Outcomes of Chimeric Antigen Receptor (CAR) T-Cell Therapy in Extranodal (EN) B-Cell Non-Hodgkin Lymphoma (NHL): Results from a Multicenter Analysis

Introduction: EN lymphoma is defined by involvement of sites outside of the lymph nodes, spleen, thymus and oropharyngeal lymphoid tissue. Certain EN sites are associated with a worse prognosis in diffuse large B-cell lymphoma (DLBCL) (Castillo et al. AJH 2014). Although CAR-T has become standard in treating patients (pts) with relapsed/refractory (R/R) LBCL, the outcomes of CAR-T in pts with EN disease are not well established.

Methods: We performed a retrospective multi-center analysis of pts with R/R aggressive B-cell NHL who had EN disease and received treatment with anti-CD19 CAR-T. EN disease could be present at diagnosis and/or at time of CAR-T. The primary objective was to determine the overall survival (OS), while secondary objectives included overall and complete response rates (ORR/CRR), progression free survival (PFS), and CAR-T toxicities. Survival outcomes were estimated using the Kaplan-Meier method and compared between groups using the log-rank test.

Results: 282 pts from 7 institutions were included in the analysis. Median age at diagnosis was 60 years (range:18-85), with a 3:2 male to female ratio. 60% (n=169) of pts had DLBCL histology, 14% (n=40) had high-grade B-cell lymphoma, 15% (n=43) had transformed indolent lymphoma, and 5% had follicular lymphoma. 42% (85 of 201 available) of pts had primary refractory disease. Median number of prior lines of therapy was 3 (range:0-9). 53% (n=150) of pts received bridging therapy prior to CAR-T. 99% (278 of 281 available) of pts received a commercial CD19 CAR (43% axicabtagene ciloleucel, 30% tisagenlecleucel, 25% lisocabtagene maraleucel, 2% brexucabtagene autoleucel), while the remaining 1% (n=3) received a bispecific or allogeneic CAR. Median follow up time was 3 years (range:0.08-7).

Median OS was 2.41 [95% CI: 1.52, 4.3] years from the day of CAR-T infusion. Survival outcomes were similar between pts with primary EN disease (i.e. the lymphoma originated in an EN site) vs secondary EN disease. Presence of EN disease at CAR-T did not predict survival (p=0.23). 13% (26 of 204 available) of pts did not have EN disease at diagnosis, but later developed EN disease at time of CAR-T. These pts had poorer survival, with median OS 0.91 [95% CI: 0.41, Not Reached (NR)] vs 3.03 [95% CI: 2.14, NR] years for patients with EN disease at diagnosis (p=0.027). Baseline characteristics were similar between the two groups including age, prior lines of therapy, cell-of-origin, and MYC alterations. Other factors associated with poorer outcomes included use of bridging therapy with median OS 1.33 [95% CI: 0.97. 2.14] vs 5.41 [95% CI: 2.84, NR] years for pts who did not receive bridging (p≤0.01), and refractory disease at time of CAR-T with median OS 1.54 [95% CI: 1.10, 2.84] vs NR [95% CI: 5.41, NR] for patients without refractory disease (p≤0.01).

Median PFS was 0.48 [95% CI: 0.30, 0.85] years, 1-year PFS was 41% [95% CI: 35, 47] and 3-year PFS was 33% [95% CI: 28, 39]. ORR was 65% [95% CI: 59, 71] and CRR was 42% [95% CI: 36, 48] at first response assessment. Cytokine release syndrome (CRS) and neurotoxicity occurred in 73% (205 of 281 available) and 38% (108 of 281 available) of pts, respectively. Rates of grade ≥ 3 of CRS and neurotoxicity were 6% (n=17) and 19% (n=54), respectively. Median durations of CRS and neurotoxicity were 3 days (range:0-57) and 5 days (range:0-188), respectively. Transfusion support (PRBC and platelets) and filgrastim beyond the first 30 days post CAR-T was needed in 18% (50 of 272 available), 20% (54 of 272 available), and 34% (93 of 272 available) of pts, respectively. Secondary malignancy occurred in 2% (5 of 204 available) of pts, including one case of T-cell large granular lymphocytic leukemia.

Conclusion:

In this large multi-center cohort study of EN LBCL undergoing CAR-T, we found no significant difference in OS based on the presence or absence of EN disease at CAR-T or whether it was primary or secondary EN lymphoma. Pts without EN disease at diagnosis who later developed EN disease at time of CAR-T had worse OS. OS, PFS, and rate of grade 3 CRS and neurotoxicity in our study (EN B-cell NHL) were comparable to the real-world outcomes of CAR-T in R/R LBCL (Jacobson et al. TCT 2024). Further analysis including comparison of single-site versus multi-site EN disease as well as multivariate analysis of factors with significant impact on survival is ongoing.