Validity assessment of the cutaneous T-cell lymphoma severity index to predict prognosis in advanced mycosis fungoides/Sezary syndrome

Background: There is a need for standardized quantitative disease assessment measures in mycosis fungoides/Sezary syndrome. In 2005, a cutaneous T-cell lymphoma (CTCL)-severity index (SI) that not only measures disease extent (on a scale of 0-75) independent of the classic TNM(B) staging system but can also be used to estimate individual 5-year survival (SR5) was reported. Objective: We sought to assess the generalizability of the CTCL-SI/SR5 equation (SR5 equation) to predict prognosis in our cohort of patients with advanced mycosis fungoides/Sezary syndrome (n = 50, photopheresis service, 1984-2001).. Methods: TNM(B) staging, CTCL-SI score (based on skin involvement, presence of tumors, lymph node/visceral/blood involvement), and SR5 (SR5 equation = 124 - 2 x [CTCL-SI]%) at initial diagnosis were calculated retrospectively and compared with overall survival by the Kaplan-Meier method. The prognostic significance of TNM(B) staging versus the CTCL-SI was determined by Cox proportional hazards models and Brier scores. Results: Patients had stage IIA to IVA disease with a median actuarial overall survival of 58 months. By disease stage, the overall 5-year survival was 70% (stage IIA), 48% (stage IIB-IIIB), and 36% (stage IVA). In our cohort, the CTCL-SI itself was predictive of overall survival (P = .028) but the SR5 equation was not predictive of survival (Brier score of 0.29). Limitations: Small sample size, single academic center population, and retrospective design are limitations. Conclusions: The CTCL-SI is a relatively simple-to-use quantitative tool that measures disease activity in all compartments (skin, nodes, blood, viscera) and has prognostic significance in multivariate analysis. The CTCL-SI may be a useful adjunct to the TNM(B) staging for tracking disease activity quantitatively in all disease compartments (skin, nodes, blood, viscera) in clinical practice and trials, but the predictive ability of the SR5 equation needs further validation at other centers in larger groups of patients. (J Am Acad Dermatol 2011;64:682-9.)