OUTCOMES OF PATIENTS WITH PRIMARY REFRACTORY AND EARLY RELAPSING LARGE B CELL LYMPHOMA ARE INFERIOR IN THE CD19 CAR T CELL THERAPY ERA

Abstract

Introduction: Although early relapse in relapsed/refractory (R/R) large B cell lymphoma (LBCL) is associated with worse outcomes, the definition of primary refractory disease (PRD) has varied. This study aimed to evaluate the association of time to relapse (TTR) on the trajectory of second line (2L) treatments and outcomes.

Methods: This retrospective study included pts aged ≥ 18 with R/R LBCL at Princess Margaret Cancer Centre (2017–2024, follow-up 01/25). TTR was categorized as PRD (progression during or at end of 1L treatment or < 3 mos), early relapse within 3–12, 12–24, or ≥ 24 mos of 1L rituximab-based therapy. Univariate and multivariable Cox regression analyses assessed the association between TTR and overall survival (OS) and progression-free survival (PFS) from first (OS1, PFS1) and second progression (OS2). Covariates included age, diagnosis, rIPI, and LDH.

Results: Of 367 pts (median age 61, 63% male) included, pts with PRD were more likely to have HGBL and GCB subtype. There were no significant differences in characteristics between relapse groups at first progression (Table 1). For 2L treatment, 86% received platinum-based salvage chemotherapy (SC), and 48% underwent ASCT. Response rates were significantly higher in TTR ≥ 24 mos (81%) compared to PRD (50%), 3–12 mos (60%), and 12–24 mos (68%) (p < 0.001). At median follow-up of 41 mos, 245 (67%) of pts progressed, and 154 (42%) died. The 5-y OS1 was 45%; median OS1 for PRD and early relapse (3–12 mos) was 23 and 24 months, respectively, compared to not reached for relapse ≥ 24 mos (12–24 mos 56 mos). Cox analysis showed PRD and early relapse ≤ 12 mos were associated with inferior OS1 and PFS1 (OS1: PRD aHR 3.8: 95% CI: 2.1–7.2, p < 0.001; 3–12 mos aHR 3.3: 1.7–6.5, p < 0.001), while 12–24 mos was not (aHR 1.7: 0.7–4.1, p = 0.25) (Figure 1). Sensitivity analysis in de novo LBCL pts did not change these findings. Of the 245 patients who progressed after 2L, 120 (49%) had PRD after 1L. PRD and early relapse were associated with refractoriness to 2L treatment (92% and 67%, vs. 50%; p < 0.001). The 5-y OS2 was 34% (median 14 mos). Cox analysis showed PRD and early relapse were associated with worse OS2 compared to ≥ 24 mos (PRD aHR 2.3: 1.3–4.3, p = 0.008; 3–12 mos aHR 2.0: 1.0–3.8, p = 0.043; 12–24 mos aHR 1.4: 0.6–3.5, p = 0.43) (Figure 2). For 3L treatment, 215 pts were considered for CAR-T (85% apheresed), and 154 (72%) were CAR-T infused. Multivariable analysis showed the combined cohort of pts with PRD/early relapse < 12 mos was associated with worse OS2 compared to relapse > 12 mos (aHR 2.2: 1.1–4.4, p = 0.022; mOS2 for early relapse 31 mos vs. NR for late relapse).

Conclusions: Our study shows that early relapse, particularly pts with PRD, continues to be associated with inferior outcomes even with 3L CAR-T. The association of relapsing timing on 2L CAR-T outcomes in an unselected population requires further investigation. Improved 1L treatments are needed to address outcome disparities in this high risk pt group.

Keywords: Aggressive B-cell non-Hodgkin lymphoma

Potential sources of conflict of interest:

A. Prica

Honoraria: Abbvie, Roche and Astra-Zeneca.

J. Kuruvilla

Consultant or advisory role: Abbvie, BMS, Gilead, Merck, Roche, Seattle Genetics, OmniaBio

Honoraria: Abbvie, Amgen, Astra Zeneca, BMS, Gilead, Incyte, Janssen, Karyopharm, Merck, Novartis, Pfizer, Roche

Other remuneration: DSMB: Karyopharm