Performance of the 7th and 8th Editions of the American Joint Committee on Cancer Staging System in Patients with Intraductal Papillary Mucinous Neoplasm - Associated PDAC: a Multi-Institutional Analysis

Background: Although several studies have validated AJCC staging systems in patients with conventional pancreatic ductal adenocarcinoma (PDAC), the applicability of these schemas to invasive Intraductal Papillary Mucinous Neoplasms (invIPMN) has not been assessed. Methods: All patients who underwent resection for invIPMN between January 1, 1996 and December 31, 2015 and had available data on the size of the invasive component (to determine T staging) and lymph node status (to determine N staging) were identified from the prospective databases of Johns Hopkins Hospital, Memorial Sloan Kettering Cancer Center and Massachusetts General Hospital. Concordance probability estimates (CPE) for the 7th and 8th editions of the AJCC staging system were calculated and recursive partitioning analysis was employed to identify optimal prognostic cutoffs for T and N. Results: 275 patients met inclusion criteria. The CPE for both the 7th and 8th editions of the AJCC schema was relatively good (0.64 in both cases) and similar for both colloid and tubular invIPMN (0.64 in both). Although both survival analysis by stage and recursive partitioning validated the utility of T1a sub-staging according to the 8th edition, we did not identify any significant prognostic difference between patients with T1b vs T1c or T2 vs T3 disease. Similarly, the distinction between N1 and N2 that was introduced by the 8th edition was not shown to improve prognostic discrimination in the present cohort. Conclusions: The 7th and 8th editions of the AJCC staging system were shown to be relatively accurate predictors of prognosis in patients with invIPMN. Although our findings support the use of these schemas in the invIPMN population, some possible areas of weakness were identified and may need to be revisited in future editions.