Reconsidering the Cutoff Value for Sensitive and Refractory Relapses in Extensive-Stage SCLC in the Era of Immunotherapy.

ABSTRACT

INTRODUCTION: Traditionally, relapsed SCLC has been classified as "sensitive" or "refractory" on the basis of cutoff values (60 or 90 d) for the duration between the last chemotherapy and disease progression. Nevertheless, these cutoff values are not derived from rigorous analytical methods, and their applicability to contemporary treatments remains uncertain. METHODS: We conducted a retrospective multicenter study on patients with extensive-stage SCLC who underwent second-line therapy after platinum-doublet chemotherapy with or without immune checkpoint inhibitor (ICI) resistance before (pre-ICI cohort) and after (post-ICI cohort) approval of combination immunotherapy. We selected the optimal platinum-free interval cutoff value with the lowest two-sided p value in the multivariable Cox regression model for second-line overall survival. The internal validity of the chosen cutoff value was assessed using twofold cross-validation. RESULTS: There were 235 and 98 patients in the pre-ICI and post-ICI cohorts, respectively. In the pre-ICI cohort, the optimal cutoff was 59 days (p = 0.0001); the hazard ratio calculated using twofold cross-validation was 1.31 (95% confidence interval 0.95-1.82]). In the post-ICI cohort, although the 60- and 90-day cutoff values could predict prognosis (60 d; p = 0.002, 90 d; p = 0.005), the optimal cutoff value was 75 days (p = 0.0002), which resulted in a median second-line overall survival of 15.9 and 5.0 months for patients with sensitive and refractory relapse, respectively (hazard ratio = 2.77, 95% confidence interval 1.56-4.93). CONCLUSIONS: We clarified the previously ambiguous cutoff values for classifying relapsed SCLC and revealed that the 75-day cutoff most accurately predicts subsequent prognosis than the traditional cutoffs in the post-ICI era.