5-Year Follow-up Analysis from ZUMA-5: A Phase 2 Trial of Axicabtagene Ciloleucel (Axi-Cel) in Patients with Relapsed/Refractory Indolent Non-Hodgkin Lymphoma

Introduction: Axi-cel is an autologous anti-CD19 chimeric antigen receptor (CAR) T-cell therapy approved for relapsed/refractory (R/R) follicular lymphoma (FL) based on the Phase 2 ZUMA-5 trial in patients (pts) with R/R indolent non-Hodgkin lymphoma (iNHL; FL and marginal zone lymphoma [MZL]). With a median follow-up of >52 mo, median progression-free survival (PFS) was 57.3 mo in pts with FL and 46.9 mo in pts with MZL. No new safety signals were observed (Neelapu et al. ASH 2023. Abstract 4868). Here, we report an updated analysis from ZUMA-5 after a median follow-up of >64 mo.

Methods: Eligible pts with R/R FL or MZL after ≥2 lines of therapy (including an anti-CD20 monoclonal antibody plus an alkylating agent) underwent leukapheresis followed by lymphodepletion and axi-cel infusion at a dose of 2×10⁶ CAR T cells/kg. The primary endpoint was overall response rate. Time-to-event outcomes were assessed per investigator in all enrolled pts. Exploratory analyses included lymphoma-specific survival using competing risk assessment (deaths unrelated to lymphoma including progression or study treatment complications were competing risks) and identification of covariates associated with ongoing response.

Results: A total of 159 pts (FL: 127, MZL: 31) were enrolled, with a median follow-up at data cutoff (March 31, 2024) of 64.6 mo (range, 32.3-81.4; FL: 65.7, MZL: 55.8). Median duration of response (DOR) was 60.4 mo (95% CI, 39.7-not estimable [NE]; FL: 60.4, MZL: not reached [NR]), with estimated 60-mo DOR of 53.4% (95% CI, 43.9-62.0; FL: 52.2%, MZL: 60.0%). Among pts who achieved a complete response (CR), 58% remained in CR at the time of data cutoff. Median PFS was 62.2 mo (95% CI, 34.9-NE; FL: 57.3, MZL: NR), with 50.4% achieving the 60-mo landmark (FL: 49.8%, MZL: 53.9%). Median PFS was NR among pts with CR (n=120; 95% CI, 62.2-NE); among pts with partial response, median PFS was 6.9 mo (n=23; 95% CI, 4.5-12.4). PFS rates at 60 mo in pts with FL were consistent regardless of high-risk characteristics, including progression <2 y from initiating first anti-CD20–containing chemoimmunotherapy (POD24). Median time to next therapy was NR in all pts with iNHL (95% CI, 38.6-NE; consistent by disease type), with a 60-mo estimated rate of 53.3% (95% CI, 45.0-60.9; FL: 54.0%, MZL: 50.9%). At data cutoff, 55% of pts (n=87) were alive with no new anticancer therapy. The median overall survival (OS) was NR (95% CI, NE-NE; consistent by disease type), and the 60-mo OS estimate was 69.0% (95% CI, 60.8-75.8; FL: 68.9, MZL: 71.1).

Among pts with FL, the 60-mo cumulative incidence of progression or lymphoma-specific death was 35.1%, and the cumulative incidence of non–lymphoma-specific deaths was 15.1%. In addition, the cumulative incidence of lymphoma-specific death at 60 mo was 15.6%, with a cumulative incidence of non-lymphoma-specific death of 15.6%. Only 4 pts progressed >24 mo post-leukapheresis. One pt progressed after the data cutoff of the 4-y analysis; no pts died of disease progression after the prior analysis.

Among treated pts (n=152; 124 FL; 28 MZL), 3 new events not related to axi-cel were reported after the 4-y analysis, including Grade 3 metastasis, Grade 1 bladder cancer, and Grade 4 myelodysplastic syndrome (a serious adverse event). One pt died of pneumonia, unrelated to axi-cel. At any time on trial, a total of 47 pts died due to progressive disease (n=14), secondary malignancies (n=6), infections (n=15), or other/unknown (n=12).

Treated pts in ongoing response at the 60-mo data cutoff had higher median post-infusion peak CAR T-cell expansion (59.41 cells/µL) and area under the curve within the first 28 d after treatment (AUC_0-28; 696.92 cells/µL×d) than those who relapsed (30.45 cells/µL and 362.68 cells/µL×d) or nonresponders (22.18 cells/µL and 269.82 cells/µL×d). The total number of infused CCR7+CD45RA+ T cells, indicative of naive phenotype, was associated with improved response.

Conclusions: After a median follow-up of more than 5 y in ZUMA-5, treatment with axi-cel continued to demonstrate durable response and long-term survival in pts with R/R iNHL. No new safety signals emerged. Over half of pts were alive at data cutoff without need of subsequent therapy. After the prior analysis, there were no lymphoma-specific deaths, and only 1 late progression event occurred in FL, indicating the curative potential of axi-cel in iNHL.

SN and JC contributed equally.