Outcomes of brexucabtagene autoleucel in patients with relapsed/refractory acute lymphoblastic leukemia with CNS involvement Open Access

Patients with relapsed and refractory (R/R) B-cell acute lymphoblastic leukemia (B-ALL) with central nervous system (CNS) involvement (CNS B-ALL) experience poor clinical outcomes and have limited efficacious treatment options.¹ Although the use of prophylactic intrathecal (IT) chemotherapy throughout frontline ALL therapy has decreased the risk of CNS relapse, it is still estimated to occur in up to 15% of patients with B-ALL.^2-4 Although effective in treating systemic R/R B-ALL, targeted therapies, such as blinatumomab and inotuzumab ozogamicin, have insufficient evidence to support efficacy in treating CNS B-ALL.^4,5 Additionally, early pivotal CD19-targeting chimeric antigen receptor (CAR) T-cell (CD19 CAR T cell) clinical trials excluded patients with B-ALL with active CNS disease.^6,7 Nonetheless, retrospective studies in pediatric and young adult patient populations have demonstrated the efficacy and safety of some commercial and investigational CD19 CAR T cells in CNS B-ALL.^8-11 For instance, a study of pooled clinical trial data demonstrated that among 55 children and adolescents receiving second-generation CD19 CAR T cells, 94% of patients achieved a complete response of CNS B-ALL.⁹ 

Brexucabtagene autoleucel (brexu-cel) is the first commercially approved CD19 CAR T-cell product for adults with R/R B-ALL, based on results of the ZUMA-3 trial.⁸ ZUMA-3 was an open-label single-arm phase 2 clinical trial demonstrating a 71% complete remission (CR) rate in patients with R/R B-ALL. The ZUMA-3 trial allowed asymptomatic CNS-2 disease (active CNS disease with detectable blasts in the cerebrospinal fluid [CSF] analysis but with <5 white blood cells [WBCs] per μL) but excluded patients with symptomatic CNS-2 or CNS-3 disease (patients with detectable blasts with >5 WBCs per μL and/or clinical signs/symptoms). Only 5 patients had CNS-2 disease at the screening phase but subsequently had no CNS disease before lymphodepletion because of CNS-directed bridging chemotherapy. Brexu-cel has also demonstrated promising efficacy in mantle cell lymphoma with secondary CNS disease.¹² However, data regarding the efficacy and safety of brexu-cel in adult patients with CNS B-ALL are lacking. In this retrospective, multi-institutional study, we report the efficacy and safety of brexu-cel in R/R B-ALL with CNS involvement.

This multicenter retrospective study uses the Real-World Outcomes Collaborative for CAR T in ALL (ROCCA) consortium database. The consortium includes data from 31 US-based academic and community cellular therapy centers. Each participating institution obtained institutional review board approval and submitted the deidentified data to a Health Insurance Portability and Accountability Act–compliant electronic database maintained by Stanford University.

Data were collected from participating ROCCA sites among adult patients (aged >18 years) with R/R B-ALL who underwent apheresis and brexu-cel infusion between October 2021 and August 2023. For this analysis, patients with CNS involvement with or without medullary disease at the time of preapheresis disease assessment were identified as the CNS group. Patients without reported CNS involvement at the time of preapheresis disease assessment were identified as the non-CNS comparator group. Patients in the non-CNS group could have had previous CNS disease but were in CNS remission at the time of preapheresis disease assessment. The ROCCA database classified CNS disease according to the Children’s Oncology Group classification¹³: CNS-1 indicates patients with no identifiable CNS disease, CNS-2 indicates active CNS disease with detectable blasts in the CSF analysis but with <5 WBCs per μL, and CNS-3 indicates patients with detectable blasts with of >5 WBCs per μL and/or clinical signs/symptoms. Two investigators (I.N.M. and G.W.R.) communicated with local investigators at participating sites to confirm CNS classification and collect additional information regarding patients with CNS disease, including patterns of CNS involvement and relapses.

Data were collected for disease response ∼1 month (day +28) after brexu-cel infusion. Medullary response was assessed by bone marrow (BM) evaluation including morphological assessment and flow cytometry. When available, measurable residual disease (MRD) assessment was reported and was performed based on the practices of the participating institutions with different used methodologies, including next-generation sequencing, flow cytometry, and quantitative reverse transcription polymerase chain reaction for BCR:ABL1 fusion in Philadelphia chromosome (Ph)⁺ ALL. Ph-like status was determined using cytogenetics and/or molecular testing based on the practices of the participating institutions. CNS status was reevaluated after bridging therapy and/or brexu-cel infusion through CSF analysis, imaging, and/or assessment of clinical signs and symptoms at the discretion of the treating team.

CR was defined as <5% blasts in the BM evaluation, and absence of peripheral blood circulating blasts and extramedullary disease. CNS response was defined as partial if improved from CNS-3 to CNS-2, and complete if CNS-1 status was achieved. Cytokine release syndrome (CRS) and immune effector cell–associated neurotoxicity syndrome (ICANS) grading were according to the American Society for Transplantation and Cellular Therapy consensus criteria.¹⁴ 

The efficacy outcomes of interest were CNS response rate, progression-free survival (PFS), overall survival (OS), and cumulative incidence of relapse. Safety outcomes of interest were CRS, ICANS, and nonrelapse mortality rates.

Data were reported as counts and percentages for categorical data. Continuous data were reported as medians and ranges. Comparisons between groups were done using the χ² test or the Fisher exact test, as appropriate. PFS was measured from time of brexu-cel infusion to time of disease progression, lack of response, relapse, or death. OS was measured from time of brexu-cel infusion to time of death. Both PFS and OS were estimated using the Kaplan-Meier method. The difference between groups was estimated using log-rank test. Kaplan-Meier curves were censored at the time of the last follow-up. Statistical analysis was performed using R (version 4.2.3) software. A P value <.05 was considered statistically significant.

Of 189 patients who received infusion with brexu-cel as a standard-of-care therapy in the ROCCA consortium database, 31 patients (16.4%) had active CNS disease (CNS-2 or CNS-3) at the preapheresis disease assessment and were included as the CNS group, whereas 158 (83.6%) had no reported CNS disease and were included as the non-CNS group.

Among the CNS group, 15 (48.4%) had CNS-2 B-ALL, and 16 (51.6%) had CNS-3 B-ALL. Of 16 patients with CNS-3 disease, 7 had radiological evidence and/or symptoms at presentation, 7 had CSF-only disease, and 2 had unknown presentations. A total of 17 patients (54.8%) had CNS disease without morphological medullary disease. Among these patients, 11 had MRD-positive or unknown CR, and 6 had MRD-negative CR. The other 14 patients had both morphological medullary and CNS disease at preapheresis assessment.

Among the CNS group, the median age was 46.5 years (range, 24-76), and more than half of patients (58.1%) were male. More patients had Ph⁺ disease (45.2%) than Ph^– disease (35.5%) and Ph-like disease (18.4%). In patients with Ph^– disease, 7 had a normal karyotype, 3 had cytogenetic abnormalities (hyperdiploidy [n = 1], t(v;14q32)/immunoglobulin H [n = 1], and other abnormalities [n = 1]), and 1 with unknown status. Additionally, 4 patients had molecular mutations reported: TP53, SLC16A3-NOTCH1, TP53/DNMT3A, and PPM1D/CDKN2A/MLL3; each respectively identified in 1 patient. Before brexu-cel infusion, patients received a median of 4 lines of therapy (range, 2-12): 18 patients (58.1%) underwent allogeneic hematopoietic stem cell transplant (allo-HSCT), 19 (61.3%) received blinatumomab, 12 (38.7%) received inotuzumab ozogamicin, and 11 (35.5%) had previously received CNS-directed radiation (ie, craniospinal, orbital, or cranial). Among patients who underwent previous allo-HSCT, 12 patients received total body irradiation as part of their conditioning regimen.

Relative to the non-CNS disease group, the CNS disease group included greater proportions of patients with Ph⁺ ALL, history of allo-HSCT, and medullary CR at preapheresis assessment. Table 1 summarizes the patients’ general characteristics in both the CNS and non-CNS groups.

Among the CNS group, 27 (87.1%) patients received bridging therapy: 17 received combination systemic (chemotherapy, tyrosine kinase inhibitors [TKIs] and/or immunotherapy) and IT therapy, 7 received IT chemotherapy only, 1 received IT chemotherapy and radiation, and 2 received systemic therapy only. Ten patients received TKI as part of their bridging therapy (8 patients received ponatinib and 2 received dasatinib). Four patients received no bridging therapy. Seventeen patients had no CNS disease assessment after bridging therapy, before lymphodepletion. Among 10 patients with postbridging CNS disease assessment, 8 patients (80.0%) cleared their CNS disease, and 2 patients did not respond to bridging therapy (1 with CNS-2 status and 1 with CNS-3 status). Among the non-CNS group, bridging was used in 103 patients (65.2%), and included chemotherapy (66.0%), immunotherapy (28.2%), TKIs (14.6%), steroids (12.6%), and IT chemotherapy (16.5%). A total of 34 (33.0%) patients received combinatorial therapy.

Twenty-three patients (74.2%) in the CNS group developed CRS (1 with grade 3-4; 3.2% overall rate of grade 3-4), and 20 patients (64.5%) developed ICANS (11 with grade 3-4; 35.5% overall rate of grade 3-4). Rates of CRS and ICANS were similar in the non-CNS group, with 84.8% developing CRS (12.0% grade 3-4) and 54.4% developing ICANS (30% grade 3-4). There was no statistically significant difference in rates of all-grade or grade 3/4 CRS and ICANS between CNS and non-CNS groups (Table 2).

Among patients with CNS-3 disease, CRS occurred in 62.5% compared with 86.7% in the CNS-2 group (P = .22). ICANS occurred in 68.8% of patients with CNS-3 disease and 60.0% in patients with CNS-2 disease (P = .72). A high rate of grade 3/4 ICANS (43.7%) was observed in patients with CNS-3 disease (26.7% in CNS-2). Of note, 6 of 7 patients (86%) with parenchymal or symptomatic CNS involvement developed grade 3/4 ICANS but none died secondary to ICANS. Higher rates of ICANS were also observed in patients with CNS disease without morphological medullary disease than patients with both CNS and medullary disease (76.5% vs 53.8%). Overall, a higher proportion of patients with medullary disease developed grade 3/4 ICANS than patients without medullary disease at time of apheresis (46.1% vs 31.2%). However, among patients with CNS-3 disease rates of grade 3/4 ICANS were similar among patients with and without medullary disease (44.4% vs 42.9%). Table 3 summarizes CRS and ICANS rates by CNS disease grade and presence of radiological evidence and/or symptoms.

Among 31 patients with CNS disease, 2 patients died secondary to ICANS or CRS. One patient with CNS-2 disease developed ICANS grade 4 necessitating an intensive care admission, and patient failed steroids, anakinra, and tocilizumab; the patient passed away on day 91 after infusion. Additionally, a patient with CNS-3 disease developed refractory grade 4 CRS. The patient was treated with dexamethasone, tocilizumab, and anakinra but had no response and eventually died on day 10 after brexu-cel.

In the CNS group, 28 of 30 evaluable patients (93.3%) had a CR in the marrow after brexu-cel, with 25 patients (83.3%) achieving MRD-negative remission. Similarly, in the non-CNS group, the rate of CR/CR with incomplete count recovery in the marrow was 89.1%, with a 68.1% MRD-negative remission rate. Figure 1A summarizes the medullary disease status at preapheresis and postinfusion assessments.

Of 31 patients in the CNS group, 24 had CNS disease assessment after brexu-cel. Of 24 patients, 21 (87.5%) patients achieved CNS-1 after brexu-cel, whereas 3 patients continued to have active CNS disease (Figure 1B). All 21 patients with CNS-1 disease also achieved CR by BM evaluation (19 with MRD-negative remission, and 2 without MRD testing). Among the patients with CNS-3 disease, 12 (75.0%) achieved CNS-1, 2 patients had no response, and 3 patients had no reported CNS evaluation. Three of 6 patients were noted to clear CNS disease with brexu-cel alone, in the absence of effective bridging therapy.

The median follow-up was 13.8 months for the entire cohort. In the CNS group, 12 patients had R/R disease after brexu-cel, with a median time to relapse of 100 days. Of 12 relapses, 5 occurred in the CNS with or without systemic disease. The remaining relapses occurred in the BM (n = 2), were extramedullary (n = 1), or had no specified site of relapse documented (n = 4). The cumulative incidence of relapse at 1 year was 39% in the CNS group compared with 35% in patients without CNS disease (P = .54). Of 12 patients who relapsed, 7 patients were reported to receive subsequent therapies of varying combinations. Among these patients, 4 patients received chemotherapy, 3 received inotuzumab, 2 received TKIs, 1 received blinatumomab, and 1 received radiotherapy. One patient was enrolled on a trial investigating the use of a CD22-targeting antibody drug conjugate.

A total of 8 patients (25.8%) in the CNS group and 56 patients (35.4%) in the non-CNS group died. In the CNS group, 4 of 8 patients died in the absence of leukemia relapse; 2 died because of CRS and ICANS, and 2 died because of infections and multiorgan failure. The 6- and 12-months PFS in the CNS group were 57% (95% confidence interval [CI], 38-73) and 47% (95% CI, 29-34), respectively. The median PFS was 263 days. There was no statistically significant difference in PFS between the CNS and non-CNS groups (P = .83; Figure 2A). The 6- and 12-month OS in the CNS group were 84% (95% CI, 65-93) and 76% (95% CI, 57-88), respectively. The median OS was not reached. There was no statistically significant difference in OS between the CNS and non-CNS groups (P = .14; Figure 2B).

Nine patients (29.0%) in the CNS group received consolidation/maintenance therapy after infusion. Among those, 6 received TKIs, 2 received allo-HSCT (along with radiotherapy in 1 patient), and 1 received donor lymphocyte infusion. Most of these patients (77.8%) had CNS-3 disease at preapheresis assessment. The main indication cited for consolidation/maintenance therapy initiation was increased risk of relapse (n = 5) such as in cases of overt MRD-positive disease at time of infusion or history of relapse after allo-HSCT. Other reasons cited were consolidative allo-HSCT after achieving second CR (n = 1), prolonging remission for a patient without planned allo-HSCT (n = 1), and institutional practice (n = 2).

Earlier studies have shown that CAR T cells can be detected in the CSF of patients with and without neurotoxicity.^15-17 However, the migration of CAR T cells into the CNS is enhanced because of increased blood-brain barrier permeability, which occurs because of CAR T-cell–related inflammatory responses and, in cases of endothelial injury, because of tumors or leukemia infiltration.^18-20 Thus, enrolling patients with CNS leukemia or lymphoma in many CAR T-cell clinical trials has been restricted or entirely excluded because of concerns of increased risk of neurotoxicity.^6-8,21 For instance, enrollment in the ZUMA-3 trial was limited to patients with asymptomatic CNS-2 disease.⁸ In this study, we described a multi-institutional experience using the CD19 CAR T-cell therapy, brexu-cel, in adult patients with B-ALL with CNS-2 and -3 disease (Figure 3). Our results showed that the toxicity of brexu-cel is similar between patients with and without CNS involvement. Additionally, we demonstrated that brexu-cel is efficacious in patients with CNS B-ALL, resulting in high rates of CNS disease remission and outcomes (both PFS and OS) comparable with those of patients without CNS involvement.

Most patients in our cohort developed CAR T-cell–related adverse events, including CRS and ICANS, which occurred at similar rates in patients with and without CNS involvement. This is consistent with previous studies reporting a high rate of CRS and ICANS after brexu-cel. For instance, 89% of patients enrolled in ZUMA-3 had CRS and 60% had neurotoxicity, with ∼25% of patients developing grade 3/4 CRS or ICANS.⁸ When looking specifically at patients with CNS B-ALL in our study, >60% of patients developed ICANS, with grade 3/4 ICANS affecting more than one-third of patients. Nevertheless, no statistically significant difference was detected when compared with patients without CNS ALL. These findings are consistent with previous studies demonstrating comparable safety of CD19 CAR T-cells in patients with CNS leukemia compared with patients without CNS involvement.^9,21,22 Leahy et al⁹ performed a pooled analysis of 5 different CD-19 CAR T-cell trials that enrolled 195 pediatric and young adult patients with ALL and lymphocytic lymphoma, among whom 18 had active CNS disease at preinfusion analysis. Using a multivariate logistic regression model, CNS disease at preinfusion assessment did not affect the risk of neurotoxicity.⁹ Similarly, other retrospective and prospective studies have shown that the risk of grade >3 ICANS in patients with CNS B-ALL disease at the time of infusion ranged between 0% and 33%.^10,11,23 Of note, most patients reported in the literature received CAR T cells that used a 4-1BB costimulatory domain, whereas brexu-cel has a CD28 costimulatory domain. CAR T cells with a CD28 costimulatory domain have a higher incidence of CRS and ICANS.^10,24,25 For instance, 5 of 8 (62%) patients (7 of whom received brexu-cel) with active CNS mantle cell lymphoma at the time of CAR T-cell infusion developed grade 3/4 ICANS.²⁶ Additionally, we have noted that 6 of 7 patients who had parenchymal disease and/or symptoms at preapheresis assessment developed grade 3/4 disease. The number is very small to draw conclusions; however, this indicates that those patients might benefit from cytoreduction strategies such as radiation before brexu-cel infusion.²⁷ 

Furthermore, the results of our study demonstrated the efficacy of brexu-cel in CNS B-ALL despite including a very heavily pretreated population. More than 80% of those evaluated after brexu-cel infusion cleared their CNS disease, including some patients who did not respond to, or receive, bridging therapy. However, more than one-third of patients relapsed, including 5 patients with confirmed CNS relapse during the follow-up period. To our knowledge, no previous studies have specifically investigated the efficacy of brexu-cel in adult patients with CNS B-ALL. Nevertheless, several retrospective studies have reported the response rates of other CAR T-cell products. Using CD19 and CD19-CD22 CAR T cells, Qi et al¹¹ reported a CNS remission rate of 85.4% in 48 patients with CNS-3 disease, including 21 patients who received no bridging. Other studies reported similar response rates, mainly in pediatrics and young adults.^9,10,23 

Additionally, active CNS disease before CAR T-cell infusion has not been shown to affect outcomes such as PFS or OS compared with patients without CNS disease in pediatrics and young adults.^9,23 These findings are consistent with our study findings, which showed that CNS disease at apheresis did not affect the PFS or OS of patients. In a univariate analysis of the ROCCA population, we previously reported that CNS/extramedullary disease did not affect PFS in adult patients with B-ALL receiving brexu-cel (hazard ratio, 0.88; 95% CI, 0.58-1.33).²⁸ Recently, a study from France reported outcomes of a cohort of 67 patients who received infusion with brexu-cel.²⁹ In their study, 10 patients had CNS-2 or -3 disease at time of apheresis and had shorter event-free survival than patients without CNS disease (odds ratio, 2.63; 95% CI, 1.12-6.13). However, the sample size was small, and the findings were limited to a univariable analysis.²⁹ 

This study has several limitations. Despite the multicenter nature of this study, the number of patients with CNS B-ALL treated with brexu-cel remains small. Thus, larger studies are needed to confirm many of our results and observations. It is also important to highlight that given this study’s retrospective nature, there was variability in practices between the different institutions including MRD testing methodology and disease assessment time points, which led to missing data points including unknown CNS status in ∼40% of patients in the non-CNS group. Additionally, there is variability in treatment patterns including previous therapies, bridging therapy, and the use of maintenance between the institutions, which makes it challenging to study its impact on outcomes. Importantly, very few patients had CNS disease assessment between apheresis and lymphodepleting therapy thus we could not determine the impact of the bridging therapy on the response rate after brexu-cel. Given the multi-institutional nature of this study, the database did not capture information such as traumatic lumbar punctures and radiological evaluation of CNS disease, particularly in patients with parenchymal CNS involvement. Moreover, the CNS burden assessment is limited to the Children’s Oncology Group classification, which does not accurately reflect the degree of parenchymal involvement and symptomatology. It also relies on cytomorphology, a technique that is less sensitive and may be less objective than flow cytometry. This method of evaluating CSF is not as established, and therefore it was not included in our data collection. However, it may be more reliable and/or predictive of outcome than traditional approaches, as other studies in newly diagnosed patients have suggested.^30-32 

In conclusion, our results demonstrate that using brexu-cel in adult patients with CNS B-ALL is feasible. Patients with CNS B-ALL at the time of apheresis had a high response rate and survival, but a significant number of patients relapsed. We also observed a high rate of ICANS, including grade 3/4 ICANS, which seems to occur regardless of CNS disease status. Our results are promising and encouraging to include patients with CNS B-ALL in future CAR T-cell clinical trials but also underscore the ongoing work needed to optimize efficacy and reduce toxicity in this population.

I.N.M. was supported by Cancer Prevention and Research Institute of Texas (CPRIT) RP210027, Baylor College of Medicine Comprehensive Cancer Training Program.

Contribution: I.N.M. and G.W.R. wrote the first draft; I.N.M., G.W.R., and A.Z. performed statistical analyses; I.N.M., G.W.R., L.S.M., and L.C.H. contributed to the study design and data interpretation; and all authors contributed to patients, provided critical revision of the manuscript, and approved the final version of the manuscript.

Conflict-of-interest disclosure: G.W.R. served on an advisory board for Kite and Autolus. R.F. reports research funding from Kite/Gilead and Novartis; advisory board member role with Kite/Gilead and Autolus; and consulting role with Sanofi. N.M. reports membership on an entity’s board of directors or advisory committees for Anthem Inc. M.B. received research funding from Novartis and Fate Therapeutics. M.M.S. reports speakers bureau role with Bristol Myers Squibb (BMS). P.S. received honoraria from Autolus Therapeutics and BMS; and reports speakers bureau role with BMS and Sanofi. C.J.L. reports consultancy with Fresenius Kabi, Sanofi, and Incyte Corp; received honoraria from Kite Pharma, BMS, Sanofi, and Kadmon; served as an advisory board member with Kite Pharma, Sanofi, and Incyte Corp; reports speakers bureau role with Kite Pharma; and received research funding from Incyte Corp. A.C.L. reports research funding from Amgen, Astellas, Autolus Therapeutics, Kadmon, Kite/Gilead, Pharmacyclics, and Talaris; and consultancy with AbbVie, Amgen, Actinium, BMS, Pfizer, Sanofi, and Takeda. S.B.T. reports speakers bureau role with BMS and Jazz Pharmaceuticals; and served on an advisory board for Autolus. J.T.L. reports consultancy with Adaptive Biotechnologies, Pfizer, Kite/Gilead, and Takeda; and membership on an entity’s board of directors or advisory committees with, and travel, accommodations, and expenses from, Adaptive Biotechnologies. C.H.O. received research funding from Electra, Novartis, Arog, Orca Bio, Jazz Pharmaceuticals, Pfizer, and Seagen. M.S. reports consultancy with Jazz Pharmaceuticals, Kite, and Autolus. J.P.S. reports consultancy with Kite/Gilead Autolus, and Genmab. D.K. reports consultancy with, and research funding from, BMS. R.D.C. received research funding from Servier, Incyte, Kite/Gilead, Amgen, Vanda Pharmaceuticals, Jazz Pharmaceuticals, Merck, and Pfizer; served on advisory committees of PeproMene Bio and Autolus; reports consultancy with, and honoraria from, Kite/Gilead, Amgen, Jazz Pharmaceuticals, and Pfizer; and reports employment (spouse) with, and stock ownership in, Seagen, within the last 24 months. B.D.S. reports research funding from Incyte, Jazz Pharmaceuticals, Kite/Gilead, and Servier; received honoraria from Pharmacyclics/Janssen, Spectrum/Acrotech, BeiGene, and Gilead Sciences; reports current employment with Moffitt Cancer Center; served on a data and safety monitoring board of, and received travel and accommodations and expenses from, Celgene, Novartis, Pfizer, Janssen, Seattle Genetics, AstraZeneca, Stemline Therapeutics, and Kite/Gilead; reports membership on an entity’s board of directors or advisory committees with PeproMene Bio; and reports consultancy with Takeda, AstraZeneca, Adaptive Biotechnologies, BMS/Celgene, Novartis, Pfizer, Amgen, Precision Biosciences, Kite/Gilead, Jazz Pharmaceuticals, Century Therapeutics, Deciphera, Autolus Therapeutics, Lilly, and PeproMene Bio. I.A. reports consultancy with Kite, Sobi, Jazz, Pfizer, Amgen, and Takeda; received honoraria from Amgen; served on an advisory board with Amgen, Pfizer, Jazz, Kite, Takeda, Syndax, Sobi, and Wugen; and received research support from AbbVie and MacroGenics. L.S.M. reports consultancy with Amgen, Pfizer, Kite, Astellas, and Autolus; received research funding from BMS, Adaptive, Kite, Autolus, Astellas, Orca Bio, and Jasper; received honoraria from Kite; and reports membership on an entity’s board of directors or advisory committees with Adaptive. L.C.H. reports consulting for, and membership on an entity’s board of directors or advisory committees of, March Biosciences; serves as a speaker for Kite/Gilead; and reports honoraria from Sanofi. The remaining authors declare no competing financial interests.

Correspondence: LaQuisa C. Hill, Center for Cell and Gene Therapy, Baylor College of Medicine, 6565 Fannin St, Ste A6-080, Houston, TX 77030; email: laquisa.hill@bcm.edu; and Lori Muffly, Division of Blood and Marrow Transplantation and Cellular Therapy, Stanford University, 780 Welch Rd CJ250B, Stanford, CA 94305-5623; email: lmuffly@stanford.edu.