Introduction

While the life expectancy for those living with multiple myeloma (MM) has steadily increased, most patients undergo a series of remissions and relapses. This has underscored the need for chimeric antigen receptor (CAR) T-cell therapy, a novel immunotherapy for MM involving genetic modification of patient T-cells. B-cell maturation antigen (BCMA)-directed CAR T-cell products idecabtagene vicleucel (ide-cel) and ciltacabtagene autoleucel (cilta-cel) demonstrated superior outcomes over standard care, leading to expanded FDA approvals [1, 2].

As improvements in supply of BCMA-directed CAR T-cell therapies catch up to the high demand, concerns persist regarding prolonged time from T-cell harvest to infusion (vein-to-vein time, V2V) [3,4,5]. In the CARTITUDE-4 study of patients with 1–3 prior lines of therapy and effective bridging therapy, 6% died before receiving cilta-cel [2]. While V2V delays have been a focus due to rapid disease progression, delays in brain-to-vein time (B2V)—from formal request for insurance approval to T-cell harvest—are also problematic [5]. In comparison to patients with Medicare, those with private insurance in the United States may experience longer B2V times because they require a single case agreement (SCA) – contracts that allow customized payment terms for specialized treatments not typically covered by insurance – with the CAR T-cell center before approval is granted.

This study analyzed B2V and V2V for CAR T-cell therapy in MM, exploring their relationship with insurance status, disease characteristics, and patient outcomes.

Methods

We conducted a retrospective chart review of 64 consecutive insured patients at the University of Chicago collected for commercial CAR T-cell therapy between April 2021 and June 2024.

Patients were stratified by insurance type (private vs Medicare). Chi-square or Fisher’s exact test were used for comparisons of categorical variables and the Mann-Whitney test for continuous variables. Outcomes of interest included B2V (time from submission of a letter to insurance for approval), V2V (time from T-cell harvest to infusion), cytokine release syndrome (CRS) incidence/grade, and immune effector cell associated neurotoxicity syndrome (ICANS) incidence/grade, progression-free survival (PFS) and overall survival (OS). Survival outcomes (PFS and OS) were estimated using the Kaplan-Meier method and compared between groups using the log-rank test. Cox proportional hazard models estimated hazard ratios (HRs) for PFS and OS and were indexed from the time of T-cell collection, though progression alone during V2V was not considered an event. Univariate and multivariate analysis with stepwise selection were performed, with a p-value < 0.05 considered statistically significant.

Results

Among the 64 patients analyzed, 40 had Medicare and 24 had private insurance which all required an SCA. Baseline characteristics were comparable across insurance types; the private group was younger in age, received a higher proportion of cilta-cel infusions, and received more bridging therapy (Table 1). In the entire cohort, 20% of patients had prior BCMA-directed therapy and 47% had aggressive disease features leading into CAR T-cell therapy.

Table 1 Patient Characteristics.
Full size table

Median B2V time for Medicare was 29.5 days and for private was 39.5 days (p = 0.2) (Table 1). B2V time was ≤15 days in 23% for Medicare vs 4% for private (p = 0.049). Median V2V time for Medicare was 62 days and 59 days for private (p = 0.8). There were no differences in B2V by year, internal vs external referral, or by private payor.

With a median follow-up time from T-cell collection of 9.5 months for the entire cohort (median 9 months for Medicare, 9.9 months for private), there were a total of 32 PFS events including 21 deaths; 5 patients died prior to receipt of CAR T due to disease progression, 6 died from non-relapse mortality, and 10 died following disease progression (Table S1). All five patients who died prior to receipt of CAR T-cell therapy had a B2V > 15 (range 28–59) days. Of the 6 patients with CAR T-related mortality, 5 (83%) had a B2V > 15 (range 16–67) days.

On univariate analysis, the following variables were significantly associated with inferior PFS and OS: receipt of ide-cel, higher ECOG performance status, presence of aggressive disease features, prior BCMA-directed therapy (PFS only), bone marrow plasmacytosis ≥60%, and ferritin > median (285 ng/mL).

Stratified by insurance type, there was an early crossing of Kaplan-Meier curves with the private group having more early survival events; the Cox proportional hazards assumption was not violated. Insurance type was not associated with PFS or OS (Fig. 1). The estimated 12-month PFS was 45% (95% CI 28–61%) for Medicare and 64% (95% CI 39–82%) for the private group (logrank p = 0.2) and the estimated 12-month OS was 61% (95% CI 42–75%) for Medicare and 81% (95% CI 57–92%) for the private group (logrank p = 0.1). Amongst patients who received cilta-cel, the 12-month PFS was 59% (95% CI 38–76%) for Medicare and 67% (95% CI 41–84%) for private (p = 0.5); the 12-month OS was 65% (95% CI 41%–82%) for Medicare and 85% (95% CI 60%–95%) for private (p = 0.1).

Fig. 1: Outcomes stratified by insurance status.
figure 1

A Progression-free survival and B overall survival stratified by insurance status with 95% confidence intervals.

Full size image

Stepwise multivariate analysis revealed that older age and cilta-cel (vs ide-cel) were associated with superior PFS and OS, while worse ECOG PS and bone marrow plasmacytosis >60% were associated with worse PFS and OS (Tables S2, S3). Receipt of prior BCMA-directed therapy was associated only with inferior PFS and receipt of bridging chemotherapy was associated only with superior OS.

Discussion

Patients with Medicare insurance had numerically shorter B2V time and a higher likelihood to achieve a B2V ≤ 15 days. Nearly all early deaths before and after CAR T occurred in patients with B2V > 15 days, irrespective of insurance type. In aggregate, neither insurance type, B2V nor V2V were associated with outcomes. Multivariate analysis revealed that younger age, receipt of ide-cel vs cilta-cel, worse ECOG PS, and BMPC ≥ 60% were associated with inferior PFS and OS; prior BCMA-directed therapy was associated with worse PFS, and bridging chemotherapy was associated with worse OS.

In our study of B2V and V2V in patients with MM receiving commercial BCMA-directed CAR T, we indexed outcomes from the time of T-cell apheresis rather than CAR T-cell infusion to better understand the impact of B2V on survival. Two studies of patients with diffuse large B-cell lymphoma receiving CD19-directed CAR T-cell therapy found that between 35%–66% of patients with private insurance required an SCA. Similar to our findings, B2V was longer for patients needing an SCA, V2V was similar, and outcomes did not differ significantly except that patients who did not ultimately receive CAR T had worse survival [6, 7]. Our multivariable analysis findings differ slightly from the MyCARe CAR T prediction model, which identified extramedullary disease or plasma cell leukemia, lenalidomide-refractoriness, high-risk cytogenetics, and increased ferritin at the time of lymphodepleting chemotherapy as independent predictors of early progression following BCMA-directed CAR T-cell therapy [8]. This may be because most patients in that analysis received ide-cel or an investigational product, and because we examined risk factors prior to apheresis rather than prior to lymphodepletion.

The most important limitation of this study is the sample size and single institution data source, which may limit the ability to detect significant between-group differences. B2V did not improve during the span of the study, but B2V may improve over time with more flexible slot allocation. Pooling data on B2V and V2V from a large sampling of CAR T centers would help to validate our findings.

While the proportion of Black patients in this study is higher (14%) than in the pivotal CAR T trials, it is still not representative of the US population and this study is likely underpowered to detect racial differences in outcomes. Considering the historically poorer outcomes for Black patients with MM, expanding the dataset to include more Black individuals may be valuable in providing insight into potential racial differences.

Efforts are needed to improve the B2V and V2V in MM. Clinicians need to be expeditious in submitting for insurance approval for CAR T-cell therapy, ideally, as soon as a patient is considered a suitable candidate. Healthcare institutions should collaborate with private insurers to transition from SCAs toward establishing CAR T-cell therapy as a standardized component of contractual coverage. Manufacturers need to improve supply and reduce slot restrictions, including offering flexible apheresis timelines which should be matched by apheresis center accommodations. Patients should avoid lymphodepleting therapies such as bendamustine prior to apheresis which may delay T-cell harvest due to inadequate T-cell numbers or function. This will also help to avoid the generation of out-of-specification products that add to V2V. Manufacturers must work harder to shorten V2V, which in our study was twice as long as in real-world lymphoma cohorts.

In conclusion, insurance type may influence B2V, and nearly all early deaths before and after receipt of CAR T occurred in patients with B2V > 15 days. We found no statistically significant association between insurance type, B2V, or V2V and overall outcomes in MM patients receiving CAR T-cell therapy. Further inclusion of a larger patient population across multiple centers is essential to validate these findings and explore racial disparities in outcomes. Future efforts should focus on optimizing insurance approval and treatment timelines, streamlining insurer agreements, and addressing logistical and manufacturing barriers to improve B2V and V2V for all patients.