Safety and efficacy of less differentiated CAR T-cell products in relapsed/refractory B-cell non-Hodgkin lymphomas: A meta-analysis

Abstract Introduction: CAR T-cell therapy has revolutionized the treatment of B-cell malignancies, particularly in heavily pretreated patients. However, despite high initial response rates, relapse occurs in 50–60% of cases and treatment-related toxicities remain major limitations. Conventional manufacturing requires a prolonged time from leukapheresis to CAR T-cell infusion (often several weeks), during which many patients experience disease progression or even die while waiting for the final product. Additionally, manufacturing processes that lead to T-cell terminal differentiation have been shown to impair long-term efficacy. To address these challenges, newer strategies to shorten vein-to-vein time, preserving naive/memory T lymphocytes, and enhancing persistence are evolving worldwide. This meta-analysis evaluates the safety and efficacy of such next-generation CAR T-cell therapies in B-cell non-Hodgkin lymphomas (B-NHL).Methods: This meta-analysis was conducted following PRISMA guidelines and was registered in PROSPERO (CRD420251083527). PubMed, EMBASE, and Cochrane Library were searched to identify clinical studies reporting outcomes of CD19-directed CAR T-cell products for B-NHL with a 4-1BB co-stimulatory domain, enriched for less differentiated T-cell phenotypes or developed using rapid manufacturing platforms. Eligible studies reported at least one safety or efficacy outcome. Safety endpoints included cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS), both any grade and grade ≥3. Efficacy outcomes included overall (ORR) and complete response rates (CRR), assessed at day 28 or day 90. Data extraction followed Cochrane standards. Meta-analyses were performed using a random-effects model in R (v4.3.2). Heterogeneity was assessed with the I² statistic.Results: Four single-arm studies comprising 144 patients were included. All studies evaluated CD19-targeted CAR T-cell therapy with 4-1BB costimulatory domain in adults with r/r B-LNH, including at least one of the following subtypes: diffuse large B-cell lymphoma, mantle cell lymphoma, follicular lymphoma, marginal zone lymphoma, or Burkitt lymphoma. One study also included patients with B-acute lymphoblastic leukemia, who were excluded from this analysis. Across studies, the reported median age ranged from 64 to 67 years (25–81), and patients were heavily pretreated, with median prior lines ranging from 2 to 3 (1–11). Two trials were international multicenter (Europe, Australia, United States, and Japan), while the other two were single-countries in the United States and India. The pooled incidence of CRS (any grade) was 40.5% (95% CI, 32.0–49.7), with grade ≥3 CRS in 5.4% (95% CI, 2.4–12.0). ICANS occurred in 13.9% (95% CI, 6.1–28.6), and grade ≥3 ICANS in 3.3% (95% CI, 1.2–8.9). Pooled ORR was 87.5% (95% CI, 80.0–92.5), and CRR was 73.5% (95% CI, 60.1–83.6), with assessments performed at day 28 or day 90. Heterogeneity, which reflects the variability of results across studies, was negligible for grade ≥3 CRS, grade ≥3 ICANS, and ORR (I² ≈ 0%), and moderate for CRR (I² = 41.5%), likely reflecting differences in response evaluation timing.Conclusions: CAR T-cell products preserving a less differentiated T-cell phenotype or produced with rapid manufacturing platforms are safe and effective for the treatment of B-LNH, with lower rates of CRS and ICANS (both any grade and grade ≥3), while maintaining high response rates compared with pivotal studies. Longer follow-up and additional prospective studies are required. However, these findings highlight the potential of next-generation platforms to enhance CAR T-cell therapy outcomes.