A Propensity Score-Matched Analysis on the Outcomes of Brexucabtagene Autoleucel from Zuma-2 Study and Allogeneic Stem Cell Transplantation from the EBMT Database in Relapsed and Refractory Post-Btki Mantle Cell Lymphoma

Introduction

Brexucabtagene autoleucel (brexu-cel), an autologous anti-CD19 CAR T-cell therapy, is approved in the US for the treatment of relapsed/refractory (r/r) mantle cell lymphoma (MCL) and in Europe for r/r MCL after ≥2 prior lines of therapies, including Bruton's tyrosine kinase inhibitor (BTKi). The pivotal ZUMA-2 trial demonstrated very good efficacy of brexu-cel in r/r MCL and tolerable safety including cytokine release syndrome and immune effector cell-associated neurotoxicity syndrome (Wang et al. J Clin Oncol. 2022). Allogeneic stem cell transplantation (alloSCT) is potentially curative but is associated with high treatment related mortality (TRM) and clinically significant graft-versus-host disease (GvHD). The optimal choice of treatment between brexu-cel and alloHCTs is unclear. We conducted a propensity score matched (PSM) analysis of brexu-cel and alloSCT in patients (pts) with r/r MCL.

Methods

Pts from ZUMA-2 were matched with alloSCT pts from the EBMT database. Selection criteria to identify pts in the EBMT registry were: r/r MCL diagnosis and receipt of alloSCT from 2010 to 2020, age ≥18 years, prior treatment with anti-CD20, prior anthracycline- or bendamustine containing regimens, and prior BTKi. We performed a 1:1 nearest neighbor matching without replacement, balanced for 5 baseline parameters: age, sex, number of prior treatment lines, time from diagnosis to cellular immunotherapy, and prior autologous stem cell transplantation (autoSCT).

The primary endpoint was overall survival (OS) from cellular immunotherapy. The following secondary endpoints were analyzed: progression-free survival (PFS), TRM, defined as incidence of deaths which are related to the treatment and occurred in absence of disease progression, and rate of GvHD. Competing risks were defined as relapse, progression or treatment-unrelated death for TRM, and death for GvHD.

The EBMT registry includes pts who actually underwent alloSCT. To characterize the dropout rate prior to cellular therapy, we conducted a multicenter intent-to-treat (ITT) analysis and identified 77 pts with r/r MCL for whom an unrelated donor search was initiated and an alloSCT was intended.

Results

Sixty-four of 68 pts from the ZUMA-2 cohort were included in the study; 4 pts were excluded because of inability to anonymize. A total of 270 pts who had undergone alloSCT and met eligibility criteria were identified in the EBMT registry. Patient characteristics between the brexu-cel and the alloSCT cohorts were imbalanced. Brexu-cel pts were older (p<0.001), had more prior lines of treatment (median 3 vs 2, p<0.001), had less prior autoSCT (41% vs 71%, p<0.001), and had longer time between diagnosis and treatment (53 vs 40 months, p=0.01). After 1:1 PSM matching, criteria were well balanced between cohorts.

Median follow-up time was 36.5 and 46.4 months for the brexu-cel and the matched alloSCT cohort, respectively. PFS and long-term OS were similar between both cohorts (Fig. 1A). However, the risk of death within the first 12 months was significantly lower in the brexu-cel cohort vs the alloSCT cohort (logistic regression with inverse probability of censoring weighting, risk ratio=0.5, 95% confidence interval [CI] 0.3-0.9, p=0.02). This was mainly attributable to the significantly higher TRM rate in the alloSCT cohort (alloSCT vs brexu-cel: cause specific hazard ratio=7.7, 95% CI 1.9-31.6, p=0.005, Fig. 1B). The cumulative incidence of acute GvHD grade ≥3 at 3 months and chronic GvHD at 24 months was 17.9% (95% CI 9.1-29.0%) and 44.3% (95% CI 30.6-57.2%), respectively.

In our ITT analysis, the dropout rate was significantly higher in the cohort intended to receive alloHCT (23%, [18/77]) vs the cohort intended to receive brexu-cel (8% [6/74]; Fisher test, p=0.01). For alloSCT, failure to proceed to cellular therapy was related to progression in 13%, change of patient's preference in 5% and ineligibility in 4%; in the brexu-cel cohort this was due to manufacturing failures in 4% and progression in 3%.

Conclusion

Brexu-cel is an effective treatment in pts with r/r MCL post-BTKi with a superior safety profile compared with alloSCT, as indicated by lower 12-month mortality, lower TRM, and absence of chronic GVHD-related morbidity. Despite efforts to match pts between the cohorts, the inherent limitations of this study, such as incongruent data sources and case selection bias, have to be considered. Additional analyses will be presented.