Objective: To determine the frequency distribution of K-ras-2 point mutation genotypes in pancreatic adenocarcinoma and to evaluate the effectiveness of K-ras-2 genotyping as a means to predict localized disease and potential long-term survival.
Design: Topographic genotyping from archival formalin-fixed, paraffin-embedded large- and biopsy-sized tissue specimens as well as cytologic fluid using polymerase chain reaction products and direct sequencing together with clinicopathologic and statistical analysis.
Setting: Tertiary care medical center with molecular diagnostics pathology laboratory.
Patients: Patients treated between 1988 and 1993 at Rhode Island Hospital, Providence, yielding 55 primary and 56 metastatic specimens of pancreatic adenocarcinoma.
Results: Each primary pancreatic adenocarcinoma was found to contain one of eight specific genotypes that was maintained in all metastatic deposits of that individual tumor. Primary adenocarcinomas confined to the pancreatic bed at diagnosis were predominantly of a normal genotype (56% [14/25]). Pancreatic adenocarcinomas progressing to distant hematogenous metastasis were almost exclusively mutated (88% [7/8]; P < .005). Patients undergoing pancreatic resection (Whipple's operation) and having a normal K-ras-2-genotype (58% [11/19]) had a significantly longer survival (21.3 months) than similar patients with mutated tumors (8.2 months).
Conclusions: The findings support the feasibility of K-ras-2 topographic genotyping to identify potentially indolent disease and suggest a potentially useful role in the preoperative evaluation of pancreatic adenocarcinoma.