Background: Conditional survival estimates provide useful prognostic information for cancer survivors. The objective of this study was to determine conditional survival estimates for melanoma patients with substages of stage III disease.
Materials and methods: A retrospective analysis of 760 patients who underwent lymphadenectomy for node-positive melanoma was conducted, and patients were stratified into substages: IIIA, IIIB, and IIIC. The 5-year conditional disease-free survival (DFS) and disease-specific survival (DSS) were calculated following lymphadenectomy using the methods of Kaplan and Meier and were reassessed for survivors on an annual basis. Multivariate Cox regression models were used to calculate adjusted conditional DFS and DSS accounting for age, gender, tumor histology, and extracapsular extension.
Results: For patients with IIIA, IIIB, and IIIC disease, 5-year conditional DSS from treatment to year 5 improved from 78% to 90%, 54% to 79%, and 39% to 78%, respectively. For 5-year conditional DFS over the same period, the estimates increased from 65% to 79%, 37% to 81%, and 26% to 92%, respectively. Male patients experienced decreased 5-year conditional DSS and DFS across all substages, with the most pronounced effect on DSS in stage IIIC. Multivariate analysis demonstrated that survival differences among stage IIIC patients based on histologic subtype and extracapsular extension decreased over time.
Conclusions: Conditional survival estimates are more optimistic and realistic for cancer survivors than traditional survival estimates over time. For node-positive melanoma survivors, 5-year conditional DFS and DSS improve significantly over time. These estimates are critical to treatment decisions and non-treatment-related planning for both clinicians and patients.