Trends in Pre-CAR-T Attrition in Relapsed/Refractory Multiple Myeloma: Prevalence, Risk Factors, and Outcomes at a Tertiary Academic Medical Center

INTRODUCTION:

BCMA-targeted chimeric antigen receptor T-cell (CAR-T) therapy is a standard approach for managing patients (pts) with relapsed/refractory multiple myeloma (MM). Many pts who are planned to receive CAR-T therapy experience attrition, defined as inability to receive therapy due to death. The prevalence and risk factors for pre-CAR-T attrition among contemporary MM pts intended to receive standard-of-care (SOC) products are not well understood.

METHODS:

We evaluated all pts evaluated by a MM specialist and waitlisted for idecabtagene vicleucel and/or ciltacabtagene autoleucel at a large tertiary academic medical center from March 2021 to March 2024. We defined attrition as death without infusion. We excluded pts from our attrition analysis set if they remained alive and had not received an infusion. Distance to our center and household income were estimated using zip code data. Multivariable (MV) models for attrition and time to infusion adjusted for number of high-risk chromosomal abnormalities (HRCAs; defined as del 17p, amp 1q, t4;14, t14;16, t14;20), hemoglobin (Hgb) and creatinine (Cr) at time of waitlisting, prior lines of therapy (LOTs), age, and ECOG score. A MV model for receipt of an out-of-specification (OOS) product adjusted for prior LOTs, and receipt of chemotherapy, bispecific antibodies, selinexor, and proteasome inhibitors.

RESULTS:

Of 185 total waitlisted pts, 89 (48%) have received a BCMA CAR-T product to date. 138 pts who received CAR-T cells and/or died were included in our attrition analysis set. Among pts in the attrition analysis dataset, the overall attrition rate was 36%, though this improved from 44% in 2021-2022 to 15% in 2023-2024. Those who received a CAR-T product had better ECOG score (0-1: 97% vs 69%, p<.001), higher Hgb (median 11.6 vs 9.4 g/dL, p<.001), and lower Cr (0.96 vs 1.12 mg/dL, p=.009). Socioeconomic and demographic factors were similar between groups. With a median follow-up duration of 17.2 months from time of waitlisting for the total waitlisted cohort, overall survival was not reached for those who received a CAR-T product, compared to 9.4 months for those who did not (p<.001).

In MV logistic regression (applied to the attrition analysis dataset), increasing ECOG score (aOR 3.06, 95% CI 1.68-6.00, p<.001) and increasing Hgb (aOR 0.64, 95% CI 0.49-0.81, p<.001) were associated with higher and lower odds of attrition, respectively. Among evaluable pts who experienced CAR-T attrition (n=49), the main reason was progressive disease (PD; 82%), whereas attrition events related to infection (10%) or manufacturing failure (8%) were less common.

In MV Cox regression (applied to the total waitlisted cohort), factors associated with longer time to infusion (TTI) included increasing ECOG score (aHR 0.63, 95% CI 0.44-0.90, p=.011) and increasing Hgb (aHR 0.87, 95% CI 0.78-0.97, p=.013). The cumulative incidence (CI) of CAR-T infusion in this population increased over time (12-month CI, 2021 vs 2022 vs 2023: 17% vs 48% vs 64%, p<.001), whereas waitlist mortality decreased (47% vs 26% vs 8.2%, p<.001). Among waitlisted pts who started a new treatment (n=99), bispecific antibody usage increased (2021-2022 vs 2023-2024: 5.6% vs 29%, p=.002) and chemotherapy usage decreased (30% vs 6.7%, p=.004). Notably, among waitlisted pts who had a product manufactured (n=87), receipt of chemotherapy after waitlisting was independently associated with OOS product infusion (aOR 6.77, 95% CI 1.22-42.1, p=.029).

CONCLUSION:

Only 48% of eligible MM pts waitlisted for SOC BCMA CAR-T therapy at our center ultimately received a product to date. Attrition rates have improved over time yet remain suboptimal. We could not confirm an impact of demographic and socioeconomic factors on attrition. Increasing ECOG score and Hgb at time of waitlisting were independently associated with pre-CAR-T attrition and TTI. Receipt of chemotherapy while on the waitlist was independently associated with receipt of an OOS product. PD was the predominant reason for waitlist attrition, consistent with other large studies of attrition over lines of therapy in MM. These data highlight the need for more effective bridging therapies and faster product manufacturing, appropriate CAR-T slot allocation, and an expansion of cellular therapy capacity at all centers.