9578 Background: The recently revised (AJCC) Staging Manual, 8th edition, introduced changes including removal of mitotic index and addition of the IIID substage. There is active debate on the utility of this revision, especially, without the inclusion of a novel prognostic biomarker, during an era of major therapeutic shifts and amidst accrual of adjuvant clinical trials for high-risk resected primary melanoma. We examined whether re-staging primary melanoma patients using the new AJCC 8 system yielded improved prognostication as compared to AJCC 7. Methods: We compared the impact of changes in staging criteria in stage I-III melanoma patients who were prospectively enrolled in a NYU clinicopathological database between January 2010 and December 2016 with active protocol-driven follow up (FU). We assessed primary tumor category (T) and nodal status (N) according to both AJCC 7 and 8. Progression free survival (PFS) and overall survival (OS) curves were generated for both editions and then stratified by substage. We analyzed discordance using Cox Regression Models. Results: 1,379 patients (56% male, mean thickness 1.6, median FU 34.8 months) were included in the analyses. All but one patient remained in the same ‘major’ stage using AJCC 7 and 8 (stage I- 998; II- 224, 225; III- 157, 156) whereas 44% of stage III substage classifications were discordant comparing AJCC 7 to 8. Despite removing mitoses as a criterion for Stage I, there was no significant change between editions in PFS/OS when evaluating major and substages of stage I. Stage IIC patients had worse PFS/OS than stage IIIA patients in AJCC 8 (PFS p = 0.04, OS p = 0.02). AJCC 8, which implemented four rather than three substages, had improved PFS prognostication (c-index = 0.59 vs 0.66, p = 0.05 for AJCC 7 vs 8). Conclusions: Our results reinforce the added value of AJCC 8 compared to 7, as removing an operator dependent variable is more practical for stage I, and increased influence of thickness/ulceration and the addition of a new substage is more prognostically informative for stage III. Nevertheless, the poor prognosis of stage IIC patients, despite nodal negative disease, continues to be an unaddressed gap within our current staging framework.

No abstract