|
|
| |||||||||||||||||||||||||||||||||||||||
Background Resection of hepatocellular carcinoma is associated with high rates of morbidity and mortality. Since intensive nutritional support can reduce the catabolic response and improve protein synthesis and liver regeneration, we performed a prospective study to investigate whether perioperative nutritional support could improve outcome in patients undergoing hepatectomy for hepatocellular carcinoma.
Methods We studied 124 patients undergoing resection of hepatocellular carcinoma. Sixty-four patients (39 with cirrhosis, 18 with chronic active hepatitis, and 7 with no associated liver disease) were randomly assigned to receive perioperative intravenous nutritional support in addition to their oral diet, and 60 patients (33 with cirrhosis, 12 with chronic active hepatitis, and 15 with no associated liver disease) were randomly assigned to a control group. The perioperative nutritional therapy consisted of a solution enriched with 35 percent branched-chain amino acids, dextrose, and lipid emulsion (50 percent medium-chain triglycerides) given intravenously for 14 days perioperatively.
Results There was a reduction in the overall postoperative morbidity rate in the perioperative-nutrition group as compared with the control group (34 percent vs. 55 percent; relative risk, 0.66; 95 percent confidence interval, 0.45 to 0.96), predominantly because of fewer septic complications (17 percent vs. 37 percent; relative risk, 0.57; 95 percent confidence interval, 0.34 to 0.96). There were also a reduction in the requirement for diuretic agents to control ascites (25 percent vs. 50 percent; relative risk, 0.57; 95 percent confidence interval, 0.37 to 0.87), less weight loss after hepatectomy (median loss, 0 kg vs. 1.4 kg; P = 0.01), and less deterioration of liver function as measured by the change in the rate of clearance of indocyanine green (-2.8 percent vs. -4.8 percent at 20 minutes, P = 0.05). These benefits were seen predominantly in the patients with underlying cirrhosis who underwent major hepatectomy. There were five deaths during hospitalization in the perioperative-nutrition group, and nine in the control group (P not significant).
Conclusions Perioperative nutritional support can reduce complications after major hepatectomy for hepatocellular carcinoma associated with cirrhosis.
Methods
Randomization
From September 1990 through June 1993, 150 consecutive patients whose hepatocellular carcinoma was considered to be resectable on the basis of radiologic studies and hepatic-function tests were randomly assigned to receive either perioperative intravenous nutritional support in addition to their usual oral diet (the perioperative-nutrition group, 75 patients) or no additional therapy (the control group, 75 patients). The principal hepatic-function test used in this study was the indocyanine green clearance test. Indocyanine green is a dye that is rapidly eliminated by a normal liver after intravenous administration. It can be used as a sensitive and dynamic measurement of liver function11. Retention of more than 10 percent of the administered dose of indocyanine green 15 minutes after intravenous injection is an indication of reduced hepatic functional reserve and has been considered by some surgeons to be a contraindication to major hepatectomy12. However, in this study we included such patients as surgical candidates.
Sixty-four patients in the perioperative-nutrition group and 60 patients in the control group underwent hepatectomy. The other patients did not undergo hepatectomy because metastatic lesions were discovered in the abdominal cavity. Only the patients undergoing hepatectomy were studied and their data analyzed. Major hepatectomy was defined as the resection of two or more liver segments. Minor hepatectomy was defined as the resection of only one segment. The study protocol was approved by the ethics committee of the University of Hong Kong. Informed consent was obtained from all the patients included in the study.
Treatment
All patients in the perioperative-nutrition group had Broviac catheters implanted in the superior vena cava by surgical cutdown of the external jugular vein for parenteral nutrition. The parenteral nutrition was given 12 hours a night for seven nights before hepatectomy and was continued around the clock for seven days immediately after hepatectomy. The nutritional therapy consisted of a solution enriched with 35 percent branched-chain amino acids, at a dosage of approximately 1.5 g of amino acid per kilogram of body weight per day, and dextrose and lipid emulsion (50 percent medium-chain triglycerides) providing 30 kcal per kilogram per day. Vitamins and trace minerals were added to the parenteral-nutrition fluid daily. The total volume of parenteral-nutrition fluid was limited to 1.75 liters per day.
The patients in the control group had their usual oral diet preoperatively. In the postoperative period, they received 5 percent dextrose and normal saline with a volume and sodium content approximately equal to those of the fluid given to the patients in the perioperative-nutrition group.
All the patients received 2 g of cefotaxime at the time of the induction of general anesthesia. Postoperatively, all patients received 2 g of cefotaxime intravenously every 12 hours for three days and 25 g of albumin intravenously daily for five days. They were encouraged to resume oral feeding as soon as bowel function returned. Fluid balance and body weights were measured daily to avoid fluid overload.
The surgical treatment was standardized. Hepatectomy was performed by two surgeons using the same technique -- i.e., preliminary control of vascular inflow and outflow, followed by transection of the liver parenchyma. All resected specimens were examined histologically to determine the stage of the disease and to diagnose cirrhosis and chronic active hepatitis.
Assessment
Before randomization, three types of measurement were made. First, an anthropometric assessment included measurements of body weight, triceps skin-fold thickness, midarm circumference, and grip strength. Second, a liver biochemical assessment included measurements of serum bilirubin and liver enzymes; prothrombin time; serum concentrations of albumin, transferrin, prealbumin, and retinol-binding protein; plasma glucose; and indocyanine green clearance, which was expressed as the percentage of the dose that was retained in the blood 15 and 20 minutes after the intravenous administration of dye. Finally, an immunologic assessment included measurements of the response of lymphocytes to stimulation with phytohemagglutinin and serum immunoglobulin levels.
Serum bilirubin and liver enzymes were measured daily for eight days after hepatectomy. Serum concentrations of transferrin, prealbumin, and retinol-binding protein were measured on postoperative days 1, 4, 6, and 8. Anthropometric measurements, the indocyanine green clearance test, determinations of serum immunoglobulins, and tests of the lymphocyte response to phytohemagglutinin were repeated on postoperative day 8.
The assessment of outcome was based on mortality in the hospital, overall postoperative morbidity, changes in anthropometric measurements, and liver and immunologic function. Mortality in the hospital was defined as death during the hospital admission for surgery. Postoperative morbidity was determined by independent observers and defined as follows. Pulmonary infection was defined as the presence of pneumonic or atelectatic changes on radiographs associated with a positive sputum culture. Wound infection was defined as erythema and induration of a wound associated with purulent discharge that was positive on bacterial culture. A subphrenic abscess was defined as a collection of pus with or without necrotic material associated with a positive bacterial culture. Central-line sepsis was defined as a positive culture of the catheter tip in the presence of a febrile episode. Pleural effusion was considered a complication when the collection of fluid caused dyspnea and tapping was required for relief. Ascites was considered a complication when the accumulation was massive, leading to dyspnea or leakage through the abdominal wound, and a diuretic agent was needed for control. The diuretic agents were prescribed by medical staff members who were not involved in the study. Renal failure was defined as an elevation of the postoperative serum creatinine to more than two times the upper limit of normal.
Statistical Analysis
The overall rate of postoperative morbidity associated with hepatectomy for hepatocellular carcinoma was 47 percent in Hong Kong3. A reduction of the rate by half was considered necessary to indicate the efficacy of perioperative nutritional support. To detect a difference of this magnitude at a level of statistical significance of 0.05 and a power of 0.80, 60 patients were required in each group13.
The Mann-Whitney U test was used to compare continuous variables. The chi-square test (with Yates' correction if there were more than 10 cells) was used to compare discrete variables. All P values of 0.05 or less were considered to indicate statistical significance (by the two-tailed test). Calculations were made with SPSS computer software (SPSS, Chicago). The important end points are presented as relative risks and 95 percent confidence intervals, which were calculated with Epi Info 6 computer software (World Health Organization, Geneva).
Results
The two groups of patients were similar in terms of age, sex, the amount of weight lost, the results of the preoperative assessment (Table 1), the incidence of underlying cirrhosis, the stage17 of disease according to the tumor-node-metastasis (TNM) classification system, and the proportion of patients undergoing major hepatectomy (Table 2). There were more patients in the perioperative-nutrition group who retained more than 10 percent of the dose of indocyanine green at 15 minutes than in the control group (67 percent vs. 47 percent, P = 0.03). Among the patients undergoing major hepatectomy, the proportion who retained more than 10 percent of the dose at 15 minutes was also higher in the perioperative-nutrition group (64 percent vs. 43 percent, P = 0.048).
|
|
|
|
|
The data were further analyzed to determine whether patients with underlying chronic liver disease benefited more from perioperative nutritional support than patients with normal livers. Patients with cirrhosis who received perioperative nutritional support had better outcomes than control patients with cirrhosis in that the overall rate of postoperative morbidity and the need for diuretic therapy to control ascites were reduced and weight loss was also less severe (Table 6). Patients with no associated liver disease who received perioperative nutritional support had a lower overall rate of postoperative morbidity than control patients, but the other benefits were not obvious.
|
Two patients had preoperative complications related to nutritional therapy (one had catheter sepsis and the other a badly positioned catheter), and two had postoperative complications (hyperglycemia and diuresis in the absence of a history of diabetes mellitus). None of the patients died for reasons related to nutritional support.
Discussion
Modern surgical techniques have reduced the operative morbidity and mortality associated with the resection of hepatocellular carcinoma, but only in patients without underlying liver disease18,19 and those with cirrhosis undergoing minor hepatectomy for small tumors20. When major hepatectomy was carried out in patients with underlying cirrhosis, the operative mortality ranged from 26 to 50 percent18,19,21,22. Major hepatectomy has therefore been considered to be contraindicated23,24 in patients with cirrhosis, especially if they retained more than 10 percent of a dose of indocyanine green 15 minutes after intravenous administration12. In Hong Kong, many patients with hepatocellular carcinoma present with large tumors associated with cirrhosis, and there is a policy of performing major hepatectomy in these patients. We have shown that perioperative nutritional support offers a benefit in terms of better protein synthesis, preserved liver function, and reduced morbidity. The benefit was most evident in patients undergoing major hepatectomy, even though the proportion who retained more than 10 percent of the dose of indocyanine green at 15 minutes was higher than that in the control group.
Perioperative nutritional support has been shown to be useful in reducing the rate of postoperative complications in severely malnourished patients25. In our study, the incidence of malnutrition (defined as the loss of more than 10 percent of usual body weight26) was 15 percent, indicating that many patients with normal nutritional status were included. This is because many patients with hepatocellular carcinoma do not lose substantial amounts of weight at the stage when their tumors are resectable. Moreover, hepatectomy is a major operation that induces a severe catabolic response, even in patients with normal nutritional status. Therefore, the degree of malnutrition was not an important consideration in the criteria for inclusion. Since many of our patients were not cachectic, preoperative nutritional support was given at night to allow patients to be fully ambulatory during the day and to compensate for the lack of nutritional intake when they underwent preoperative investigation.
A solution enriched with branched-chain amino acids was chosen for this study because it is anticatabolic,27 promotes hepatic, muscle, and plasma protein synthesis in patients with chronic liver diseases,28,29 and accelerates liver regeneration in animals30. Since patients with cirrhosis often have glucose intolerance and insulin resistance,31 branched-chain amino acids may also be useful in providing energy to the liver and peripheral tissues,32 especially in the immediate postoperative period, when liver function is abnormal and oral intake is lacking.
To avert glucose intolerance, a lipid emulsion was included in the regimen. However, the usual preparation of lipid emulsion consists of long-chain triglycerides, which are considered inadvisable for patients with cirrhosis who have impaired synthesis of apoprotein C-II,33 albumin, hepatic triglyceride lipase,34 and carnitine35. Medium-chain triglycerides, which have a reduced dependence on albumin and apoprotein C-II for breakdown36 and on carnitine for intracellular metabolism,37 were therefore included. An additional advantage of medium-chain triglycerides is that they are readily oxidized by all body tissues and little is deposited in the liver37. We also demonstrated previously that the mixture of long-chain and medium-chain triglycerides could be rapidly eliminated from the blood in patients with cirrhosis38.
There have been few reports of perioperative nutritional support for patients undergoing hepatectomy. Reilly et al.39 showed that postoperative nutritional therapy could improve muscle function in patients undergoing liver transplantation, allowing early weaning from respirators. However, Kanematsu et al.40 showed that the postoperative administration of a solution enriched with branched-chain amino acids did not reduce the incidence of hepatic coma. In that study, the possible benefits of such a solution on nutritional variables were not evaluated. In the present study, we demonstrated that perioperative nutritional support could lead to less weight loss and to higher serum concentrations of proteins with a short half-life as well as to higher concentrations of glucose and urea. The patients' skin-fold thickness, midarm circumference, and grip strength were not altered by the nutritional support. Probably these measurements were insensitive to short-term changes. Like Kanematsu et al.,40 we found the incidence of fatal hepatic coma to be the same between the two groups of patients. We identified the cause of hepatic coma as an error in surgical technique that led to massive hemorrhage. In this situation, hepatic coma progressed rapidly and was not reversed by nutritional support.
The reduction in postoperative morbidity in this study did not shorten the hospital stay. Although there was a significant difference between study groups in the incidence of septic complications, there was no major difference in postoperative immunologic measures. The majority of the septic complications were due to pulmonary infection.
In conclusion, perioperative nutritional support in the form of a solution enriched with branched-chain amino acids, dextrose, and medium-chain triglycerides is beneficial in reducing postoperative morbidity. It is indicated in patients undergoing major hepatectomy, especially when the liver is cirrhotic.
Supported by the Croucher Foundation, Hong Kong.
Source Information
From the Department of Surgery, University of Hong Kong, Queen Mary Hospital, Hong Kong, where reprint requests should be addressed to Professor Fan.
References
| |||||||||||||||||||||||||||||||||||||||
Related Letters:
Physicians and Managed Care
Weintraub M. I., Helsel E. V., Marino J. T., Iglehart J. K.
Extract |
Full Text
N Engl J Med 1995;
332:1173-1174, Apr 27, 1995.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
