FDA Approval Summary: Brexucabtagene Autoleucel for Treatment of Adults With Relapsed or Refractory B-Cell Precursor Acute Lymphoblastic Leukemia

Abstract

In October 2021, the FDA approved brexucabtagene autoleucel (brexu-cel), a CD19-directed chimeric antigen receptor (CAR) T-cell therapy, for the treatment of adult patients with relapsed or refractory B-cell precursor acute lymphoblastic leukemia (B-ALL). Approval was based on the phase II portion of ZUMA-3, a single-arm, open-label, multicenter trial that evaluated a single infusion of brexu-cel, preceded by lymphodepleting chemotherapy with cyclophosphamide and fludarabine, in this population. Efficacy was established on the basis of complete remission (CR) within 3 months after infusion and the duration of CR (DOCR). Among 54 patients in the efficacy analysis population, the CR rate was 52% (95% CI: 38, 66) with a median time-to-response of 56 days. With a median follow-up for responders of 7.1 months, the median DOCR was not reached. For all leukapheresed patients in the phase II portion of this trial (n = 71), the CR rate was 41% (95% CI: 29, 53). Among the 78 patients treated with the approved dose of brexu-cel, serious adverse reactions occurred in 79% and fatal adverse reactions occurred in 5% and included cerebral edema and infections. Cytokine release syndrome occurred in 92% (grade ≥3, 26%) and neurologic toxicities occurred in 87% (grade ≥3, 35%), leading to implementation of a risk evaluation and mitigation strategy (REMS). Postmarketing study with 15 years of follow-up will further evaluate long-term safety in adult patients with relapsed or refractory B-ALL.

This article summarizes the FDA review of the data supporting the approval of brexucabtagene autoleucel for the treatment of adult patients with relapsed or refractory B-cell precursor acute lymphoblastic leukemia.

Implications for Practice.

Brexucabtagene autoleucel is a new treatment option for adult patients with r/r B-ALL. The risk of serious and life-threatening adverse reactions necessitates careful patient selection, close monitoring, and a REMS to help assure safe use.

Introduction

Adult patients with relapsed or refractory B-cell precursor acute lymphoblastic leukemia (r/r B-ALL) have an unfavorable prognosis, which is influenced by disease biology, patient characteristics, and prior therapy. Duration of first complete remission (CR), age, white blood cell count at diagnosis, refractoriness to prior therapy, number of relapses, and subsequent allogeneic hematopoietic stem cell transplantation (HSCT) are known prognostic factors for survival outcomes.^1,2 Treatment approach includes the use of cytotoxic chemotherapy, HSCT, and novel targeted agents such as blinatumomab or inotuzumab; however, the latter do not induce long-term remission and are largely dependent on subsequent allogeneic HSCT. Tisagenlecleucel, an autologous anti-CD19 chimeric antigen receptor (CAR) T-cell therapy, was approved for the treatment of r/r B-ALL in pediatric and young adult patients up to 25 years of age.³

Brexu-cel is an autologous, CD19-directed CAR T-cell therapy created by collecting a patient’s T cells, transducing them with a retroviral vector encoding the CAR, and infusing the modified cells back into the patient (Table 1). The chimeric receptor in brexu-cel includes CD3ζ and CD28 intracellular co-stimulatory domains that promote T-cell activation. It is produced through a manufacturing process that removes malignant cells, reducing the potential for ex vivo activation and exhaustion of anti-CD19 CAR T cells during the manufacturing process.^4,5

Table 1.

Brexucabtagene autoleucel background information.

Pharmacologic class	CD19-directed, genetically-modified autologous T-cell immunotherapy
Mechanism of action	Following anti-CD19 CAR T-cell engagement with CD19-expressing target cells, the CD28 and CD3-zeta co-stimulatory domains activate downstream signaling cascades that lead to T-cell activation, proliferation, acquisition of effector functions and secretion of inflammatory cytokines and chemokines. This sequence of events leads to apoptosis and necrosis of CD19-expressing target cells
Pharmacokinetics	Following infusion at the dose of 1 × 106 CAR-positive viable T cells per kg body weight, brexu-cel exhibited an initial rapid expansion phase, achieving maximal expansion 14 days following infusion and persisting in the peripheral blood for a median of 53.3 days (0 -535 days). Cytokine and chemokine levels peaked within the first 8 days and generally returned to baseline within 28 days
Approval in B-ALL	For the treatment of adult patients with r/r precursor B-cell ALL (2021)
Other approvals a	For the treatment of adult patients with r/r mantle cell lymphoma (2020)

Under accelerated approval.

Brexu-cel was originally approved in July 2020 for adult patients with r/r mantle cell lymphoma (MCL) after 2 or more lines of systemic therapy (Table 1). CD19-directed CAR T-cell therapies have been approved for other malignancies such as aggressive and indolent lymphomas; however, there is no approved CAR T-cell therapy for adults with ALL. In October 2021, after a priority review, FDA approved brexu-cel for the treatment of adult patients with r/r B-ALL. Herein, we summarize the FDA clinical review and regulatory considerations regarding this licensing application.⁶

Trial Design

ZUMA-3 (NCT 02614066) was a single-arm, multicenter, open-label phase I/II study that evaluated brexu-cel in adults with primary refractory B-ALL, first relapse following a remission lasting ≤12 months, r/r B-ALL after second-line or higher therapy, or r/r B-ALL at least 100 days after allogeneic HSCT. The study excluded patients with active or serious infections, active graft-vs-host disease (GVHD) or taking immunosuppressive medications within 4 weeks prior to enrollment, and any history of central nervous system (CNS) disorders, including CNS-2 disease with neurologic changes and CNS-3 disease irrespective of neurological changes. Patients with Philadelphia chromosome-positive (Ph+) disease were eligible to enroll in the study.

Treatment consisted of lymphodepleting (LD) chemotherapy (fludarabine 25 mg/m² i.v. daily on days −4, −3, and −2; cyclophosphamide 900 mg/m² i.v. on day −2) followed by a single intravenous infusion of brexu-cel on day 0 which denotes the day of CAR T-cell infusion. The trial permitted bridging therapy between leukapheresis and LD chemotherapy at the discretion of the treating physician. All treated patients were hospitalized for the CAR T-cell infusion and for a minimum of 7 days afterward. Patients with Ph+ B-ALL who achieved CR could resume treatment with tyrosine kinase inhibitors 2 months after brexu-cel infusion.

The phase II portion of ZUMA-3 served as the pivotal trial in the application, and the dose of brexu-cel administered in the phase II portion was based on the experience in phase I. There were 45 patients treated with brexu-cel in phase I, including 16 patients at the target CAR T-cell dose of 0.5 × 10⁶/kg, 23 patients at 1 × 10⁶/kg, and 6 patients at 2 × 10⁶/kg. The brexu-cel dose of 1 × 10⁶ CAR T cells/kg (with a maximum of 1 × 10⁸ CAR T cells) was identified by the applicant as the lowest dose on the plateau of efficacy and with manageable toxicity, so this dose was carried forward to phase II.

The primary endpoint of ZUMA-3 phase II pivotal cohort was overall complete remission (OCR), defined as the combination of CR and CR with incomplete hematologic recovery (CRi) as determined by central review. The analysis was to be conducted when all patients completed the Month 6 disease assessment. The target accrual was 50 patients. The study was powered to exclude an OCR rate of 40% when the true OCR was 65% in all patients who received an infusion of brexu-cel with 93% power and a 1-sided alpha level of 0.025. FDA analyzed the raw data from response evaluations to adjudicate all responses.

Results

Efficacy Analysis Populations

In the phase II portion of ZUMA-3, the 71 patients who were leukapheresed and enrolled comprised the all leukapheresed population (see Supplementary Fig. S1). Of these 71 patients, 16 did not receive brexu-cel due to intercurrent adverse events, manufacturing failures, or failure to meet eligibility. Manufacturing failure was observed in 8% of leukapheresed patients. The applicant’s efficacy analysis focused on the remaining 55 patients. FDA’s primary efficacy analysis population was limited to 54 patients who were enrolled and treated in ZUMA-3 phase II study and who had documented disease at baseline following bridging therapy, which excluded one patient who had no documented disease after bridging therapy.

The characteristics of FDA’s efficacy analysis populations are described in Table 2. In the efficacy analysis population (n = 54), the median age was 40 years (range: 19-84 years), with slight predominance of men over women, and a predominance of Caucasians. As the study was multiregional, 76% of patients were from the US and the remaining patients were from several European countries. Among prior therapies, 43% of patients were previously treated with allogeneic HSCT, 46% with blinatumomab, and 22% with inotuzumab. Twenty-six percent of patients had primary refractory disease and 26% had Ph+ disease. No patient had CNS involvement prior to LD initiation.

Table 2.

ZUMA-3 phase II characteristics of the efficacy analysis populations.

Characteristic	Efficacy analysis population N = 54 n (%)	All leukapheresed population N = 71 n (%)
Demographics
Median agein years(range)	40 (19-84)	44 (19-84)
≥65 years old	8 (15)	11 (15)
Sex
Male	33 (61)	41 (58)
Female	21 (39)	30 (42)
Race
White	36 (67)	51 (72)
Other	9 (17)	9 (13)
Missing	4 (7)	4 (6)
Asian	3 (6)	4 (6)
Black or African American	1 (1.9)	2 (2.8)
American Indian or Alaska Native	1 (1.9)	1 (1.4)
Ethnicity
Not Hispanic or Latino	41 (76)	57 (80)
Hispanic or Latino	11 (20)	12 (17)
Missing	2 (3.7)	2 (2.8)
Baseline ECOG disease status
0	16 (30)	18 (25)
1	38 (70)	53 (75)
Disease Characteristics
Baselineaextramedullary disease	6 (11)	9 (13)
Philadelphia chromosome+	14 (26)	19 (27)
Prior blinatumomab	25 (46)	33 (46)
Blinatumomab as last prior therapy	12 (22)	13 (18)
Prior inotuzumab	12 (22)	16 (23)
Prior allogeneic HSCT	23 (43)	28 (39)
Disease status at enrollment
Primary refractory disease	14 (26)	17 (24)
First untreated relapse	4 (7)	6 (8)
Second or later untreated relapse	11 (20)	15 (21)
Refractory relapse	25 (46)	33 (46)

Baseline is defined as following bridging and prior to lymphodepleting chemotherapy.

Abbreviations: ECOG, Eastern Cooperative Oncology Group; HSCT, hematopoietic stem cell transplantation.

For the 54 patients in the efficacy analysis population, the median time from leukapheresis to product delivery to the treatment site was 16 days (range: 11-39 days) and the median time from leukapheresis to product infusion was 29 days (range: 20-60 days). Fifty (93%) patients received bridging therapy between leukapheresis and lymphodepleting chemotherapy to control disease burden.

Efficacy

The applicant reported that an OCR was achieved by 39 (70.9%; 95% exact Clopper-Pearson confidence interval [CI]:CI 57, 82) of the 55 patients treated with brexu-cel in the phase II portion of ZUMA-3. Since the lower bound of the 95% CI exceeded the prespecified limit of 40%, they concluded that the primary objective was met. Excluding the one patient who did not have evidence of active disease at treatment baseline and using FDA-adjudicated responses, FDA identified an OCR in 35 (64.8%; 95% CI: 50.6, 77.3) of 54 patients and confirmed that the primary objective was met.

However, as CRi is not a validated clinical endpoint for the treatment of ALL, the efficacy of brexu-cel was established on the basis of CR within 3 months after infusion and on the duration of CR. The key efficacy results for the efficacy analysis population and the all leukapheresed population are shown in Table 3. In the efficacy analysis population, 52% (95% CI: 38, 66) of the 54 patients achieved CR. Supplementary Table S1 shows the CR rates by subgroup. The median time to CR was 56 days (range: 25 to 86 days).

Table 3.

ZUMA-3 phase II efficacy outcomes.

	Efficacy analysis populationa N = 54	All leukapheresed populationa N = 71
Achieved CR rate within 3 months from infusion of brexu-cel, n (%) [95% CI]	28 (52) [38, 66]	29 (41) [29, 53]
Median duration of CR (months) [95% CI] (range in months)	NR [9.6, NE] (0.03+, 16.07+)	13.6 [9.4, NE] (0.03+, 16.07+)
Median follow-up for CR in months	7.1 (0.03+, 16.1+)	5.0 (0.03+, 16.1+)

See Supplementary Fig. S1.

A + sign indicates a censored value.

Abbreviations: CR, complete remission; NE, not estimable; NR, not reached.

All efficacy evaluable patients had potential follow-up for ≥10 months with a median actual follow-up time of 12.3 months (range: 0.3 to 22.1 months), the median duration of CR was not reached. For patients who achieved CR within 3 months after infusion, the probability of CR lasting at least 12 months was 62% (95% CI: 35, 80). Notably, 24% of patients achieving CR underwent subsequent allogeneic HSCT while in remission, and their durations of CR were thus censored at their last evaluable disease assessments prior to the transplant.

Safety

Safety Overview

The safety population was comprised of the 78 patients enrolled and treated in ZUMA-3 phase I or phase II with brexu-cel at a target dose of 1 × 10⁶ anti CD19 CAR T cells/kg (see Supplementary Fig. S1). Table 4 summarizes the most common adverse reactions (ARs) in the safety population (n = 78). Any-grade ARs affecting ≥20% of patients, excluding laboratory terms, included fever, cytokine release syndrome (CRS), hypotension, encephalopathy, tachycardia, nausea, chills, headache, fatigue, febrile neutropenia, diarrhea, musculoskeletal pain, hypoxia, rash, edema, tremor, infection with pathogen unspecified, constipation, decreased appetite, and vomiting.

Table 4.

Selected adverse reactions in the safety population^a.

Adverse reactions by categoryb,c	Safety population (N = 78)
	All grades%d	Grade 3 or higher%
Blood and lymphatic system
Febrile neutropenia	35	35
Coagulopathy	17	5
Cardiac
Tachycardia	63	6
Arrythmia	15	1
Gastrointestinal
Nausea	41	1
Diarrhea	32	6
Abdominal pain	19	0
Constipation	24	0
Vomiting	21	3
General
Fever	96	38
Chills	40	0
Edema	29	5
Fatigue	37	1
Pain	13	1
Immune system
CRS	92	26
Infection
Infection with pathogen unspecified	28	22
Bacterial infection	15	8
Fungal infection	13	5
Metabolism and nutrition
Decreased appetite	22	1
Musculoskeletal
Musculoskeletal pain	32	5
Muscular weakness	14	1
Nervous system or psychiatric
Encephalopathy	63	27
Headache	38	1
Tremor	29	1
Delirium	18	5
Dizziness	13	1
Insomnia	13	0
Anxiety	12	0
Respiratory
Hypoxia	31	23
Cough	12	0
Dyspnea	12	1
Skin
Rash	31	0
Vascular
Hypotension	69	33
Hemorrhage	13	4
Hypertension	13	6

Adverse reactions (≥10% any grade or ≥5% grade ≥ 3), excluding laboratory terms.

Includes grouped preferred terms. See Supplementary Table S2.

Baseline for adverse reactions was assessed immediately prior to brexucabtagene autoleucel infusion.

Toxicities were graded using NCI CTCAE version 4.03 except for CRS which was graded per Lee et al.⁷

Abbreviation: CRS, cytokine release syndrome.

Grade ≥3 ARs occurred in 97% of patients, including febrile neutropenia (35%), neurologic toxicities (35%), infections (30%), and CRS (26%). Grade 3 or higher cytopenias not resolved by day 60 following brexu-cel infusion occurred in 11%, with lack of resolution of neutropenia in 9% and thrombocytopenia in 6%. Other key grades 3 and 4 laboratory abnormalities included hypophosphatemia (47%), increased alanine aminotransferase (31%), increased aspartate aminotransferase (23%), hyperglycemia (22%), and hypocalcemia (22%).

Serious adverse events (SAEs) occurred in 79% of patients and included CRS, febrile neutropenia, hypotension, encephalopathy, fever, infections with pathogen unspecified, hypoxia, tachycardia, bacterial infection, respiratory failure, seizure, diarrhea, dyspnea, fungal infection, viral infection, coagulopathy, delirium, fatigue, hemophagocytic lymphohistiocytosis/macrophage activation syndrome (HLH/MAS), musculoskeletal pain, edema, and paraparesis. Four patients had fatal ARs: one with cerebral edema and 3 with infections (sepsis and fungal pneumonia). Three patients had ongoing CRS events at the time of death.

ARs of Special Interest

CRS and neurologic toxicity are leading safety concerns of brexu-cel. In the ZUMA-3 study, CRS of any grade, based on modified Lee criteria,⁷ occurred in 72 (92%) patients, including a 26% incidence of Grade ≥3 CRS. Manifestations of CRS in ≥20% of patients included fever, hypotension, tachycardia, headache, chills, hypoxia, fatigue, and nausea. The median time to CRS onset was 5 days (range: 1-12 days), with a median time to first grade ≥3 CRS of 7 days (range: 1-15 days). The incidence of CRS onset within the first 7 days after brexu-cel infusion was 90%. The median duration of CRS was 8 days (range: 2-63 days).

Neurologic toxicity, as defined by FDA (see regulatory considerations), was reported in 68 (87%) patients, with 35% incidence of grade ≥3 neurologic toxicity. Six patients had ongoing neurologic events at the time of death. The most common neurologic toxicities included encephalopathy, headache, tremor, delirium, dizziness, insomnia, and anxiety. The median time to onset of any neurologic toxicity was 7 days (range: 1-51 days), with 55% of patients presenting within the first 7 days of product infusion. Median time to first Grade ≥3 neurologic toxicity was 8 days (range: 4-24 days) and median duration of symptoms was 15 days (range: 1-397 days). Median time to resolution was 22 days (range: 5-402 days). In total, 4% of patients developed neurologic toxicity without developing CRS.

The instructions for prevention and treatment of CRS and neurologic toxicity were revised several times during the conduct of ZUMA-3 Study (eg, use of tocilizumab for prevention of CRS, administration of corticosteroids for lower grades of CRS and neurologic events).⁶

Among the 78 patients in the ZUMA-3 safety population, 65 (83%) received tocilizumab. Systemic steroids were administered in 59 (76%) patients. Fifty-five (71%) patients were treated with systemic corticosteroids and tocilizumab to manage toxicities.

Clinical Pharmacology

Following infusion at the dose of 1 × 10⁶ CAR-positive viable T cells per kilogram body weight in patients with r/r B-ALL in ZUMA-3 (n = 78), brexu-cel exhibited an initial rapid expansion phase followed by a bi-phasic decline. Peak levels of anti-CD19 CAR T cell (C_max) were achieved around 15 days following infusion (T_max). Brexu-cel persisted in peripheral blood with the median persistence time of 53.5 day (range: 0-535 days). Brexu-cel pharmacokinetic parameters are shown in Table 5.

Table 5.

Comparison of pharmacokinetic parameters of Brexu-cel by key efficacy and safety outcomes (Brexu-cel dose level: 1 × 10⁶ CAR-positive viable T cells/kg).

		Median exposure metric (range)
Outcome	n	C max (cells/μL)	AUC0-28d (days*cells/μL)	T max (days)
Overalla
All patients	66	24 (0, 2777)	243 (0, 20451)	15 (7,32)
Phase II only	50	21 (0, 1533)	221 (0, 19390)	15 (7,32)
Efficacyb
CR/CRi	32	38 (1, 1533)	424(14, 19390)	15(8.0, 30.0)
No CR/CRi	17	0.9 (0, 184)	8(0, 889)	15 (7.0, 32)
P-valuec		.0006	.0003	NS
CRSd
Grades 0-1	18	6 (0, 87)	79 (0, 693)	15 (7, 29)
Grades 2-5	39	46 (0, 1533)	461 (0, 19390)	15 (8, 32)
P-valueb		.001	.0003	NS
Neurotoxicityd
Grades 0-1	18	14 (0, 87)	134 (0, 693)	15 (8, 32)
Grades 2-5	39	82 (0, 1533)	390 (0, 19390)	15 (7, 29)
P-valuec		.04	.02	NS

Only patients who had evaluable pharmacokinetic information are included.

In the Phase II patients only.

Nominal P-value by Wilcoxon rank sum test.

Phases I and II patients with modified toxicity management.

Abbreviations: AUC_0-28d, area under the curve from time 0 to day 28; C_max, maximum concentration; T_max, time to maximum concentration.

In the phase I portion of ZUMA-3, there was no clear dose-exposure response (C_max and AUC_0-28d) within the doses ranging from 0.5 × 10⁶ to 2 × 10⁶ CAR T cells/kg. Median values of brexu-cel exposure (C_max and AUC_0-28d) were highest in subjects treated at 1 × 10⁶ CAR T cells/kg dose level with modified toxicity management.⁸ In a multivariate analysis of baseline demographic and disease factors for impact on exposure metrics in the 78 patients treated with 1 × 10⁶ CAR T cells/kg (phase I and phase II), only marrow blast percentage at screening and IFN-g production in co-culture (brexu-cel potency) were associated significantly with C_max and with AUC_0-28.

Table 5 provides the T_max, C_max, or AUC_0-28 by the outcomes of CR/CRi, occurrence of grades 2-5 CRS, and occurrence of grades 2-5 neurotoxicity for patients treated with 1 × 10⁶ CAR T cells/kg. There was a difference in C_max and AUC_0-28 by subgroup for all 3 outcomes, but there was no difference for T_max by subgroup. In multivariate analysis, no demographic, disease, or product characteristics other than brexu-cel exposure metrics were associated with CR/CRi, grades 2-5 CRS, or grades 2-5 neurotoxicity for the patients treated with 1 × 10⁶ CAR T cells/kg.

After brexu-cel infusion, substantially higher brexu-cel expansion (C_max and AUC_0-28d) was observed in responding patients (CR + CRi), compared with non-responding (non-CR/CRi) patients. Higher brexu-cel exposure was associated with grade 2 or higher CRS and grade 2 or higher neurological events. However, due to high inter-subject variability of brexu-cel pharmacokinetic parameters (C_max and AUC_0-28d), there were overlaps of brexu-cel exposure between these subgroups.

Regulatory Considerations

In the pivotal portion of ZUMA-3, adults with r/r B-ALL treated with brexu-cel had a CR rate of 52% (95% CI: 38, 66) within 3 months from infusion, and the median duration of CR was not reached with a median of 7.1 months of follow-up. FDA considers CR with durability for determination of clinical benefit for regulatory decision-making in a population with acute leukemia who have relapsed or refractory disease after multiple standard treatments or when there is no available therapy. Based on the risk-benefit analysis (Table 6), the high CR rate with durable response in ZUMA-3 in conjunction with the manageable safety profile justified regular approval for brexu-cel. Importantly, of the 78 patients treated with brexu-cel, 12 (15%) were 65 years of age and over. No overall differences in safety or effectiveness were observed between these patients and younger patients.

Table 6.

FDA benefit-risk Analysis of brexu-cel for R/R B-ALL.

Parameter	Summary
Unmet medical need	Adult patients with relapsed or refractory B-ALL do not survive without treatment with long-term survival of <1%. B-ALL after second or subsequent relapse or refractory to initial induction chemotherapy is highly resistant to salvage chemotherapy based on prior exposure to standard of care chemotherapy and allogeneic HSCT. R/R B-ALL has a poor prognosis with standard of care therapy including allogeneicHSCT and prognosis is influenced by disease biology, patient characteristics, and prior therapy.
Clinical benefit	Of 54 patients in the primary efficacy population, OCR rate was 65% (95% CI: 51, 77), and CR rate was 54% (95% CI: 40, 67). With a median follow-up of 7.1 months (0.03+, 16.07+), the median duration of CR was not reached. CR rate within 3 months of infusion of brexu-cel was 52% (95% CI: 38, 66).
Risks	• The USPI for brexu-cel has Warnings and Precautions for CRS, neurologic toxicity, HLH/MAS, serious infections, prolonged cytopenias, hypogammaglobulinemia, and secondary malignancies. • Of 78 patients in the safety population who were treated with brexu-cel: ­-  ARs in ≥20%, excluding laboratory terms, included fever, CRS, hypotension, encephalopathy, tachycardia, nausea, chills, headache, fatigue, febrile neutropenia, diarrhea, musculoskeletal pain, hypoxia, rash, edema, tremor, infection with pathogen unspecified, constipation, decreased appetite, and vomiting. ­- Grade ≥3 ARs occurred in 97%; the most common grade ≥ 3 ARs of interest included prolonged cytopenias (61%), febrile neutropenia (35%), neurologic toxicities (35%), infections (30%), and CRS (26%). ­- Any grade of CRS occurred in 92% patients, neurologic toxicity in 87% ­- Serious ARs occurred in 79% patients and 5% of patients had fatal ARs which included cerebral edema and infections.
Uncertainties	• Long-term safety after treatment with brexu-cel, particularly for secondary malignancies, is undefined.
Conclusions	• The ZUMA-3 study provides substantial evidence of effectiveness in adult patients with r/r B-ALL, with an acceptable safety profile. • CRS and neurologic toxicity can be life-threatening or fatal, supporting boxed warnings in labeling and a REMS.

One potential challenge with the review was that the endpoint definition did not prespecify the observation time for the response assessment to be used in the analysis of the efficacy. In general, responses to induction chemotherapy are expected to occur within 42 days from the start of therapy for treatments of acute leukemia. Delay in count recovery has been observed after CAR T-cell treatments, but the risks from prolonged cytopenias could at least partially negate the clinical benefit. Prior precedent for CAR T-cell therapy for ALL used CR within 3 months from infusion,⁹ and that maximum observation time was applied to FDA’s analysis of efficacy in ZUMA-3 as well.

With targeted small molecules, the optimal dose may be established by first finding the optimal exposure for the clinical outcomes and then confirming the dose that provides such exposure in most patients. That approach is not possible for CAR T cells, since exposure is impacted by in vivo expansion which may not be related to dose.¹⁰ For brexu-cel for r/r ALL, we relied on dose-response in a pooled analysis rather than exposure-response to confirm the dose; however, due to the small number of patients treated at doses other than 1 × 10⁶ CAR T cells/kg, there remains some uncertainty as to whether the dose of brexu-cel is optimal. As data accumulate with more CAR T-cell experience, it may be possible in the future to apply mechanistic pharmacokinetic-pharmacodynamic models to improve the precision of CAR T-cell dosing.¹¹

It should be noted that this approval does not establish how brexu-cel should be placed in the adult r/r B-ALL treatment paradigm. The CR rate achieved with brexu-cel is clearly higher than expected with conventional combination chemotherapy or single-agent use of the recently-approved targeted therapies for this patient population. On the basis of registry or single-arm trials, only allogeneic HSCT¹² or combinations of chemotherapy with a targeted agent¹³ provide for CR rates comparable to those demonstrated with brexu-cel in ZUMA-3. However, the duration of response after these types of therapies is known to be limited.¹⁴ In ZUMA-3, several patients underwent allogeneic HSCT while in remission following brexu-cel infusion, however, their DOCR was not overestimated given the censoring at the last evaluable disease assessment prior to transplant. Nevertheless, due to the limitation of this statistical approach which reduces the follow-up and may render the DOCR less accurate, the ultimate duration of response with brexu-cel remains not known. At the present time, blinatumomab is the only targeted therapy for ALL that has a demonstrated survival benefit in comparison to conventional chemotherapy. Additional randomized trials are warranted to establish the optimal sequencing of therapies and potential role of a consolidative allogenic HSCT for r/r B-ALL based on safety and efficacy considerations.

CRS was reported and graded as a syndrome using a modification of the grading system proposed by Lee et al.⁷ It should be noted that in the original Lee grading scale, neurologic ARs were included as symptoms of CRS. In the applicant’s modification of the Lee grading scale, neurologic ARs were reported separately. Except for CRS which was graded as a syndrome per the modified Lee et al⁷ criteria, all other ARs including the individual symptoms associated with CRS were graded using the National Cancer Institute (NCI) Common Terminology Criteria for Adverse Events (CTCAE) criteria v4.03. FDA’s review strategy of identifying additional CRS cases included searching for events such as fever, hypotension, or hypoxia between Day 0 and Day 30 in the patients who were not recognized by the applicant as having CRS. FDA also looked for patients who may have received tocilizumab, vasopressors, intravenous fluids or oxygen and who were not in the applicant’s CRS database. Cytokine and laboratory data (ferritin, C-reactive protein, IL-6 levels) were also used for supportive evidence. Patients identified as having isolated symptoms such as hypotension without other symptoms suggestive of CRS were not considered.

The applicant defined neurologic toxicities as per Topp et al¹⁵ and reported that 25% of the 55 patients treated in the ZUMA-3 phase II cohort had a grade 3 or higher neurological event.⁴ In contrast, neurologic toxicities were defined more broadly by FDA, and occurred according to this definition in 35% of the safety population. FDA’s definition included all events classified as nervous system disorders or psychiatric disorders, excluding sleep disorders and disturbances and peripheral neuropathies. However, ARs such as insomnia when occurring in setting of other neurologic events were included in FDA’s definition. Furthermore, other ARs under other body systems (system organ classes) that were not classified as neurologic toxicity, but that overlapped with other neurologic events, were also included in FDA’s definition of neurologic toxicity. Isolated events (not overlapping with other neurologic or psychiatric symptoms) or ARs that started relatively late following brexu-cel infusion and were unlikely to represent neurologic toxicity were not included in this definition. Because of differences in definitions of neurologic toxicities and observation times, any comparisons across studies should be made with caution.

Notably, because the toxicity management guidance for CRS and neurologic toxicity was modified during the conduct of ZUMA-3 study, the impact of the change in management over time was assessed in the safety review.⁶ However, because of the small number of patients, the majority of the safety analyses included a pool of all patients treated with the brexu-cel target dose, without regard to the version of the safety management instructions in use at the time of treatment.

Risk factors for CRS and neurologic toxicity from brexu-cel have not been established. In registrational trials of brexu-cel, keeping in mind the caveats of cross-trial comparisons, we noted a higher incidence of Grade ≥3 CRS in patients with r/r B-ALL (26%) than was reported for patients with r/r MCL (18%). Interestingly, the incidence of neurologic toxicity was similar in both populations. Specifically, any grade neurologic toxicity was observed in 87% of patients with ALL and 81% of patients with MCL, while grade ≥3 neurologic toxicity occurring in 35% of patients with ALL compared to 37% of patients with MCL. Because of the limited number of patients treated in the clinical trials which may not reflect the rates observed in clinical practice, a definitive conclusion on the difference in toxicity rates between the 2 diseases may not be made.

CAR T-cell-associated hemophagocytic lymphohistiocytosis (carHLH), clinically resembling HLH/MAS but occurring in recipients of CAR T-cell therapies, is an increasingly recognized severe toxicity with unclear underlying pathophysiology.^16,17 Although carHLH and CRS are both hyperinflammatory in etiology and have overlapping clinical manifestations, it is not clear that effective treatment would be the same, and there is currently great interest in establishing diagnostic criteria. carHLH was reported in 4% of patients on ZUMA-3 and was thus identified as a potential risk for inclusion in the “Warning and Precautions” of the U.S. Prescribing information (USPI).

During the ZUMA-3 study, life-threatening and fatal adverse reactions caused by brexu-cel were mitigated by mandated site and investigator training, careful site selection and monitoring, instructions for early detection and management of the most serious complications. The life-threatening and fatal ARs warranted warnings in the labeling, including a boxed warning for CRS and neurologic toxicity, and a REMS with elements to assure safe use (ETASU).

The genetic modification of brexu-cel triggers an additional safety concern. Generation of replication-competent retrovirus during the manufacturing process for brexu-cel is a theoretical safety concern. Additionally, insertional mutagenesis due to vector integration is a potential risk for inducing secondary malignancies. Integration of the vector into the patient’s cells might inadvertently activate a cellular protooncogene or disrupt a tumor suppressor gene, leading to malignant transformation events. Long-term safety following brexu-cel treatment remains important due to the limited duration of follow-up on ZUMA-3 and the potential risk of insertional mutagenesis from the retrovirus. Therefore, a postmarketing requirement (PMR) was issued that requires 15 years of follow-up for all patients treated with brexu-cel to assess its long-term toxicities.

Conclusion

Patients with B-ALL that is refractory to induction therapy or have experienced multiple relapses have few meaningful treatment options. The results of the ZUMA-3 study support the determination that a single dose of brexu-cel has substantial efficacy as demonstrated by high CR rate and durability of CR in adult patients with r/r B-ALL. Because of the life-threatening nature of the disease in the indicated population, the adverse reactions of CRS and neurologic toxicity, if managed appropriately and with a stringent risk mitigation plan in place, represent toxicities that are acceptable from a benefit-risk perspective in the intended population.

Funding

None declared.

Conflict of Interest

The authors indicated no financial relationships.

Author Contributions

Conception/design: N.B, and D.P.. Collection and/or assembly of data: N.B., D.P., X.W., and X.L. Data analysis and interpretation: All authors. Manuscript writing: All authors. Final approval of manuscript: All authors.