Early time to relapse as a survival prognosticator in nodal mature T-cell lymphomas: results from the PETAL consortium

We assessed the overall survival (OS) impact of time to relapse (TTR) in multinational cohorts with independent observational and randomized validation. Patients with nMTCL with frontline complete response were assigned to TTR12 (≤12 months) or without TTR12 based on time to progression or next therapy. OS analyses included modified landmark (m-LM), standard landmark (s-LM), and time-dependent Cox (td-Cox), adjusting for age, histology, and prognostic index for T-cell lymphoma (PIT) score. Across 452 patients, 165 (36.5%) had TTR12, 181 (40%) relapsed at ≥12 months, and 106 (23.5%) remained relapse-free. TTR12 conferred worse OS using m-LM (hazard ratio [HR], 2.14; 95% confidence interval [CI], 1.58-2.90; P< .001), s-LM (HR, 1.92; 95% CI, 1.39-2.66; P< .001); and td-Cox (HR, 5.81; 95% CI, 2.94-11.46; P< .001). Results were consistent in the independent validation cohorts. TTR12 consistently conferred worse OS irrespective of frontline stem cell transplantation or PIT score, in peripheral T-cell lymphoma, not otherwise specified (m-LM: HR, 2.32; 95% CI, 1.51-3.55; P< .001; s-LM: HR, 2.10; 95% CI, 1.33-3.31; P = .001), anaplastic large cell lymphoma (m-LM: HR, 3.34; 95% CI, 1.18-9.50; P = .023; s-LM: HR, 2.96; 95% CI, 1.02-8.81; P = .046), and angioimmunoblastic T-cell/T follicular helper cell lymphoma (m-LM only: HR, 1.92; 95% CI, 1.15-3.21; P = .013). Second-line novel therapies improved OS (second-line start to death) vs chemotherapy in TTR12 only (HR, 0.60; 95% CI, 0.37-0.97; P = .038; without TTR12: HR, 0.82; 95% CI, 0.51-1.32; P = .407). TTR12 serves as a prognostic and potential OS surrogate marker, supporting stratification of new risk groups and need for their differential treatment.