Patterns of Use and Outcomes of Novel Agents in Patients with Relapsed or Refractory Large B-Cell Lymphoma: A Single-Center Retrospective Analysis

Introduction

In addition to chimeric antigen receptor T-cell therapy (CAR-T), 6 novel agents are approved in the United States for relapsed/refractory (r/r) large B-cell lymphoma (LBCL): polatuzumab (pola) in combination with bendamustine (B) and rituximab (R), tafasitamab plus lenalidomide (tafa/len), loncastuximab (lonca), selinexor, and two CD3xCD20 bispecific antibodies (BsAb) (epcoritamab and glofitamab). These agents are used before CAR-T (as definitive or bridging/holding therapy), after progression (PD) on CAR-T, and in patients (pts) ineligible for CAR-T. However, their patterns of use and outcomes in these different settings are not well defined. Here we report our single-center retrospective experience with these agents in r/r LBCL.

Methods

We identified consecutive pts treated with ≥1 dose of pola, lonca, tafa/len, and/or CD20 BsAb used as definitive or bridging/holding therapy for r/r LBCL between January 2019 and March 2024. A treatment (tx) was considered definitive when intended to be continued until standard course completion, PD, or intolerance. Event-free survival (EFS) was calculated from tx initiation to either the start of new tx, PD, or death; pts without events were censored at the last follow-up.

Results

We identified 132 pts. Baseline characteristics were: median age of 66 years (range 24-86), 60% male, 79 pts (60%) had diffuse large B-cell lymphoma, 21 (16%) high-grade B-cell lymphoma, 24 (18%) transformed indolent, and 8 (6%) others. Cell-of-origin by Hans Criteria was germinal-center (GC) in 73 (62%) pts and non-GC in 45 (38%) (missing n=14). Most pts (78%) had primary refractory or early (≤12 months) relapse disease post first-line tx. The median number of lines of tx (LOT) was 5 (range 2-17), including CAR-T in 80 (61%) pts and autologous stem cell transplant in 19 (15%) pts.

Pola (alone or with R±B) was used in 91 pts (69%), lonca in 45 (34%), tafa/len in 39 (30%), and BsAb in 35 (27%, epcoritamab n=24, glofitamab n=7, mosunetuzumab n=3, other n=1). BsAb (100%), tafa/len (95%), and lonca (89%) were mainly used as definitive tx whereas pola was used as bridging/holding tx in 36 pts (40%). The best overall (ORR) and complete (CR) response rates for each were: pola 46% and 22%, lonca 44% and 22%, tafa/len 30% and 14%, and BsAb 55% and 27%, respectively. The median and 1-year EFS of the novel agent when used as definitive tx were: pola (n=55) 4.2 months and 24%; lonca (n=40) 2.7 months and 6%; tafa/len (n=37) 2.5 months, and 14%; and BsAb (n=35) 4.0 months and 28%, respectively.

Based on the timing of novel agent(s) administration in relation to CAR-T, we divided pts into 3 overlapping cohorts: novel agent before CAR-T (pre-CAR-T, n=30), novel agent after CAR-T (post-CAR-T, n=62), and pts who did not receive CAR-T (no-CAR-T, n=52). Only 2 pts (7%) received a novel agent as definitive tx before CAR-T. Reasons for not receiving CAR-T (≥1 possible) were: poor performance status/comorbidities (n=35, 67%), patient/physician preference (n=14, 27%), rapid PD (n=8, 15%), psychosocial barriers (n=5, 10%), lack of insurance coverage (n=1, 2%), and other (n=1).

Pts in the no-CAR-T cohort were older (median 73 years vs 62 years post-CAR-T vs 68 years pre-CAR-T), less likely to have primary refractory or early relapse (69% vs 82% post-CART vs 90% pre-CAR-T), and received fewer LOT (median 3 vs 7 post-CAR-T vs 5 pre-CAR-T). In the no-CAR-T cohort, the first novel agent used was pola (n=27, 52%), tafa/len (n=15, 29%), BsAb (n=5, 10%), and lonca (n=5, 10%). For the first novel agent used, the best ORR and CR rate were 44% and 22% (missing n=2), and the median and 1-year EFS were 3.4 months and 20%, respectively. In the post-CAR-T cohort, the median LOT after CAR-T was 1 (range 1-3). The most used novel agent immediately after CAR-T was pola (n=25, 40%), then lonca (n=14, 23%), BsAb (n=13, 21%), and tafa/len (n=10, 16%). For the first novel agent after CAR-T, the best ORR and CR rate were 54% and 31% (missing n=1), and the median and 1-year EFS were 3.1 months and 17%, respectively.

Conclusions

Our study shows poor outcomes with currently approved novel agents, including CD20 BsAb, in post-CAR-T and CAR-T-ineligible pts with modest CR rates and short EFS, and highlights the need for more effective tx options in r/r LBCL. Pola was commonly used as holding/bridging tx before CAR-T, whereas the use of lonca, tafa/len, and CD20 BsAb was mainly limited to post-CAR-T and CAR-T-ineligible patients.