乌鸦传媒

SAN DIEGO -- In a propensity-matched comparison of two CAR T-cell therapies, ciltacabtagene autoleucel (cilta-cel, Carvykti) demonstrated better efficacy than idecabtagene vicleucel (ide-cel, Abecma) for relapsed/refractory multiple myeloma, but had a higher rate of severe toxicity.

A chart review of patients treated outside of clinical trials showed that 70% of patients treated with cilta-cel achieved complete or better response compared with 47% of patients treated with ide-cel (P<0.001). Overall response rate (ORR) was 89% with cilta-cel and 79% with ide-cel (P=0.11). By intention-to-treat and as-treated analyses, cilta-cel was associated with better progression-free survival (PFS) and overall survival (OS). Non-relapse mortality (NRM) did not differ significantly.

Severe cytokine release syndrome (CRS) occurred more than twice as often with cilta-cel, although the rate was low in both groups. Delayed neurotoxicity and infections also occurred significantly more often with cilta-cel. A trend toward a higher rate of second malignancy with cilta-cel did not reach statistical significance, reported Doris Hansen, MD, of the Moffitt Cancer Center in Tampa, Florida, at the American Society of Hematology meeting.

"I do think that the results must be interpreted with caution," said Hansen. "We did identify that patients treated with cilta-cel had better responses, particularly depth of response, which was a complete response... . In terms of survival, we see a clear difference. We see significantly superior progression-free and overall survival for cilta-cel in both the intention-to-treat and infused cohorts compared to ide-cel."

"Propensity-score matching was performed across several high-risk subgroups, and we identified that cilta-cel was associated with superior progression-free survival [in almost all cases]... . The results did remain largely consistent in the sensitivity analysis."

Noting that "this is not a randomized controlled trial," Hansen acknowledged limitations of the analysis, including "unknown unknowns or biases inherent in real-world data."

"We do hold that this study does aid clinical disciplinary patient counseling, as well as guiding selection of CAR T-cell products," she added.

During a discussion that followed the presentation, an unidentified member of the virtual audience asked Hansen when she might prefer to use ide-cel.

"If you have patients with a poor ECOG performance status, we can see a statistical significance there," she responded. "I think that treatment is personalized, and I think patient factors, comorbidities, must be taken into account. I think patient selection should be personalized on those factors."

Ben Derman, MD, of the University of Chicago, noted that the survival curves (PFS and OS) appeared to cross early during follow-up. "Is that because there might be a higher non-relapse mortality early on?"

"During the first 5 months or so, we did see a higher NRM for cilta-cel, mostly due to infections," said Hansen. "There were also some severe CRS cases, and I do believe that's likely what led to that initial crossover of the PFS curves."

Survival in multiple myeloma has improved significantly with the advent of novel, more potent immunotherapies, including CAR T-cell therapy. In 2021 ide-cel became the first FDA-approved CAR T-cell therapy for multiple myeloma, followed a year later by cilta-cel.

The two products share some similarities, but also have notable differences, said Hansen. For example, ide-cell has a single B-cell maturation antigen binding domain versus two with cilta-cel. They also vary with respect to active domain structure, as well as cell dose, kinetics, and immune-mediated toxicities.

The two products' safety and efficacy have not been compared in a prospective randomized trial. In the absence of category 1 data, Hansen and colleagues performed a comparative analysis of real-world experience with the two CAR T-cell therapies.

The analysis included 641 patients treated at 19 U.S. centers from April 2021 through December 2022. The intention-to-treat population comprised 386 patients scheduled to receive ide-cel and 255 who had planned treatment with cilta-cel. Subsequently, 55 patients were not infused, 36 in the ide-cel group and 19 in the cilta-cel group. The most common reason for not receiving the product was disease progression.

Investigators used inverse probability treatment weighting, a propensity score matching approach to balance measured confounders across treatment groups. Because the study period began before approval of cilta-cel, investigators calculated a patient's probability of receiving cilta-cel, and weighting was assigned to each patient on the basis of likelihood to receive cilta-cel. The analysis adjusted for more than a dozen demographic, clinical, and disease covariates.

Several baseline characteristics differed between the two groups. More patients treated with cilta-cel had performance status 0-1 (89% vs 86%, P=0.03) and they were more likely to receive a low dose of the therapy (4% vs 1%, P=0.04). Patients in the ide-cell group more often received fludarabine/cyclophosphamide lymphodepleting therapy (91% vs 81%, P<0.001), had a significantly higher ferritin level (345 vs 208 ng/mL, P<0.001), and were more likely not to receive bridging therapy (28% vs 24%) and less likely to receive bridging therapy that resulted in partial response or greater (10% vs 21%, P=0.003).

The primary outcomes of the analysis were safety, response rate, and survival (PFS and OS)

The safety analysis identified three statistically significant differences between the treatment groups, all occurring more often with cilta-cel: CRS grade ≥3, 5% vs 2%, P<0.001; delayed neurotoxicity, 10% vs 0.6%, P<0.001; and infections, 47% vs 35%, P<0.001. Additionally, more patients treated with cilta-cel developed second malignancies but the difference did not achieve statistical significance (9% vs 5%, P=0.11). Overall, NRM did not differ between groups (HR 1.24, P=0.49) but cilta-cel had a higher trajectory early in follow-up before the survival curves came together between 10 and 20 months.

Efficacy endpoints all favored cilta-cel. In addition to best response and ORR, the PFS and OS hazards were reduced by 57% and 47%, respectively, in the intention-to-treat analysis (95% CI 0.34-0.55, P<0.001 and 95% CI 0.40-0.73, P<0.001). Considering only the patients who received infusion as planned, the PFS hazard for cilta-cel versus ide-cel was 0.48 (95% CI 0.36-0.63, P<0.001) and the OS hazard was 0.67 (95% CI 0.46-0.97, P=0.03).

Sensitivity analyses yielded results consistent with the primary analyses, showing a 54% reduction in the PFS hazard (95% CI 0.34-0.63, P<0.001) and a 34% reduction in the OS hazard (95% CI 0.44-0.99, P=0.05).

Comment