AIM: We reported a 6 to 10 times higher incidence of gallstone disease in curatively gastrectomized patients for cancer (Gastroenterology 104: A362, 1993) and a possible factor for the lithogenesis to be bile infection (Gastroenterology 106: A340, 1994). The aim of this study is to experimentally investigate whether bile infection and postgastrectomy gallstone disease is related with methods of gastrectomy for cancer, i.e., a conservative pylorus-preserving gastrectomy (PPG) and a radical Billroth II subtotal gastrectomy (B-II). METHOD: Five mongrel dogs underwent PPG with the truncal vagi and the hepatic branch intact and 6 dogs underwent B-I1 with wide bursectomy and bilateral truncal vagotomy. Other 6 dogs undergoing only laparotomy (sham operation) were served as the controls. The dogs were maintained with an ordinary laboratory canine diet for 1 year. Gallbladder (GB) bile was collected by sterile fine needle puncture during laparotomy at the time of surgery, 6 months and 12 months. The obtained bile was cultured for bacteriology and analyzed for chemical constituents. Bile acids were quantified by GLC. The presence or absence of gallstone (GS) was consecutively examined by ultrasound at 2 week intervals and by inspection of the aspirated bile. GS's were examined with infrared spectral analysis for classification. RESULTS: The PPG and control dogs incurred no bile infection nor gallstone disease. Three Of the 6 B-II dogs incurred bile infection (E. coli, Bacteroides and Enterococcus). GS's (black stones containing calcium bilirubinate) developed in 3 of the 6 B-II dogs (1 identified at 6 months and 2 at 12 months). Two of the 3 B-II dogs with GS were with bile infection but one without. Total bile acids and bile acid fractions did not significantly differ among the three animal groups. Free bile acids was detected in all of the B-II dogs and in 2 of the 5 PPG dogs but in none of the control dogs. Lithogenicity calculated was within the micellar zone for all dogs. CONCLUSION: Calcium bilirubinate stone developed only in the B-II gastrectomized dogs. Positive bacterial culture and detection of free bile acids were most common in the B-II dogs, which suggested that (1) bile infection is involved in postgastrectomy gallstone disease, (2) B-II subtotal gastrectomy is more liable to bile infection and gallstone disease and (3) PPG may preserve a better sphincter of Oddi motor function, thereby preventing retrograde biliary infection. COMPARATIVE STUDIES OF EPIDERMAL GROWTH FACTOR, EPIDERMAL GROWTH FACTOR RECEPTOR, K-RAS, AND P53 IN GALLBLADDER CARCINOMAS WITH AND WITHOUT AN ANOMALOUS JUNCTION OF PANCREATICOBILIARY DUCT. Keiji Hanada, Masaki Itoh, Kiyomu Fujii, Akira Tsuchida, Shouhei Ishimaru, Manabu Hirata, Toshiyasu Iwao, and Goro Kajiyama. First Dept. of Internal Medicine, Hiroshima University School of Medicine, Hiroshima, Japan. AIMS. Previously, we reported that mutations of Kras and p53 contributed to carcinogenesis in gallbladder mucosa with an anomalous junction of the pancreaticobiliary duct (AJPBD). In this study, we examined the expression of epidermal growth factor (EGF) and epidermal growth factor receptor (EGFR), and mutations of K-ras or p53in gallbladder carcinoma with and without AJPBD. METHODS. We used l0 gallbladder carcinomas with AJPBD (Group A), and 25 gallbladder carcinomas without AJPBD (Group B). EGF and EGFR were examined using immnohistochemistry. Genomic DNA was extracted from routinely processed tissues. Polymerase chain reaction (PCR) restriction fragment length polymorphism was performed for mutations of K-ras. Mutations of p53 were examined using PCR single strand conformation polymorphism. RESULTS. The percentage of immunoreactivity for EGF and EGFR in Group A (70%) was higher than that in Group B (40%). The incidence of mutation of Kras in Group A (30%) was higher than that in Group B (7%). As for p53, although mutations were found in only exon 7 and 8 in Group A, they were found in exon 5, 6, and 7 in Group B. CONCLUSIONS. i) The autocrine loop of EGF and EGFR may play an important role in tumor growth of gallbladder carcinomas associated with AJPBD. 2) Mutations of K-ras may strongly contribute to carcinogenesis of gallbladder carcinomas associated with AJPBD. 3) Mutation spots of p53 in gallbladder carcinomas associated with AJPBD may be different from that in gallbladder carcinomas without AJPBD.