fein AM, Knaus WA et al. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Chest 1992; 101: 1644–1655. 24 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: 205–213. 25 Kondrup J, Rasmussen HH, Hamberg O, Stanga Z, Ad Hoc ESPEN Working Group. Nutritional risk screening (NRS 2002): a new method based on an analysis of controlled clinical trials. Clin Nutr 2003; 22: 321–336. 26 The Veterans Affairs Total Parenteral Nutrition Cooperative Study Group. Perioperative total parenteral nutrition in surgical patients. N Engl J Med 1991; 325: 525–532. 27 Hübner M, Cerantola Y, Grass F, Bertrand PC, Schäfer M, Demartines N. Preoperative immunonutrition in patients at nutritional risk: results of a double-blinded randomized clinical trial. Eur J Clin Nutr 2012; 66: 850–855. 28 Schulz KF, Grimes DA. Sample size calculations in randomized trials: mandatory and mystical. Lancet 2005; 365: 1348–1353. 29 Austin PC. Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples. Statist Med 2009; 28: 3083–3107. 30 Plank LD, McCall JL, Gane EJ, Rafique M, Gillanders LK, McIlroy K et al. Pre- and postoperative immunonutrition in patients undergoing liver transplantation: a pilot study of safety and efficacy. Clin Nutr 2005; 24: 288–296. 31 Plank LD, Mathur S, Gane EJ, Gillanders LK, McIlroy K, McCall JL. Perioperative immunonutrition in liver transplantation: results of a double-blind randomized controlled trial. Clin Nutr 2010; 5 (Suppl2): 3. 32 Hendry PO, van Dam RM, Bukkems SF, McKeown DW, Parks RW, Preston T et al. Randomized clinical trial of laxatives and oral nutritional supplements within an enhanced recovery after surgery protocol following liver resection. Br J Surg 2010; 97: 1198–1206. 33 Ishikawa Y, Yoshida H, Mamada Y, Taniai N, Matsumoto S, Bando K et al. Prospective randomized controlled study of short-term perioperative oral nutrition with branched chain amino acids in patients undergoing liver surgery. Hepatogastroenterology 2010; 57: 583–590. 34 Meng WC, Leung KL, Ho RL, Leung TW, Lau WY. Prospective randomized control study on the effect of branched-chain amino acids in patients with liver resection for hepatocellular carcinoma. Aust N Z J Surg 1999; 69: 811–815. 35 Tzeng CW, Katz MH, Fleming JB, Pisters PW, Lee JE, Abdalla EK et al. Risk of venous thromboembolism outweighs post-hepatectomy bleeding complications: analysis of 5651 National Surgical Quality Improuvement Program patients. HPB (Oxford) 2012; 14: 506–513. 36 Kooby DA, Stockmann J, Ben-Porat L, Gonen M, Janargin WR, Dematteo R et al. Influence of transfusion on perioperative and long-term outcome in patients following hepatic resection for colorectal metastases. Ann Surg 2003; 237: 860–870. 37 Marimuthu K, Varadhan KK, Ljungqvist O, Lobo DN. A meta-analysis of the effect of combinations of immune modulating nutrients on outcome in patients undergoing major open gastrointestinal surgery. Ann Surg 2012; 255: 1060–1068. 38 Osland E, Hossain MB, Khan S, Memon MA. Effect of timing of pharmaconutrition (immunonutrition) administration on outcomes of elective surgery for gastrointestinal malignancies: a systematic review and meta-analysis. JPEN J Parenter Enteral Nutr 2014; 38: 53–69.
European Journal of Clinical Nutrition (2014) 964 – 969
