Beyond the CAR-T horizon in mantle cell lymphoma Open Access

In this issue of Blood Advances, Epstein-Peterson et al¹ report an international, multicenter retrospective cohort of patients with mantle cell lymphoma (MCL) who developed progression of disease (POD) after commercial CD19-directed chimeric antigen receptor T-cell (CAR-T) therapy, which is, to our knowledge, the largest series of such patients to date.

Despite the potent activity of CAR-T against Bruton tyrosine kinase inhibitor (BTKi)–exposed relapsed/refractory MCL, long-term follow-up of the pivotal ZUMA-2 study shows a median progression-free survival (mPFS) of 25.3 months and a continuing pattern of relapse, suggesting cure remains elusive for most patients.^2,3 Therefore, the clinical challenge of managing POD after CAR-T therapy for MCL is expected to become increasingly commonplace.

Among 387 patients who received CAR-T for relapsed/refractory MCL across 15 centers, the authors focus on 135 patients (40%) with subsequent POD. As expected, this subgroup was highly enriched for established adverse biological features before CAR-T infusion, including Ki67 ≥30% (67%), TP53 mutations (68%), blastoid/pleomorphic morphology (44%), refractoriness to last therapy (57%), and multiple prior treatments (median, 3; range, 1-9), including frequent bendamustine exposure (62%). Although baseline data for the full patient cohort (n = 387) are not available, several of these parameters are comparable in frequency to the pre–CAR-T characteristics of the cohorts in ZUMA-2 and a recent US consortium analysis. This indirect comparison suggests that many of these risk factors are similarly common among patients who require cellular therapy, rather than being clear predictors of CAR-T therapy failure. Two potential exceptions, however, may be TP53 mutation status and refractoriness to last therapy, which appear to be more frequent among this POD cohort compared to unselected patients receiving CAR-T within trials or standard of care (see table). This aligns with the findings of Wang et al, who reported inferior PFS after brexucabtagene autoleucel (brexu-cel) for patients with relapsed/refractory MCL harboring a TP53 aberration, as well as those with high MIPI, Ki67, complex karyotype, and high-risk morphology.⁴ Overall, it seems that the efficacy of CAR-T in MCL continues to be undermined by traditional risk factors, perhaps most strikingly TP53 aberration. Extensive study of resistance mechanisms to CAR-T therapy in MCL will be required to distinguish causality from correlation and acquisition from expansion, ideally informed by single-cell multiomic analyses of paired prerelapse and postrelapse samples.⁵ However, from this study at least, we learn that CD19 antigenic loss is an uncommon cause of therapeutic failure (10%).

CAR-T use in the POD cohort reflected current practice, with predominant use of brexu-cel (87%) and fludarabine-cyclophosphamide lymphodepletion (87%). The complete response rate (CRR) to CAR-T infusion was 64%, broadly consistent with unselected cohorts (CRR, 67%-82%).^2,3 The aggregate survival outcomes after POD post CAR-T therapy, put simply, are dismal. The mPFS and median overall survival (mOS) were 2.5 and 5.4 months, respectively, with a 12-month overall survival (OS) of 36%. Lack of response to CAR-T therapy, earlier POD, and the traditional risk factors of age, central nervous system disease, TP53 aberrancy, heavier prior treatment, and disease bulk were all significantly associated with inferior survival on univariate analysis. Overall, these sobering data expose an area of high unmet clinical need and frequent mortality.

There are, however, grounds for cautious optimism.

Firstly, this cohort is systematically biased toward inclusion of patients with earlier POD, with relapses occurring at a median of 6 months after CAR-T infusion and almost entirely within 1 year (87%). Given the expected mPFS of 16.4 to 25.3 months among unselected patients,^2,3 the aggregated outcomes of patients in this study likely represent a pessimistic view, representing the more aggressive spectrum of post–CAR-T disease. This caveat is important, because the interval from CAR-T infusion to POD was highly predictive of OS, with very poor outcomes for patients with POD <3 months, most of whom likely had no response (18-month OS, <10%), and relative optimism if POD occurred beyond 6 months (18-month OS, ∼50%). Although longer-term data are needed, it seems likely that patients with POD occurring at the more typical 1- to 2-year interval after CAR-T therapy may have better outcomes than the patients reported in this study.

Secondly, treatment with T-cell–engaging bispecific antibodies or novel combinations (BTKis, venetoclax, lenalidomide, or bortezomib ± anti-CD20 monoclonal antibodies) may offer some hope, each achieving a CRR of 67% in small subsets (13 and 9 patients, respectively). Less encouragingly, the mOS after bispecific antibodies was only 8.3 months. These findings are consistent with emerging data in diffuse large B-cell lymphoma, which suggest activity of bispecifics in patients with POD after CAR-T infusion,⁶ particularly in cases of later relapse.⁷ By contrast, chemoimmunotherapy and single targeted agent regimens (pirtobrutinib or venetoclax) were infrequently effective, mirroring their modest monotherapy efficacy observed in covalent BTKi–exposed MCL.^8,9 The apparent superiority of bispecifics and novel combination therapies may be confounded by clinical trial participation and/or selection bias and are based on limited numbers. For now, they represent the most promising therapies among limited available choices.

A final therapeutic consideration is allogeneic stem cell transplantation (alloSCT). Among the 8 patients who underwent alloSCT in response after post–CAR-T POD, 6 remain alive at short follow-up (9 months). Other modest data sets support the curative potential of alloSCT in the era of modern MCL therapies, including for patients with TP53-mutated disease^10-12; however, this potentially toxic path to cure will be viable for only a small minority of patients. Given the significant failure rate of CAR-T therapy among patients with TP53-aberrant disease, it remains unclear when alloSCT should best be pursued within their treatment journey, but it should not be forgotten as an option.¹³ 

In summary, relapse after CAR-T therapy for MCL will become increasingly common, and at least for cases of early POD, many patients are currently unsalvageable. Although those with later relapse may benefit from bispecifics and novel agent combinations, there remains a critical need for improved and ultimately curative therapies for patients with high-risk disease.

Conflict-of-interest disclosure: J.F.S. reports consultancy/advisory role for and honoraria from AbbVie, Acerta, Celgene, Genentech, Janssen, Roche, Sunesis, and Takeda; speakers bureau participation for AbbVie, Celgene, and Roche; research funding, travel accommodations, and expenses from AbbVie, Janssen, Celgene, and Roche; and expert testimony provided for Roche. T.E.L. reports honoraria from AbbVie; employment with Walter and Eliza Hall Institute, which receives milestone and royalty payments related to venetoclax; and is a recipient of a share in royalty payments paid to the Walter and Eliza Hall Institute.