A population-based multistate model for diffuse large B-cell lymphoma-specific mortality in older patients

Background Despite effective therapies, outcomes for diffuse large B-cell lymphoma (DLCBL) remain heterogeneous in older individuals due to comorbid diseases and variations in disease biology. Methods Using the Surveillance, Epidemiology, and End Results (SEER)-Medicare database, the authors conducted a multistate survival analysis of 11,780 patients with DLBCL who were aged >= 65 years at the time of diagnosis (2002-2009). Cox proportional hazards models were used to specify the impact of prognostic factors on overall survival and cause-specific deaths, and the Aalen-Johansen estimator was used to project the course of DLBCL over time with or without standard therapy with rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP). Results Advanced age (hazard ratio [HR] for ages 71-75 years: 1.25; HR for ages 76-80 years: 1.46; HR for ages 81-85 years: 1.88; and HR for age >= 86 years: 2.26), DLBCL stage (HR for Ann Arbor stage II: 1.28; HR for stage III: 1.54; and HR for stage IV: 1.95), Charlson Comorbidity Index (CCI) >= 1 (HR for CCI of 1, 1.15; and HR for CCI >1, 1.37), and not being married (HR, 1.12) were associated with an increased risk of DLBCL-specific death. Being female (HR, 0.91) and of higher socioeconomic status (HR, 0.91) were associated with a lower risk of DLBCL-related mortality after therapy. For patients treated with R-CHOP (3610 patients), the risk of death due to DLBCL was 14.0% and 18.6%, respectively, at 2 and 5 years of treatment and plateaued afterward, confirming a 5-year "cure" point while receiving R-CHOP among older patients. Conclusions Conducting a survival analysis over a large data set, the current study evaluated competing risks for death within a multistate modeling framework, and identified age, sex, and CCI as risk factors for DLBCL-specific and other causes of death.