An updated UK real-world analysis of factors affecting outcome following brexucabtagene autoleucel for relapsed or refractory adult acute lymphoblastic leukemia Free

Background The UK Adult ALL CAR-T panel - established May 2023 - promotes equity of access and consistent eligibility assessment for nationally commissioned CAR-T therapies. This collaborative platform supports robust collection and assessment of real-world data (RWD). We previously reported RWD around use, toxicity and outcomes of adult ALL patients (pts) approved for brexu-cel therapy in the 1^st 12mo of nationally funded access. Here we provide an updated analysis of UK CAR-T use for adult ALL and evaluate factors influencing outcome following brexu-cel infusion.

Methods All pts approved for brexu-cel by panel between May 2023-April 2025 and with sufficient data were included. Retrospective data from 20 accredited immune effector cell centres were anonymised, collected on a standardised proforma and collated to support RWD analysis.

Results Of 134 pts assessed, 92 were eligible for brexu-cel by national criteria (ITT cohort). Indications were post-allo-SCT relapse (60; 65.2%); 1^st relapse within 12mo of CR1 (11; 12.0%); 2^nd/subsequent relapse (7; 7.6%); relapse in non-SCT candidate (4; 4.4%); refractory disease (10; 10.8%).

Median age (ITT) was 50.5y (IQR 40.5-59y), with 21 (22.8%) >60y. One-third (30; 33.3%) had tyrosine kinase activating fusions; 17 (18.9%) and 7 (7.8%) had v high and high-risk cytogenetics, respectively (per revised UKALL14 genetic risk). Median prior therapy lines were 2 (1-5), including prior allo-SCT, blinatumomab and inotuzumab in 61 (66.3%), 26 (28.3%) and 21 (22.8%), respectively.

Of approved cases, 83 (88%) underwent apheresis (2 failed apheresis; 1 failed manufacture). Ultimately, 68 (73.9% of approved; 85% with product) underwent brexu-cel infusion. Median time from approval to infusion was 57.5d (IQR 43.5-74.5); 95.6% received bridging (84.6% 1 line only). At infusion 48 (73.8%) were in CR, with 53 (77.9%) having BM blasts <5% and 16 (24.6%) confirmed MRD negative.

At D30 and D90 post-infusion, 58 (95.1%) and 49 (98%) were in CR, with 52 (85.2%) and 41 (82%) MRD negative, respectively. Of responders, 79% (95% CI 65.1-87.9) maintained response at 6mo. Five (7.4%) went on to allo-SCT. Grade 3+ CRS and ICANS occurred in 2.9% and 23.5% of cases, respectively.

With median 12mo FU [95% CI 6.7-14.5] from infusion, 6mo relapse-free (RFS) and overall survival (OS) was 73.9% and 85.5%, respectively. Estimated median RFS was 17.1mo (95% CI 9.9-NR); median OS was not reached.

Disease burden at infusion predicted outcome, with est median RFS for pts with ≤5% and >5% blasts of 20.1mo [95% CI 13.5-NR] and 4.2mo [95% CI 0.0-9.9] respectively (log-rank p<0.001). MRD-negativity at infusion was associated with improved RFS (median NR v 11.7mo [95% CI 7.3-20.0]; est 12mo RFS 90.0% [95% CI 47.3-98.5] v 50.0% [95% CI 33.6-64.3]; log-rank p=0.014). Prior EMD predicted worse RFS (median 4.7mo [95% CI 0-13.5] v 19.2mo [95% CI 11.3-NR]; log-rank p=0.004) and OS (median 9.0mo [95% 0.9-NR] v NR; log-rank p=0.021).

Prior blinatumomab was associated with worse RFS (median 11.3mo [95% CI 3.3-NR] v 19.2mo [95% CI 9.9-NR]; log-rank p=0.046) and OS (median 14.2mo [95% CI 3.6-NR] v NR; log-rank p=0.042). OS was not impacted by prior inotuzumab exposure (any ino p=0.375; ino as bridging p=0.573). Relapse within 6mo of prior allo-SCT predicted poor outcomes post CAR-T (v relapse ≥6mo post-SCT: RFS p<0.001; OS p=0.015).

In univariate Cox regression models, prior EMD (RFS HR 3.18 [1.38–7.33]; OS HR 3.22 [1.13–9.21]), >5% blasts (RFS HR 5.17 [2.35–11.39]; OS HR 4.04 [1.55–10.54]), and post-SCT relapse within 6mo (RFS HR 6.51 [2.16-19.63]; OS HR 4.70 [1.20-18.41]) predicted inferior outcomes. Adjusted for age, all retained significance (EMD: RFS HR 3.21 [1.39–7.43]; OS HR 3.04 [1.05–8.79]; >5% blasts: RFS HR 5.31 [2.39–11.82]; OS HR 4.30 [1.63–11.47]; post-SCT relapse <6mo: RFS HR 6.53 [2.12-20.08]; OS HR 4.50 [1.07-18.93]).

Conclusion Here we present one of the largest RWD analyses of brexu-cel use in adult ALL. Brexu-cel was deliverable, tolerable, and efficacious, in a cohort notably differing from that trialled in ZUMA-3, including a higher proportion post-SCT, older median age, and more in CR at infusion. Outcomes were strongly influenced by disease burden at infusion and EMD status: the latter may be relevant in the frontline blinatumomab era (in which patterns of relapse may differ). Early post-SCT relapse (<6mo) was also associated with worse CAR-T outcomes, highlighting need for risk-adapted strategies in this subgroup.