A population-based validation of the AJCC melanoma staging system

Abstract

7500 Background: A major revision of the AJCC stages for melanoma was implemented in 2002 after validation in populations from cancer centers and cooperative groups. Methods: In this study, the National Cancer Institute's Surveillance, Epidemiology and End Results (SEER) cancer registry was used to validate the new AJCC staging system. 36,190 patients with primary invasive cutaneous melanoma diagnosed between 1988 and 2000 were identified. Survival rates computed from stage-specific Kaplan Meier curves (time to melanoma-specific death) were compared to the rates from 14,670 patients in the original validation study (Balch, et. al. JCO 19:3635–48, 2001). Cox proportional hazard models were used to identify significant prognostic factors. Results: 68% of SEER melanomas were ≤ 1.00 mm and 5% were > 4.00 mm compared respectively to 39% and 10% in the original study (p<0.001). As in the original study, Stage I, II and III melanomas with ulceration had lower survival rates, equivalent to the next higher stage. The 10-year survival rates and standard errors (SE) are presented below for stages with ≥ 900 SEER patients and from the original study. SEER rates were significantly higher than those in the original study in 15 of the 16-stage categories and notably so in T1a patients. For lesions ≤1.00, 1.01–2.00 and 2.01–4.00 mm, level, ulceration, age, sex, and site were statistically significant prognostic factors. SEER associations between survival and the prognostic factors were stronger than those in the original study, except for ulceration in lesions 2.01–4.00 mm. Conclusions: While this population-based study validates the new AJCC stages, it emphasizes that risk estimates are population specific. No significant financial relationships to disclose.