Chimeric antigen receptor T cell (CAR-T) therapies have revolutionized the treatment of B-cell malignancies. CAR-T therapies are now standard-of-care for relapsed B-cell acute lymphoblastic leukemia (B-ALL), several non-Hodgkin’s B-cell lymphomas, and multiple myeloma. Ongoing trials are expanding CAR-T therapies to earlier-line settings and other B-cell malignancies. As with all novel therapies and treatment modalities, it is important to consider how outcomes vary for different demographic groups. These associations offer windows into both underlying disease biology and disparities and inequities in our healthcare systems, which are potentially exacerbated with therapies of high cost and complexity. In this review, we highlight clinical data on associations between age, race, and ethnicity on outcomes after CAR-T therapy for B-cell malignancies; we also discuss disparities in the representation of CAR-T clinical trials and access post-approval, which in some cases limit our understanding of how demographic factors interact with outcomes. We refer collectively to minority health populations (MHP) defined by the US Office of Management and Budget and US National Institutes of Health1 as individuals who self-identify as American Indian or Alaska Native, Asian, black/African American, and Native Hawaiian or other Pacific Islander or Hispanic/Latino ethnicities.

B-cell acute lymphoblastic leukemia

Tisagenlecleucel was the first Food and Drug Administration (FDA)-approved CAR-T therapy and has changed the treatment landscape of relapsed and refractory B-ALL. With complete remission (CR) rates greater than 80% and event-free survival (EFS) greater than 50%, tisagenlecleucel has offered the opportunity to substantially improve outcomes for this otherwise difficult-to-treat disease.2 While initially only available in the setting of a clinical trial, tisagenlecleucel is now available commercially (in the real-world setting). In addition, there are previous and ongoing trials of additional CAR-T constructs against B-ALL such as CD22-directed CAR-T therapy.3

Association of outcomes with race and ethnicity

In the modern era prior to immunotherapy, survival for MHP of children, adolescents, and young adults (AYA) with B-ALL was substantially lower than non-Hispanic white (NHW) patients.4 CAR-T therapy for B-ALL can overcome chemotherapy resistance and is relatively agnostic to cytogenetics.5 Given the known contribution of the Philadelphia chromosome-like genetic alterations to chemotherapy resistance and its increased prevalence among Hispanic6 and AYA patients, CAR-T therapy has the potential to mitigate or eliminate previously reported outcome disparities observed in these patient populations. Two groups have explicitly evaluated the association of Hispanic ethnicity with outcomes following receipt of CAR-T therapy, and both found similar responses among Hispanic patients (CR 63% vs 68%, p=0.33; median overall survival (OS) 14.6 vs 12.8 months, p=0.47, compared with the overall study population) (CR 81% vs 89%, p=0.19; 1-year OS 72% vs 71%, p=0.62) compared with NHW patients.7 8 Interestingly, a significantly lower proportion of Hispanic patients received allogeneic hematopoietic stem cell transplant (allo-SCT) prior to CAR-T therapy compared with non-Hispanic patients (35% vs 64%, p=0.004),7 and Hispanic patients had shorter time from diagnosis to infusion (mean: 34.4 vs 46.4 months, p=0.04).8 Notably, one of these studies found that Hispanic patients were more likely to experience severe cytokine release syndrome compared with NHW patients even after adjusting for leukemia disease burden and age (OR 4.5; p=0.001); this effect was particularly pronounced among Hispanic patients who were overweight or obese.7

The ability to evaluate racial disparities in response to CAR-T therapy for B-ALL is limited by small numbers of black, Asian, and Native American/American Indian patients included in CAR-T clinical trials9 and real-world studies (table 1). The limited available data suggests that these minority populations may experience worse outcomes relative to NHW patients. For example, among patients treated with commercial tisagenlecleucel, black patients had a lower CR rate and lower OS compared with other patients (CR 57% vs 86%, p=0.07; 1-year OS 43% vs 73%, p=0.03).10 In contrast to Hispanic patients who were noted to be referred early for CAR-T therapy, black patients had received more lines of therapy, had had more pre-CART relapses, and had higher rates of pre-CAR-T allo-SCT compared with other patients. These data suggest that black patients may have later access in their disease course to commercial CAR-T therapy. This study also found that black patients were disproportionately represented among non-infused patients (36% vs 6%, p=0.005) compared with patients of other races.10 While underpowered to detect statistical differences, Faruqi et al report that 60% of black patients and 71% of Asian patients had died by the end of the study period.7

Table 1

Race and ethnicity of patients included in B-ALL CAR-T trial and real-world data sets

Access to CAR-T clinical trials

Despite these findings of equal efficacy, Hispanic patients are under-represented in referrals from external hospitals for CAR-T cell trials.11 In addition, Hall et al found that patients whose preferred language was Spanish (8% vs 22%, p=0.006), as well as those who were publicly insured (38% vs 65%, p=0.001), were under-represented among those referred from an external hospital compared with those that were treated primarily at a clinical trials consortium site and enrolled in a CAR-T trial.11

Association of outcomes with socioeconomic status

Investigations into the association of social determinants of health (SDOH) with response and survival following CAR-T therapy have been limited by the lack of collection of SDOH data in CAR-T trials. Using insurance type as a proxy for socioeconomic status, a recent study showed that CR rate and OS for children with B-ALL treated on CD19-directed CAR T clinical trials or with commercial tisagenlecleucel is comparable for those that are publicly and privately insured (CR 92% vs 94%, p=0.33; 3-year OS 74% vs 68%, p=0.48).12 Using a contextual neighborhood measure of socioeconomic status (called the Child Opportunity Index), Newman et al demonstrated that response and OS were similar between those living in low-opportunity neighborhoods compared with those living in high-opportunity neighborhoods (CR 90% vs 95%, p=0.50; 3-year OS 73% vs 71%, p=0.72).12 Importantly, in multivariable analysis, patients who received CAR-T from low-opportunity neighborhoods experienced a twofold hazard of relapse (HR 2.3, p=0.006), though these relapses were more likely to retain CD19 as an antigen target when compared with relapses among high neighborhood opportunity patients who received CAR-T.12

We have focused our discussion here on children and young adults based on the available literature and approval of tisa-cel in this population. 4 years after the approval of tisa-cel, however, brexucabtagene autoleucel (bruxu-cel) was also approved by the FDA for B-ALL in the first relapse for patients greater than 18 years of age based on results of the ZUMA-3 trial. As noted in table 1, the representation of MHPs in ZUMA-3 was relatively low.13

In summary, while tisagenlecleucel has changed the landscape of B-ALL treatment with impressive remission rates, its impact on disparities remains an open question. CAR-T therapy is given as a single infusion, unlike front-line pediatric regimen-inspired therapy for ALL that requires adherence to home oral chemotherapy.14 CAR-T therapy for B-ALL thus has the potential to ameliorate a known source of outcome disparities. Research reveals comparable responses between Hispanic and NHW patients receiving CAR-T therapy, yet disparities in access and referral to CAR-T trials are an ongoing concern. Under-representation of MHPs in clinical trials and real-world data sets, coupled with their poorer outcomes, highlight the complex interplay of timing, access to care, and survival outcomes. As we move toward wider adoption and utilization of CAR-T therapy in B-ALL, a concerted effort is needed to ensure equitable access and utilization for diverse and vulnerable populations across the age spectrum, with the ultimate goal of improving survival outcomes for all patients with B-ALL.15

B-cell non-Hodgkin’s lymphomaAssociation of outcomes with race and ethnicity

Four anti-CD19 CAR-T therapies are now FDA-approved for non-Hodgkin’s B-cell lymphoma, and many others are under evaluation in clinical trials. As with B-ALL, these effective therapies have highlighted access barriers.16–24

Access to CAR-T clinical trials for MHP with lymphoma is reduced as observed in the phase 3 clinical trials of CAR-T products versus standard-of-care autologous stem cell transplant (auto-SCT) in patients with large B-cell lymphoma (table 2). In the Zuma-7 phase 3 clinical trial evaluating axicabtagene ciloleucel (axi-cel) versus auto-SCT, 5% of patients were black or African Americans and, considering ethnicity, only 5% of patients were Hispanic/Latino.21 In the BELINDA phase 3 clinical trial (tisagenlecleucel vs auto-SCT) only 3% of included patients were black or African Americans and 8% were Hispanic/Latino.23 In the TRANSFORM phase 3 clinical trial evaluating lisocabtagene maraleucel in second line versus auto-SCT, only 3% of the treated patients declared to consider themselves black or African American.16 Moreover, under-representation of MHP was also observed in CAR-T trials for indolent lymphomas. The ZUMA-5 phase 2 clinical trial, which evaluated axi-cel in indolent lymphomas, included only 3% of black or African American patients.18 Brexu-cel is available for the treatment of relapsed/refractory mantle cell lymphoma.22 25 However, the ZUMA-2 clinical trial did not disclose the race and ethnicity representation of enrolled patients. The limited representation of MHP patients in these clinical trials has precluded analysis of differential efficacy in these subpopulations.

Table 2

Race and ethnicity of patients in pivotal B-cell non-Hodgkin’s lymphoma chimeric antigen receptor-T trials

These data highlight limited access to clinical trials among MHP. In a recent article, Shahzad and colleagues evaluated the geographical and racial disparities in accessing CAR-T and bispecific antibody trials for diffuse large B-cell lymphoma by evaluating the distribution of clinical trial sites around the USA and the prevalence of MHP living in those counties. They observed that 91.9% of unique study sites were localized in counties characterized by white predominant population and only 8.1% of the sites were located in predominantly black or African American counties (p=0.009).26 The disproportionate geographic distribution of clinical trial sites in predominantly white counties can impact the population of patients able to travel to and access these trials and likely has downstream effects on the distribution of expertize to deliver these therapies after they are approved. These data strongly suggest that more inclusive site selection for clinical trial accessibility should be promoted.

We recently investigated access to CAR-T therapy for MHP at two geographically different institutions, characterized by different prevalence of MHP living in the area (Philadelphia, Pennsylvania and Portland, Oregon). We observed that in the Philadelphia area, despite preserved access to a tertiary specialized center for the treatment of large B-cell lymphoma, the proportion of MHP patients receiving CAR-T therapy for large B-cell lymphoma is reduced compared with the proportion of MHP receiving any treatment at the same institution (6.7% vs 15.7%). This trend was not observed for the Portland area, probably due to the different prevalence of MHPs (lower than the US average), which could preclude an appreciation of small differences. However, when we compared the outcomes of MHP patients to non-MHP patients, we observed similar clinical outcomes (CR rate 66.7% vs 42.0%, median progression-free survival (PFS) 6.3 vs 3.2 months, median OS not reached vs 26 months) and toxicities (any grade cytokine release syndrome (CRS) 66.7% vs 63.3%, immune effector cell neurotoxicity syndrome (ICANS) 13.3% vs 22.1%) in the two groups, respectively.27 These data showing similar efficacy of CAR-T therapy for lymphoma are encouraging and support the need for policies to ensure equitable access to these therapies among MHP.

Association of outcomes with age

Associations of age with outcome after CAR-T therapies in B-cell lymphoma have been systematically examined both in the pivotal trials of CAR-T therapies and in post-approval real-world data sets. In a post hoc analysis of the phase 2 ZUMA-1 trial, similar response rates and OS were observed in patients ≥65 years of age compared with younger patients. Adverse event profiles were also similar in the two groups, except for higher rates of grade ≥3 delirium (11% vs 0%) and encephalopathy (30% vs 21%) in the older patients.28

Several real-world reports have analyzed post-approval experiences with CAR-T in older patients.29–35 In particular, an analysis of the Cellular Immunotherapy Data Resource, managed by the Center for International Blood and Marrow Transplant Research, observed similar outcomes in commercial patients treated with tisa-cel and axi-cel in patients ≥65 years of age compared with younger patients. Also, no significant differences in the risk of developing treatment-related toxicities were observed in the two groups, including all grade CRS (OR 1.41, 95% CI 1.02 to 1.94), grade ≥3 CRS (OR 1.52, 95% CI 1.00 to 2.31), ICANS (OR 1.77, 95% CI 1.39 to 2.26), and grade ≥3 ICANS (OR 1.38, 95% CI 1.04 to 1.82).30

A more recent study used a Medicare claims database and focused on patients ≥65 and observed that age ≥75, use of bridging therapy, and Charlson Comorbidity Index ≥5 were independently associated with inferior EFS. Nevertheless, a significant proportion of patients aged ≥75 experienced clinically meaningful benefit, however, as indicated by 24-month EFS of 25%. Importantly, the use of Medicare data identified that <20% of patients with diffuse large B-cell lymphoma, and only 13% of those ≥75, received CAR-T therapy among those who received any third-line therapy. It is uncertain what mix of access barriers and clinical factors restricted use to such a small fraction of patients with late-line lymphoma, though access to CAR-T therapy has likely improved since the 2020 data cut-off for this study. Nonetheless, this low utilization raises the possibility that reported outcomes in trials and real-world studies may not be generalizable to the full breadth of elderly patients with lymphoma.

Two additional noteworthy studies analyzed patients with lymphoma aged >70 treated with CAR-T therapy. Ram et al conducted a matched control study of 41 patients with a mean age of 76 to a younger cohort with a mean age of 55 treated primarily with tisagenlecleucel.33 Overall response rates (63% vs 78%) and PFS (median 3.6 months vs not reached) were numerically worse in the elderly group, though these differences were not statistically significant in this small cohort; there were no meaningful differences in toxicity. Berning et al, reported 172 patients with diffuse large B-cell lymphoma treated with axi-cel or tisagenlecleucel analyzed according to age ≥70 or <70.32 No significant differences were noted in either efficacy or toxicities between the groups.

The more recent randomized clinical trials that showed the benefit of axi-cel and lisocabtagene maraleucel versus standard salvage chemotherapy in earlier lines of diffuse large B-cell lymphoma therapy also examined age-based subgroups.16 21 There was no suggestion in either study of a significant interaction between age and the benefit conferred by CAR-T therapy.

Collectively, these studies do not show major differences in outcomes among older patients with B-cell lymphoma receiving CAR-T therapy. Though select studies identified higher neurologic toxicity risk, all demonstrated clinically meaningful long-term survival fractions. Moreover, randomized trials have not shown major effect heterogeneity in age-based subgroups. There are significant caveats to these conclusions, however. Clinical trial eligibility criteria and selection bias among investigators likely excluded many frailer elderly patients and those with significant comorbidities, and these same selection biases likely operate to some extent in the post-approval setting, as indicated by the low proportion of elderly Medicare patients with late-line lymphoma who received CAR-T therapy. Furthermore, most studies dichotomized age at 65 years, though there is significant heterogeneity among patients >65 years of age; very few studies have separately analyzed patients >75 years of age. Finally, few studies took into account comorbidity and geriatric evaluation to understand the real contribution of age to CAR-T therapy outcomes.

Multiple myeloma

There are currently two CAR-T therapies commercially available for multiple myeloma: Idecabtagene vicleucel (ide-cel) and ciltacabtagene autoleucel (cilta-cel). Both are FDA-approved for patients with relapsed/refractory multiple myeloma after >1 (cilta-cel) or >2 (ide-cel) prior lines of therapies based on randomized trials showing superior PFS compared with standard alternatives.36 37 Unlike B-ALL or diffuse large B-cell lymphoma, where initial therapy is curative in a substantial portion of patients, nearly all patients with multiple myeloma progress after first-line therapy and require sequential lines of therapy. Most patients with multiple myeloma will therefore be candidates for CAR-T therapy at some point in their disease course.

Association of outcomes with race and ethnicity

Differences in outcome by race or ethnicity are especially salient in multiple myeloma, where black patients have a twofold higher risk of myeloma,38 are 4 years younger when they develop it,39 comprise 20% of new myeloma diagnoses while representing only 14% of the total population, but comprise only 3–5% of patients on clinical trials.40 Comparatively, Asian patients have a lower incidence of myeloma than white patients, and Hispanic patients are slightly higher than white patients (per 100,000: Asian 3.8, white 6.2, Hispanic 6.7, black 12.4).41 Moreover, monoclonal gammopathy of undetermined significance, the indolent precursor to multiple myeloma, is more common in black patients overall and six times higher among those aged 40–49 years (3.26% in blacks vs 0.53% in whites; p=0.0013).42

Representation of MHP in the pivotal ide-cel and cilta-cel prospective clinical trials varied (table 3). The phase 1/2 study of ide-cel did not report participant race and ethnicity.43 Among the other pivotal studies, the representation of black, Hispanic, and Asian participants was all <10%.

Table 3

Race and ethnicity of patients in pivotal multiple myeloma chimeric antigen receptor-T trials

Following their report of post-approval outcomes in patients treated with ide-cel at 11 US centers,44 the US Myeloma CAR-T Consortium published a separate analysis of 207 patients’ outcomes according to race and ethnicity.45 Patients self-identified their race as follows: 72% (n=149) were NHW, 17% (n=36) non-Hispanic black, and 11% (n=22) were Hispanic. No Asian patients were included. Hispanic patients were younger than black or white patients (57.0 vs 64.5 vs 65 years, respectively, p=0.07). Black patients had higher median levels of ferritin compared with Hispanic and NHW patients (680.5 vs 362.0 vs 307.0 ng/mL, respectively, p=0.08) and C-reactive protein (3.5 vs 1.0 vs 0.8, respectively, p=0.02) before ide-cel infusion. There were no other major differences in baseline patient characteristics based on race and ethnicity.

For outcomes, overall response rate (ORR), PFS and OS were evaluated. After 9.3 months of follow-up, there was a notable difference in ORR between the groups showing lower responses in Hispanic patients compared with black and white patients (59% vs 86% vs 86%, respectively, p=0.01). There were no statistically significant differences observed in PFS or OS (p=0.49 and p=0.99, respectively). There was, however, a suggestion of decreased PFS in Hispanic patients compared with black patients and white patients (median 4.2 vs 6.5 vs 8.5 months, respectively). Median OS was not reached in any group. Black patients were more likely to develop any grade CRS compared with Hispanic and NHW patients (97% vs 77% vs 85%, respectively, p=0.04) and had a longer median length of hospital stay (13.5 vs 9.0 vs 8.0 days, respectively, p=0.006). No differences in severe CRS or ICANS were noted.

Though this study did look at a real-world population in which 75% of patients would not have been eligible for the KarMMa registration trials for ide-cel and included a larger proportion of MHP than the registration studies for ide-cel and cilta-cel, the absolute number of black and Hispanic patients was still low, and no Asian patients were reported. There was therefore limited power to detect differences from the NHW population and leaves open the question of how generalizable the data are to the broader American population. Studies using the larger real-world data sets, such as the Cellular Immunotherapy Data Resource, may be more informative. Data on therapeutic outcomes by race and ethnicity for different myeloma therapies in other settings have shown instances of better and worse outcomes for MHP, and the underlying causes are unclear. With emerging data on response to CAR-T therapy being related to immune status, such as T-cell fitness and diversity (PMID: 36026513), differences in response to CAR-T according to race/ethnicity may be due to underlying differences in immune status; however, much like all other information related to race/ethnicity, rigorous data on CAR-T therapy is lacking. Therefore, race or ethnicity should not be used in clinical decision-making or counseling of patients regarding risk of or prognosis after CAR-T therapy. This paucity of data highlights the need to improve the inclusion of MHP to CAR-T clinical trials.

Association of outcomes with age

Multiple myeloma predominantly affects the elderly, with the average age at diagnosis standing at 69 years, and approximately 40% of new cases emerging in individuals aged 75 and above.46 Advances in myeloma therapy, reflected in improved survival rates for patients with multiple myeloma,47 48 have been notable, extending benefits even to those ineligible for high-dose chemotherapy and auto-SCT.49 Despite enhanced OS, multiple myeloma remains incurable, leading to relapse in most patients, necessitating additional therapeutic interventions. Consequently, the population of older patients with myeloma experiencing relapsed and refractory multiple myeloma (RRMM) is anticipated to increase progressively.

Clinical trials are a critical avenue for accessing novel therapies, but elderly patients are historically under-represented in multiple myeloma (MM) trials.50 Frailty, reduced functional status, and medical comorbidities are common in older patients with multiple myeloma, leading to concerns over the tolerability and treatment delivery of the newest therapies.51 Thus, the consideration of age’s impact on CAR-T therapy outcomes in older patients with RRMM remains fundamental in the comprehensive treatment approach for this patient population.

Available data on older patients with myeloma undergoing anti-B cell maturation antigen (BCMA) CAR-T therapy is limited to single-center retrospective studies and post hoc analyses of clinical trials (table 4). Overall, this data suggests favorable efficacy and safety outcomes in patients aged ≥70 compared with younger patients.

Table 4

Clinical outcomes according to age in MM in pivotal multiple myeloma chimeric antigen receptor-T trials and real-world data sets

Similarly, healthcare resource utilization was comparable among a real-world cohort of 141 patients aged ≥65 years who receive commercial CAR T cells for lymphoma or MM.52 38 patients (31.6%) had a diagnosis of multiple myeloma and 14 (36.8% of patients with myeloma) were 75 years or older. 23 patients receive ide-cel (60.5%) and 15 cilta-cel (39.5%). The analysis of the whole study group showed similar rates of CRS, grade ≥3 CRS, ICANS, grade ≥3 ICANS (16.9% vs 21.1%, p=0.48), and infection within the first 30 days post CAR-T therapy in patients >75 years compared with those age 65–74. No significant differences in 30-day hospital readmissions, 30-day intensive care unit admissions or median length of stay were observed. The OS rates for patients 65–74 versus ≥75 years were similar at 30 days (93.4% vs 95.4%, p=0.73), 90 days (83.3% vs 85%, p=0.82) and 1 year (51.7% vs 58.5%, p=0.54). Notably, the most common cause of death in both groups was progressive disease (72.7% and 66.7%), respectively.

In contrast to age, frailty has been linked to inferior efficacy following anti-BCMA CAR-T therapy in multiple myeloma. In a retrospective multicenter analysis, Davis et al53 evaluated 136 patients with multiple myeloma treated with CAR-T cell therapy (76.5% ide-cel and 23.5% cilta-cel) using the simplified frailty index (score based on age, Eastern Cooperative Oncology Group performance status, and Hematopoietic Cell Transplantation Comorbidity Index). Frail was defined as a score ≥2. 83 (61%) patients were considered frail at the time of CAR-T infusion. The frail group had higher proportions of renal insufficiency (18% vs 6%, p=0.04) and received less treatment with cilta-cel (16% vs 36%, p=0.012). With a median follow-up of 7 months, the median PFS (6.9 vs 11.1 months, p=0.028) and median OS (14 months vs not reached, p=0.025) were inferior in the frail compared with the non-frail group. The incidence of any-grade CRS (76% vs 79%, p=0.52), grade ≥3 CRS (8% vs 4%, p=0.071) were similar in both groups. Patients in the frail group had a higher incidence of any-grade ICANS (39% vs 17%, p=0.008), but similar grade ≥3 ICANS (11% vs 2%, p=0.09). Though CAR-T therapy was tolerated in frail patients, the inferior efficacy seen in the frail group could be related to the CAR-T product and a more aggressive disease biology leading to worsening functional performance.

In conclusion, the current data suggest similar clinical outcomes and healthcare resource utilization in elderly patients with multiple myeloma receiving CAR-T therapy. Future investigations should incorporate frailty assessments and aim to identify barriers to CAR-T therapy access and risk factors associated with suboptimal outcomes in the older patient with myeloma population.