CADTH Reimbursement Review

Brexucabtagene Autoleucel (Tecartus)

Sponsor: Gilead Sciences Canada, Inc.

Therapeutic area: Mantle cell lymphoma

This multi-part report includes:

Clinical Review

Pharmacoeconomic Review

Ethics Review

Clinical Review

Executive Summary

An overview of the submission details for the drug under review is provided in Table 1.

Introduction

Mantle cell lymphoma (MCL) is an aggressive subtype of B-cell non-Hodgkin lymphoma that develops in the outer edge of a lymph node (i.e., the mantle zone) and is defined by the overexpression of cyclin D1 due to translocation (11;14)(q13;q32), making up less than 10% of all non-Hodgkin lymphoma diagnoses.^1-7 Mantle cell lymphoma primarily affects men and is usually diagnosed in patients older than 60 years.^5,6 Canadian-specific incidence and prevalence estimates of MCL are limited; however, it is estimated that 500 to 600 new cases of MCL are diagnosed each year.^8,9 Patients are most often diagnosed with advanced stage disease, showing generalized lymphadenopathy and extranodal involvement of the blood, bone marrow, spleen, and gastrointestinal tract, resulting in swollen painless lymph nodes, headache, weakness, loss of appetite, nausea and/or vomiting, abdominal pain or bloating, fatigue, and general B symptoms including fever, weight loss, and night sweats.^4,5 Diagnosis is confirmed by lymph node or bone marrow biopsy with immunohistochemistry or flow cytometry showing the presence of B-cell surface markers (CD20, CD19, CD5), cyclin D1 protein overexpression, t(11;14) translocation, or overexpression of the SOX11 transcription factor.^2,4,8

Despite frequent diagnoses at the advanced stage, front-line treatments are generally associated with high response rates (objective response rate [ORR] = 95%; complete response [CR] = 53%) and extension of survival^10-14; however, most patients will eventually relapse and require further therapy, generally with shorter periods of remission for each subsequent line of therapy.¹⁵ Front-line treatment consists of chemoimmunotherapy with a rituximab-containing regimen (rituximab, dexamethasone, and cytarabine [R-DHA]; rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone [R-CHOP] or rituximab, dexamethasone, cytarabine, and cisplatin [R-DHAP]; Nordic regimen; cyclophosphamide, vincristine, doxorubicin, and dexamethasone alternating with high-dose methotrexate and cytarabine [hyper-CVAD] plus rituximab; rituximab, bendamustine, and cytarabine [R-BAC]; bortezomib, rituximab, cyclophosphamide, doxorubicin, and prednisone [VR-CAP]; or bendamustine plus rituximab), followed by autologous stem cell transplant (SCT) in younger, fit patients. In patients not fit for SCT, chemoimmunotherapy and rituximab maintenance are the only treatment options in the first line.¹⁶ There is no established standard of care for patients in the relapsed or refractory (R/R) setting, and treatment options are based on prior therapies and response to treatment.⁹ Further treatment options for patients who experience a short response to prior treatment include additional chemoimmunotherapy or a Bruton tyrosine kinase (BTK) inhibitor or ibrutinib or acalabrutinib.¹⁶ Response to BTK inhibitors is associated with decreased efficacy in patients who have received more prior lines of therapy. Following failure of BTK inhibitors in the second line, treatment options consist of re-treatment with previously unused chemoimmunotherapy, BTK inhibitor, or palliative care.¹⁶

Brexucabtagene autoleucel (Tecartus) is a single-dose, autologous T-cell product manufactured from patients’ individual leukapheresis material. Brexucabtagene autoleucel was submitted as a priority review for a new drug (cell therapy) with pre–Notice of Compliance status (Notice of Compliance received on June 8, 2021). The objective of this report was to perform a systematic review of the beneficial and harmful effects of brexucabtagene autoleucel cell suspension in a patient-specific single infusion bag for IV infusion, with a target of 2 × 10⁶ chimeric antigen receptor (CAR) T cells per kilogram per the Health Canada indication for the treatment of adult patients with R/R MCL after 2 or more lines of systemic therapy including a BTK inhibitor

Stakeholder Perspectives

The information in this section is a summary of input provided by the patient groups who responded to CADTH’s call for patient input and from clinical experts consulted by CADTH for the purpose of this review.

Patient Input

One patient group, Lymphoma Canada, provided input for the review of brexucabtagene autoleucel, based on information gathered through an anonymous online survey of patients with MCL that was circulated between October 2020 and January 2021. Thirty-three patients provided input on their experience with MCL.

Patients reported that the symptoms of MCL — such as enlarged lymph nodes, fatigue, gastrointestinal symptoms, aches and pains, and high or low white blood cell (WBC) count — had the greatest impact on their quality of life and affected their ability to travel, exercise, concentrate, perform daily activities like household chores, and perform regular duties like work or volunteering. Patients’ mental and emotional well-being was negatively impacted, and many patients experienced stress, anxiety or worry, and difficulty sleeping.

All patients surveyed rated faster remission and longer life as the most important outcomes for a new therapy. Other important outcomes included control of disease and symptoms, improved quality of life, and improved blood counts. Having choice in their treatment selection was rated as very important, with a large majority of patients agreeing that there is a need for more effective therapy options.

Most patients would accept a treatment with known and potentially serious adverse effects if it was recommended by their doctor. The majority of patients were willing to tolerate adverse effects of a new treatment if they were short-term.

Clinician Input

Input From Clinical Experts Consulted by CADTH

Clinical experts came to the consensus that at the time of relapse or failure of BTK inhibitor, patients have a short duration of response and a poor life expectancy of around 6 months, and therefore new treatments that provide durable responses are needed for this population. The experts agree that the population eligible for treatment with brexucabtagene autoleucel would be small and would reflect the inclusion criteria of patients in the pivotal trial. Brexucabtagene autoleucel would be used post-ibrutinib or in those who are intolerant of ibrutinib. This would also include anyone with a suboptimal response to ibrutinib or acalabrutinib. Clinical experts agreed that if a patient starting BTK inhibitor therapy is a potential candidate for brexucabtagene autoleucel, this should be considered early, as once response to BTK inhibitors is lost, patients will rapidly deteriorate. Experts agreed that R/R MCL patients who have previously received chemoimmunotherapy and a BTK inhibitor and have a suitable Eastern Cooperative Oncology Group Performance Status (ECOG PS) (0 to 2) and adequate organ function would be readily available to receive this treatment, with no specific subgroups that would likely benefit more at this time.

Clinical experts agree that patients eligible for treatment would be identified by the treating specialist in hematology or oncology at the time of first relapse of MCL and that patients should be prioritized based on individual need, with those progressing on BTK inhibitors being higher need than those currently stable or responding to BTK inhibitor treatment. Experts believe that the least suitable patients for brexucabtagene autoleucel would be patients with a very high disease burden that is rapidly progressing and those who cannot tolerate further chemoimmunotherapy therapy, those with a poor performance status, and those with multiple comorbidities.

The clinical experts noted that in clinical practice, a combination of clinical exam, bloodwork (complete blood count and lactate dehydrogenase), and imaging (CT/PET) would be used to assess response to therapy, and response would be followed up post-infusion at 1 month and 3 months, and then as needed. Given that brexucabtagene autoleucel is a single-dose treatment, clinical experts stated that discontinuation is not possible, and any patients for whom the drug has been manufactured but who experience rapid decline would not receive treatment.

Clinician Group Input

Two physician groups provided input for this submission: the Ontario Health (Cancer Care Ontario [CCO]) Hematology Disease Site Drug Advisory Committee and a group of 7 lymphoma experts in Canada whose input was coordinated by Lymphoma Canada.

The experts noted that the current data support the role of this therapy as a single agent in patients treated with multiple therapies who have disease progression following primary chemoimmunotherapy and BTK inhibitor therapy. Brexucabtagene autoleucel would replace treatment options in the third line or later, including palliative chemotherapy; intensive chemotherapy or experimental treatments for select patients; or SCT in young, fit patients.

Drug Program Input

The Provincial Advisory Group (PAG) provided questions considered important for decision-making:

• Are there specific subgroups of patients with R/R MCL who are more or less likely to benefit from brexucabtagene autoleucel?

• If there is limited capacity to offer this treatment, how would you prioritize which patients should be offered brexucabtagene autoleucel? How would you select which ones should be treated in Canada versus out of country?

• Is there evidence to support re-treatment with brexucabtagene autoleucel for R/R MCL if a patient responds then subsequently relapses or if the initial response is suboptimal?

• PAG seeks guidance on the provision of brexucabtagene autoleucel in patients whose central nervous system (CNS) lymphoma or viral infection is being actively treated.

The clinical experts consulted by CADTH stated that the characteristics of patients who should receive brexucabtagene autoleucel align with the approved indication. With regard to capacity, the clinical experts stated that patients would likely remain treated in their home province, but, for those who want to receive treatment elsewhere, the decision would be made on a case-by-case basis. PAG identified a potential time-limited need to cover brexucabtagene autoleucel in patients who are on other therapies for R/R MCL, including immunochemotherapies and salvage chemotherapy, and would like confirmation that patients who are doing well on these therapies should not be switched to CAR T-cell therapy. If switching is an option, PAG would like to understand under what circumstances it would be preferred as opposed to waiting until disease progression. The clinical experts consulted by CADTH stated that there are currently no data available to suggest re-treatment or sequencing of treatment for brexucabtagene autoleucel; however, switching would not be warranted unless there was disease progression on a patient’s current therapy.

PAG noted the potential for “indication creep” in the use of brexucabtagene autoleucel in first-line treatment for MCL and in B-cell non-Hodgkin lymphomas or leukemias for which CAR T-cell therapy is not available. PAG also noted there may be creep to patients who have stable disease or partial response (PR) to prior therapies or who are at high risk of early relapse. Since the Health Canada–approved indication includes prior BTK inhibitor therapy as a requirement, treating patients who have not received this prior therapy would be considered indication creep. Experts agreed that indication creep is a possibility; however, there is currently no evidence for treatment in patients who are BTK inhibitor naive, and the place in therapy may change as more data become available.

Clinical Evidence

Pivotal Studies and Protocol Selected Studies

Description of Studies

The ZUMA-2 study was a phase II, multi-centre, single-arm, open-label study evaluating the efficacy and safety of brexucabtagene autoleucel at a target dose of 2 × 10⁶ anti-CD19 CAR T cells per kilogram (with a maximum dose of 2 × 10⁸ anti-CD19 CAR T cells for patients more than 100 kg) in patients with R/R MCL whose disease had progressed on anthracycline- or bendamustine-containing chemotherapy, an anti-CD20 antibody, and a BTK inhibitor (ibrutinib and/or acalabrutinib). Eligible patients were adults (≥ 18 years) with pathologically confirmed MCL and documentation of either cyclin D1 overexpression or presence of t(11;14) that is relapsed or refractory, as defined by disease progression after the last regimen, or failure to achieve PR or CR to the last regimen. Patients should have received up to 5 prior treatment regimens that must have included all of the following: anthracycline or bendamustine-containing chemotherapy, anti-CD20 antibody, and ibrutinib or acalabrutinib. The primary objective of the ZUMA-2 study was ORR, defined as a CR or PR using central assessment per the Lugano classification.¹⁷ Key secondary outcomes included best objective response (BOR), duration of response (DOR), progression-free survival (PFS), overall survival (OS), and health-related quality of life (HRQoL). Harms evaluated included adverse events (AEs), including those of special interest to this review, such as cytokine release syndrome (CRS) and neurologic AEs.¹⁸

Seventy-four patients were enrolled and leukapheresed. Sixty-nine patients received lymphodepleting chemotherapy, and 68 patients received brexucabtagene autoleucel, making up the modified intention-to-treat (mITT) and safety populations. Of the patients who received brexucabtagene autoleucel, 38% received bridging therapy. The most commonly administered bridging therapies were ibrutinib (n = 14 [21%]) and dexamethasone (n = 12 [18%]). Patients in the full analysis set (FAS) were leukapheresed a median of 16.0 days (range = 5 days to 274 days) after study screening, and the median time from screening to leukapheresis was also 16.0 days (range = 5 days to 274 days). The median time from leukapheresis to administration of brexucabtagene autoleucel was 27.0 days (range = 19 days to 134 days).¹⁸

The median age of included patients was 65.0 years (range = 38 years to 79 years), and the majority of patients were male (84%) and White (93%). Fifty-five patients (81%) had received 3 or more prior regimens. Twenty-nine patients (43%) relapsed after prior autologous SCT; the remaining patients had either relapsed after their last therapy for MCL (n = 12 [18%]) or were refractory to their last therapy for MCL (n = 27 [40%]).

Efficacy Results

The ORR in the inferential analysis set (IAS) at the primary data cut-off (July 24, 2019) was 93% (95% confidence interval [CI], 83.8% to 98.2%), which was significantly higher than the pre-specified historical control rate of 25% (P < 0.0001), and the CR rate was 67% (95% CI, 53.3% to 78.3%). As of the December 31, 2019, data cut-off, the ORR was 92% (95% CI, 81.6% to 97.2%), which was also higher than the historical control (P < 0.0001), and the proportion of patients with CR was 67% (95% CI, 53.3% to 78.3%). The median time to CR or PR using the Lugano classification was 1.0 month (range = 0.8 months to 3.1 months), and the median time to achieve a CR was 3.0 months (range = 0.9 months to 9.3 months).

As of the July 24, 2019, data cut-off, and with a median follow-up time for DOR of 8.6 months, the median DOR was not reached (95% CI, 8.6 months to not estimable). At the December 31, 2019, data cut-off, with a median follow-up time for DOR of 14.1 months, the median DOR was also not reached (95% CI, 13.6 months to not estimable).

As of the July 24, 2019, data cut-off (12.3 months follow-up), the median PFS in the IAS was not reached (95% CI, 9.2 to not estimable). Kaplan–Meier estimates of PFS rates at 6 months and 12 months were 77.0% and 60.9%, respectively. As of the updated analysis, the median PFS was also not reached (95% CI, 9.6 to not estimable), with a median follow-up of 16.8 months.

The median OS was not reached (95% CI, 24.0 to not estimable) as of the July 24, 2019, data cut-off (12.3 months follow-up). The OS rates at 6 and 12 months were 86.7% and 83.2%, respectively. At the December 31, 2019, data cut-off, the median OS was not reached (95% CI, not estimable to not estimable), and the OS rates at 6 and 12 months were identical to those at the primary data cut-off.

Harms Results

At least 1 treatment-emergent adverse event (TEAE) was reported in all patients in the safety analysis set, of which 99% had AEs that were Grade 3 or higher. Eleven patients (16%) had Grade 3 TEAEs, 52 patients (76%) had Grade 4 TEAEs, and 4 patients (6%) had Grade 5 TEAEs. The most common Grade 3 or higher AEs were anemia and decreased neutrophil count (50%) and decreased WBC count (40%).

Serious AEs were recorded for 68% of patients. The most common serious AEs (SAEs)were encephalopathy and pyrexia (22% each), followed by hypotension (16%). The most common Grade 3 or higher SAEs were encephalopathy (18%) and hypotension and hypoxia (12% each). Two patients had Grade 5 SAEs of B-cell lymphoma and died due to disease progression. In general, the incidence of SAEs was slightly lower in the updated analysis. The most common Grade 3 or higher SAEs were encephalopathy (16%), pneumonia (which increased to 13% from 7%), and hypotension (12%).

As of the December 31, 2019, data cut-off, 18 patients had died: 16 (24%) due to progressive disease and 2 related to AEs (organizing pneumonia and staphylococcal bacteremia, both deemed related to lymphodepleting chemotherapy). Most deaths occurred more than 3 months after brexucabtagene autoleucel infusion (14 of 18 deaths).

Notable harms identified in the protocol of this review were CRS, which was reported in 62 patients (91%). The most common Grade 3 or higher CRS symptoms were hypotension (25%), hypoxia (19%), and pyrexia (11%). The median time to onset of CRS was 2 days following brexucabtagene autoleucel infusion, and it lasted a median of 11 days (range = 1 day to 50 days). As of the July 24, 2019, data cut-off, CRS had resolved in all 62 patients.

Neurologic AEs of any grade occurred in 43 patients (63%). The most common Grade 3 or higher neurologic events were encephalopathy (19%), confusional state (12%), and aphasia (4%). Twenty-two patients (32%) had serious neurologic events of any grade; 18% were Grade 3, and 9% were Grade 4.

The median onset of a neurologic event following brexucabtagene autoleucel infusion was 7 days. In patients whose neurologic AEs had resolved, the median duration was 12 days (range = 1 day to 567 days). As of the July 24, 2019, data cut-off, neurologic events had resolved in all but 6 patients.

Critical Appraisal

The ZUMA-2 study was a single-arm study that lacked a comparator arm, which increases the risk of bias in the estimation of treatment effects due to the potential for confounding related to placebo response, fluctuations in health status, and other unidentified prognostic factors that could affect subjectively assessed outcomes. The use of well-validated measurement scales for assessing patient-reported outcomes was appropriate and might ameliorate bias in the measurement of these outcomes; however, it does not remove the potential for bias related to lack of comparison. The follow-up time was considered appropriate for assessing response to treatment; however, it was noted to be immature for survival outcomes.

Primarily, the ZUMA-2 study did not include any Canadian patients. According to the panel of clinical experts, the eligibility criteria based on absolute neutrophil counts, platelet counts, and absolute lymphocyte counts were believed to be higher than would be seen in the general population. Moreover, they noted that the majority of patients had an ECOG PS of 0 (65%), indicating a healthier population, and only 10% of patients had bulky disease at baseline. Lastly, only 38% of patients received bridging therapy in preparation for brexucabtagene autoleucel infusion, which the experts believed to be low. Together, it is unclear whether the included population was healthier than would be typically seen in these patients in clinical practice; thus, generalizability of the results to the typical Canadian population may be limited.

Indirect Comparisons

Description of Studies

The sponsor submitted a matching-adjusted indirect comparison (MAIC) that compared the efficacy of brexucabtagene autoleucel to standard of care treatments in terms of OS, PFS, and tumour response outcomes. The analysis was informed by a systematic literature review that identified 9 uncontrolled, mainly retrospective, open-label studies that provided outcome data in patients with R/R MCL who received treatment following BTK inhibitor therapy (N = 12 to 73; median follow-up range = 3.2 months to 38 months). The subsequent therapies reported in the trials included lenalidomide-based treatments, venetoclax, R-BAC or rituximab in combination with bendamustine, bortezomib, and dexamethasone (R-iBVD), and mixed treatments (various chemo-immunotherapies or systemic therapies).

For the MAIC, a logistic propensity score model was used to estimate patient weights for the ZUMA-2 trial, so that the weighted mean baseline characteristics of the ZUMA-2 patients (FAS N = 74) matched the pooled mean characteristics of the standard of care studies. Pairwise indirect comparisons were then conducted using the weighted ZUMA-2 data and pooled outcome data for standard of care studies (n = 2 to 8).

Efficacy Results

The primary MAIC analyses for OS reported a hazard ratio (HR) of 0.18 (95% CI, 0.09 to 0.38; effective sample size = 36.2), a PFS HR of ||, (95% CI, ||||||||||||; effective sample size = 16.3), and an objective response odds ratio of 7.91 (95% CI, 2.35 to 26.62; effective sample size = 29.5).

Harms Results

The sponsor-submitted MAIC did not assess safety outcomes for brexucabtagene autoleucel.

Critical Appraisal

The key limitation of the MAIC is the assumption that absolute outcomes can be predicted from the covariates included in the model (i.e., that every effect modifier and prognostic factor are accounted for).²⁰ This assumption is largely considered impossible to meet, and the failure of this assumption leads to an unknown amount of bias in the unanchored estimate.²⁰ For the sponsor-submitted indirect treatment comparisons (ITCs), some prognostic factors were excluded from the model or may have been incompletely specified due to missing data from the clinical trials. The effective sample size was small for all outcomes (16 to 36), which suggests poor population overlap and unstable estimates. The results for the tumour response outcome lacked precision and showed wide 95% CIs. Uncertainty in the results of the unanchored MAIC are compounded by the inclusion of lower quality comparator trials and clinical heterogeneity across studies. Due to the limitations of the ITC, no conclusions on the comparative efficacy of brexucabtagene autoleucel can be drawn from the MAIC.

Other Relevant Evidence

No long-term extension studies or other relevant studies were included in the sponsor’s submission to CADTH.

Conclusions

Evidence from the single-arm, open-label ZUMA-2 trial suggests that CAR T-cell therapy with brexucabtagene autoleucel at a target dose of 2 × 10⁶ anti-CD19 CAR T cells per kilogram is associated with statistically significant rates of objective response relative to a historical control. Moreover, brexucabtagene autoleucel was associated with substantial improvements in survival (PFS and OS) and improved HRQoL over 6 months, which are outcomes that patients have identified as relevant. The results of the ZUMA-2 trial may potentially be biased due to the inability to control for confounding and the unblinded assessment of subjective outcomes. As well, the included population may be more fit than the general Canadian population with MCL who would be eligible for treatment, which could limit the generalizability of the results. The observed benefits of brexucabtagene autoleucel also need to be weighed against the associated harms, including serious CRS and neurologic AEs, which result in further intensive care unit (ICU) admission. Data on long-term outcomes of therapy with brexucabtagene autoleucel are needed.

Given the lack of head-to-head studies for brexucabtagene autoleucel, the sponsor submitted an unanchored MAIC that provided indirect evidence of the efficacy of brexucabtagene autoleucel compared with standard of care therapies in patients with R/R MCL. However, due to limitations of the MAIC analysis methods, inclusion of lower quality comparator trials, clinical heterogeneity between studies, lack of complete covariate data to inform the patient weights, and poor population overlap, no conclusions can be drawn from the MAIC.

Introduction

Disease Background

Mantle cell lymphoma is an aggressive subtype of B-cell non-Hodgkin lymphoma that develops in the outer edge of a lymph node (i.e., the mantle zone) and is defined by the overexpression of cyclin D1 due to translocation (11;14)(q13;q32) in approximately 85% of patients.^1-5 In Western countries, MCL makes up less than 10% of all non-Hodgkin lymphoma diagnoses.^2-4,6,7 The annual incidence of MCL is approximately 0.5 per 100,000 people.^3,21,22 In Canada, there are approximately 500 to 600 new cases of MCL diagnosed each year,^8,9 occurring more frequently in men (3:1), and it is usually diagnosed in patients aged 60 to 70 years.^5,6

The exact underlying cause of MCL is unknown. The molecular hallmark of MCL is the chromosomal translocation t(11;14)(q13;q32), which transposes the CCND1 gene at 11q13 to the immunoglobulin heavy chain of chromosome 14q32, leading to cyclin D1 overexpression. However, the transcription factor sex-determining region, Y-box 11 (SOX11), may be used as a diagnostic marker for the rare cases of MCL that are cyclin D1 negative.^3,23,24 Mutations of TP53 are also associated with aggressive disease and poor outcomes in MCL.^16,25 WHO recognizes 2 types of MCL: (a) nodal MCL (80% to 90% of cases) with unmutated immunoglobulin heavy chain, SOX11 overexpression, and nodal and extranodal involvement, which is generally more aggressive, and (b) non-nodal leukemic MCL (10% to 20% of cases) with mutated immunoglobulin heavy chain, no SOX11 mutation, increased WBC count, and splenomegaly, which typically has a more indolent behaviour.^2,26

Generally, patients with MCL have multiple symptoms involving lymph nodes throughout the body, which include generalized lymphadenopathy and extranodal involvement of the blood, bone marrow, spleen, and gastrointestinal tract, resulting in non-specific symptoms including swollen painless lymph nodes, headache, weakness, loss of appetite, nausea and/or vomiting, abdominal pain or bloating, fatigue, and general B symptoms including fever, weight loss, and night sweats.^4,5 Often, patients with early stages of the disease are asymptomatic, and MCL is usually diagnosed in later stages of the disease, which are characterized by an aggressive clinical course and a poor prognosis. Additional complications from MCL progression include cytopenias (neutropenia, anemia, and thrombocytopenia), and gastrointestinal, pulmonary, and CNS complications.²

Diagnosis of MCL is suspected by a hematologist or hematopathologist through the initial workup and laboratory assessments (complete blood count, serum lactate dehydrogenase) and imaging (PET/CT scans) and is confirmed by lymph node or bone marrow biopsy with immunohistochemistry or flow cytometry showing the presence of B-cell surface markers (CD20, CD19, CD5), cyclin D1 protein overexpression, t(11;14) translocation, or overexpression of the SOX11 transcription factor.^2,4,8 No specific risk factors or predispositions have been identified.² However, MCL has been associated with Borrelia infection, living or working on farmhouses, familial MCL, and autoimmune diseases.^27-30

The Mantle Cell Lymphoma International Prognostic Index (MIPI) is used to determine prognosis and treatment plans in MCL patients, incorporating scores for age, performance status, normalized lactate dehydrogenase level, and WBC counts, which influence MCL prognosis (Table 3). Patients are assigned to a low-risk, intermediate-risk, or high-risk category based on the cumulative number of points assigned to the individual’s prognostic factors. Patients with 0 to 3 points are considered to be at low risk, patients with 4 to 5 points are intermediate risk, and patients with 6 to 11 points are high risk.⁴ A modification of the MIPI also includes the Ki-67 proliferative index.^3,4

The advent of newer therapies has improved outcomes in the MCL population. From onset of the disease, the estimated median OS is 2.4 years to 4.25 years, while in patients with R/R MCL, the median OS is 1 year to 2 years.^9,31

Quality of life in MCL is scarcely reported in published literature.³² Aside from the general psychological distress associated with a cancer diagnosis, symptoms are often associated with reduced quality of life, particularly difficulty with daily activities, and the resulting burden of B symptoms. Respondents to an unpublished survey indicated that anxiety and/or depression, fatigue, gastrointestinal symptoms, aches and pains, and blood cell counts were the symptoms that had the greatest impact on their lives, interfering with their ability to work, travel, exercise, and perform day-to-day activities.

Standards of Therapy

Mantle cell lymphoma is incurable with standard treatment for advanced disease, which is guided by the patient’s age and comorbidities. Patients are often diagnosed with advanced disease, which has a more aggressive clinical course and is associated with poor prognosis.^2,7 Despite this, front-line treatments in MCL are generally associated with high response rates (ORR = 95%; CR = 53%) and extension of survival.^10-14

Younger, fit patients are treated with aggressive induction chemoimmunotherapy, followed by autologous SCT and rituximab maintenance.^9,16 Preferred chemoimmunotherapy options include¹⁶:

• R-DHA (rituximab, dexamethasone, and cytarabine) + platinum chemotherapy (carboplatin, cisplatin, or oxaliplatin)

• alternating R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone) and R-DHAP (rituximab, dexamethasone, cytarabine, and cisplatin)

• the Nordic regimen (rituximab plus cyclophosphamide, vincristine, doxorubicin, and prednisone, alternating with rituximab plus high-dose cytarabine)

• hyper-CVAD (cyclophosphamide, vincristine, doxorubicin, and dexamethasone alternating with high-dose methotrexate and cytarabine) + rituximab

• bendamustine + rituximab.

Older, less fit patients who are not appropriate for SCT are typically treated with chemoimmunotherapy and rituximab maintenance. Less aggressive chemoimmunotherapy options include bendamustine plus rituximab, VR-CAP, R-CHOP, lenalidomide plus rituximab, and R-BAC, followed by maintenance rituximab until progression or intolerance.¹⁶

Although initial treatment strategies achieve high response rates, most patients will eventually relapse and require further therapy, which generally results in shorter periods of remission with each subsequent line of therapy.¹⁵ There is no established standard of care for patients in the R/R setting, and treatment options are based on prior therapies and response to treatment.⁹ Further treatment options for patients that experienced a short response to prior treatment include additional chemoimmunotherapy or a BTK inhibitor or ibrutinib or acalabrutinib.¹⁶ Ibrutinib is the only BTK inhibitor publicly funded in Canada for MCL.¹⁵ However, response to BTK inhibitors is associated with decreased efficacy in patients who have received more prior lines of therapy.^15,33,34 Following failure of BTK inhibitors in the second line, treatment options consist of re-treatment with previously unused chemoimmunotherapy, BTK inhibitor, or palliative care, with poor median survival of 1 year or less.^35-40

Drug

Brexucabtagene autoleucel (Tecartus, also KTE-X19) is an autologous T-cell product manufactured from leukapheresis and administered after lymphodepleting chemotherapy, and optional bridging therapy, as a single-dose treatment in a patient-specific infusion bag, containing a suspension of anti-CD19 CAR T cells at a target dose of 2 × 10⁶ anti-CD19 CAR T cells per kilogram.

Brexucabtagene autoleucel is a CD19-directed genetically modified autologous T-cell immunotherapy that targets CD19-expressing cancer cells and normal B cells. Anti-CD19 CAR T cells engage with CD19-expressing cells, and co-stimulatory domains CD28 and CD3-zeta activate downstream signalling cascades that lead to T-cell activation, proliferation, acquisition of effector functions, and secretion of inflammatory cytokines and chemokines, leading to the killing of CD19-expressing cells.⁴¹

The drug under review has not been previously reviewed by CADTH and was submitted as a priority review for a new drug (cell therapy) as pre–Notice of Compliance status. The Health Canada Notice of Compliance was received on June 8, 2021. Brexucabtagene autoleucel has already been approved for use by the FDA under accelerated approval based on ORR and durability of response for the treatment of adult patients with R/R MCL⁴² and is currently under review by the National Institute for Health and Care Excellence. The submitted indication for review is identical to the Health Canada indication: treatment of adult patients with R/R MCL after 2 or more lines of systemic therapy including a BTK inhibitor.⁴³ The main characteristics of brexucabtagene autoleucel are summarized in Table 4.

Stakeholder Perspectives

Patient Group Input

This section was prepared by CADTH staff based on the input provided by patient groups.

About the Patient Groups and Information Gathered

One patient group, Lymphoma Canada, provided input for the review of brexucabtagene autoleucel. Lymphoma Canada is a national Canadian registered charity whose mandate is to empower the lymphoma community through education, support, advocacy, and research. It collaborates with patients, caregivers, health care professionals, and other organizations and stakeholders to promote early detection, find new and better treatments for lymphoma patients, help patients access those treatments, learn about the causes of lymphoma, and find a cure (www.lymphoma.ca).

The information submitted was gathered through an anonymous online survey of patients with MCL that was circulated between October 2020 and January 2021 using Lymphoma Canada’s email contact database, social media, US and Canadian cancer society message boards, physician specialists, and international lymphoma organizations’ contacts. Thirty-three patients provided input on their experience with MCL, including 1 US patient who had received brexucabtagene autoleucel therapy. Of the 25 patients who provided demographic data, 24 were from Canada (15 male and 9 female patients). All patients were over 44 years of age (age range 45 to 54 years: 5 patients; 55 to 64 years: 6 patients; 65 to 74 years: 11 patients; 75 to 84 years: 2 patients).

Disease Experience

The symptoms that patients experience due to MCL are variable and can change over time as the disease progresses. Respondents reported that enlarged lymph nodes (63% of patients), fatigue (75%), gastrointestinal symptoms (50%), aches and pains (68%), and high or low WBC count (50%) impacted their current quality of life. Patients with MCL stated that their symptoms had the greatest impact on their ability to travel, exercise, concentrate, perform daily activities like household chores, and perform regular duties like work or volunteering.

Receiving a diagnosis of MCL negatively impacted patients’ mental and emotional well-being. All respondents (N = 33) rated that 1 or more symptoms affected their quality of life, including stress of diagnosis (88%), anxiety or worry (79%), and difficulty sleeping (33%). As years pass from a patient’s MCL diagnosis, different mental and emotional aspects related to their disease may impact their quality of life, but anxiety, stress, and difficulty sleeping remained the most common issues among respondents.

As described by 1 patient: “I am extremely fortunate to have indolent MCL, but it can be difficult managing the anxiety of never knowing when or if I become sick and how aggressive my disease will be if I convert. It’s like living with a bomb sometimes.”

Experience With Treatment

Of the 25 respondents who provided information about their experience with MCL treatments, 76% required immediate treatment, while 24% remained in watch-and-wait status. Of the 25 respondents, 29% received more than 1 line of treatment following MCL relapse. The most commonly reported first-line treatment (44% of respondents) was the chemoimmunotherapy regimen R-CHOP, with 36% of patients receiving SCT. Of those who received more than 1 line of treatment, most received BTK inhibitors such as acalabrutinib or clinical trial drugs.

The most commonly reported adverse effects of MCL treatments included fatigue, hair loss, thrombocytopenia, diarrhea, nausea, anemia, neutropenia, confusion or memory loss, mouth sores, cough, skin rash or itching, constipation, and infections. Most patients (15 of 25 respondents) found fatigue, nausea, vomiting, hair loss, and neurocognitive effects such as brain fog or headaches were the most difficult adverse effects to tolerate. When asked about the impact of different aspects of their treatment on daily living, respondents noted that treatment-related fatigue, the duration of infusions, infusion reactions, and other late adverse effects of treatment had the most significant impact on their quality of life.

Treatment also had a financial impact on patients, with 36% reporting having to miss work, and 24% experiencing financial burdens related to drug costs.

One patient from the US, aged 65 to 74 years, had experience with brexucabtagene autoleucel through a clinical trial. This patient reported not being prepared for the numerous tests involved in determining treatment eligibility; however, the patient was moderately prepared and understood the various steps of CAR T-cell treatment, including the blood extraction process, wait times between extraction and infusion, the potential need for bridging therapy, and post-infusion monitoring. The patient experienced adverse effects of neutropenia, thrombocytopenia, and anemia, but was not admitted to the hospital for management of these adverse effects, and these adverse effects did not last for longer than 2 months post-treatment. The treatment burden (e.g., number of clinic visits, CAR T-cell extraction and infusion, adverse effect management, and travel and related costs) and emotional impact (e.g., worry or concern over potential adverse effects or the possibility of relapse) were rated as having a minor negative impact on this patient’s quality of life. Overall, the patient rated their experience with this treatment as positive.

Improved Outcomes

All patients surveyed rated faster remission and longer life as the most important outcomes for a new therapy. Other important outcomes included control of disease and symptoms (79%), improved quality of life (58%), and improved blood counts (58%). The majority of patients (58%) were willing to tolerate adverse effects of a new treatment if they were short-term, while the remaining respondents were unsure. Most patients (78%) would accept a treatment with known and potentially serious adverse effects if it was recommended by their doctor. Having choice in their treatment selection was rated as very important, with a large majority of patients (88%) agreeing that there is a need for more effective therapy options.

As described by 1 patient: “My great concern is that I am running out of treatment options. If I am to live, I need more options. It's really that simple.”

Clinician Input

Input From Clinical Experts Consulted by CADTH

All CADTH review teams include at least 1 clinical specialist with expertise in the diagnosis and management of the condition for which the drug is indicated. Clinical experts are a critical part of the review team and are involved in all phases of the review process (e.g., providing guidance on the development of the review protocol; assisting in the critical appraisal of clinical evidence; interpreting the clinical relevance of the results; and providing guidance on the potential place in therapy). In addition, as part of the brexucabtagene autoleucel review, a panel of 3 clinical experts from across Canada was convened to characterize unmet therapeutic needs, assist in identifying and communicating gaps in the evidence that could be addressed through the collection of additional data, promote the early identification of potential implementation challenges, gain further insight into the clinical management of patients living with a condition, and explore the potential place in therapy of the drug (e.g., potential reimbursement conditions). A summary of this panel discussion is presented below.

Unmet Needs

There is currently no curative option readily available in this patient population. The clinical experts agreed that at the time of relapse or failure of BTK inhibitor, patients have a short DOR and a poor life expectancy. Once patients stop second-line ibrutinib or acalabrutinib, life expectancy is around 6 months, so new treatments are needed for this population. Treatments are needed to control the disease and provide patients with a durable response, as a substantial number of patients will not respond to chemoimmunotherapy at this point.

Place in Therapy

This therapy best fits into treatment of relapsed MCL. Clinical experts noted that it is important that patients are treated with chemoimmunotherapy at diagnosis, followed by autologous SCT (if eligible) and maintenance rituximab, as patients can have long durations of remission after initial therapy. Therapy with BTK inhibitors only works for a limited time period, and brexucabtagene autoleucel would be used after ibrutinib or acalabrutinib or in those who are intolerant of ibrutinib or acalabrutinib. This would also include anyone with a suboptimal response to ibrutinib or acalabrutinib. Clinical experts agreed that if a patient starting BTK inhibitor therapy is a potential candidate for brexucabtagene autoleucel, this should be considered early as once response to BTK inhibitors is lost, patients rapidly deteriorate. One clinical expert noted that given the cost of ibrutinib, if the intention is to give brexucabtagene autoleucel, it does not seem logical to continue BTK inhibitor therapy beyond what is necessary. It would not make sense for brexucabtagene autoleucel to be reserved for patients who are intolerant of other therapies. Treatment with brexucabtagene autoleucel would result in increased surveillance in real-world practice, with increased monitoring of any signs of relapse to ensure that patients rapidly receive treatment with brexucabtagene autoleucel.

Patient Population

In line with the clinical trial, the clinical experts considered ideal patients to be those with R/R MCL who have previously received chemoimmunotherapy and a BTK inhibitor. Moreover, patients should have a suitable ECOG PS (0 to 2) and adequate organ function. Experts noted that patients with a p53 mutation tend to have poor outcomes and may benefit the most in terms of improvement of survival.

Patients eligible for treatment would be identified by the treating specialist in hematology or oncology at the time of first relapse of MCL. Clinical experts agree that patients should be prioritized based on individual need, with those progressing on BTK inhibitors being higher need than those currently stable and/or responding to BTK inhibitor treatment. It was noted by all the clinical experts that this would be a limited number of patients, and prioritization for treatment within Canada compared to out of country is not of concern. One clinician noted that there will be patients who refuse to accept treatment due to travel requirements. Moreover, manufacturing capacity and COVID-19 implications play a role in whether patients would be willing to accept treatment within their province, out of province, or out of country.

The least suitable patients for brexucabtagene autoleucel were thought to be patients with a very high disease burden that is rapidly progressing and those who cannot tolerate further chemoimmunotherapy therapy, those with a poor performance status, and those with multiple comorbidities. In line with the trial, patients with cardiac or CNS disease were excluded; however, 1 clinical expert noted that perhaps if these comorbid conditions were controlled, the patients could be considered. In addition to the clinical aspects that render patients ineligible, clinical experts noted that patients without a caregiver, patients who cannot or do not want to travel to receive treatment, and patients who will not survive longer than 4 weeks to 8 weeks are unlikely to receive treatment with brexucabtagene autoleucel.

Assessing Response to Treatment

The clinical experts noted that in clinical practice, a combination of clinical exam, bloodwork (complete blood count and lactate dehydrogenase), and imaging (CT/PET) would be used to assess response to therapy. Unlike clinical trials, imaging may not be as frequent after initial response to therapy has been documented, as it typically does not change disease management. Clinical experts also noted that repeat bone marrow biopsy or measurement of minimal residual disease is not common practice in MCL. Typically, patients would be followed up with at 1 month, at 2 months, and then as needed until best response to treatment is determined.

More data are currently needed to determine which patients are more likely to exhibit a response; however, experts did note that patient factors, including performance status and organ function, are generally known to influence response to treatment. No subgroup data from the trial provide insight to this.

Discontinuing Treatment

Clinical experts agreed that discontinuing treatment for CAR T-cell therapy is not applicable as it is a 1-time treatment, and once the therapy is initiated, it is not able to be stopped. However, patients who become unwell before infusion, or those who experience rapid clinical decline, may be deemed unable to proceed with infusion. However, this would be considered before treatment with brexucabtagene autoleucel.

Clinical experts agreed that there was no evidence of re-treatment for brexucabtagene autoleucel at present.

Prescribing Conditions

Clinical experts agree that brexucabtagene autoleucel must be administered in a setting supervised by specialists such as a hematologist and, possibly, a transplant hematologist to determine the use of CAR T-cell therapy versus allogeneic SCT. Experts noted that such a setting may consist of academic or tertiary centres that already deliver cellular therapies (bone marrow transplant) with the ability to provide urgent intensive care and on-call neurology. At this time, treatment should only be conducted in a centre that offers transplantation to ensure that standard operating procedures and a multi-disciplinary team are available. One expert noted that there is some evidence that the treatment can be conducted in the outpatient setting, assuming that outpatient after-hour care is available (i.e., an assessment and infusion clinic in addition to a 24-hour emergency department).

Clinician Group Input

This section was prepared by CADTH staff based on the input provided by clinician groups.

Two physician groups provided input for this submission: the Ontario Health (CCO) Hematology Disease Site Drug Advisory Committee and a group of lymphoma experts in Canada whose input was coordinated by Lymphoma Canada.

Cancer Care Ontario’s (CCO’s) Drug Advisory Committee provides timely evidence-based clinical health system guidance on drug-related issues in support of CCO’s mandate, including the Provincial Drug Reimbursement Programs and the Systemic Treatment Program.

Lymphoma Canada, a national not-for-profit organization for Canadian lymphoma and chronic lymphocytic leukemia patients, coordinated the group clinician response from leading experts in lymphoma across Canada.

The information submitted was based on joint discussions at a CCO Drug Advisory Committee meeting and based on collated clinician responses to the template questions based on research results, clinical experience, and understanding of patient needs and challenges.

Unmet Needs

For patients who have failed multiple lines of therapy, clinicians may attempt to access unfunded targeted therapies or enrol patients in clinical trials. Allogeneic SCT has been employed for younger patients, who typically have disease progression following primary chemoimmunotherapy and BTK inhibitor therapy. However, clinical experts stated that this intensive approach is only available to the minority of patients who have a donor and are of an appropriate age and fitness. Palliative chemotherapy may be the only option for some patients.

Clinical experts agree that there is an unmet need for effective treatments that produce clinical responses and remission and may prolong life in patients with R/R MCL. Existing options benefit only a fraction of patients (35% to 75%) and typically do not offer durable responses (approximately 6 months to 18 months). Many of these treatments must be administered indefinitely, and toxicities may adversely affect quality of life. For example, lenalidomide may cause fatigue, gastrointestinal upset, and cytopenia, which predispose patients to infections. Venetoclax also is associated with some toxicities, including early tumour lysis syndrome. Unfortunately, the median survival in this population is quite short post–ibrutinib failure, typically between 4 months and 6 months. Patients with p53 mutation have an average PFS of 4 months, versus 12 months in patients without the p53 mutation.

Place in Therapy

Brexucabtagene autoleucel is a novel patient-specific targeted therapy that augments the immune system’s ability to control the cancer. The experts noted that the current data support the role of this therapy as a single agent in patients treated with multiple prior lines of therapy who have disease progression following primary chemoimmunotherapy and BTK inhibitor therapy. Brexucabtagene autoleucel would replace treatment options in the third line or later, including palliative chemotherapy; intensive chemotherapy or experimental treatments for select patients; or SCT in young, fit patients.

Clinical trials to determine the efficacy and toxicity of brexucabtagene autoleucel when administered earlier in the disease course are required to support earlier integration in the treatment pathway. To date, this therapy has not been evaluated against standard front-line approaches or in a large cohort of BTK inhibitor–naive patients. The population of MCL patients who are typically treated from an exclusively palliative approach, who may be older and frail, is not expected to change if CAR T-cell therapy is available.

Patient Population

The experts noted that brexucabtagene autoleucel should be reserved for patients who have progressed after having received standard chemoimmunotherapy and BTK inhibitor therapy (unless these are contraindicated). Progression would be identified by the treating hematologist or oncologist by standard clinical testing (imaging, laboratory findings). Candidates for brexucabtagene autoleucel therapy would include patients who are younger (although age is not a specific criterion), without comorbidity, and with good performance and fitness status. In addition, patients must express CD19, but this is almost universal in MCL, and patients must have adequate numbers of circulating T lymphocytes to allow generation of CAR T-cell product.

The experts identified patients least suitable for treatment with brexucabtagene autoleucel as including those whose disease could not be controlled in the short-term to allow them to proceed to CAR T-cell therapy. Experts noted that patients with comorbid illnesses that may increase their risk of sepsis, cytokine release–related complications, or neurologic complications immediately after the T-cell reinfusion may be less suitable for this treatment. Such comorbidities could include difficult to manage diabetes or diabetic complications, chronic renal failure with impairments of creatinine clearance or on dialysis, or significant symptomatic cardiomyopathies. Patients with obvious uncontrolled infections would not be acceptable candidates. Patients with active CNS lymphoma may not be good candidates unless the disease is controlled and stable.

Currently, it is not possible to identify patients who are most likely to exhibit a response to treatment.

Assessing Response to Treatment

Clinically meaningful results to therapy as stated by the clinical experts include stabilization of disease or objective response to therapy. These responses would usually be associated with improvement in constitutional or organ-related symptoms. Success with this treatment should ensure improved quality of life and independence in the activities of daily living. Durable responses would be important, given the logistical difficulties and expense of treatment.

Standard clinical parameters, including those obtained through CT scans and possibly PET scans, would be used to document clinical response and remission. Bloodwork and assessments of organ function and the hematologic profile would also be important. Response to treatment should be assessed radiologically post-treatment and several months again post-treatment. Ongoing imaging may be dependent on symptoms and the results of the previous testing, clinical findings, and laboratory results.

Discontinuing Treatment

Progression of disease or a recurrence of symptomatology would indicate treatment failure. It would be appropriate to consider initiating a new treatment at that time. As a single infusion therapy, the main question around discontinuation would be for patients who have disease control issues before T-cell infusion.

Prescribing Conditions

Experts agree that patients who would be candidates for this therapy would be identified by the treating hematologist or oncologist and that treatment should be administered in a tertiary referral cancer centre that has experience and infrastructure for cell therapies or autologous SCT. Currently, this would be centres with CAR T-cell experience, which typically would be regional academic transplant programs.

Additional Considerations

It was noted by experts that tocilizumab may be required to manage CRS in some patients. Also, bridging chemotherapy may be needed for some patients to control the disease before proceeding with brexucabtagene autoleucel.

Drug Program Input

The drug programs provide input on each drug being reviewed through CADTH’s reimbursement review processes by identifying issues that may impact their ability to implement a recommendation. The implementation questions and corresponding responses from the clinical experts consulted by CADTH are summarized in Table 5.

Clinical Evidence

The clinical evidence included in the review of brexucabtagene autoleucel is presented in 3 sections. The first section, the systematic review, includes pivotal studies provided in the sponsor’s submission to CADTH and Health Canada, as well as those studies that were selected according to an a priori protocol. The second section includes indirect evidence from the sponsor and indirect evidence selected from the literature that met the selection criteria specified in the review. The third section includes sponsor-submitted long-term extension studies and additional relevant studies that were considered to address important gaps in the evidence included in the systematic review.

Systematic Review: Pivotal and Protocol Selected Studies

Objectives

To perform a systematic review of the beneficial and harmful effects of brexucabtagene autoleucel cell suspension in a patient-specific single infusion bag for IV infusion, with a target of 2 × 10⁶ CAR T cells per kilogram, for the treatment of adult patients with R/R MCL previously treated with a BTK inhibitor.

Methods

Studies selected for inclusion in the systematic review included pivotal studies provided in the sponsor’s submission to CADTH and Health Canada, as well as those meeting the selection criteria presented in Table 6. Outcomes included in the CADTH review protocol reflect outcomes considered to be important to patients, clinicians, and drug plans.

Of note, the systematic review protocol presented below was established before the granting of a Notice of Compliance from Health Canada.

The literature search was performed by an information specialist using a peer-reviewed search strategy according to the PRESS (Peer Review of Electronic Search Strategies) checklist (https://www.cadth.ca/resources/finding-evidence/press).

Two CADTH clinical reviewers independently selected studies for inclusion in the review based on titles and abstracts, according to the predetermined protocol. Full-text articles of all citations considered potentially relevant by at least 1 reviewer were acquired. Reviewers independently made the final selection of studies to be included in the review, and differences were resolved through discussion.

Published literature was identified by searching the following bibliographic databases: MEDLINE All (1946‒) via Ovid and Embase (1974‒) via Ovid. The search strategy comprised both controlled vocabulary, such as the National Library of Medicine’s MeSH (Medical Subject Headings) and keywords. The main search concept was brexucabtagene autoleucel. Clinical trial registries were searched: the US National Institutes of Health’s clinicaltrials.gov, WHO’s International Clinical Trials Registry Platform search portal, Health Canada’s Clinical Trials Database, Canadian Partnership Against Cancer Corporation’s Canadian Cancer Trials, and the European Union Clinical Trials Register.

No filters were applied to limit the retrieval by study type. Retrieval was not limited by publication date or by language. Conference abstracts were excluded from the search results. See Appendix 1 for the detailed search strategies.

The initial search was completed on January 25, 2021. Regular alerts updated the search until the meeting of the CADTH pan-Canadian Oncology Drug Review Expert Committee on May 13, 2021.

Grey literature (literature that is not commercially published) was identified by searching relevant websites from the Grey Matters: A Practical Tool for Searching Health-Related Grey Literature checklist (https://www.cadth.ca/grey-matters).⁴⁴ Included in this search were the websites of regulatory agencies (the FDA and the European Medicines Agency). Google was used to search for additional internet-based materials. See Appendix 1 for more information on the grey literature search strategy.

Findings From the Literature

Thirty-three studies were identified from the literature for inclusion in the systematic review, while 9 potentially relevant reports from other sources were retrieved for scrutiny (Figure 1). In total, 1 report from 1 study was included in the review.

The included study is summarized in Table 7. A list of excluded studies is presented in Appendix 2.

Figure 1: Flow Diagram for Inclusion and Exclusion of Studies

Description of Studies

One study met the inclusion criteria for this review. The ZUMA-2 study was a phase II, multi-centre, open-label study evaluating the efficacy and safety of brexucabtagene autoleucel (KTE-X19) in patients with R/R MCL whose disease had progressed on anthracycline- or bendamustine-containing chemotherapy, an anti-CD20 antibody, and a BTK inhibitor (ibrutinib and/or acalabrutinib).

ZUMA-2 was an open-label trial. Patients were assigned a unique patient identification number used to identify the patient throughout the study; it was used on all study documentation related to the patient. The study schema for the ZUMA-2 trial is shown in Figure 2.

Figure 2: ZUMA-2 Trial Schema

CAR = chimeric antigen receptor.

Source: ZUMA-2 clinical trial protocol.¹⁸

This study was conducted at 33 centres in the US, France, the Netherlands, and Germany. No Canadian sites were included.

Key eligibility criteria for the ZUMA-2 trial are summarized in Table 7. Eligible patients included adults diagnosed with R/R MCL who had received prior chemoimmunotherapy, anti-CD20 monoclonal antibody, and a BTK inhibitor. Patients were excluded if they had any measurable CNS disease or a history of allogeneic SCT or autologous SCT within 6 weeks of the planned infusion.¹⁸

Two cohorts were included in the ZUMA-2 trial to evaluate the efficacy of brexucabtagene autoleucel: Cohort 1, which aimed to treat up to 90 patients with a target dose of 2 × 10⁶ anti-CD19 CAR T cells per kilogram, and Cohort 2, which aimed to treat up to 40 patients with brexucabtagene autoleucel at a target dose of 0.5 × 10⁶ anti-CD19 CAR T cells per kilogram, a 4-fold lower dose of brexucabtagene autoleucel. For the purposes of this report, we will only focus on Cohort 1, for whom the target dose of brexucabtagene autoleucel specified in the systematic review protocol (Table 6) was evaluated, which is the target dose in the funding request. Ninety-one patients were enrolled in the ZUMA-2 study: 74 patients in Cohort 1 at a target dose of 2 × 10⁶ anti-CD19 CAR T cells per kilogram, and 17 patients in Cohort 2. Sixty-eight patients in Cohort 1 were treated with brexucabtagene autoleucel.¹⁸

The last observation and data cut-off for the ZUMA-2 trial was July 24, 2019, representing a median follow-up of 12.3 months (range = 7.0 months to 32.3 months) for the primary efficacy analysis.^18,45 An updated analysis as of December 31, 2019 was provided, representing a median follow-up of 17.5 months (range = 12.3 months to 37.6 months).^19,46

Populations

Inclusion and Exclusion Criteria

Eligible patients were adults (≥ 18 years) who were required to have pathologically confirmed MCL — with documentation of either cyclin D1 overexpression or presence of t(11;14) — that is relapsed or refractory, as defined by disease progression after the last regimen, or failure to achieve a PR or CR to the last regimen. Patients were required to have had prior treatment with up to 5 regimens that included all of the following: anthracycline or bendamustine-containing chemotherapy, anti-CD20 antibody, and ibrutinib or acalabrutinib. The original protocol did not specify the requirement for prior BTK inhibitors, and the inclusion criteria were modified at the June 22, 2018, protocol amendment to include ibrutinib and acalabrutinib as required prior therapies in all prospective patients. Therapy with a BTK inhibitor was not required to be the last line of therapy before trial entry, and patients were not required to have disease that was refractory to BTK inhibitor therapy. Additional inclusion criteria were an ECOG PS of 0 or 1, absolute neutrophil count greater than or equal to 1,000/μL, platelet count greater than or equal to 75,000/μL, absolute lymphocyte count greater than or equal to 100/μL, and adequate renal, hepatic, pulmonary, and cardiac function.¹⁸

Key exclusion criteria were a history of allogeneic SCT or autologous SCT within 6 weeks of study drug infusion. Patients were also excluded if they had detectable cerebrospinal fluid malignant cells or brain metastases, or a history of CNS lymphoma, cerebrospinal fluid malignant cells, or brain metastases; a history or presence of CNS disorders or any autoimmune disease with CNS involvement; or a history of myocardial infarction, cardiac angioplasty or stenting, unstable angina, active arrhythmias, or other clinically significant cardiac disease within 12 months before enrolment.¹⁸

Baseline Characteristics

The baseline and demographic characteristics of the ZUMA-2 trial are summarized in Table 8 for patients included in the safety analysis or mITT set, which included 68 patients. The median age of included patients was 65.0 years (range = 38 years to 79 years), and the majority of patients were male (84%), and the majority were White (93%). The majority of patients were from the US (91%), with 3, 2, and 1 patients from France, the Netherlands, and Germany, respectively.¹⁸

The majority of patients had classical MCL subtypes (n = 40 [59%]). Seventeen patients (25%) had blastoid MCL, while 4 (6%) had the classical MCL subtype of pleomorphic MCL. Based on the simplified MIPI, 28 patients (41%) were classified as low risk, 29 patients (43%) were classified as intermediate risk, and 9 patients (13%) were classified as high risk. Among the 49 patients with evaluable samples, the median percentage of tumour cells expressing Ki-67 was 65.0% (range = 1% to 95%), and 47 of patients (69%) had confirmed CD19 positivity.¹⁸

Fifty-five patients in Cohort 1 (81%) received 3 or more prior regimens. Twenty-nine patients (43%) relapsed after prior autologous SCT; the remaining patients had either relapsed after their last therapy for MCL (12 [18%]) or were refractory to their last therapy for MCL (27 [40%]). Per the protocol, all patients had received an anti-CD20 antibody and a BTK inhibitor. Forty-two treated patients (62%) had disease that did not respond to BTK inhibitor therapy (primary refractory disease), and 18 (26%) had a relapse after having an initial response while receiving BTK inhibitor therapy.¹⁸

Other prior treatments included anthracycline-based therapy (49 [72%]), bendamustine (37 [54%]), proteasome inhibitors (25 [37%]), and lenalidomide (19 [28%]).¹⁸

Interventions

Leukapheresis

All patients underwent leukapheresis to obtain leukocytes for the manufacturing of brexucabtagene autoleucel. A minimum of 12 L to 15 L of leukapheresis material was processed, with a goal of obtaining approximately 5 × 10⁹ to 10 × 10⁹ mononuclear cells.¹⁸

Bridging Therapy

If necessary, and at the discretion of the investigator, bridging therapy was considered for any patient, and particularly for patients with high disease burden at screening. Bridging therapy was administered after leukapheresis and was completed at least 5 days before the initiation of lymphodepleting chemotherapy. It was not intended for treatment purposes, only to ensure that a patient remained eligible for the brexucabtagene autoleucel infusion. The following bridging therapy regimens were permitted: dexamethasone 20 mg to 40 mg or equivalent orally or IV daily for 1 day to 4 days, ibrutinib at a dose of 560 mg daily or most recent dose if there had been a dose adjustment, or acalabrutinib 100 mg every 12 hours or most recent dose if there had been a dose adjustment.¹⁸

Conditioning Chemotherapy

Conditioning (lymphodepleting) chemotherapy consisting of fludarabine at a dose of 30 mg/m²/day and cyclophosphamide at a dose of 500 mg/m²/day was administered for 3 days on days −5, −4, and −3 to induce lymphocyte depletion. Lymphodepleting chemotherapy was supplied by the investigative site and to commence only when the brexucabtagene autoleucel was manufactured and available at the administrative site.¹⁸

Brexucabtagene Autoleucel

The ZUMA-2 study evaluated 2 doses of brexucabtagene autoleucel, an autologous T-cell product manufactured from individual patients’ leukapheresis material. Patients in Cohort 1 were to receive a target dose of 2 × 10⁶ anti-CD19 CAR T cells per kilogram, with a maximum dose of 2 × 10⁸ anti-CD19 CAR T cells for patients 100 kg or more. Patients in Cohort 2 were to receive a target dose of 0.5 × 10⁶ anti-CD19 CAR T cells per kilogram, with a maximum dose of 0.5 × 10⁸ anti-CD19 CAR T cells for patients 100 kg or more.¹⁸ As previously stated, only the patients from Cohort 1 are of interest to this review.

The individual, patient-specific product in single-use infusion bags was cryopreserved and returned to the treatment facility in a liquid nitrogen dry shipper. The product was kept frozen until the patient was ready for treatment and stored in the bag in the vapour phase of liquid nitrogen. Patients were to receive a single IV infusion of brexucabtagene autoleucel on day 0, after receiving the 3-day lymphodepleting chemotherapy regimen on day –5 through day –3, with a 2-day rest period between the completion of lymphodepleting chemotherapy and the brexucabtagene autoleucel infusion. The volume of brexucabtagene autoleucel infused, the thaw start and stop times, and the brexucabtagene autoleucel administration start and stop times were recorded. The following medications were administered 1 hour before infusion: acetaminophen 650 mg, and diphenhydramine 12.5 mg to 25 mg IV or 25 mg orally. If the patient did not meet the criteria for infusion, the administration of brexucabtagene autoleucel was to be delayed until the events resolved. If the infusion was delayed by more than 2 weeks, the lymphodepleting chemotherapy was to be repeated. Patients were hospitalized for treatment with brexucabtagene autoleucel and were to remain in the hospital for a minimum of 7 days after treatment, unless a longer stay was required. Patients who achieved a PR or CR had the option to receive a second course of lymphodepleting chemotherapy and brexucabtagene autoleucel if their disease subsequently progressed more than 3 months after the initial brexucabtagene autoleucel infusion, provided that the relapse was confirmed to be CD19 positive.¹⁸

Concomitant Medications

Investigators could prescribe medications or treatments deemed necessary to provide adequate supportive care, including prophylactic antimicrobials, growth factor support, and routine anti-emetic prophylaxis. Targeted concomitant medications consisting of gamma globulin, immunosuppressives, anti-infective drugs, and vaccinations were to be recorded for 24 months or until disease progression, whichever occurred first.¹⁸

Corticosteroid therapy at a pharmacologic dose (≥ 5 mg/day of prednisone or equivalent doses of other corticosteroids) and other immunosuppressive drugs were to be avoided for 7 days before leukapheresis and 5 days before the brexucabtagene autoleucel infusion, unless these treatments were used for bridging therapy, and were also to be avoided for 3 months after brexucabtagene autoleucel infusion, unless used to manage brexucabtagene autoleucel–related toxicities. Other medications that could interfere with the evaluation of brexucabtagene autoleucel, such as nonsteroidal anti-inflammatory agents, were also to be avoided for the same period unless medically necessary. Treatment for the patient’s lymphoma other than what was defined or allowed in the protocol, such as chemotherapy, immunotherapy, targeted agents, radiation, high-dose corticosteroids, and other investigational agents, was prohibited except as needed for treatment of disease progression after brexucabtagene autoleucel infusion.¹⁸

Outcomes

A list of efficacy end points identified in the CADTH review protocol that were assessed in the clinical trials included in this review is provided in Table 9. These end points are further summarized in the sections that follow. A detailed discussion and critical appraisal of the outcome measures is provided in Appendix 3.

Objective Response Rate

The primary efficacy end point was the ORR, defined as the incidence of CR or PR using central assessment per the Lugano classification. The original protocol had a primary end point of ORR per investigator assessment, which was amended to the Lugano classification based on independent radiological review on November 13, 2017. To maintain consistency, all subsequent patients enrolled in Cohort 1 were assessed by the investigators per the International Working Group (IWG) 2007 criteria as a secondary outcome. Patients were to have their first PET/CT scan to assess for disease 4 weeks after infusion of brexucabtagene autoleucel. Scans were repeated every 3 months, and as needed for patients who displayed symptoms suggestive of disease progression. Evaluations of bone marrow were also needed to confirm CR.^17,18

Best Objective Response

Incidence of BOR (CR, PR, stable disease, progressive disease, and not evaluable) were calculated using central assessment per the Lugano classification,¹⁷ and investigator assessment per the IWG 2007 criteria.^18,47

Duration of Response

Duration of response was evaluated only for patients who had an objective response (CR or PR) and was defined as the time from the first objective response to disease progression or death, using disease assessments (central and investigator assessment).¹⁸

Progression-Free Survival

Progression-free survival was defined as the time from the date of brexucabtagene autoleucel infusion to the date of disease progression or death from any cause using both central assessment and investigator assessment. In the FAS, PFS was defined as the time from the enrolment date to the date of disease progression or death from any cause.¹⁸

Overall Survival

Overall survival was defined as the time from the date of brexucabtagene autoleucel infusion to the date of death from any cause. In the FAS, OS was defined as the time from enrolment to the date of death from any cause.¹⁸

Health-Related Quality of Life: EuroQol 5-Dimensions Questionnaire

Patients’ HRQoL was assessed using the EuroQol 5-Dimensions questionnaire (EQ-5D) at screening (for baseline scores), week 4 (± 3 days), month 3 (± 1 week), and month 6 (during the long-term follow-up period) before any other assessments or procedures were performed. The EQ-5D was added to the schedule of assessments following a protocol amendment on August 23, 2016.¹⁸ For each health dimension in the EQ-5D, patients were instructed to select the severity level (no problems, slight problems, moderate problems, severe problems, or extreme problems) that best described their health status on the day that the questionnaire was administered. The EQ-5D also included the EuroQol Visual Analogue Scale (EQ VAS), in which patients rated their overall health status from 0 (representing “the worst health you can imagine”) to 100 (representing “the best health you can imagine”).¹⁸ A detailed discussion and critical appraisal of the EQ-5D measure is provided in Appendix 3.

Safety

Patients underwent safety assessments throughout the study. Investigators were to report all AEs and SAEs that occurred from enrolment through 3 months after brexucabtagene autoleucel infusion. Targeted AEs (e.g., neurologic, hematological, infections, autoimmune disorders, and secondary malignancies) were monitored and reported for 24 months after the brexucabtagene autoleucel infusion or until disease progression, whichever occurred first. All AEs were considered treatment emergent, defined as having had an onset on or after the brexucabtagene autoleucel infusion. Investigators were to identify CRS as a syndrome via case report form that was specifically designed to record CRS and to grade its severity according to a modification of the grading system proposed by Lee and colleagues.⁴⁸ All SAEs were reported from screening to 3 months after infusion with brexucabtagene autoleucel. Adverse events were investigator reported and were coded with the Medical Dictionary for Regulatory Activities, version 22.0, and the severity of AEs was graded using Common Terminology Criteria for Adverse Events, version 4.03. Cytokine release syndrome, neurologic AEs, cytopenias, infections, and hypogammaglobulinemia were of special interest in the ZUMA-2 trial protocol.¹⁸

Statistical Analysis

Sample Size and Power Calculation

Up to approximately 130 patients with R/R MCL were enrolled and treated in 2 cohorts. Cohort 1 was to include at least 60, and up to approximately 80, patients who received brexucabtagene autoleucel at a dose of 2 × 10⁶ anti-CD19 CAR T cells per kilogram (an additional 10 patients were treated with axicabtagene ciloleucel; these patients are reported separately from the results of ZUMA-2, as per the protocol amendment on June 22, 2018). The same protocol amendment clarified that the primary analysis was conducted after 60 patients in Cohort 1 were treated with brexucabtagene autoleucel and had the opportunity to be assessed for response 6 months after the week 4 disease assessment. A sample size of 60 patients in Cohort 1 had at least 96% power to distinguish between an active therapy with a true response rate of 50% or higher and a therapy with an ORR of 25% or less, with a 1-sided alpha level of 0.025.¹⁸

Interim and Final Analyses

Four interim analyses were performed for Cohort 1 and were reviewed by an independent data and the safety monitoring board (DSMB). The interim analyses for Cohort 1 included¹⁸:

• Interim analysis 1, conducted after 10 patients in this cohort had been treated with anti-CD19 CAR T cells and had had the opportunity to be followed for 30 days. The DSMB reviewed these data for safety only.

• Interim analysis 2, conducted after 20 patients in this cohort had had the opportunity to be evaluated for response 3 months after treatment with anti-CD19 CAR T cells. The DSMB reviewed these data for safety and futility and made study conduct recommendations based on the risk versus benefit of anti-CD19 CAR T-cell treatment. A rho (parameter = 0.30) beta spending function was used to allocate the beta level between the futility analysis at Cohort 1, interim analysis 2, and the primary efficacy analysis. The nonbinding futility boundary for this interim analysis would have been reached if no more than 5 of the 20 patients achieved a response. The criteria for futility were not met, and the DSMB recommended that the study continue.

• Interim analysis 3, reviewed by manufacturer to assess the accumulating data of efficacy and safety. This analysis was performed after 38 patients in Cohort 1 had had the opportunity to be evaluated for response 6 months after the anti-CD19 CAR T-cell infusion.

• Interim analysis 4, which occurred after 44 patients in this cohort had had the opportunity to be followed for 30 days after the anti-CD19 CAR T-cell infusion. The DSMB reviewed these data for safety only and focused on the first 6 patients treated with brexucabtagene autoleucel when enrolment in this cohort resumed in June 2018.

Two data cut-off analyses were included in this report. The original data cut-off date for the ZUMA-2 trial was July 24, 2019, representing a median follow-up of 12.3 months.¹⁸ The updated analysis data cut-off date was December 31, 2019, representing a median follow-up of 16.8 months.¹⁹ No additional patients were enrolled between the first and second data cut-offs, and the second data cut-off solely represents a longer follow-up time.

Primary End Point Analysis

Originally, Cohort 1 was designated as the nonpivotal cohort (November 13, 2017, protocol amendment), which was subsequently changed as the primary end point of ORR was updated to be based on the Lugano classification by an independent radiology review. In the fifth protocol amendment on June 22, 2018, it was clarified that Cohort 1 was again the pivotal cohort, and the primary analysis would be conducted after 60 patients treated with brexucabtagene autoleucel had been evaluated for response 6 months after their week 4 disease assessment (IAS).

The primary efficacy end point of ZUMA-2 was ORR, defined as the incidence of CR or PR using central assessment per the Lugano classification,¹⁷ which was updated from investigator review using the 2007 IWG criteria as of the November 13, 2017, protocol amendment.¹⁸ Confidence intervals for the ORR were calculated using the Clopper-Pearson method (an exact interval), Wilson method (sensitivity analysis), Agresti-Coull method (sensitivity analysis), and modified Jeffrey method (sensitivity analysis). Originally, the protocol’s hypothesis was that the response rate was greater than 20%.⁴³ The protocol amendment as of June 22, 2018, specified that the hypothesis was that brexucabtagene autoleucel ORR using central assessment would be significantly higher than the pre-specified historical control rate of 25%, derived from 2 retrospective studies.^35,37 This hypothesis was tested in the IAS of Cohort 1 at the 1-sided significance level of 0.025 using an exact binomial test.¹⁸ The primary end point of the study was not controlled for multiple comparisons.

Secondary End Point Analysis

For consistency before the protocol amendment, ORR was also calculated using investigator assessment and the 2007 IWG criteria. Incidence of BOR was calculated using the same methods used for ORR, with CR, PR, stable disease, progressive disease, and not evaluable as best response to treatment per the Lugano classification by central assessment and investigator assessment per the 2007 IWG criteria. Secondary outcomes were not tested against the historical control nor were they controlled for type I error. Concordance between central and investigator assessments were conducted for ORR and BOR using a kappa statistic and 2-sided 95% CI.¹⁸

Duration of response estimates were determined using the Kaplan–Meier approach and derived using disease assessments (per central and investigator assessment). The DOR for patients who had a new anticancer therapy (including SCT) was censored at the last evaluable disease assessment date before the initiation of the new anticancer therapy. The follow-up time for DOR was estimated using the reverse Kaplan–Meier approach described by Schemper and Smith (1996).⁴⁹ A sensitivity analysis of DOR was conducted in which disease assessments obtained after SCT (for patients who received an SCT while in a brexucabtagene autoleucel–induced response) were used in the derivation of DOR.¹⁸

Kaplan–Meier plots, estimates, and 2-sided 95% CIs were generated for PFS, and the proportion of patients alive and progression-free at 3-month intervals was estimated. The number of patients censored and the reasons for censoring were summarized. Patients who were alive as of the data cut-off date and who had not met criteria for progression were censored at their last evaluable disease assessment date. Patients who went on to receive a new anticancer therapy (including SCT) while in response were censored at their last evaluable disease assessment date before receiving the anticancer therapy. A sensitivity analysis was conducted in which disease assessments obtained after SCT (for patients who received an SCT while in a brexucabtagene autoleucel–induced response) were included in the derivation of PFS.¹⁸

For OS, patients who had not died by the data cut-off date were censored at the last date they were known to be alive or at the data cut-off date, whichever was earlier.¹⁸

EQ-5D and EQ VAS scores were summarized at baseline and at each post-treatment visit using descriptive statistics.

Safety analyses included all the patients who had received treatment and were summarized using descriptive statistics.

Secondary and exploratory end points of the study were not controlled for multiplicity.

Subgroup Analyses

A priori subgroup analyses of the ORR, DOR, PFS, and OS were conducted for the following baseline covariates. Subgroups were analyzed to evaluate the robustness and consistency of treatment effects found overall.

• ECOG PS at baseline

• Age at baseline (< 65 years, ≥ 65 years)

• Sex

• Race

• R/R subgroup

• Morphologic characteristics

• Ki-67 index

• CD19 +

• t(11;14)

• Cyclin D1 overexpression

• Disease stage and extent

• s-MIPI

• Number of prior regimens

• Prior BTK inhibitors

• Prior therapy regimens

• Tumour burden

Analysis Populations

The following analysis populations were defined in ZUMA-2¹⁸:

• The FAS, which included all enrolled patients that underwent leukapheresis (n = 74). This analysis set was used for the summary of patient disposition, as well as for analyses of ORR and other key efficacy end points (BOR, DOR, PFS, and OS).

• The IAS, which consisted of the first 60 patients treated with brexucabtagene autoleucel in Cohort 1 who had had the opportunity to be evaluated for response 6 months after the week 4 disease assessment after brexucabtagene autoleucel infusion. This analysis set was used for efficacy analyses in Cohort 1 and the hypothesis testing of the primary end point ORR at the time of the primary analysis (n = 60).

• The safety analysis set and mITT analysis sets, which were identical and included all patients treated with any dose of brexucabtagene autoleucel (n = 68).

Results

Patient Disposition

ZUMA-2 was a single-arm, open-label, phase II clinical trial. Table 10 summarizes the disposition of enrolled patients. Seventy-four patients were enrolled and leukapheresed in Cohort 1 of ZUMA-2. In the FAS, brexucabtagene autoleucel was successfully manufactured for 96% of patients. Sixty-nine patients (93%) received lymphodepleting chemotherapy, and 68 patients (92%) received brexucabtagene autoleucel, making up the mITT and safety populations. One patient was not treated with brexucabtagene autoleucel after receiving lymphodepleting chemotherapy due to ongoing active atrial fibrillation. Of the patients who received brexucabtagene autoleucel, 38% received bridging therapy. As of the July 24, 2019, data cut-off, the median potential follow-up time (calculated as the time from KTE-X19 infusion to the data cut-off date) from the brexucabtagene autoleucel infusion was 11.6 months (range = 1.9 months to 32.3 months) in the FAS (12.3 months in the IAS), and 16 of the 68 patients who received brexucabtagene autoleucel had died: 14 due to progressive disease, and 2 due to AEs (organizing pneumonia and staphylococcal bacteremia).¹⁸

As of the December 31, 2019, data cut-off, the median potential follow-up time from brexucabtagene autoleucel infusion was 16.8 months in the FAS, and 18 patients (24%) had died.¹⁹

Patients in the FAS were leukapheresed a median of 16.0 days (range = 5 days to 274 days) after study screening, and the median time from screening to leukapheresis was also 16.0 days (range = 5 days to 274 days). The median time to from leukapheresis to administration of brexucabtagene autoleucel was 27.0 days (range = 19 days to 134 days). As per the study protocol, patients were to remain in hospital for at least 7 days following brexucabtagene autoleucel infusion. The mean duration of hospitalization was 21.2 days (SD: 14.9) (median = 15 days; range = 8 days to 87 days).¹⁸

Exposure to Study Treatments

Exposure to study treatments is summarized in Table 11. Bridging therapy was administered to patients at the discretion of the treating investigator. A total of 25 patients (37%) received bridging therapy. The most commonly administered bridging therapies were ibrutinib (n = 14 [21%]) and dexamethasone (n = 12 [18%]).¹⁸

All patients received the planned total body surface area–adjusted dose of cyclophosphamide (1,500 mg/m²), and patients received a median total body surface area–adjusted dose of fludarabine of 90 mg/m² (range = 69 mg/m² to 90 mg/m²). One patient received a total body surface area–adjusted dose of fludarabine that was reduced to 69 mg/m² due to elevated creatinine. All other patients received within 10% of the planned total dose.¹⁸

The median weight-adjusted dose of brexucabtagene autoleucel was 2.0 × 10⁶ anti-CD19 CAR T cells per kilogram. Two patients did not receive within 10% of the target dose, 1 of whom had progressive disease soon after leukapheresis and received a dose of 0.6 × 10⁶ in lieu of being leukapheresed a second time. This patient was not among the first 60 patients to be treated with brexucabtagene autoleucel and, therefore, was not included the IAS.¹⁸

Efficacy

Only those efficacy outcomes and analyses of subgroups identified in the review protocol are reported below.

Primary Efficacy End Point(s)

ORR Using Central Assessment: Primary Analysis Data Cut-Off — July 24, 2019

The ZUMA-2 trial met its primary efficacy end point of ORR per central assessment in the IAS. A summary of ORR results in the IAS and FAS is included in Table 12. The ORR in the IAS was 93% (95% CI, 83.8% to 98.2%) versus the pre-specified historical control rate of 25% (P < 0.0001), and the CR rate was 67% (95% CI, 53.3% to 78.3%). Twenty-four patients (57%) who initially had PR or stable disease went on to achieve a CR after a median of 2.2 months (range = 1.8 months to 8.3 months).¹⁸

In the FAS, the ORR was higher than the pre-specified historical control rate at 85% (95% CI, 75.0% to 92.3%), with a CR rate of 59% (95% CI, 47.4% to 70.7%).¹⁸

ORR Using Central Assessment: Updated Analysis Data Cut-Off — December 31, 2019

As of the longer-term data cut-off, with a median follow-up of 16.8 months (FAS), 55 patients (92%) in the IAS achieved an ORR (Table 12), which differed from the primary analysis (93%) due to a change in the disease response assessment for 1 patient. As with the primary analysis, the 92% ORR observed in the IAS and || ORR in the FAS were significantly higher than the pre-specified historical control rate of 25%.¹⁹

Subgroup Analysis: Primary Analysis Data Cut-Off — July 24, 2019

Subgroups of interest outlined in the review protocol (Table 6) for ORR in the IAS are summarized in Table 13. Across all subgroups of interest, the ORRs ranged from 80% to 100%, compared with the ORR of 93% that was observed for all patients.¹⁸

Subgroup Analysis: Updated Analysis Data Cut-Off — December 31, 2019

Subgroup data for the IAS as of the December 31, 2019, data cut-off were consistent with the results observed for subsets in the primary analysis.¹⁹

Secondary Efficacy End Points

ORR Using Investigator Assessment: Primary Analysis Data Cut-Off — July 24, 2019

In the IAS, the ORR using the investigators’ assessment of response per IWG 2007 criteria was 88% (95% CI, 77.4% to 95.2%), and the CR rate was 70% (95% CI, 56.8% to 81.2%).¹⁸

DOR: Primary Analysis Data Cut-Off — July 24, 2019

Duration of response in the IAS and FAS is summarized in Table 14. The median time to CR or PR using the Lugano classification was 1.0 month (range = 0.8 months to 3.1 months), and the median time to achieve a CR was 3.0 months (range = 0.9 months to 9.3 months). As of the July 24, 2019, data cut-off and a median follow-up time for DOR of 8.6 months (reverse Kaplan–Meier approach), the median DOR was not reached. Thirty-nine patients (65%) were censored due to ongoing response, receipt of non-SCT anticancer therapy or allogeneic SCT, and consent withdrawal. Among those who were not censored, 14 patients (25%) had disease progression and 3 patients (5%) died. The median DOR was not reached in the 40 patients who achieved a CR, as 52% of patients had an ongoing CR, and only 7 patients had disease progression. For the 16 patients who achieved a PR, the median DOR was 2.2 months (95% CI, 1.4 months to not estimable). Six patients were censored: 3 (5% of the IAS) had an ongoing PR, 2 started a non-SCT anticancer therapy, and 1 withdrew consent. Of those not censored, 7 patients had disease progression and 3 patients died.¹⁸

In the FAS, the median DOR was not reached with a median follow-up of 8.1 months (Table 14). Forty-five patients (71%) were censored: 39 (53%) in the FAS had an ongoing response (62% of patients with objective response), 3 started a non-SCT anticancer therapy, 2 had an allogeneic SCT, and 1 withdrew consent. Of the 18 patients who experienced progression or death, 15 had disease progression and 3 died.¹⁸

Sensitivity analysis of DOR was conducted in patients who received an allogeneic or autologous SCT while in response and were censored at their last evaluable disease assessment date after SCT. As of the July 24, 2019, data cut-off, 1 patient had received an allogeneic SCT while in remission induced by brexucabtagene autoleucel. Excluding this patient, and with a median follow-up of 8.6 months, the median DOR was not reached. Thirty-eight patients were censored compared to the 39 censored in the primary analysis: 34 (57%) had an ongoing response, 3 started a non-SCT anticancer therapy, and 1 withdrew consent.¹⁸

DOR: Updated Analysis Data Cut-Off — December 31, 2019

With a median follow-up time for DOR of 14.1 months, the median DOR was still not reached (95% CI, 13.6 months to not estimable), and the proportion of responders remaining in response at 6 months and 12 months was 77.2% and 65.6%, respectively. In patients who achieved a CR, the median DOR was still not reached (95% CI, 14.4 to not estimable), with a median follow-up time of 13.8 months (95% CI, 11.4 months to 26.5 months). Similarly, in the FAS, the median DOR was |||||||||||| (95% CI, ||||||||||||||||||||||) after 13.8 months follow-up (Table 15), and the proportion of responders remaining in response was similar to that of the IAS.¹⁹

PFS: Primary Analysis Data Cut-Off — July 24, 2019

Results of PFS for the IAS and FAS are summarized in Table 16. As of the July 24, 2019, data cut-off (12.3 months follow-up), the median PFS in the IAS was not reached (9.2 to not estimable) (Figure 3a). Kaplan–Meier estimates of PFS rates at 6 months, 12 months, and 24 months were 77.0%, 60.9%, and 56.9%, respectively. In the FAS, the median PFS was not reached (9.9 to not estimable) and the 6-month, 12-month, and 24-month PFS rates were 75.2%, 55.6%, and 51.9%, respectively.¹⁸

Sensitivity analysis of PFS was conducted for patients who received an allogeneic or autologous SCT while in response and were censored at their last evaluable disease assessment date after SCT. As of the July 24, 2019, data cut-off, 1 patient had received an allogeneic SCT while in a brexucabtagene autoleucel–induced remission and was censored from the analysis. Similar to the overall IAS, the median PFS in the remaining patients was not reached with a median follow-up of 12.3 months, and the 6-month and 12-month PFS rate estimates were 77.2% and 59.4%.¹⁸

PFS: Updated Analysis Data Cut-Off — December 31, 2019

Results of PFS for the IAS and FAS as of the December 31, 2019, data cut-off are summarized in Table 16. As of the updated analysis, the median PFS in the IAS was still not reached (95% CI, 9.6 to not estimable; Figure 3b), and the Kaplan–Meier estimates of PFS rate at 6 months and 12 months were similar to those at the primary data cut-off, at 76.8% and 62.2%, respectively. The 24-month and 33-month PFS rate estimates were 55.5% and 50.5%, respectively.

In the FAS, the median PFS was ||||||||||||, and the 6-month,12-month, 24-month, and 33-month PFS rates were ||||||||||||||||||||||||||||||||||||||, respectively.¹⁹

Figure 3: PFS Using Central Assessment per Lugano Classification: IAS

CI = confidence interval; IAS = inferential analysis set; NE = not estimable; PFS = progression-free survival.

Note: (a) Kaplan–Meier curve of PFS on the July 24, 2019, data cut-off; (b) Kaplan–Meier curve of PFS on the December 31, 2019, data cut-off.

Source: ZUMA-2 Clinical Study Report¹⁸; ZUMA-2 updated analysis.¹⁹

Progression-free survival rate estimates for patients in the IAS with CR and PR are summarized in Table 17. As of the July 24, 2019, data cut-off, the PFS rates at 6 months, 12 months, and 24 months were 100%, 76.6%, and 70.2% in patients with CR, and the median PFS was not reached (95% CI, 14.5 to not estimable). Among patients who achieved a PR, the PFS rate estimates at 6 months and 12 months were each 25.9%, and the median PFS was 3.1 months (95% CI, 2.3 to not estimable).¹⁸

As of the December 31, 2019, data cut-off, the median PFS was not reached for patients who achieved a CR (95% CI, 15.3 months to not estimable) and was identical to the median PFS reported for patients with PR in the primary analysis.¹⁹

Subgroup Analysis: Primary Analysis Data Cut-Off — July 24, 2019

Subgroups of interest outlined in the review protocol (Table 6) of the IAS for PFS are summarized in Table 18. Across subgroups of interest to the review, the PFS rate ranged from 69% to 100%, which was comparable to the 77% PFS rate observed for the overall IAS population.¹⁸

OS: Primary Analysis Data Cut-Off — July 24, 2019

Overall survival results for the IAS and FAS are summarized in Table 19. As of the July 24, 2019, data cut-off (12.3 months follow-up), the median OS was not reached (95% CI, 24.0 to not estimable) (Figure 4a) in the IAS, and the OS rates at 6 months, 12 months, and 24 months were 86.7%, 83.2%, and 66%, respectively. In the FAS, the 6-month, 12-month, and 24-month OS rates were 83.2%, 77.1%, and 64.3%, and the median OS was not reached (95% CI, 21.1 to not estimable) (Figure 4b).¹⁸

OS: Updated Analysis Data Cut-Off — December 31, 2019

Results of OS for the IAS and FAS as of the December 31, 2019, data cut-off are summarized in Table 19. As of the updated analysis, the median OS in the IAS was still not reached (Figure 4c), and the Kaplan–Meier estimates of OS at 6 months and 12 months were 86.7% and 83.3% and in the 24-month and 36-month estimates were 68.8%. These were comparable to the estimates in the primary analysis. In the FAS, the median OS was also ||||||||||||, and based on Kaplan–Meier estimates, the survival rate was |||||||||||||||||||||||||||||||||||||||||| at 6 months, 12 months, 24 months, and 36 months, respectively.¹⁹

Figure 4: OS Kaplan–Meier Estimates: IAS and FAS

CI = confidence interval; FAS = full analysis set; IAS = inferential analysis set; NE = not estimable; OS = overall survival.

Note: (a) OS in the IAS (data cut-off July 24, 2019); (b) OS in the FAS (data cut-off July 24, 2019); (c) OS in the IAS (data cut-off December 31, 2019); (d) OS in the FAS (data cut-off December 31, 2019).

Source: ZUMA-2 Clinical Study Report¹⁸; ZUMA-2 updated analysis.¹⁹

Overall survival rate estimates for patients in the IAS with a BOR of CR and PR are summarized in Table 20. Among patients who achieved a CR, OS rate estimates at 6 months and 12 months were 100% and 97.2%, and the median OS was not reached. Among patients who achieved a PR, OS rate estimates at 6 months and 12 months were 62.5% and 56.3%, and the median OS was 19.9 months (3.8 months to not estimable).¹⁸

As of the December 31, 2019, data cut-off, the median OS for patients in the IAS with a CR was not reached and the median OS for patients with a PR was 12.6 months (95% CI, 3.3 months to not estimable). In the FAS, the median OS for patients with a CR was not reached and the median OS for patients with a PR was 13.5 months (95% CI, 5.4 months to not estimable).¹⁹

Subgroup Analysis

Subgroups of interest outlined in the review protocol (Table 6) of the IAS for OS are summarized in Table 21. At the July 24, 2019, data cut-off, patients had a median follow-up of 12.3 months, and 53% of patients had at least 12 months of follow-up. Across evaluable subgroups of interest, the OS rate at month 12 (where the OS rate could be estimated) ranged from 71% to 100%, which was comparable to the 83% rate for the overall IAS population.¹⁸

Health-Related Quality of Life

A summary of the proportion of patients in the safety analysis set (N = 68) who completed the questionnaire for each of the 5 domains of the EQ-5D is shown in Table 22. A total of 62 patients (91%) completed the questionnaire at study screening. The majority of responses were available at baseline or screening, with 66% to 95% of patients reporting no health problems on any of the 5 domains, with the highest percentages of no health problems observed for self-care (95%), mobility (85%), and usual activity (82%).¹⁸

The proportion of patients reporting severe health problems generally increased through week 4, and the largest decrease in percentage points of patients reporting no health problems between screening and week 4 was observed in the usual activities domain (39 percentage points), with increases in slight and moderate problems doing usual activities of 10 and 17 percentage points, respectively. Conversely, by month 3, increases of up to 26 percentage points were seen from week 4, and increases in the proportion of patients reporting no health problems increased at 3-month and 6-month assessments, demonstrating an improved HRQoL over time.¹⁸ Given the limited number of questionnaire respondents at longer follow-up durations, trends in HRQoL relative to earlier time points cannot be interpreted.

The median EQ VAS score was 85.0 (range = 45 to 100) at screening, which decreased to 78.0 (range = 38 to 100) at week 4, followed by higher median scores at month 3 (83.0; range = 40 to 100) and month 6 (90.0; range = 20 to 100). The proportion of patients with a decrease of 10% or more in EQ VAS scores relative to screening was 50% at week 4, 29% by month 3, and 12% by month 6,¹⁸ suggesting that HRQoL decreases following infusion but improves over time. Inferences on the trend for improvement in HRQoL over time cannot be made given the decrease in the number of respondents at longer time points (65 at screening versus 42 at 6 months).

Harms

Adverse Events

AEs: Primary Analysis Data Cut-Off — July 24, 2019

At least 1 TEAE was reported in all patients, of which 67 (99%) had AEs that were Grade 3 or higher. In the IAS, 11 patients (16%) had Grade 3 TEAEs, 52 patients (76%) had Grade 4 TEAEs, and 4 patients (6%) had Grade 5 TEAEs. The most common Grade 3 or higher AEs were anemia and decreased neutrophil count (50%) and decreased WBC count (39%). The most common AEs by preferred term were pyrexia (94%), anemia (68%), and decreased platelet count (53%); these AEs are summarized in Table 23.

Adverse events related to brexucabtagene autoleucel that occurred in at least 15% of patients are listed in Table 24. A total of 21 patients (31%) had brexucabtagene autoleucel–related AEs that were Grade 3, and 32 patients (47%) had such AEs that were Grade 4. The most common brexucabtagene autoleucel–related AEs that were Grade 3 or higher were decreased WBC count (31%), anemia (28%), and decreased neutrophil count (26%).¹⁸

AEs: Updated Analysis Data Cut-Off — December 31, 2019

Treatment-emergent AEs as of the December 31, 2019, data cut-off are summarized in Table 25. In the updated analysis, the most common Grade 3 or higher AEs were slightly higher than the primary analysis and included decreased neutrophil count (||), anemia (||), and decreased WBC count (||). The most common Grade 5 AEs by preferred term in the updated analysis included B-cell lymphoma (　|　), staphylococcal bacteremia (||), and organizing pneumonia (||).¹⁹

Serious Adverse Events

SAEs: Primary Analysis Data Cut-Off — July 24, 2019

Serious AEs occurring in greater than or equal to 5% of patients in the safety analysis set of the ZUMA-2 trial are summarized in Table 26. Serious AEs were recorded for 68% of patients. The most common SAEs were encephalopathy and pyrexia (22% each), followed by hypotension (16%). The most common Grade 3 or higher SAEs were encephalopathy (18%) and hypotension and hypoxia (12% each). Two patients had Grade 5 SAEs of B-cell lymphoma, and these patients died due to disease progression.¹⁸

SAEs: Updated Analysis Data Cut-Off — December 31, 2019

The incidence of SAEs in the updated analysis is summarized in Table 27. Grade 3, Grade 4, and Grade 5 SAEs were observed in ||, ||, and 6% of patients, respectively. Similar to the primary analysis, the most common SAEs in the current analysis were pyrexia (||), encephalopathy (||), and hypotension (||), although incidence was slightly lower. The most common Grade 3 or higher SAEs in the current analysis were encephalopathy (||); pneumonia, which increased to || from ||; and hypotension (||).¹⁹

Mortality

As of the July 24, 2019, data cut-off, 16 patients (24%) had died, of whom 13 (19%) died due to progressive disease, 2 (3%) died due to AEs (organizing pneumonia and staphylococcal bacteremia, both deemed related to lymphodepleting chemotherapy), and 1 (1%) was listed as “other” as the cause of death was unknown; this was later changed to “progressive disease.” The majority of deaths occurred more than 3 months after brexucabtagene autoleucel infusion (12 of 16 deaths).¹⁸

Two additional deaths occurred between the primary analysis and the updated data cut-off of December 31, 2019, for a total of 18 deaths (26%). The 2 additional deaths were attributed to progressive disease (n = 16 [24%]) and occurred more than 3 months after infusion.¹⁹

Notable Harms/AEs of Special Interest

Cytokine Release Syndrome

Cytokine release syndrome and symptoms by preferred term are summarized in Table 28 and were all considered to be related to treatment with brexucabtagene autoleucel. Sixty-two patients (91%) had CRS, of which the majority were Grade 2 (47%). No patient had Grade 5 CRS. The most common CRS symptoms of any grade were pyrexia (100%), hypotension (56%), and hypoxia (37%). The most common Grade 3 or higher CRS symptoms were also hypotension (25%), hypoxia (19%), and pyrexia (11%). The median time to onset of CRS was 2 days after the brexucabtagene autoleucel infusion, with a median duration of 11 days (range = 1 day to 50 days). As of the July 24, 2019, data cut-off, CRS had resolved in all 62 patients.¹⁸

As of the December 31, 2019, data cut-off, no additional CRS symptoms had occurred.¹⁹

Neurologic AEs

Table 29 summarizes neurologic events in the safety analysis set. Forty-three patients (63%) had at least 1 neurologic event of any grade. The most frequently occurring neurologic events were Grade 3 (22%), while 6 patients (9%) had Grade 4 neurologic events, and no patients had a Grade 5 neurologic event. The most common neurologic events of any grade were tremor (35%), encephalopathy (31%), and confusional state (21%). The most common Grade 3 or higher neurologic events were encephalopathy (19%), confusional state (12%), and aphasia (4%). A total of 22 patients (32%) had serious neurologic events of any grade; 18% were Grade 3, and 9% were Grade 4. The most common serious neurologic event was encephalopathy (22%), followed by confusional state (7%) and aphasia (4%).

The median onset of a neurologic event following brexucabtagene autoleucel infusion was 7 days, and in patients whose neurologic events had resolved, the duration of neurologic events was a median of 12 days (range = 1 day to 567 days). Three patients had neurologic events beyond 200 days that were attributed to lymphodepleting chemotherapy and brexucabtagene autoleucel. As of the July 24, 2019, data cut-off, neurologic events had resolved in all but 6 patients. Neurologic events for 2 of these patients had not resolved at the time of death: 1 patient had Grade 2 nonserious agitation (deemed related to brexucabtagene autoleucel) and Grade 3 serious confusional state (deemed related to lymphodepleting chemotherapy and brexucabtagene autoleucel), and 1 patient had Grade 2 nonserious hyperesthesia. Ongoing neurologic events for the remaining 4 patients were Grade 1 or Grade 2.¹⁸

No additional neurologic events occurred between the primary analysis and the December 31, 2019, data cut-off, and no changes occurred in the most common neurologic events of any grade. The median time to resolution remained the same between data cut-offs; however, the mean time to resolution of neurologic events was 61.9 days in the updated analysis compared to 53.1 days in the primary analysis.¹⁹

Other Harms Outcomes

Immunoglobulins were included as concomitant medications regardless of whether they were administered before or after the hospital discharge date following treatment with brexucabtagene autoleucel. A total of 22 patients (32%) were treated with immunoglobulins.¹⁸

Hypogammaglobulinemia was a TEAE of interest in the ZUMA-2 trial. As of the July 24, 2019, data cut-off, 13 patients (19%) in the safety analysis set had hypogammaglobulinemia, of which 11 (16%) were considered Grade 2 and the majority occurred in patients 65 years and older (8 [21%]).¹⁸ There were no additional cases of hypogammaglobulinemia as of the updated analysis.¹⁹

Critical Appraisal

Internal Validity

ZUMA-2 was the only individual study included in the review and is the only prospective clinical trial in the R/R MCL setting post–BTK inhibitor treatment. This phase II, multi-centre, single-arm trial aimed to assess the efficacy and safety of brexucabtagene autoleucel in the treatment of R/R MCL when patients have progressed on prior chemotherapy, anti-CD20 antibody, and BTK inhibitor therapy. The following points were noted in the appraisal of the internal validity of the ZUMA-2 trial.

As a single-arm study, the ZUMA-2 trial lacks a comparator arm, which increases the risk of bias in the estimation of treatment effects due to the potential for confounding related to placebo response, fluctuations in health status, and other unidentified prognostic factors that could affect subjectively assessed outcomes. The open-label, single-arm design can increase the risk of bias in reporting of outcomes that are subjective in measurement and interpretation, such as response, HRQoL, and AEs. Outcomes such as mortality are less likely to be affected. The risk of bias due to the open-label design may be unavoidable given the rarity of the indication and unmet need in a population where larger head-to-head randomized controlled trials are not methodologically or ethically feasible. The potential for this bias was also reduced by using an independent central assessment for key study outcomes such as ORR, DOR, and PFS.

The primary end point of the ZUMA-2 trial was ORR per central assessment using the Lugano classification. The original study protocol used the primary end point of ORR based on investigator assessment using IWG 2007 criteria, which was changed to the Lugano classification by independent radiology review at the fourth protocol amendment on November 13, 2017. For the clinical response outcome of ORR, a historical control of 25% was applied as of the June 22, 2018, protocol amendment based on 2 retrospective studies published before the ZUMA-2 protocol (2015 and 2016), which demonstrated that patients with R/R MCL who had 3 or more prior lines of therapy before receiving the BTK inhibitors had ORRs to salvage therapy of approximately 25%. The ZUMA-2 study demonstrated an overall ORR of 93% (95% CI, 83.8% to 98.2%), which was significantly superior to the historical control. While the lower limit of the CI was significantly higher than the historical control rate, clinical experts consulted by CADTH noted that the 25% rate is largely underestimated, which has the potential to bias the comparison of the results in favour of brexucabtagene autoleucel. However, the expected response with standard of care would be dependent on how many prior lines of therapy have been received and so may be variable. Objective response and CR were also analyzed using investigator assessment. The concordance rate demonstrated agreement between investigator assessment and central assessment, with concordance rates of 95% (κ = 0.70; 95% CI, 0.39 to 1.00) and 90% (κ = 0.77; 95% CI, 0.60 to 0.94) for ORR and CR, respectively. The Lugano classification is preferred in B-cell lymphomas compared to the IWG 2007 criteria⁵⁰; however, it is unclear what impact the use of IWG 2007 criteria in the investigator assessment compared to the Lugano classification may have had on the direction or magnitude of results.

Health-related quality of life outcomes were not considered in a multiplicity adjustment and had significant missing data at later time points (i.e., 6 months), impacting the interpretability of trends over time and creating potential for bias in those that remain; the outcomes may therefore not be reflective of the overall population (i.e., patients who did not continue to complete assessments or who died tended to have poorer HRQoL). Therefore, the suggestion that HRQoL improves over time cannot be made. Also, no index scores for the EQ-5D were provided, and as such the net treatment effect across domains could not be captured.

Follow-up time (both potential and actual) as of the primary data cut-off was considered appropriate for assessing response and safety outcomes associated with brexucabtagene autoleucel. As per the protocol amendment on June 22, 2018, the primary analysis of the ZUMA-2 study was performed when 60 patients had been followed for 6 months after their 4-week assessment for ORR. However, the short median follow-up for the primary data cut-off of 11.6 months and 12.3 months in the FAS and IAS, respectively, was considered insufficient and immature for key survival outcomes of PFS and OS. It was acknowledged, however, that the median PFS and median OS were not reached in the IAS with 12.3 months or 16.8 months of follow-up and were considered impressive according to the clinical experts consulted by CADTH given the stage of disease and typical survival rates seen in this population.

Subgroups were defined a priori and were considered of clinical interest according to the experts consulted for this review. The subgroups were analyzed to evaluate the robustness and consistency of treatment effects found overall. However, many of the specific subgroups had limited sample sizes, for which results were imprecise, as reflected by the wider CIs, limiting the interpretability of the results. Furthermore, some subgroups (notably Ki-67 index and CD19 positivity) only reflected subpopulations of the overall study population, as there were missing data for certain covariates at baseline. The Ki-67 index is a known prognostic factor in MCL, and it has previously been shown that patients with Ki-67 greater than 50% have lower ORRs and CR rates.⁵¹ The Ki-67 index was evaluated in 3 ways (one of which was data driven using a median split rather than clinically relevant thresholds). It is unclear whether the analysis of patients with greater than or less than 30% or 50% thresholds were specified a priori, and the rationale for these cut-offs was unclear.

External Validity

The inclusion and exclusion criteria of ZUMA-2 aimed to enrol patients who have received second-line or later therapy for R/R MCL with prior treatments consisting of chemotherapy, anti-CD20 antibody, and BTK inhibitor therapy. There was a lack of Canadian sites in the ZUMA-2 trial, as well as no Canadian patients included in the trial. All co-interventions were readily available in Canadian clinical practice, and while there is no clear reason to assume major differences in the standards of practice and the patient population between the US and Canada, the lack of representation of Canadian patients could potentially reduce the generalizability to Canadian settings if practice variation exists. Brexucabtagene autoleucel was successfully manufactured for 96% of patients. The majority of patients in the ZUMA-2 trial received the target dose per the Health Canada indication (N = 66 [97%]). It is uncertain whether the time to infusion shown in the trial (median 27 days from leukapheresis) reflects what would be seen in the real world.

The outcomes used to inform the efficacy are clinically meaningful to the assessment of improvements in survival and are reflective of clinical practice in Canada. Despite the relevance of the efficacy end points to clinicians and patients, HRQoL — an outcome of particular importance to patients — was only assessed using the EQ-5D, which may not fully capture the impact of symptoms and treatment in R/R MCL.

The included population required patients to have an absolute neutrophil count greater than or equal to 1,000/μL, a platelet count greater than or equal to 75,000/μL, and an absolute lymphocyte count greater than or equal to 100/μL, which the clinical experts believed to be high; they noted that patients in a clinical setting would likely have lower counts than this, potentially impacting generalizability. They also noted that the ZUMA-2 trial population may represent a population that is generally healthier than would be found in clinical practice settings. As part of the inclusion criteria, patients were excluded if they had evidence of CNS lymphoma, or active CNS disease, and a history of significant cardiac disease. The clinical experts consulted by CADTH believed that should these aspects be under control, then these patients would also be eligible for treatment with brexucabtagene autoleucel.

Overall, the demographic and baseline characteristics of the included population generally reflect Canadian practice; however, the clinical expert panel noted that the ECOG PS of included patients may be lower than is seen in clinical practice, as 65% had an ECOG PS of 0, and they expect more ECOG PS 2 patients in the real world. Additionally, they noted that only 10% of patients had bulky disease at baseline, which given the disease stage and line of therapy, they would expect to be higher. Lastly, the proportion of patients who received bridging therapy was only 38%, which clinical experts considered to be low; they expected a higher proportion to require this. Together, although the ZUMA-2 population was reflective of the funding request and included R/R MCL patients who had failed BTK inhibitors, the selected patients may represent a less sick population and may not be generalizable to the typical Canadian patient at this stage of the disease.

Indirect Evidence

Objectives and Methods for the Summary of Indirect Evidence

The objective of this section is to provide an appraisal and summary of indirect evidence for brexucabtagene autoleucel versus other therapies in patients with R/R MCL. A review of indirect evidence was undertaken due to the lack of comparative clinical trials for brexucabtagene autoleucel.

A focused literature search for ITCs dealing with MCL was run in MEDLINE All (1946–) on January 25, 2021. No filters were applied to limit the retrieval by study type. Retrieval was not limited by publication date or by language. Articles were screened by 1 researcher for ITCs that met the patient, intervention, comparator, and outcome criteria listed in Table 6. In addition, the sponsor-submitted ITC was reviewed.

The literature search identified 5 articles, but none evaluated the efficacy or safety of brexucabtagene autoleucel in patients with MCL. The sponsor-submitted ITC, which was used to inform the pharmacoeconomic model, was appraised and summarized.⁴³

Description of Indirect Comparison

The sponsor submitted a MAIC that compared the efficacy of brexucabtagene autoleucel to standard of care treatments in patients with R/R MCL who had previously been treated with BTK inhibitors.⁴³

Methods of ITC

Objectives

The sponsor-submitted report had 2 objectives. The first was to conduct a systematic review and meta-analysis of studies to obtain estimates of the absolute treatment effects for OS, PFS, and tumour response of interventions considered standard of care for the treatment of R/R MCL after BTK inhibitor therapy. The second objective was to conduct a MAIC to estimate the relative treatment effects for brexucabtagene autoleucel versus standard of care treatments in patients with R/R MCL who had previously been treated with BTK inhibitors.

Study Selection Methods

The meta-analysis and ITC were informed by a systematic literature review conducted in 2018, with updates performed in 2019 and 2020. The study selection criteria and review methods have been summarized in Table 30. The population included in the literature review was broader than the population of interest for the ITC, and only studies that provided data for patients with R/R MCL who had received BTK inhibitor therapy (either the full study population or subgroup analysis) were included in the ITC.

ITC Analysis Methods

A feasibility assessment was conducted to review the standard of care studies identified in the systematic literature review. Its objective was to evaluate if the distribution of treatment, outcomes, study, and patient characteristics that may affect treatment response were sufficiently similar to permit pooling of the standard of care studies and if these characteristics were comparable to the ZUMA-2 trial of brexucabtagene autoleucel. The primary outcomes of interest were OS and PFS, and secondary outcomes were tumour response (objective response, CR, or PR). Only standard of care studies that reported outcome data for the target population that were suitable for meta-analysis were included.

Meta-Analysis of Standard of Care Studies

Data from the standard of care studies were pooled to obtain estimates of the absolute treatment effects for OS, PFS, and tumour response. For survival outcomes, Kaplan–Meier curves for OS or PFS, or individual patient data that could be used to construct the Kaplan–Meier curve, were required for pooling. Data from the Kaplan–Meier curves were digitized using Digitizelt software, and the number of patients at risk over time was extracted. Survival and censoring times for each study were created based on an algorithm proposed by Guyot et al.⁵³ Commonly used survival distributions (i.e., exponential, log-normal, log-logistic, Weibull, Gompertz, gamma, and generalized gamma) were fitted to the reconstructed individual patient data from each study, and the best-fitting distribution was selected based on the Akaike information criterion and visual inspection of the fitted and observed survival data. The model parameter estimates for OS and PFS for each trial were then pooled using a Bayesian multivariate meta-analysis model (modified methods based on Achana et al. [2014]⁵⁴). Bayesian models were run using 2 chains, with the first iterations discarded as burn-in (number of iterations not specified). No information was provided on prior distributions or model convergence diagnostics. Fixed- and random-effects models were used to estimate the pooled treatment effects, and the deviance information criterion was used to select the best-fitting model. Five sensitivity analyses were conducted for OS that excluded trials based on differences in the intervention or outcome definition (see Table 31 for details). No sensitivity analyses were conducted for PFS.

The meta-analysis of tumour response outcomes was based on the proportion of patients reporting objective response, CR, or PR in the standard of care studies. A frequentist generalized linear mixed model was used to estimate the pooled response rate outcomes. Both fixed- and random-effects models were estimated. Between-study statistical heterogeneity was assessed based on I² values, where values greater than 50% were considered high heterogeneity and where random-effects models were preferred. Six sensitivity analyses were conducted for tumour response outcomes (see Table 31 for details). Clinical heterogeneity was assessed through the feasibility study.

The mean baseline patient characteristics of the standard of care studies were pooled as a weighted average.

Naive ITC and MAIC

The use of unanchored MAIC methods to conduct the ITC were necessary, as the single-arm design of ZUMA-2 precluded the use of other anchored ITC methods. The ITC’s authors stated that since there is no standard therapeutic approach for patients who have relapsed following BTK inhibitor therapy, comparison to a control group that included various subsequent treatments was justified.

For the MAIC, a logistic propensity score model was used to estimate patient weights for the ZUMA-2 trial so that the weighted baseline mean characteristics of ZUMA-2 matched the pooled mean characteristics of the standard of care studies (Table 31). Covariates used for weighting were identified based on a literature review and input from clinical experts on prognostic factors in patients with R/R MCL. From the literature search, the following variables were identified to be prognostic based on univariate or multivariate Cox regression models: number of prior therapies, duration of prior BTK inhibitor therapy, response to prior BTK inhibitor therapy (ORR), MIPI, ECOG PS, bulky disease, primary refractory disease and lactate dehydrogenase. These variables, and all other baseline characteristics reported in the trials, were reviewed and ranked by clinical experts from the sponsor, and 12 variables were selected (number of prior therapies, prior autologous SCT, duration of prior BTK inhibitor therapy, response to BTK inhibitor therapy, MIPI, blastoid morphology, Ki-67 (≥ 30%, ≥ 50%), disease stage, prior ibrutinib therapy, male sex, extranodal disease, and bone marrow involvement). This list of covariates was validated by experts from the UK and Canada. After reviewing the results of initial analyses that were based on the full list of covariates, a decision was made to include only the 6 most relevant characteristics, as the ITC’s authors stated this “led to more conservative and clinically plausible results while also giving a higher effective sample size.”⁴³ Key opinion leaders were consulted, and the following covariates (in rank order) were included in the final propensity score model: number of prior therapies; prior autologous SCT; duration of prior BTK inhibitor therapy; response to prior BTK inhibitor therapy; blastoid morphology: and Ki-67 (≥ 30%, ≥ 50%). All but 1 of the standard of care studies were missing baseline data for at least 1 key characteristic used to determine patient weights in the MAIC: 3 studies were missing data for 2 covariates, 3 studies were missing data for 3 or 4 covariates, and 1 study was missing data for 6 covariates. To address the missing baseline data, a number of assumptions were made. The weighted average for specific characteristics was assumed to apply to any trials that were missing that parameter. For 3 trials where the population of interest was a subgroup and only the overall study population characteristics were reported, it was assumed that the overall population characteristics (i.e., the characteristics of those who did or did not receive therapy post–BTK inhibitor therapy) were representative of patients who had received subsequent therapy. Not all studies reported baseline characteristics at the start of subsequent therapy (e.g., at diagnosis or at start of BTK inhibitor therapy). It was assumed that the differences in the time of measurement did not have an impact on prognosis.

For OS and PFS, the same parametric survival function used in the meta-analysis was fitted to the weighted individual patient data from ZUMA-2, and the best-fitting model was determined based on the Akaike information criterion. The best-fitting parametric survival models for the weighted brexucabtagene autoleucel data and the pooled standard of care studies were then pooled in a pairwise indirect comparison to provide an estimate of the relative treatment effects of brexucabtagene autoleucel versus standard of care therapy. Results were reported as the average HR and 95% CI. For the tumour response outcomes, treatment effects were estimated for the weighted ZUMA-2 data versus pooled standard of care data using weighted contingency table methods and reported as odds ratios and 95% CIs.

The sponsor-submitted report also includes a naive ITC that compared unadjusted data from ZUMA-2 to the pooled outcomes from the standard of care studies.

For the ITC, follow-up time was truncated at 39 months, which was the longest follow-up time available for the ZUMA-2 study. The mean survival was defined as the area under the curve of the survival function from 0 months to 39 months. Based on the December 2019 update in the ZUMA-2 study, the median potential follow-up time was 21.7 months (from infusion to data cut-off). Overall, 88% of patients had completed 12 months of potential follow-up time, 41% had completed 24 months, 41% had completed 30 months, and 2 had completed 36 months.¹⁹ Three ZUMA-2 populations were estimated: the FAS (N = 74), the IAS (N = 60), and the safety set (N = 68). Effective sample size and baseline characteristics before and after weighting were reported for each scenario. The analyses were conducted using R software (version 3.6.1; meta, metaphor, and flexsurv packages), with Markov chain Monte Carlo methods implemented using JAGS software (version 4.3.0).

Results of ITC

Summary of Included Studies

Fifteen studies in patients with R/R MCL post–BTK inhibitor therapy were identified in the systematic review. Two studies were excluded due to small sample size (≤ 5 patients), and 1 was excluded as it evaluated ibrutinib as bridging therapy, which was not an intervention of interest.

The feasibility assessment included the ZUMA-2 trial and 11 other studies of standard of care treatments that were identified in the literature review. Based on the feasibility assessment, 2 trials were excluded from the meta-analysis because patients had received CAR T-cell therapy, which was not considered standard of care, and 1 trial was excluded in which there was substantial overlap between the patients enrolled with another study³⁸ included in the analysis. Thus, 9 trials were included in the ITC (Table 32). The trials comprised 7 retrospective observational studies,^{35,36,38-40,55,56} a single arm from a randomized controlled trial,⁵⁷ and ZUMA-2, which was a prospective uncontrolled phase II clinical trial.⁴⁵ One study was available only as a poster, and thus limited data were available.⁵⁶ Another study was published as a letter to the editor.⁵⁵ All other studies were published in full text. Individual patient data were available for the ZUMA-2 study.

Across the 9 included studies, the sample size ranged from 12% to 73% of patients who had received subsequent therapy following BTK inhibitor treatment. The median number of prior therapies that patients had received for MCL ranged from 2 to 4. In the standard of care studies, the proportion of patients who stopped BTK inhibitor therapy due to progression ranged from 40% to 100%, and in 5 studies, 6% to 25% of patients stopped due to intolerance. For 4 studies, the subsequent therapies included mixed treatments, which consisted of various chemoimmunotherapies or systemic therapies (see Table 32). In the other 4 trials, the subsequent therapies included lenalidomide-based treatments, venetoclax, R-BAC, or R-iBVD, each reported in 1 study.

The median age of patients was 65 years for the ZUMA-2 trial and ranged from 63 years to 71 years in the other studies. Patients were predominantly male (standard of care: range = 72% to 85%; ZUMA-2: 84%). The proportion of patients with blastoid morphology ranged from 12% to 31% in the standard of care studies (not reported for 3 trials), compared with 25% for the ZUMA-2 study. In the ZUMA-2 trial, 88% of patients had Ki-67 greater than 30%, and in the standard of care studies that percentage ranged from 63% to 92% (not reported for 4 studies). For 3 trials,^36,56,57 the population of interest consisted of a subgroup of the overall study population, and no baseline patient characteristics were reported for these patients.

Four standard of care trials provided OS outcome data,^35,38,40,55 2 trials reported PFS data,^40,55 and 8 trials^{35,36,38-40,55-57} reported tumour response outcome data that were suitable for analysis.

In the ZUMA-2 trial, survival outcomes were defined as the time from the date of brexucabtagene autoleucel infusion or leukapheresis to the date of death from any cause (OS), or until disease progression or death from any cause (PFS). Definitions of OS and PFS varied across the standard of care studies. One trial³⁸ defined OS from the date of last ibrutinib infusion until death or date of last follow-up, and 2 studies^35,40 defined survival from the start of subsequent therapy until death from any cause or the date last known to be alive. One study reported that OS and PFS were “calculated in the standard fashion,” from the start of subsequent therapy.⁵⁵

Tumour response outcomes were based on the Lugano classification in 3 studies^40,45,55 and the IWG definition in 4 studies^35,36,39,57 (classification system not reported in 2 studies^38,56). Tumour response outcomes were investigator-assessed in all the standard of care studies. Although the ZUMA-2 study reported both centrally adjudicated response (Lugano) and investigator-assessed tumour response outcomes, only central adjudication results were included in the MAIC. The definition of study baseline also varied for studies reporting only tumour response outcomes. The baseline was the start of BTK inhibitor therapy in Dreyling et al. (2016)⁵⁷ and Epperla et al. (2017),³⁶ whereas baseline was the start of subsequent therapy for Wang et al. (2017)³⁹ and Regny et al. (2019).⁵⁶

Results

Table 33 includes a summary of the results of the meta-analysis of survival outcomes for the standard of care studies. For both OS and PFS, the random-effects models showed better fit than fixed-effects models, based on the deviance information criterion. The median OS was 9.3 months (95% CI, 7.4 months to 11.6 months) and the mean OS was 15.4 months (95% CI, 12.2 months to 18.9 months) for the analyses that included all 4 standard of care studies. The point estimate for the median OS ranged from 8.0 months to 10.3 months, and the mean OS ranged from 14.0 months to 16.1 months across the sensitivity analyses conducted (Table 33).

The estimate of the 1-year survival rate was 42% (95% CI, 34% to 49%), and 2-year survival was 22% (95% CI, 15% to 28%), based on the random-effects meta-analysis of the 4 standard of care studies (primary analysis). Across the other sensitivity analyses, the 1-year survival rate ranged from 38% to 45% and the 2-year survival rate ranged from 19% to 23%.

The median PFS was 6.6 months (95% CI, 4.6 to 9.3), and the mean PFS was 11.3 months (95% CI, 7.3 to 16.3), based on pooled data from 2 standard of care studies. The PFS rate was 30% (95% CI, 19% to 42%) at 1 year and 14% (95% CI, 6% to 23%) at 2 years.

For the tumour response outcomes, the meta-analysis of all standards of care studies reported an ORR of 42% (95% CI, 27% to 59%), a CR of 21% (95% CI, 11% to 38%), and a PR of 19% (95% CI, 13% to 38%) (random-effects model; Table 34). Results varied across the 7 sensitivity analyses conducted, with ORR point estimates (random effects) that ranged from 26% to 56%, depending on which studies were included in the analyses. The majority of analyses showed high heterogeneity, with 14 of 24 analyses reporting I² values greater than 50% (range = 0% to 93%).

Matching-Adjusted ITC

Only the results that included the FAS population of the ZUMA-2 study have been summarized in this report. The FAS population may be considered a more conservative estimate as it includes all patients eligible for CAR T-cell therapy who underwent leukapheresis (i.e., those who did and did not receive brexucabtagene autoleucel). In Table 35, the baseline characteristics of the ZUMA-2 FAS population have been summarized for the primary analyses of OS, PFS, and ORR, showing the differences before and after weighting to match the standard of care studies. After weighting, the studies appear to be well balanced for the 6 characteristics that were included in the propensity model (marked in bold in Table 35); however, a number of differences between the brexucabtagene autoleucel and comparator studies are evident in other characteristics that were identified as prognostic factors for OS in MCL (e.g., MIPI risk level and extranodal involvement, and prior ibrutinib BTK inhibitor therapy). Effective sample size was substantially reduced from 74 patients to 36.2 (OS), 16.3 (PFS), and 29.5 (objective response) for the primary analyses, which suggests there were important differences between patients included in the ZUMA-2 and standard of care studies.

The observed and weighted OS curves for the FAS population of the ZUMA-2 trial are shown in Figure 5. Weighted data are shifted toward better survival. The effective sample size of the weighted survival function was 36 patients at the start of follow-up, and after 21 months the curve is predicted based on 9 or fewer patients. Figure 6 shows the fitted parametric survival curve for the weighted brexucabtagene autoleucel data versus pooled standard of care studies for OS. The pairwise indirect comparison reported an average HR of 0.18 (95% CI, 0.09 to 0.38) for the primary analysis, which included all standard of care studies (n = 4 trials). Table 36 reports the results of the unadjusted (naive) and MAIC sensitivity analyses for OS, which show that the point estimates and 95% CI consistently favour brexucabtagene autoleucel over standard of care treatments. The effective sample size for brexucabtagene autoleucel ranged from 18.9 patients to 40.8 patients, depending on which standard of care studies the ZUMA-2 data were matched to.

Figure 5: Unadjusted and Matching-Adjusted Kaplan–Meier Curves for Overall Survival With Brexucabtagene Autoleucel: ZUMA-2 FAS

FAS = full analysis set; KTE-X19 = brexucabtagene autoleucel; MAIC = matching-adjusted indirect comparison.

Note: Matched to all studies reporting overall survival Kaplan–Meier curves.

Source: Sponsor’s submission to CADTH.⁴³

Figure 6: MAIC Parametric Survival Curve for Overall Survival: Brexucabtagene Autoleucel Versus Standard of Care — ZUMA-2 FAS

CI = confidence interval; FAS = full analysis set; HR = hazard ratio; KTE-X19 = brexucabtagene autoleucel; MAIC = matching-adjusted indirect comparison.

Note: Matched to all studies reporting overall survival Kaplan–Meier curves.

Source: Sponsor’s submission to CADTH.⁴³