Predictors of Recurrence and Survival in High-Stage Primary Cutaneous Squamous Cell Carcinoma: A Recursive Partitioning Analysis.

Purpose/Objective(s) High stage cutaneous squamous cell carcinoma (CSCC) has an elevated risk of recurrence (5-25%). Current staging systems stratify patients rely on tumor factors only. We sought to develop an RPA incorporating patient-, tumor-, disease-, and treatment-related factors to identify prognostic subgroups for high-stage CSCC patients receiving definitive therapy. Materials/Methods All high-stage CSCCs, defined as Brigham & Women's Hospital (BWH) tumor (T) staging system T2B and T3, treated with curative intent were included. Recurrence-free survival (RFS) and overall survival (OS) were estimated using the Kaplan-Meier method and compared by log rank test. Recursive partitioning analysis (RPA) was performed to identify prognostic factors for RFS and OS. Results A total of 444 patients (76% male, 95% Caucasian, median age 74 years) with 523 primary tumors were included. Median follow up was 31 months (range 1-194). 31% of patients were immunosuppressed (organ transplant 13%, chronic lymphocytic leukemia (CLL) 14%, and other reasons 3%). 432 tumors were BWH T2b and 91 were BWH T3, of whom 12% were node-positive, 32% node-negative, and 56% unknown nodal status. 33% of patients experienced recurrence. Immunosuppression, recurrent tumor, perineural invasion (PNI), and lymphovascular space invasion (LVSI) were associated with significantly worse RFS (median RFS 23 months, 95% CI, 19-28), while immunosuppression, lymph node positivity, and extranodal extension (ENE) were associated with significantly worse OS (median OS 40 months, 95% CI, 33-46). RPA identified 4 prognostic subgroups with distinct RFS patterns (P 1 lymph node positive (N = 48, 2-year OS: 54%; 95% CI, 41-71%), and class IV included age ≥ 78 years and immunosuppression from CLL (N = 34, 2-year OS: 37%; 95% CI, 23-57%). Conclusion An RPA classification system was created for patients with high-stage primary CSCC treated with definitive intent that identifies 4 prognostic subgroups associated with recurrence-free survival and overall survival based on age, LVSI, PNI, burden of nodal metastasis, and immunosuppression status. This has implications for patient counselling, clinical trial design and post-treatment surveillance.