Quality-adjusted survival analysis of neurologic events with ciltacabtagene autoleucel (Cilta-cel) vs standard of care (SOC) in patients (Pts) with lenalidomide-refractory multiple myeloma (MM) who received 1–3 prior lines of therapy (LOT): Cartitude-4 trial population (Pop)

Abstract

Abstract Introduction: A single infusion of cilta-cel demonstrated significant improvement in overall survival (OS) vs SOC in pts with lenalidomide-refractory MM after 1–3 prior LOT in the CARTITUDE-4 trial (NCT04181827). At median follow-up of 33.6 mo, cilta-cel reduced risk of death by 45% vs SOC (P=0.0009), reinforcing its potential for durable benefit. A previously reported Quality-Adjusted Time Without Symptoms or Toxicity (Q-TWiST) analysis using CARTITUDE-4 data demonstrated a 32.1% relative gain in quality-adjusted survival with cilta-cel vs SOC based on all grade 3/4 adverse events (AEs), including neurologic AEs (an infrequent but recognized risk with chimeric antigen receptor T-cell [CAR-T] therapies). Here, we report a focused sensitivity analysis assessing the impact of neurologic AEs on the quality-adjusted survival benefit of cilta-cel vs SOC. Methods: Pts from CARTITUDE-4 intent-to-treat (ITT; cilta-cel, N=208; SOC, N=211) and as-treated pop (cilta-cel, N=176; SOC, N=208) with maximum follow-up of 45 mo were included. For this analysis, OS was partitioned into 3 health states: TOX (time with grade 3/4 neurologic AEs before disease progression), TWiST (time without grade 3/4 neurologic AEs before disease progression), and REL (survival time after progression). TOX was defined across 2 scenarios focused on neurologic AEs: base case scenario including all grade 3/4 neurologic AEs related or unrelated to CAR-T, and a sensitivity scenario including only CAR-T–related grade 3/4 neurologic AEs. Neurologic AEs were defined as nervous system or psychiatric disorders that were not designated as CAR-T–related. CAR-T–related neurologic AEs considered in this analysis included movement and neurocognitive AEs, cranial nerve palsy, and peripheral neuropathy. The sensitivity scenario was only conducted for the as-treated pop. Both treatment-emergent and non–treatment-emergent events were included. Q-TWiST was calculated as the utility-weighted sum of mean durations in each state, using conventional utility weights: TWiST (1.0), TOX (0.5), REL (0.5). Relative Q-TWiST gains of 10–15% were considered clinically important. Results: Pts treated with cilta-cel demonstrated significantly longer survival time without grade 3/4 neurologic AEs (TWiST) vs those receiving SOC. In the ITT pop, the mean PFS without grade 3/4 neurologic AEs was 30.4 mo for cilta-cel vs 17.7 mo for SOC in the base case. In the as-treated pop, the mean PFS without grade 3/4 neurologic AEs was 38.0 mo vs 25.1 mo with cilta-cel vs SOC. In the sensitivity analysis, the mean PFS without CAR-T–related grade 3/4 neurologic AEs was 35.7 mo vs 18.0 mo with cilta-cel vs SOC. In the base case scenario, the mean PFS with grade 3/4 neurologic AEs (TOX) was 0.07 mo vs 0.15 mo with cilta-cel vs SOC (P=0.169). In the sensitivity scenario, the mean PFS with grade 3/4 CAR-T–related neurologic AEs for cilta-cel vs SOC was 0.05 mo vs 0 mo (P=0.185). In both scenarios, cilta-cel demonstrated significantly greater Q-TWiST gains vs SOC. In the ITT pop, the mean Q-TWiST gain was +8.7 mo in the base case, representing a 34.6% relative improvement in Q-TWiST (95% CI, 5.7–11.6; P<0.001). In the as-treated pop, corresponding mean Q-TWiST gains were +13.0 mo, representing a 51.6% improvement (base case; 95% CI, 10.3–15.6; P<0.001) and +12.9 mo, representing a 51.1% improvement (sensitivity scenario; 95% CI, 10.2–15.6; P<0.001). Conclusions: Given that neurologic AEs, though infrequent, are a recognized consideration with CAR-T therapies such as cilta-cel, this focused Q-TWiST analysis evaluated the specific impact of these events on quality-adjusted survival. In both base case and sensitivity scenarios, pts treated with cilta-cel experienced statistically significant and clinically meaningful gains in quality-adjusted survival vs SOC, with gains exceeding the 10–15% threshold generally considered important in oncology. In the ITT pop, a significant improvement was observed in time without neurologic AEs for cilta-cel vs SOC. In contrast, differences in time with neurologic AEs were not statistically significant and represented a small proportion of total survival time. These findings are consistent with results from a previous Q-TWiST analysis that incorporated a broader set of AEs, including neurologic AEs and cytokine release syndrome. Overall, these data further demonstrate the benefit of treating pts with lenalidomide-refractory MM with cilta-cel as early as first relapse.