Preoperative platelet count predicts survival after resection of periampullary adenocarcinoma

Background/Aims: Thrombocytosis or thrombocytopenia have been shown to act as negative predictors of outcome for various solid tumors. No such effect is known for periampullary cancer. The preoperative peripheral blood platel et count impacts on outcome after resection of pancreatic and other periamp ullary adenocarcinomas. Methodology: Clinicopathologic information, treatment aspects, and outcome parameters of patients undergoing pancreatectomy at City of Hope Cancer Cen ter were retrospectively collected and tabulated. The impact of the preoper ative platelet count on postoperative recovery, disease-free survival, and overall survival was analyzed. Results: Between 1988 and 1998, 65 patients underwent partial or total panc reatectomy at City of Hope Cancer Center, 49 of whom had a diagnosis of pan creatic or periampullary adenocarcinoma. There were 26 females and 23 males , with a median age of 64 years (range: 24-86). Median preoperative platele t count was 308 (x10(9)/L; range: 104 to 547). Diagnoses were pancreatic (n = 28), duodenal (n = 12), and bile duct/ampullary cancer (n = 9). Procedur es included pancreatoduodenectomy (n = 42), distal pancreatectomy (n = 4), and total pancreatectomy (n = 3). Six patients underwent a splenectomy. A l ower preoperative platelet count was correlated to a shortened prothrombin time (P = 0.02), and a positive resection margin (P = 0.01), but not operat ive blood loss or transfusion requirements. Postoperative complications and hospital stay were not affected by the platelet count. Preoperative platel ets of <300 were associated with a decreased median overall survival (13 us . 33 months, P = 0.02) and disease-free survival (11 vs. 29 months, P = 0.0 2), at a median follow-up of 14 months (18 for survivors). On multivariate analysis, the platelet count remained a significant predictor of survival i n addition to grade, perineural invasion, the primary tumor size, and the s urgeon. Conclusions: Based on these retrospective data, a lower preoperative platel et count correlates with inferior, a higher count with superior survival ou tcome after resection of periampullary cancer. The mechanism is unclear, bu t may relate to general factors (bone-marrow suppression or hypersplenism f or low platelets, systemic antitumor mediators for high platelets) or plate let-specific effects (platelet influence on tumor angiogenesis or metastati c efficiency). The preoperative thrombocyte count should be considered a pa rameter with potential clinical significance in prospective clinical studie s of periampullary neoplasms.