Older patients with primary central nervous system lymphoma (pcnsl): real world (rw) outcomes of post‐induction therapy in the modern era

Introduction: Treatment of older patients (pts) with PCNSL is challenging due to the prevalence of comorbidities, frailty, and effective delivery of chemotherapy (CT). Optimal induction CT or consolidation for older PCNSL pts are unknown. Moreover, there are a paucity of large-scale prognostication studies available. We analyzed post-induction treatment patterns and outcomes with prognostication across 19 academic centers. Methods: We conducted a large, RW retrospective study of newly diagnosed PCNSL pts (1/2008-1/2019) ages > 60 years (yrs). Among 525 older pts in the full data set (David KA et al, ASH 2020), 363 pts achieved either complete or partial remission (CR/PR) with induction. Survival rates were estimated by Kaplan-Meier with differences assessed by log rank test. Univariate associations were derived via Cox model with variables p < 0.05 entered stepwise into multivariate (MVA) models. Results: Among 363 pts in PR or CR, mean age was 70 yrs (60-88). 50 (14%) pts underwent consolidative autologous stem cell transplant (ASCT), with mean age 66 yrs (60-77) and ECOG PS 0-1 in 72% ASCT pts. 18 pts (mean age 68, range 62-75) received consolidative radiation therapy (RT). Post-induction maintenance was given to 22% of pts (mean age 72 yrs, 60-86); the most common regimens were temozolomide (28%), methotrexate (19%), and lenalidomide (25%). Among all 363 pts, median progression-free survival (PFS) was 61 months (95% CI 41-72, Figure 1A) and median overall survival (OS) was 70 months (95% CI 56-90, Figure 1B). In the full data set of 525 pts on MVA analysis, advancing age & worse ECOG PS were associated with inferior PFS, while advancing age, hypoalbuminemia, CIRS-G score & worse ECOG PS were associated with inferior OS. On MVA analysis among the 363 CR/PR pts, factors associated with inferior PFS were advancing age (HR 1.050, p = 0.0004) and anemia (HR 1.13, p = 0.0228), with advancing age (HR 1.057, p = 0.0006), impaired creatinine clearance (1.009, p = 0.0381), and worse ECOG PS (p = 0.0049) associated with inferior OS. The 3-yr PFS among pts undergoing ASCT was 72% vs. 46% among those who did not (p = 0.002) with 3-yr OS 81% vs. 64%, respectively (p = 0.02); these survival improvements persisted when stratified for the aforementioned prognostic factors (Figure 1C/D). Among pts who received maintenance therapy, 3-yr PFS was 65% vs. 44% with no maintenance (p = 0.02), with 3-yr OS of 83% vs. 61%, respectively (p = 0.0007)), with both remaining significant on stratification (Figure 1E/F). Consolidative RT did not improve survival (data not shown). Conclusions: Collectively, older pts with PCNSL have suboptimal outcomes, and improved therapy is needed. Among those with response to induction, consolidative ASCT may improve outcomes, although this constituted a minority of pts. Maintenance treatment also appeared to improve outcomes. Future studies should focus on optimal maintenance regimens for patients not fit for ASCT. Survival PFS (A) and OS (B) are depicted for older PCNSL pts with CR or PR after induction. Autologous SCT was associated with improved PFS (C) and OS (D). Maintenance theraphy resulted in improved PFS (E) and OS (F). P values stratified on PFS for advancing age and anemia and for advancing age and anemia and for advancing age, impaired creatinine, and worse ECOG performance status on OS Keywords: Aggressive B-cell non-Hodgkin lymphoma, Extranodal non-Hodgkin lymphoma No conflicts of interest pertinent to the abstract.