Early and late complications of pylorus-preserving pancreatoduodenectomy in Japan 1998.

Early (within 1 month after operation) and late (more than 1 month after surgery) complications after pylorus-preserving pancreatoduodenectomy (PpPD) were analyzed in 1066 Japanese patients collected from 74 authentic institutions in Japan. As early postoperative complications after PpPD, delayed gastric emptying was evident in 46% of patients, pancreatoenterostomy leakage in 16%, intra-abdominal infection in 14%, cholangitis in 8.9%, hepaticojejunostomy leakage in 4.7%, intra-abdominal hemorrhage in 3. 5%, upper gastrointestinal hemorrhage in 3.2%, and duodenojejunostomy leakage in 2.0%. Delayed gastric emptying resolved 1-24 months after PpPD (mean, 3.1 months). The direct operative mortality (death within 1 month after the operation) was 2. 4%. Univariate and multivariate analysis of pancreatoenterostomy leakage showed that male sex (P = 0.0151) and soft consistency of the pancreas (P < 0.0001) were independent significant factors. Univariate analysis of delayed gastric emptying showed that establishment of gastrostomy (P < 0.0001), length of the preserved duodenum (P = 0.0406), gastric juice output (P = 0.0001), length of gastric tube placement (P < 0.0001), and administration of cisapride (P = 0.0059) were significant variants. As late complications, stomal ulcer was evident in 3.6% of patients, cholangitis in 6.7%, and liver abscess in 1.2%. Glucose intolerance appeared in 61 patients, resolved in 15, showed no change in 170, was absent in 695, and was ameliorated in 17. As a result, the dosage of hypoglycemic agents or insulin showed no change in 187 patients, decreased in 16, and increased in 52. Diabetes appeared 0-42 months after PpPD (mean, 102 months). When present, diabetes deteriorated 0-36 months postoperatively (mean, 6.3 months). Univariate analysis of the appearance or deterioration of diabetes showed that diabetes occurred more frequently in the following patients; those with Billroth I reconstruction compared with those with Billroth II (P = 0.0041), those with pancreatogastrostomy vs those with pancreatojejunostomy (P = 0.0229), those with pancreatogastrostomy vs those with end-to-side pancreatojejunostomy (P = 0.0165), and those with total tube drainage vs those with pancreatico-whole thickness anastomosis (P = 0.0392); a high American Society of Anesthesiologist (ASA) score (P = 0.0211) and pancreatoenterostomy leakage (P = 0.0361) were also significant factors. Postoperative body weight loss (>3 kg) was evident in 62% of patients. Body weight loss reached a maximum 4.2 +/- 5.8 months after PpPD (mean, 6.0 kg) and returned to the preoperative level 4.8 months thereafter. These results suggest that PpPD has been performed safely in Japan, the operative mortality being 2.4%. However, delayed gastric emptying was evident in 46% of the patients and pancreatoenterostomy leakage in 16%. Impairment of glucose tolerance occurred in about 10% of patients more than 1 month after PpPD. Therefore, during the early postoperative period, patients should be closely monitored for pancreatoenterostomy leakage and delayed gastric emptying and in the late postoperative period, glucose tolerance should be carefully followed-up.