Session Item

There are several acceptable consolidation strategies for primary central nervous system lymphoma (PCNSL) after induction chemoimmunotherapy like whole-brain radiotherapy [≤24Gy reduced-dose, RD-WBRT vs >24Gy standard-dose, SD-WBRT], cytarabine or autologous stem cell transplant [ASCT]. We analyzed patterns of contemporary use and associated outcomes in a large cohort.

Patients with PCNSL (diffuse large B-cell subtype) from 1983-2020 were identified from an institutional database. Longitudinal consolidation patterns were analyzed using Fisher exact test. Consolidation strategies and outcomes were stratified by decade and MSKCC recursive partitioning analysis (RPA) class 1 (age<50), 2 (age≥50, KPS≥70) or 3 (age≥50, KPS<70). Clinicodemographic associations with consolidation strategy were analyzed by multinomial logistic regression. Progression free (PFS) and overall survival (OS) were analyzed by Kaplan Meier and proportional hazards from consolidation date.

Of 645 evaluated, 559 were eligible and 385 (69%) received consolidation. Median follow-up was 7.4 years and median OS was 4.8 years. Median age was 63 years (IQR 54-72); most (n=321, 57%) were RPA class 2. Over the study period, there was significant change in consolidation patterns with declining use of WBRT plus cytarabine (61% in 1990s vs. 12% in 2010s) and rising utilization of ASCT (4% in 1990s vs. 32% in 2010s) and cytarabine alone (27% in 1990s vs. 52% in 2010s) [Fig A, p<0.001]. This temporal evolution was seen across RPA classes, with increased cytarabine usage for greater RPA in modern years (Fig B-D). On multivariable logistic regression, compared to RPA class 1, class 2 was significantly less likely to receive ASCT (OR 0.23, p=0.001), RD-WBRT (OR 0.31 p=0.02) or SD-WBRT (OR 0.08, p<0.001) vs. cytarabine. Those with partial response to induction were also less likely to receive ASCT (OR 0.36, p=0.02). Among the 351 with complete response to consolidation, on multivariable analysis, only receipt of R-MPV induction was associated with better PFS (HR 0.5 p=0.006). There was no significant PFS association with any consolidation strategy. RPA class 3 was associated with poorer OS (HR 1.9, p=0.03). For RPA class 1 and 2, among all consolidation strategies, patients treated with ASCT and RD-WBRT had similar, best OS outcomes (Fig E-G). For RPA class 1, median OS was not reached for ASCT or RD-WBRT and 2.5 years post cytarabine. For RPA class 2, median OS was 9.4 years post ASCT, 6.7 years post RD-WBRT and cytarabine and 3.0 years post cytarabine alone.

There has been significant change in consolidation approach for PCNSL with decreased utilization of WBRT plus cytarabine and increased utilization of cytarabine alone. However, RD-WBRT is associated with excellent outcomes across RPA classes. In light of preliminary data from the ongoing RTOG 1114 trial demonstrating improved PFS and no early neurotoxicity signal, RD-WBRT could be more strongly considered, particularly for higher RPA classes.