Paper Details

ABSTRACT

Introduction: CD19 CAR T-cell therapy with axicabtagene ciloleucel (axi-cel) may produce long term remissions in patients with R/R LBCL. Radiation therapy is often delivered as bridging therapy to reduce tumor burden prior to CAR T-cell infusion, but the optimal dose and strategy remain poorly defined. Previously we identified that lesions receiving lower dose bridging radiation are at risk of in-field relapse following CAR T-cell therapy. We therefore conducted a prospective clinical trial that aims to deliver a definitive dose of radiation to all identified sites of disease as bridging prior to CAR T-cell therapy (NCT06104592/CABI – “comprehensive ablative bridging irradiation”).

Methods: This single center prospective investigator-initiated clinical trial enrolls patients receiving standard-of-care axi-cel for R/R LBCL and variants. Subjects are required to have PET-avid disease such that all active disease sites can safely receive radiation. Subjects are treated with 37.5 Gy in 15 fractions to all PET-avid sites, delivered between apheresis and the start of lymphodepletion (LD). Systemic bridging therapy, with the exception of corticosteroids, is prohibited. The primary endpoints are progression-free survival (PFS) and adverse events. Responses were assessed by Lugano criteria. Prophylaxis with dexamethasone 10 mg PO on days 0, 1 and 2 was at the discretion of the treating physician. CAR T-cell expansion was by flow cytometry. Peripheral blood phenotyping was by single cell CITEseq.  

Results: As of submission, 24 subjects were enrolled. One subject required systemic chemotherapy due to progression while receiving radiation and came off study before axi-cel infusion. Among the 23 subjects who received axi-cel, baseline clinical characteristics include a median age 58 (range 25-76; 35% age 65 or older), 13% double hit, 52% with elevated LDH, 30% transformed from indolent lymphoma, and 65% treated in 2nd line. Upon radiation simulation, 35% had advanced stage disease and 73% had a bulky lesion ≥ 7.5 cm. Median follow-up from axi-cel is 189 days (range 4-872).

First, restaging was performed at the completion of radiation prior to the start of LD. Among 23 subjects completing radiation, ORR and CR to radiation alone was 78% and 26%, respectively. All PD events during radiation bridging occurred outside of the radiation fields. Then, disease response was assessed following axi-cel. Among 15 subjects evaluable at a minimum of 6 months follow-up, best ORR following axi-cel was 87% with a best response of CR in 73%, and 11/15 (73%) remain alive without progression.

Toxicity post-axi-cel is evaluable in 22 subjects. Any grade CRS occurred in 100% with maximum grades 1, 2, and 3 in 13/22 (59%), 7/22 (32%) and 2/22 (9%), respectively. Any grade ICANS occurred in 15/22 (68%), with maximum grades 1-4 in 5/22 (18%), 1/22 (5%), 9/22 (41%), and 1/22 (5%). Among subjects who received dexamethasone prophylaxis, grade 3+ ICANS was noted in 3/9 (33%) compared to 7/13 (54%) in patients without prophylaxis. No cases of secondary malignancy or NRM have been identified to date. AEs specifically attributed to radiation were either local to the tumor/radiation site (i.e. hoarseness) or were lymphopenias with possible attribution, and radiation was generally well-tolerated.

Correlative studies on the first 13 subjects identify a peak CAR T-cell expansion more than 2-fold higher than a historical control group receiving axi-cel as standard-of-care (n=150). Single cell CITEseq was performed on peripheral blood mononuclear cells prior to radiation, following radiation, and at day +7 following axi-cel in trial subjects (n=6) and compared to a carefully selected control group of patients (n=9). Overall, myeloid cell inflammation was low following radiation, but expanding CAR+ cells demonstrated a greater CD4/Th1 predominance in CABI subjects.

Conclusions: Comprehensive ablative bridging irradiation is effective to reduce tumor burden prior to CAR T-cell therapy. Following radiation, CAR T-cell expansion is increased, including Th1 subsets. By decreasing tumor burden and increasing expansion, our data suggests that we have achieved a favourable in vivo effector:target ratio resulting in high response rates and durable remissions. With increased expansion, high grade ICANS was frequent. Longer follow-up will determine if the trial has met its primary PFS endpoint and further analysis is planned to elucidate the mechanism of increased expansion.