Curative Potential of Axicabtagene Ciloleucel (Axi-Cel): An Exploratory Long-Term Survival Assessment in Patients with Refractory Large B-Cell Lymphoma from ZUMA-1

Background: Axi-cel is an autologous anti-CD19 chimeric antigen receptor (CAR) T-cell therapy approved for patients with relapsed or refractory (R/R) large B-cell lymphoma (LBCL). Long-term results from ZUMA-1, a multicenter, single-arm, Phase 1/2 study of axi-cel in patients with refractory LBCL, demonstrated sustained overall survival (OS), with a median of 25.8 months and a 5-year estimate of 43% (Neelapu SS, et al. Blood. 2023). Initial assessments of disease-specific survival (DSS) in ZUMA-1 suggested axi-cel may be curative for a subset of patients. However, endpoints defining curative potential in patients with LBCL are not yet clearly established. Here, we report exploratory analyses from the Phase 2 portion of ZUMA-1 using proposed endpoints to measure cure along with survival analyses with up to 6 years of follow-up.

Methods: Eligible adults in the pivotal Phase 2 Cohorts (1 and 2) of ZUMA-1 had refractory LBCL (diffuse LBCL, primary mediastinal B-cell lymphoma, or transformed follicular lymphoma). Patients were leukapheresed at enrollment, then received lymphodepletion and axi-cel infusion (2×10⁶ CAR T cells/kg; Neelapu et al. N Engl J Med. 2017). Exploratory analyses to assess curative potential included lymphoma-related event-free survival (LREFS), duration of complete response (DOCR), and DSS. LREFS was defined as the time from axi-cel infusion until disease progression, initiation of new lymphoma therapy, autologous stem cell transplantation, or death due to progression, whichever was earliest. DSS was defined as time from axi-cel infusion to death due to progression or axi-cel–related adverse events (AEs). Outcomes were analyzed via Kaplan-Meier (KM) estimates and competing risk analyses. Landmark analyses were used to explore prognostic potential of disease status at Week 4 and Months 3, 6, 12 and 24.

Results: At a median follow-up of 63.1 months in the 5-year analysis (data cutoff on August 11, 2021), median LREFS was 5.8 months (95% CI, 3.4-13.9) and the 60-month LREFS rate by KM estimate was 33.5%. Among patients who achieved a complete response (CR) as best response (n=59), median DOCR was not reached (95% CI, 12.9 months-not estimable). Median LREFS among those with a CR was not reached (95% CI, 18.4 months-not estimable) and the 60-month LREFS rate was 56.8%. Among those with a CR at Months 12 and 24, 60-month estimated rates of LREFS and ongoing CR were >80% (Table 1). All 3 lymphoma-related events that occurred after Month 24 in those with a CR were disease progression. Among those with a partial response at Week 4 (n=33) and Month 3 (n=10), estimates of 60-month LREFS were 27.3% and 45.0%, respectively.

The data cutoff date for the 6-year survival analysis was August 11, 2022. Median OS was 25.8 months (95% CI, 12.8-63.7). In a competing risk assessment of OS, the cumulative incidence of death at 72 months was 61.6%. Between Months 24 and 72, there appeared to be an increase of 6% in the cumulative incidences of death due to progression (40.6% and 46.6%, respectively) and other reasons (5.0% and 11%, respectively), while deaths due to axi-cel–related AEs remained 4.0% at each timepoint. Among patients with a best response of a CR, the cumulative incidence rates of death at Month 24 and Month 72 were 20.3% and 27.2% for progression, 3.4% (both timepoints) for axi-cel–related AEs, and 0% and 10.4% for other reasons, respectively. Those who had a CR at Months 12 and 24 had a 72-month cumulative incidence of death of 18.4% and 14.2%, respectively (Table 2). Additionally, those who had a CR at 12 and 24 months had 72-month KM-estimated DSS of 94.4% and 100%, respectively. Among those with a partial response at Week 4 and Month 3, 72-month cumulative incidences of death were 72.7% and 50.0%, respectively.

Conclusion: In this post hoc analysis of ZUMA-1 with up to 6 years of follow-up, axi-cel had long-term LREFS in a substantial proportion of patients, with a 5-year rate of 34% (57% among patients who achieved a CR). Additionally, a CR at 12 and 24 months may be predictive of extended OS. Among patients who achieved a CR, deaths due to axi-cel–related AEs mainly occurred before Month 24. These results demonstrate the curative potential of axi-cel in a subset of patients with R/R LBCL.