CD22-directed CAR-T cells induce complete remissions in high-risk follicular and Mantle Cell Lymphoma with limited neurotoxicity

Abstract Introduction CD19-directed CAR-T cells (CAR19) induce responses in relapsed/refractory (R/R) mantle cell (MCL) and follicular lymphoma (FL). However, CAR19 in MCL is associated with high rates of ICANS (between 9-31% grade 3+ in pivotal trials) and limited durable response. ICANS also limits outcomes for CAR19 in FL and no CAR is approved for progression of disease within 24 months of initial therapy (POD24). In a Ph1 trial in large B cell lymphoma (LBCL) after CAR19, CD22-directed CAR T-cells (CAR22) manufactured at Stanford demonstrated 2-year progression free survival (PFS) of 52% and rare ICANS (any grade: 11%, grade 3+: 0%) (Frank et al Lancet 2024). CAR22 was therefore granted Breakthrough Therapy designation by FDA. We hypothesize CAR22 will be safe and effective for R/R MCL and FL. Methods We initiated a single-institution Ph1b trial (NCT06340737) of autologous CAR22 (lentiviral vector, single cistron CD22.BB.z-CAR) using the same manufacturing process at Stanford as previously reported (Frank et al Lancet 2024). This manufacturing process was distinct from the multicenter Ph2 trial of CAR22 (NCT05972720) that was terminated early due to low durable responses and high rates of grade 3+ IEC-HS. Primary outcomes include manufacturing feasibility, maximum tolerated dose (MTD) and ORR per Lugano. Secondary outcomes include response duration and survival. Exploratory outcomes include CAR expansion, measurable residual disease (MRD) and antigen expression. Eligibility criteria for FL include relapse after ≥ 2 lines of therapy or POD24 and for MCL include relapse after ≥ 2 lines including a BTKi and immunochemotherapy. Prior CAR19 was allowed if infused >30 days before CAR22 and ≥5% of circulating CD3+ cells expressed CAR19. CD22 expression on tumor cells by flow cytometry or immunohistochemistry (IHC) was required. Results Between April 2024 and July 2025, 12 patients were enrolled and manufactured successfully (FL: n = 7, MCL: n = 5). 1 MCL patient withdrew consent and was not infused. The 11 infused patients were highly refractory receiving a median of 4 prior lines of therapy (range 1-9). 7 of 11 received prior CAR19 for transformed FL (n = 3) or R/R MCL (n = 4). 4 of 7 FL patients had POD24. All 4 MCL patients had high risk molecular features including del17p and/or TP53 mutations. 11 patients received 1 × 106 CAR+ T cells/kg with median vein-to-vein time of 24 days. 10 of 11 had grade 1-2 CRS with median onset on D+1 (range: 0-10) and duration of 3 days (range: 1-5). 1 FL patient developed grade 2 IEC-HS which resolved with steroids. 1 FL patient developed grade 1 ICANS that resolved in 1 day without treatment. There has been no non-relapse mortality to date. With median follow up of 105 days (range 28 - 452), ORR was 100% and 91% achieved CR. 6 of 7 FL patients remain in CR. 1 FL patient relapsed 1 year after CAR22 with loss of B cell aplasia at D+180 and significant reduction in CD22 expression by IHC. All 4 MCL patients responded at D+28 (CR: n = 3, PR: n = 1). 3 of 4 had undetectable MRD by peripheral blood ClonoSEQ. 1 had substantial reduction but detectable MRD (547,441 vs 1 clonal cell / 106 nucleated cells on D+28 post-CAR22). 3 MCL patients relapsed after CAR22, 1 with persistent MRD and 2 with recurrent MRD positivity before overt progression. 2 MCL patients had CD22 evaluation by QuantiBrite; 1 retained CD22 at low levels both pre- and post-relapse and 1 had substantial reduction in CD22 upon relapse. CAR22 expansion by flow cytometry peaked between D+10 and D+14 and was not significantly different between FL and MCL patients (median peak 522 vs 15.8 cells/uL, p = 0.23). Conclusion CAR22 therapy was feasible, safe at the previously identified MTD and induced responses in all R/R FL and MCL patients. We observed no cases of grade 3+ CRS or IEC-HS and no cases of grade 2+ ICANS. These clinical outcomes may reflect the use of the same manufacturing platform as was used in the Ph1 trial of CAR22 in LBCL, which demonstrated durable responses and low rates of ICANS. While CAR22 achieved high ORR, durable remissions were more frequent in FL than MCL. TP53 aberrations and relapse after prior CAR19 may represent high-risk MCL subgroups where close molecular surveillance and post-CAR maintenance should be evaluated. Accrual continues for 2 additional cohorts: LBCL and other CD22+ lymphomas including Burkitt's, Waldenstrom's, marginal zone and hairy cell leukemia. Updated outcomes will be presented at the meeting.