Gamma Secretase Inhibition May Mitigate the Impact of Low BCMA Target Density in Relapsed/Refractory Multiple Myeloma: Results from a Comparative Analysis of Two Phase 1 Clinical Trials

The safety and clinical activity of FCARH143, an autologous CAR-T cell product with a human BCMA-specific scFv and a 4-1BB costimulatory domain, has been evaluated in adult patients (pts) with relapsed/refractory multiple myeloma (R/R MM) in two single-arm phase I trials with similar eligibility criteria (contemporaneously accruing at the same center), FH9952 (n=18) and FH9762 (n=25), at dose levels of 50-450 ×10⁶ CAR⁺ cells. Pts on FH9952 received the γ-secretase inhibitor (GSI), crenigacestat, during a pre-CAR-T run-in (3 doses 48 h apart) followed by post-infusion dosing 3x weekly for up to 9 doses. GSIs have been shown to enhance in vivo CAR-T activity by increasing target BCMA density on malignant plasma cells. Our group previously reported that this strategy was well tolerated and successfully increased BCMA density (Cowan, A. Lancet Oncol. 2023). In this analysis, we performed a cross-trial comparison to understand the impact of the GSI and assessed long-term outcomes among high-risk subgroups.

Select baseline characteristics between trials (FH9952 vs FH9762) were as follows: age (median 66 vs 64 yrs), ECOG 2+ performance status (12% vs 5.6%), presence of a high risk chromosomal abnormality (HRCA: 56% vs 72%; del17p, t4;14, t14;16, t14;20), ISS stage (median 2.5 vs 2.5), prior lines of therapy (median 9.5 vs 8.0), time from diagnosis to CAR-T infusion (median 5.5 vs 5.1 yrs), BMPC % at screening (45% vs 60%), EMD+ (28% vs 44%), baseline tumor BCMA level (median 620 vs 620 ABCs), prior autologous (89% vs 80%) or allogeneic stem cell transplant (HCT; 11% vs 20%), and prior anti-BCMA exposure status (39% vs 12%).

With a median follow-up of 4.8 yrs for the combined cohort (n=53), the OS and PFS were 2.7 and 1.1 years, respectively. Median OS and PFS were shorter for pts with plasma cell leukemia (PCL; OS: 1.0 vs 2.8 yrs, p=.019; PFS: 0.5 vs 1.3 yrs, p=.064), extramedullary disease (EMD; OS: 1.4 vs 4.2 yrs, p=.002; PFS: 0.5 vs 1.7 yrs, p=.045), and prior BCMA exposure (OS: 1.0 vs 3.5 yrs, p=.005; PFS: 0.5 vs 1.7 yrs, p=.0001); but not for pts with a HRCA (OS: 2.7 vs 2.3 yrs, p=.81; PFS: 1.5 vs 1.0 yrs, p=.72).

For FH9952 vs FH9762, the median OS (3.1 vs 2.6 yrs, p=.48) and PFS (0.9 vs 1.3 yrs, p=.72) were similar. In univariate (UV) Cox regression, factors associated with OS included prior BCMA exposure (HR 3.02, 95% CI 1.35-6.79, p=.007), EMD (HR 3.06, 95% CI 1.44-6.50, p=.004), BMPC % (HR 1.19, 95% CI 1.03-1.37, p=.015), PCL (HR 5.10, 95% CI 1.12-23.2, p=.035), and soluble BCMA at day +60 (HR 1.52, 95% CI 1.22-1.90, p<.001); and factors associated with PFS included prior BCMA exposure (HR 4.20, 95% CI 1.89-9.36, p<.001), EMD (HR 1.93, 95% CI 1.00-3.72, p=.049), BMPC % (HR 1.26, 95% CI 1.09-1.45, p=.001), and day +60 soluble BCMA (HR 1.69, 95% CI 1.27-2.25, p<.001).

Among BCMA-naïve pts (n=33), the median OS was significantly longer for FH9952 vs FH9762 (NR vs 2.3 yrs, p=.033), though PFS differences did not achieve statistical significance (2.6 vs 1.4 yrs, p=.11); 2-yr point estimates were 64% vs 36% for PFS and 91% vs 59% for OS. In UV Cox regression restricted to BCMA-naïve pts, FH9952 was associated with lower hazard of OS (ref: FH9762; HR 0.32, 95% CI 0.11-0.96, p=.043), but not PFS (ref: FH9762; HR 0.64, 95% CI 0.30-1.39, p=.3). In an exploratory multivariate analysis that included the receipt of GSI and screening BCMA levels (modeled as a restricted cubic spline) with an interaction term, we observed increasing HRs of PFS and OS at low BCMA densities (<760 ABC) in pts who did not receive a GSI.

Rates of ASTCT defined CRS (G1+: 94% vs 84%, p=.4; G3+: 11% vs 4%, p=.6) and ICANS (G1+: 39% vs 20%, p=.2; G3+: 11% vs 0%, p=.2) did not achieve statistical significance, but were numerically higher for FH9952 compared to FH9762.

The BCMA CAR-T product, FCARH143, demonstrated clinical activity and manageable toxicity in two heavily pretreated R/R MM cohorts. Pts with known high-risk features (i.e., PCL, EMD, prior BCMA exposure) and increasing day +60 soluble BCMA experienced worse long-term outcomes. Among BMCA-naïve pts, those who received a GSI experienced significantly longer OS compared to those who received FCARH143 alone. Receipt of GSI may have the largest effect on PFS/OS in patients with low baseline BCMA tumor density levels. Limitations of this study include a small sample size and retrospective analysis. A randomized prospective study is needed to confirm a survival benefit of incorporating a GSI into treatment with FCARH143, or other BCMA CAR-T products.