PD is associated with a high incidence of postoperative complications even when performed at a high-volume center. While the mortality rate can be reduced to less than 4%, the incidence of postoperative complications continues to range from 35% to 50% in most series. Most patients with pancreatic tumors present with significant weight loss due to anorexia and malabsorption, and they may to undergo a period of inadequate oral intake for up to 10 days after surgery [9, 10].
In the last decade, several clinical and experimental studies have reported on the beneficial effects of perioperative enteral nutrition, especially early postoperative enteral feeding, over parenteral and delayed enteral nutrition under critical conditions [1–5]. While the precise mechanisms through which early enteral feeding exerts its positive actions on the outcome are still unclear, the preservation of the integrity of gut structure/function, balanced intestinal microflora, and the maintenance of an effective local and systemic immunocompetence, have been strongly implicated [1–5, 8]. Despite these theoretical and clinical advantages, many surgeons remain committed to a postoperative period of "bowel rest", which has long been hypothesized, but never demonstrated to reduce the risk of anastomotic leak. The role of artificial nutrition in affecting morbidity after major pancreatic resection has been markedly neglected. In 1994, Brennan et al. published the first trial on postoperative nutritional support in patients undergoing PD, and they reported that routine postoperative TPN could not be recommended . Since then, many reports have indicated the effect of postoperative enteral nutrition after PD [9, 10].
In our unit, which is a high-volume centre of pancreatic surgery, all patients received enteral feeding after surgery. However, there was no guideline for postoperative nutritional support in our unit. Before performing this study, we retrospectively examined the nutritional aspects and postoperateive complaications of 30 patients who underwent PD, including PpPD (not published). In the data obtained from these examinations, the ratio of enteral nutrition drop-outs was high (34.6%), which could be primarily attributed to diarrhea in the early postoperative period, although previous studies reported that approximately 90% of the enterally fed patients reached the full nutritional regimen within 4 days after digestive surgeries, such as esophageal resection and gastrointestinal resection [1–5]. We considered that in PD procedure encompassing the lymph node and/or including ganglion dissection around the celiac artery and the superior mesenteric artery, diarrhoea is a frequent complication. In addition, in our previous studies, there were no significant differences in the nutritional and immunological parameters and the clinical outcomes due to the volume of enteral feeding. Since the volume of enteral feeding was unstable and insufficient, almost all patients tended to undergo prolonged central venous line replacement (median, 14.0 days; range, 5–21 days); consequently, a number of patients showed catheter fever (30.8%). In a recent investigation, it was revealed that the high occurrence of infection-related complications did not result from the route of nutrition (total parenteral nutrition), but was caused by hyperglycemia . Strict blood-glucose control with insulin could lead to the prevention of infectious complications . On the basis of these findings, we considered enteral nutrition combined with parenteral nutrition as a better mode of postoperative nutritional support.
In this clinical pilot study, we primarily aimed to determine the ideal procedure of postoperative nutritional support that would ensure that the patients who underwent PD received sufficient caloric intake without dropping out. In the previous studies, gastrointestinal complications were observed in an unexpectedly high proportion of the patients who received a standard enteral preparation; these complications commonly consisted of nausea, vomiting, abdominal distension, and diarrhea [4, 8–10]. Moreover, in several recent postoperative studies on selected patients with esophageal, gastrointestinal, and pancreatic diseases, there has been considerable variation in the results describing the tolerance to enteral feedings [1, 3–5, 8–10, 12]. In these studies, enteral feeding was considered to have been successfully established in more than 70% of the patients. More than 80% of the patients received >600 kcal/day from a standard enteral diet. Almost all that patients achieved a feeding rate of >40 ml/hour. On the basis of these data, we set 600 kcal/day as the maximal dose of enteral feeding in the EN + PN group, and we selected Isocal, which contains sufficient dietary fiber and medium-chain fatty acids, both of which contributed to the reduction in the occurrence of diarrhea as the enteral diet in this study.
We divided the 17 patients into 2 groups according to the mode of postoperative nutritional support: the EN and EN + PN groups. There were no significant differences between the baseline profiles of the 2 groups. Although the routes of administration of the diet were different, the patients of both groups had a similar total caloric intake, and there was no significant difference between the nutritional analysis in the 2 groups. Consistent with these findings, there was no difference between the 2 groups in terms of weight loss on POD 21. The number of patients PD was more in the EN + PN group, while the number of patients who underwent PpPD was more in the EN group. However, we did not consider that this factor would cause any bias in the evaluation of postoperative nutrition, because the jejunal tubes were inserted from the aboral portion of the gastrojejunal anastomosis (i.e. the enteral diet did not pass through the preserved stomach).
In the subanalysis of immunological function, there were no significant differences in the parameters indicating cellular immunity, i.e. leukocytes and lymphocytes counts and the T cell subpopulation. Serum immunoglobulin plays an important role in host humoral immunity. Although the serum levels of IgA and IgM dropped remarkably in all the patients after the operation, they recovered quickly in the EN + PN group and were significantly higher than those in the EN group. However, because of the small number of patients in the present study, it is unclear whether this findings suggests an improvement in the postoperative immunological status. However, there was not inflammation or infection in the patients in the EN + PN group.
In the subanalysis of biochemical parameters, there were no significant differences in any of the parameters, excluding those of hepatic function, between the 2 groups. The postoperative increase in ALT and lactate levels and decrease in the ChE level in the EN +PN group could not be clarified. However, we considered that these changes were caused by TPN, since it has been reported that patients receiving TPN usually show mild-to-moderate elevations in transaminase and alkaline phosphatase levels and hepatic steatosis or portal triaditis on biopsy; the steatosis is reversible, if TPN is administered for a short period [13, 14].
Clinically, our most suggestive finding was that more patients of the EN group dropped out of enteral feeding, mainly due to diarrhea and abdominal distention. The patients of the EN + PN group received parenteral nutrition for a longer duraton than the EN group (7.7 ± 1.1 days vs. 12.0 ± 1.5 days, p = 0.0418); however, it was also demonstrated that there was no significant difference in the occurrence of catheter-associated infections (3/8 in the EN group vs. 2/9 in the EN + PN group, p = 0.4902) under conditions where the central venous route was removed as early as possible.
We suggest that enteral feeding combined with parenteral nutrition may be as safe as total enteral nutrition, which has been reported as the standard method, for ensuring proper completion of postoperative nutrition. Moreover, EN + PN can be a more suitable mode of postoperative nutrition for the patients who have undergone PD. We discontinued our study because many patients in the EN group had to discontinue enteral feeding. However, because of the small number of patients in this study, further studies are required for complete elucidation of these findings.