Safety and efficacy of CD19/CD22/CD20 multi-targeted CAR-T therapy in patients with refractory/relapsed B-cell non-Hodgkin lymphoma

Abstract

Abstract Background CAR-T therapy has significantly improved response rates in refractory/relapsed B-cell non-Hodgkin lymphoma (R/R B-NHL). However, for high-risk refractory patients who have failed multiple lines of therapy, including those with CAR-T treatment failure or disease progression, as well as those with antigen loss or downregulation, novel therapeutic approaches remain urgently needed. Based on this, we conducted a clinical study of CD19/CD22/CD20 multi-targeted CAR-T therapy to explore the safety and efficacy of combined infusion of CD19/CD22 dual-targeted CAR-T and CD19/CD20 dual-targeted CAR-T in patients with R/R B-NHL. Methods This first-in-human study is an open-label, multicenter, investigator-initiated trial (IIT, NCT07093086). After lymphodepleting chemotherapy, patients received sequential infusions of CD19/CD22 dual-targeted CAR-T (dose range: 2.8×106 to 1.0×107 CAR-T cells per kilogram of body weight) and CD19/CD20 dual-targeted CAR-T (dose range: 3.5×106 to 1.0×107 CAR-T cells per kilogram of body weight) in two consecutive days. The primary endpoint was to assess the safety and preliminary efficacy. Cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS) were graded according to ASTCT 2019 criteria. Efficacy was assessed according to Lugano 2014 criteria. Secondary endpoints included pharmacodynamics and pharmacokinetic evaluation. Results From September 20, 2024 to June 30, 2025, a total of 12 patients with R/R B-NHL were enrolled, with a median age of 53 years (range: 26-67), 83% of whom were male. All patients had highly refractory disease with failure of multiple prior therapy lines, including 7 cases of diffuse large B-cell lymphoma (DLBCL), 2 cases of EBV-positive DLBCL, 2 cases of transformed follicular lymphoma (TFL), and 1 case of high-grade B-cell lymphoma (HGBL). 58% (7/12) had an ECOG performance status of 2, indicating poor functional status. 92% (11/12) were stage IV, 75% (9/12) had an IPI score ≥3, 67% (8/12) had TP53 mutations, 58% (7/12) had Ki67>75%, 58% (7/12) were double-expressor lymphomas, and 17% (2/12) were double/triple-hit lymphomas. The median number of prior therapy lines was 3 (range 2-5), with 58% (7/12) having received ≥3 lines. 17% (2/12) had undergone prior autologous hematopoietic stem cell transplantation. 50% (6/12) had received prior CAR-T therapy, including 5 cases targeting CD19 alone and 1 case with CD19/CD22 dual targeting. Among these, 2 patients were treated with commercial products (relmacabtagene autoleucel or axicabtagene ciloleucel), while the remaining 4 participated in IIT. Previous CAR-T outcomes included 50% complete response (CR), 33% partial response (PR), and 17% progressive disease (PD), with a median duration of response (DOR) of 5.6 months (1.2-17.6). 92% (11/12) had extranodal involvement, including 50% (6/12) with ≥2 extranodal sites. 58% (7/12) were refractory and 42% (5/12) were relapsed. 83% (10/12) received bridging therapy after leukapheresis for this multi-targeted CAR-T therapy. 75% (9/12) developed CRS following multi-targeted CAR-T therapy, with grade 3 CRS occurring in 17% (2/12) that resolved with supportive care. No ICANS of any grade was observed. No CAR-T-related deaths occurred. The best overall response rate (BOR) was 75% (9/12), including 67% CR and 8% PR. Among 6 patients with prior CAR-T failure, BOR reached 83% (5/6), with 67% CR and 17% PR. Notably, one patient who had only weak CD22 expression and complete loss expression of CD19/CD20 after prior CD19/CD22 CAR-T therapy still achieved CR following this CD19/CD22/CD20 multi-targeted CAR-T therapy. Pharmacokinetic analysis revealed a median Tmax of 14 days (range: 7-21), median Cmaxof 51,600 copies/μg DNA (range: 1,400-324,000), and median AUC0-28 of 365,639 (range: 12,278-2,682,795). Responders demonstrated significantly higher Cmax (88,500 vs. 15,400) and approximately 10-fold greater AUC0-28 (1,100,604 vs. 105,494) compared to non-responders, indicating stronger in vivo expansion. Notably, CD19/CD20 dual-targeted CAR-T cells showed greater expansion than CD19/CD22 CAR-T. Conclusion The CD19/CD22/CD20 multi-targeted CAR-T therapy demonstrated a favorable safety profile and encouraging efficacy in highly refractory NHL patients failing multiple lines of therapy, including those with prior CAR-T failure, antigen loss/downregulation. These promising findings warrant further investigation in expanded cohorts with long-term follow-up.