Comparison of clinical outcomes in multiple myeloma patients treated with chimeric antigen receptor (CAR)-T cell therapy across clinical trials: Intention-to-treat vs. per-protocol analyses

Abstract Background: CAR T-cell therapy is an effective treatment for relapsed/refractory multiple myeloma (MM). CAR T-cell therapy is regarded as a “treatment pathway,” but outcomes are typically reported using per-protocol (PP) analyses from the time of infusion, excluding patients who undergo apheresis but may not ultimately receive CAR T-cell therapy. This may introduce bias, overestimate benefits, and hinder comparisons across studies and products. Here, we report a systematic meta-analysis of clinical trials on CAR T cell therapy, mainly examining outcomes in both PP and intent-to-treat (IIT) analyses and analyze trends over time. Methods: A systematic literature search was conducted according to PRISMA guidelines in PubMed, Embase, and Cochrane databases through April 2025. We included phase I, II, and III clinical trials evaluating CAR T-cell therapy targeting B-cell maturation antigen (BCMA), G protein-coupled receptor class C group 5 member D (GPRC5D), or multiple antigens for multiple myeloma (MM). The primary endpoints were overall response rate (ORR) and complete response rate (CRR) in the PP vs. IIT patient population. Pooled ORR, CRR, and their 95% confidence intervals (CI) were estimated using a random-effects model. Additionally, trends based on year of enrolment (before 2020 and 2020 and beyond) and study site (US vs non-US) were also analyzed. Results: A total of 46 clinical trials (phase I/II n=44 and phase III n=2) enrolling 2,561 patients were included. The median age was 59 years (range: 57-61). The median number of prior lines was 4 (range:4-6), 76% and 27% of patients had triple-class and penta-class refractory disease, respectively. Twelve studies (26%) were conducted in the US, while the majority (n=33; 73%) were non-US studies. US patients were older (61 vs. 58 years) and more heavily pretreated (6 vs. 4 prior lines). Among enrolled patients, 99.6% patients underwent apheresis, and 90.4% received CAR T-cell infusion. Infusion rates remained stable before and after 2020 (89.4% vs. 88.8%) but were higher in non-US studies compared to US studies (93.6% vs. 86.2%). The main reasons for not receiving infusion were disease progression (58.7%) and death (17.8%). Pooled response rates were higher by PP vs. ITT analysis: ORR 86% (95% CI, 82–91%) vs 78% (73–82%) and CRR 56% (50–62%) vs 49% (43–55%), respectively (p0.001; I² 88%). Outcomes remained stable post-2020 for ORR (ITT: 83.19%→83.62%; PP: 88.89%→90.07%), while CRR improved in PP populations (55.56%→60%). Response rates were consistently higher in non-US studies, notably ORR (PP: 90.91% vs 73%) and CRR (ITT: 55.56% vs 35.29%), likely reflecting differences in patient populations and products. At a median follow-up of 13.7 months, the median PFS was 12.7 months. Conclusions: This meta-analysis shows a significant difference in outcomes between ITT and PP analyses in CAR T-cell trials, with about 10% attrition from apheresis to infusion, consistent over time. While the CRR per PP has improved, likely due to better CAR T-cell products, the steady outcomes per ITT highlight the ongoing need to address clinical and logistical barriers that prevent a subset of patients from receiving the CAR T-cell product. Non-US studies had higher ORR and CRR than US ones, reflecting differences in trial design and populations, which warrants cautious interpretation