Outpatient Brexucabtagene Autoleucel in B-Cell Acute Lymphoblastic Leukemia and Mantle Cell Lymphoma

Brexucabtagene autoleucel (brexu-cel) is an effective therapeutic option for relapsed/refractory (R/R) mantle cell lymphoma (MCL) and B-cell acute lymphoblastic leukemia (ALL) based on the results of the ZUMA-2 and ZUMA-3 trials, respectively [1, 2]. Several reports describe the feasibility of outpatient administration of chimeric antigen receptor (CAR)-T cell therapy [3]. However, there is limited data supporting outpatient administration of brexu-cel [4, 5], a CAR product associated with higher reported rates of toxicity compared to others in clinical trials [2]. Here, we report our experience with outpatient administration of brexu-cel infusion as compared to inpatient.

We conducted a retrospective study of adults who received commercial brexu-cel for B-ALL and MCL at City of Hope (COH) during November 1, 2020–March 31, 2024. The study was approved by the COH institutional review board (#23837). Prior to implementing our COH outpatient-CAR program in July 2022, all patients received CAR-T infusion as inpatient. Since July 2022, patients received CAR-T as outpatient but could still receive it as inpatient based on the treating physician's discretion. Tables S1–S3 describe outpatient procedures for brexu-cel selection and administration. In this study, patients were divided into two groups, the inpatient group and the outpatient group, based on their initial brexu-cel infusion setting assignment. To improve the comparability between groups, we applied propensity score (PS) nearest matching method (ratio = 1, caliper = 0.1) with ECOG performance status, Severe4 comorbidity index [6], and disease burden as the matching variables. These were selected because of their relationship with the tendency of receiving CAR-T inpatient and worse post-infusion prognosis based on clinical experience. The presence of severe comorbidities was defined as a Severe4 score ≥ 1 [6]. Disease burden was categorized as bulky for bone marrow blasts ≥ 5% and/or the presence of extramedullary or CNS disease for B-ALL or intermediate/high MIPIb for MCL at the time of LD. Outcome analyses were performed on the post-matching cohort. The primary endpoint was cumulative incidence of 100-day non-relapse mortality (NRM). Secondary endpoints included best complete response (CR) rate, overall survival (OS), progression-free survival (PFS), incidence of cytokine release syndrome (CRS), immune effector-cell associated neurotoxicity syndrome (ICANS), incidence and timing of inpatient admission in the outpatient group, grade ≥ 3 infections within the first 100 days, grade ≥ 3 cytopenias, and hospital resource utilization (HRU). More statistical method details are described in Table S4.

Sixty-six patients received brexu-cel, including 35 inpatient and 31 outpatient per the initial treatment setting assignment. After PS matching, the final sample included in the outcome analyses consisted of 52 patients, 26 in each group (Figure S1). The inpatient group consisted of 17 B-ALL and 9 MCL patients, while the outpatient group comprised 19 B-ALL and 7 MCL patients. The distribution of ECOG performance status, Severe4 comorbidities, and disease burden at LD was balanced (Table S5).

The median follow-up among survivors was 5.5 months (range: 1.5–31.1) post-brexu-cel infusion. Three in the inpatient group experienced NRM (one on day 9 due to immune effector cell-associated hemophagocytic lymphohistiocytosis, and two on days 171, and 584 due to infection). One in the outpatient group experienced NRM on day 19 due to infection. One out of 26 in each group experienced NRM within 100 days, and the 100-day NRM was 3.8% (95% CI: 0.26%–17%) in both groups (p = 1) (Figure 1A). The best CR/CRi rate for B-ALL was 100% (95% CI: 79%–100%) in the inpatient and 89% (95% CI: 67%–99%) in the outpatient group, and the best CR rate for MCL was 67% (95% CI: 30%–93%) in the inpatient group and 83% (95% CI: 36%–100%) in the outpatient group (Figure 1B,C). The 6-month PFS was 71% (95% CI: 45%–86%) in the inpatient group and 60% (95% CI: 36%–77%) in the outpatient group (p = 0.3) (Figure 1D). The 6-month OS was 75% (95% CI: 49%–89%) in the inpatient group and 87% (95% CI: 63%–96%) in the outpatient group (p = 0.5) (Figure 1E).

No statistically significant differences were observed between the inpatient and outpatient groups in CRS, 81% versus 85% for any grade (p > 0.9) and 12% versus 8% for grade ≥ 3 CRS, respectively (p = 0.7). Similarly, no differences were observed in ICANS: 62% versus 50% for any grade (p = 0.4) and 27% versus 27% for grade ≥ 3 (p > 0.9), respectively (Table S6). Twenty-seven percent in the inpatient and 31% in the outpatient group received prophylactic steroids (p = 0.8). Notably, more patients in the outpatient group received prophylactic anakinra than patients in the inpatient group (31% vs. 8%, p = 0.035) which might be a result of our institutional standard change starting 11/2023 [7]. No statistically significant differences were observed in the incidences of early (100% vs. 85%, p = 0.11), prolonged (60% vs. 42%, p = 0.2), or late (40% vs. 36%, p = 0.8) grade ≥ 3 cytopenias between the inpatient and outpatient groups (Table S7). Thirty-one percent (n = 8) in the inpatient and 19% (n = 5) in the outpatient group experienced grade ≥ 3 infections within the first 100 days (p = 0.3), among which 88% (n = 7) in the inpatient group and 60% (n = 3) in the outpatient group were nosocomial infections (Table S8).

The median length of hospitalization during the first 100 days since LD initiation was 23 days (range: 13–48) in the inpatient and 9 days (range: 0–52) in the outpatient group (p < 0.001). All patients in the inpatient group (per initial treatment setting assignment) received brexu-cel infusion inpatient, while 22 (85%) patients in the outpatient group were admitted after LD initiation, including 3 for brexu-cel infusion (2 for fever during LD and 1 for pain control during LD due to underlying malignancy) at 0, 1, and 1 day from LD initiation, respectively, and 19 for brexu-cel-related toxicity (17 for CRS, 1 for hypotension due to dehydration, and 1 for dyspnea) at a median of 9 days (range: 6–17) from LD initiation. Thirty-five percent in the inpatient group and 50% in the outpatient group received packed red blood cells transfusions, among which a median of 4 (range: 2–30) and 2 (range: 1–10) units were used (p = 0.011). Eighty-five percent in the inpatient group and 65% in the outpatient group were given granulocyte colony stimulating factor, among which a median of 6 (range: 1–34) and 2 (range: 1–17) doses were used (p = 0.014). More HRU details are available in Table S9.

Our results suggest that outpatient brexu-cel administration is feasible and does not compromise efficacy or treatment safety, with comparable rates of NRM at 100 days and toxicity measures compared to the inpatient administration. Nonetheless, outpatient administration was associated with lower HRU, as the outpatient group spent a median of 14 fewer days in the hospital and received fewer G-CSF doses in the first 100 days. To our knowledge, the three largest studies to date showing the practicality of outpatient administration were conducted by Linhares et al. in the OUTREACH study, Ly et al. at Johns Hopkins, and Dholaria et al. at Vanderbilt [4, 5]. The OUTREACH study was a phase 2 collaborative study between community sites that successfully demonstrated the feasibility of lisocabtagene maraleucel with close monitoring [8]. At Johns Hopkins, 47 patients received outpatient-CAR; however, only eight patients received brexu-cel, and only three had B-ALL. At Vanderbilt, 13 patients received outpatient-CAR, but only 4 received brexu-cel. Moreover, neither study provided a matched comparison between patients receiving inpatient versus outpatient as presented in this study. Finally, data on outpatient brexu-cel administration in the B-ALL adult only population has been limited [3]. B-ALL comprised the majority (69%) of our study population.

While our report presents novel findings that may serve as a foundation for future studies to improve the delivery of outpatient-CAR, we note several limitations. These include the retrospective design, the single-center nature, and small sample size. We also highlight the inherent differences between patients who were assigned to receive brexu-cel infusion as inpatient versus outpatient in the latter period after we launched the COH outpatient-CAR program, as patients with worse baseline disease characteristics and worse prognosis tended to be assigned as inpatient. We attempted to mitigate this limitation by applying the PS nearest matching method, so the two groups in comparison were balanced regarding ECOG performance status, Severe4 comorbidities, and disease burden at LD. On the other hand, the PS nearest matching method discarded unmatched units or less similar matches (e.g., 14 out of 66 patients were unmatched in the current study), which reduced the sample size and potentially statistical power, but also reduced the generalizability of the study findings as the matched sample may no longer represent the broader population. In addition, residual confounding may exist. For instance, caregiver availability also influenced the decision for outpatient-CAR, but this data is not readily available. Finally, although patients spent fewer days hospitalized in the outpatient group, we were not able to measure other important outcomes such as caregiver cost and financial burden in a retrospective fashion. Further investigation into these matters is needed in subsequent studies.

In conclusion, initiating LD and brexu-cel in the outpatient setting followed by expectant monitoring appears safe, feasible, and produces similar efficacy as inpatient delivery without an increase in NRM or toxicities. While 85% of these patients were eventually admitted, HRU was reduced. Future studies with a larger sample size and longer follow-up are needed to confirm our findings.

This study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki, and the protocol was approved by the City of Hope Review Board. All participants provided informed consent prior to participation.

J.H.B.—received fees for consulting or advisory role from Kite/Gilead; and has received research funding from Kite/Gilead, Janssen, Regeneron, CARGO, and Genentech/Roche. I.A.—Ad Board KiTE, Autolus, JAZZ, Takeda, Pfizer, adaptive, Amgen, Syndax. A.D.—AstraZeneca: Consultancy, Research Funding; AbbVie: Consultancy; BeiGene: Consultancy; Genentech: Consultancy; Nurix: Consultancy, Research Funding; MorphoSys: Consultancy; Incyte: Consultancy; TG Therapeutics: Consultancy, Research Funding; Bayer: Consultancy, Research Funding; Takeda: Research Funding; MEI Pharma: Research Funding; ADCT: Consultancy; Bristol Meyers Squibb: Consultancy, Research Funding; Cyclacel: Research Funding; GenMab: Consultancy, Research Funding; Janssen: Consultancy. T.J.P.—Pharmacyclics: Consultancy; AbbVie: Research Funding; Lymphoma & Myeloma Connect: Honoraria; ADC Therapeutics: Consultancy; TG Therapeutics: Consultancy; Genmab: Consultancy; Celgene: Consultancy; Kite/Gilead: Consultancy; Curis: Consultancy; Gilead Sciences: Consultancy; Bayer: Consultancy, Research Funding; Genentech: Consultancy; Incyte: Consultancy; Pharmacyclics/Janssen: Research Funding; Seattle Genetics: Consultancy, Honoraria. M.M.—Novartis: Consultancy; Synethkine: Consultancy; SeaGen: Consultancy, Speakers Bureau; ADC Therapeutics: Consultancy; AstraZeneca: Consultancy; BMS: Research Funding; Incyte: Research Funding; Beigene: Research Funding; Genentech: Research Funding. P.B.K.—Ad Board/Consulting Fees: Ascentage, Daiichi Sankyo, BMS, Novartis, Takeda, Safety Review Committee: Treadwell Therapeutics, Research Support: Bayer. L.E.B.—AstraZeneca, Mustang Therapeutics, Merck: Research Funding; ADC Therapeutics, AstraZeneca, AbbVie, F. Hoffmann-La Roche Ltd., Genentech Inc., Genmab, Janssen, Regeneron: Consultancy. A.A.—Seagen: Consultancy; Kite, a Gilead Company: Consultancy. The rest of the authors have no disclosures to report.