Crossing the Rubicon: when pancreatic resection with curative intent ends in an R2 status. Impact of "desmoplastic pseudo-pancreatitis" and anatomical site of irresectability.

AIM: To analyze the impact of pancreatitis-mimicking, concomitant alterations on intraoperative assessment of curative resectability, the anatomical site of irresectability, and outcome after nonintentional R2 resection in pancreatic cancer. METHODS: Of 1,099 patients subjected to pancreatic resection for cancer, 40 (4%) underwent R2 resection (group A). The site where tumors turned out to be irresectable and the coincident presence of potentially misleading, fibro-desmoplastic alterations were analyzed. Outcome after resection was compared with 40 bypass patients matched for age, gender, histopathology, and use of additive chemotherapy (group B). RESULTS: R2 resection was due to misjudgment regarding resectability in 38 patients (95%) and to uncontrollable hemorrhage in 2 patients (5%). Group A patients had significantly longer operative times (P <0.0001), required more blood units (P <0.0001), and had longer hospital stay than group B patients (P = 0.049). Despite a significantly higher relaparotomy rate of 20% (n = 8) in group A versus 5% (n = 2) in group B, perioperative mortality was equal (n = 2, each). Median survival was 11.5 months in group A and 7.5 months in group B (P = 0.014). "Pancreatitis-like" lesions were assessed in 70% (n = 28/40, group A) and 25% (10/40, group B; P = 0.014). The superior mesenteric artery proximal to its jejunal branches was the most likely site of irresectability (60%), followed by its peripheral course (22.5%) and the lower aspects of the celiac trunk (17.5%). CONCLUSIONS: Concomitant "pancreatitis-like" alterations hamper the assessment of local resectability in pancreatic cancer. Although palliative resection results in elevated perioperative morbidity compared with bypass procedures, mortality is equal, while survival is prolonged.