Does size matter in pancreatic cancer? Reappraisal of tumour dimension as a predictor of outcome beyond the TNM

Objective: To explore the role of pancreatic ductal adenocarcinoma (PDAC) size on surgical and survival outcomes. Background: Tumor size is a prognostic factor for the majority of solid cancers, but the role for PDAC in predicting survival is not well delineated affecting the reliability of tumor node metastasis system classification. Methods: Between 1998 and 2012, 1507 patients with PDAC underwent resection at University of Verona Hospital and the Massachusetts General Hospital. All data were collected prospectively. Tumor size has been measured both at imaging and gross pathology. Results: Among the tumors measured, 21.5% were <20mm and 78.5% >20 mm. Larger tumors were associated with higher Ca19.9, T3-T4 and N1, higher grade, perineural invasion, R1 resections, more positive lymph nodes, and higher lymph node ratios (P < 0.05). Tumours <20mm showed a better prognosis (33 vs 23 months; P < 0.01), but worse surgical results with higher pancreatic fistula (21.1% vs 14.6%; P < 0.01) and mortality rates (1.5% vs 0.3%; P = 0.04). PDAC size was associated with worse prognosis (hazard ratio 1.26, P = 0.02), together with Ca19.9, grading, and N1. When measured at imaging, tumor size was underestimated (median 23 vs 30mm; P < 0.01) and did not influence prognosis Conclusions: PDAC size >20 mm, measured at gross pathology, correlates with surgical outcomes and is an independent predictor of poor prognosis. Given that imaging underestimates size by about 20%, perhaps tumors that measure >20mm at imaging should be considered for neoadjuvant treatment regardless of resectability.