Age at diagnosis shapes stage and survival in diffuse large B-cell lymphoma: A 22-year, population-based study of 94,598 SEER patients

Abstract

Abstract Background: Diffuse large B-cell lymphoma (DLBCL), the most common non-Hodgkin lymphoma, shows a sharp rise in incidence after age 60. While rituximab-based chemoimmunotherapy and CD19-targeted therapies have improved outcomes in fit patients, older adults remain underrepresented in trials. We analyzed 94,598 patients in SEER (2000–2021) to evaluate how age influences stage at presentation and survival, and whether modern advances have narrowed or widened age-associated disparities. Methods: Patients with first-primary DLBCL (ICD-O-3 code 9680/3) were identified in SEER-17. Stage was derived using Ann Arbor (through 2015) and Combined Summary Stage (2016–2021), mapping localized, regional, and distant disease to stages I, II/III, and IV, respectively. Age at diagnosis was grouped as ≤19, 20–39, 40–59, 60–74, and ≥75 years. Covariates included sex, race/ethnicity (non-Hispanic White [NHW], non-Hispanic Black [NHB], Hispanic, Other), and calendar era (2000–2010 vs 2011–2021). Overall survival (OS) was assessed using Kaplan–Meier and Cox models; lymphoma-specific survival used Fine–Gray competing-risk regression. Analyses included era-stratified models, age-by-stage interactions, and sensitivity analysis excluding cases with missing stage. Significance was set at P < 0.05. Results: Among 94,598 patients (median age 65; 44 % female), stage IV disease increased with age: 18 % (≤19 y), 26 % (≥75 y). Five-year OS declined across age groups: 87.6 %, 79.4 %, 67.2 %, 53.4 %, and 28.4 % (P < 0.001). In multivariable Cox analysis, age was the strongest prognostic factor (global P ≈ 6 × 10⁻²²): adjusted hazard ratios (HRs) for OS vs ≤19 y were 1.8 (20–39 y), 3.6 (40–59 y), 6.8 (60–74 y), and 14.5 (≥75 y) (all P < 10⁻⁹). Stage IV was independently adverse (HR 1.72), female sex was protective (HR 0.85), and NHB ethnicity conferred excess mortality (HR 1.28). The modern era improved OS (HR 0.81), yet the HR for age ≥75 y rose to 16.2 in 2011–2021, indicating widening disparities. Fine–Gray analysis of 13,675 lymphoma deaths confirmed similar age gradients (sub-HR 14.7 for ≥75 y). Age-by-stage interaction was modest (attenuation in stage III; P_interaction = 0.01). Excluding cases with missing stage (8 %) did not affect estimates. Conclusions: Age at diagnosis remains the dominant determinant of survival in DLBCL, independent of stage, sex, race, or era. While modern therapies have improved outcomes, patients aged ≥75 y still experience 14-fold excess mortality relative to pediatric/AYA patients. These disparities have widened in the CAR-T and bispecific antibody era. Our findings highlight the urgent need for geriatric-focused trials, equitable access to novel therapies, and biologic research to address immunosenescence and age-related tumor heterogeneity.