CADTH Reimbursement Review

Tisagenlecleucel (Kymriah)

Sponsor: Novartis Pharmaceuticals Canada Inc.

Therapeutic area: Relapsed or refractory follicular lymphoma

This multi-part report includes:

Clinical Review

Pharmacoeconomic Review

Ethics Review

Stakeholder Input

Clinical Review

Executive Summary

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

Table 1: Background Information on Application Submitted for Review

CAR = chimeric antigen receptor; NOC/c = Notice of Compliance with conditions.

Introduction

Follicular lymphoma (FL) is the second most common subtype of non-Hodgkin lymphoma (NHL) in Western countries and is characterized by follicular clusters or patterns of follicle centre B cells in a lymph node, bone marrow, or spleen.^1-3 FL is a relapsing and remitting disease, characterized by recurrent disease progression, shorter remission periods, and decreased survival (i.e., overall survival [OS] or progression-free survival [PFS]) with each treatment course.^4-7 Even though FL manifests as an indolent clinical course over many years in general, most patients eventually develop increasingly resistant disease, resulting in patients with relapsed or refractory (r/r) FL having reduced treatment options and poor prognoses.^4,8 Many patients with indolent FL may survive for decades, although a significant proportion of patients experience aggressive disease, which results in approximately 20% of affected patients dying 2 to 3 years after diagnosis.⁹ Although Canada-specific mortality data for FL could not be identified, in 2022, it was projected that there will be 3,000 deaths due to NHL in Canada.¹⁰ Patients with r/r FL experience significant health-related quality of life (HRQoL) impairment.^11,12 The overall prevalence of FL was estimated to be 1 per 3,000 people.¹³ Despite being a rare disease, FL is considered to be among the most prevalent of the lymphomas due to its chronicity and incurability in its advanced stages.^4,5,14,15 The overall incidence of FL is also low, with rates ranging from 2.2 to 3.5 per 100,000 new cases per year in Asia, Australia, Europe, and the US.¹⁶ According to a survey conducted in the US, the percentage of patients relapsing and receiving subsequent treatments was estimated to be 41%, 40%, and 30% for patients who received 1 line, 2 lines, and 3 lines of treatment, respectively.¹⁷

Staging of FL describes the extent to which the disease has spread in the body: stage I FL involves 1 node or a group of adjacent nodes or single extranodal lesions without nodal involvement; stage II FL involves 2 or more nodal groups on the same side of the diaphragm and limited contiguous extranodal involvement; stage III FL involves nodes on both sides of the diaphragm and nodes above the diaphragm with spleen involvement; stage IV FL has additional noncontiguous extralymphatic involvement.¹⁸ Grading of FL is determined by the number of large FL cells (also known as centroblasts) and reflects the aggressiveness of the tumour: grade 1 FL has 0 to 5 centroblasts per high power field (HPF); grade 2 FL has 6 to 15 centroblasts per HPF; and grade 3 FL has more than 15 centroblasts per HPF. Grade 3 FL has been subdivided into grade 3a (centrocytes are present) and grade 3b (solid sheets of centroblasts present).¹⁸ Grade 1, 2, and 3a FLs are generally considered to be low grade or slow growing, whereas grade 3b FL is faster growing.

The main goals of treatment are to cure the lymphoma in patients with stage I to II FL, and to extend remission in patients with stage III to IV FL.^19,20 Once a diagnosis of FL is confirmed, the gold standard for the management of asymptomatic patients with indolent FL is watchful waiting.¹⁷ For small, localized symptomatic FL, radiotherapy is considered the standard of care (SOC).^21,22 For grade 1, 2, and 3a FL, the preferred chemoimmunotherapy regimen is bendamustine plus rituximab (BR).^20,23,24 In frail and older adults, rituximab monotherapy, a chemotherapy-free approach, is the preferred first-line regimen, according to European and North American guidelines.^25,26 However, some Canadian centres do not have access to rituximab monotherapy. Instead, these physicians would keep patients on BR for several treatment cycles or for as long as possible. Beyond first-line treatment, there is currently no gold standard for the care of the r/r FL population. In terms of treatment options for second-line regimens for r/r FL, combined immunochemotherapy is the preferred treatment strategy, such as obinutuzumab, cyclophosphamide, doxorubicin hydrochloride, vincristine sulphate, and prednisone (O-CHOP).²⁰ Hematopoietic stem cell transplant (SCT) may be considered in young and fit patients with no comorbidities in the second-line setting. Autologous stem cell transplant (auto-SCT) is more common than allogeneic stem cell transplant (allo-SCT) in this population. However, only a small subset of patients with FL would be eligible for transplant in the second-line setting.

Patients with r/r FL in the third-line setting and beyond are heavily pretreated and have advanced disease. A heterogenous mix of immunochemotherapy (for most patients) and SCT (for a minority of patients) are the current treatment options in this hard-to-treat population. In recent years, chimeric antigen receptor (CAR) T-cell therapy has emerged as another form of immunotherapy for the treatment of blood cancers, including lymphomas. Although promising results have been reported for the CAR T-cell therapies for advanced stage lymphomas, severe adverse effects related to CAR T-cell therapy, such as cytokine release syndrome (CRS) and neurologic toxicities, have also been reported.²⁷

Tisagenlecleucel (Kymriah) is a second-generation autologous CAR T-cell therapy directed at the cell-surface protein CD19, which is only expressed on B cells or their precursors and not on hematopoietic stem cells or non-B-cell tissues. On December 19, 2022, tisagenlecleucel received a Notice of Compliance with conditions (NOC/c) from Health Canada for adults with r/r grade 1, 2, or 3a FL after 2 or more lines of systemic therapy.²⁸ The recommended dose of tisagenlecleucel for the indicated population is 0.6 to 6.0 × 10⁸ CAR-positive viable T cells (nonweight based) suspended in 1 or more patient-specific infusion bags for a single-dose, one-time, IV administration.²⁸

Stakeholder Perspectives

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

Patient Input

One patient group, Lymphoma Canada (LC), provided input to the tisagenlecleucel submission. LC is a national Canadian registered charity. It collected input from patients with r/r FL through an online anonymous survey conducted from November 2022 to January 2023. Of the 44 respondents, only 1 had experience with tisagenlecleucel. In addition, LC obtained patient feedback from a French patient organization called Ensemble Leucemie Lymphomes Espoir (ELLyE). ELLyE collected feedback from 162 patients on CAR T-cell therapy, including 19 who had experience with tisagenlecleucel from January 2019 to September 2020 in France.

Based on the patient input, FL has significant negative impact on a patient’s physical and psychosocial well-being, affecting everyday life, work, and family.

Patients indicated that there is a need for more therapeutic options that provide longer disease remission, lead to a longer life span, and improve quality of life (QoL). They also indicated that all patients in Canada should have easy access to new therapies.

Patients who had experience with tisagenlecleucel indicated that the drug is effective; adverse events (AEs) were reported but were manageable.

Clinician Input

Input From Clinical Experts Consulted by CADTH

The clinical panel indicated that for patients with FL, the most important treatment goals are to prolong survival (for both OS and PFS) and improve QoL. However, patients with r/r FL relapse after frontline therapies or are refractory to available treatments, which subsequently affects their long-term PFS and QoL. In addition, some patients may not tolerate current treatments well, due to AEs or complications associated with SCT.

The clinical panel noted that many factors (e.g., patient characteristics, previous treatments, treatment effects and toxicity, whether a treatment is reimbursed by drug plans, disease progression and transformation, and patient preference) need to be considered before deciding which treatment to provide. Watch and wait is a common approach for many patients with FL, even after disease relapse. Patients who need active treatments typically receive bendamustine or a rituximab-based therapy (such as BR; rituximab, cyclophosphamide, vincristine sulphate, plus prednisone [R-CVP]; rituximab, cyclophosphamide, doxorubicin hydrochloride [hydroxydaunorubicin], vincristine sulphate [Oncovin], plus prednisone [R-CHOP]; or rituximab plus lenalidomide [R2]). Patients with relapsed disease after the treatment with chemoimmunotherapy, particularly those who progress within 2 years, may receive auto-SCT if they are suitable candidates. After all these treatments, some patients maintain the indolent status and some transform to large cell lymphoma. The clinical panel suggested that tisagenlecleucel be used as a third-line treatment or beyond for patients with r/r FL. There are not many options available for patients at this stage.

The clinical panel indicated that although a more selective population would be suitable for treatment with CAR T-cell therapy in clinical trials of r/r FL, in practice, CAR T-cell therapy can be used in a broader patient population (for example, patients with certain comorbidities or disease status). In clinical practice, suitable patients can be identified based on clinical judgment, which combines medical history, laboratory and imaging findings, and often a lymph node biopsy. The panel noted that patients who are likely to benefit from other available treatments or who have an Eastern Cooperative Oncology Group Performance Status (ECOG PS) of 3 or higher are least suitable for tisagenlecleucel. The panel also noted that there is not a specific patient characteristic that can be used to predict which patient will respond better to tisagenlecleucel and which will not.

The panel indicated that in clinical practice, patients are evaluated and followed in a manner similar to that described in the clinical trials of FL. Remission and survival are measured. Physical exams and imaging exams are routinely conducted to assess the patient’s response to CAR T-cell therapy.

The panel suggested that meaningful responses to treatment with tisagenlecleucel would be not only a high complete remission rate (CRR), but also durability of treatment response and long-term PFS and OS. The panel noted that after CAR T-cell therapy, clinicians will assess treatment response (e.g., with CT scan) at 3 months, or sooner if needed.

The panel emphasized that a multidisciplinary team, which involves hematologists, infectious disease specialists, neurologists, an intensive care unit (ICU) team, and other specialists, is required to diagnose, treat, and monitor patients who receive tisagenlecleucel and to ensure the safe and effective delivery of this treatment.

Clinician Group Input

Two clinician groups provided input for the review of tisagenlecleucel: Cell Therapy Transplant Canada, and the Ontario Health (Cancer Care Ontario) Hematology Cancer Drug Advisory Committee.

In general, the clinician group input was consistent with the input provided by the experts consulted by CADTH for the tisagenlecleucel review. It also suggested that tisagenlecleucel be used in patients with r/r FL who are not eligible to receive an allo-SCT or auto-SCT as third-line therapy or beyond. In addition, CD19 CAR T-cell therapy, including tisagenlecleucel, will only be considered in patients without significant organ dysfunction, according to the clinician group.

The clinician group noted that assessment of response to treatment should be based on the standard lymphoma response criteria, and clinical exams and imaging scans, such as CT and PET, should be performed. Outcomes such as remission rates, PFS, patient safety, as well as HRQoL should be measured.

Drug Program Input

Input was obtained from the drug programs that participate in the CADTH reimbursement review process. The following were identified as key factors that could potentially impact the implementation of a CADTH recommendation for tisagenlecleucel:

• considerations for initiation of therapy

• considerations for prescribing of therapy

• generalizability

• care provision issues

• system and economic issues.

The clinical experts consulted by CADTH provided advice on the potential implementation issues raised by the drug programs.

Clinical Evidence

Pivotal Studies and Randomized Controlled Trial Evidence

Description of Studies

One clinical study, ELARA (formerly known as Study E2202),²⁹ was included in the systematic review. The ELARA study (N = 98) is a phase II, open-label, single-arm study that evaluated the efficacy and safety of tisagenlecleucel in patients with r/r grade 1, 2, or 3a FL after 2 or more lines of systemic therapy. The primary end point was CRR assessed by independent review committee (IRC) through 24 months. Secondary end points included overall response rate (ORR), OS, PFS, duration of response (DoR), and HRQoL (assessed by Functional Assessment of Cancer Therapy-Lymphoma [FACT-Lym], 36-Item Short Form Survey [SF-36], and 3-Level EQ-5D [EQ-5D-3L]). Data up to 30 months of follow-up were available at the time of this review (data cut-off date was March 29, 2022). The median age observed in the overall r/r FL population was 57 years (range, 29 to 73 years). More males (68.1%) were enrolled than females (31.9%), and most patients (84.0%) were white. Almost all patients (97.0%) had a baseline ECOG PS of 0 or 1. Most patients also had grade 1 or 2 disease (90.4%) and stage III or IV disease (80%). Enrolled patients received a median of 4 (range, 2 to 13) prior lines of treatments. Of the 98 included patients, 77.6% were refractory to their last line of antineoplastic therapy. The proportion of patients who progressed in the 24 months after first-line anti-CD20 monoclonal antibody (mAb)–containing therapy was 64.9%.

Efficacy Results

At the data cut-off date of March 29, 2022, among the 97 patients who were treated with tisagenlecleucel, the CRR was 68.1% (95% confidence interval [CI], 57.7% to 77.3%), ORR was 86.2% (95% CI, 77.5% to 92.4%), and partial response rate (PRR) was 18.1% (95% CI, not reported) at 24-month follow-up per IRC assessment. The results from local assessment were consistent with the IRC assessment. Median DoR was not reached with tisagenlecleucel at the respective median follow-up times in the ELARA trial.

In the ELARA trial, median OS was not reached at the 24-month follow-up (cut-off date of March 29, 2022). Thirteen deaths had occurred in the study, and the OS rate was 87.7% (95% CI, 78.3% to 93.2%) at 24 months and 82.6% (95% CI, 70.2% to 90.2%) at 30 months.

The median PFS per IRC was not reached at the time of the data cut-off (March 29, 2022), and there were 38 PFS events in total (disease progression or death). The PFS rate was 77.8% (95% CI, 67.7% to 85.1%) at 6 months, 67.2% (95% CI, 56.3% to 75.9%) at 12 months, and 57.4% (95% CI, 46.2% to 67.0%) at 24 months.

Over time, there were no notable changes in the proportion of patients reported to have improved, stable, or deteriorated FACT-Lym and SF-36 scores after tisagenlecleucel infusion. Overall, 70% to 88% of patients reported no deterioration in HRQoL based on FACT-Lym and SF-36 scores at 12 months, with similar trends observed at 18 months and 24 months. Results of the EQ-5D-3L visual analogue scale (VAS) score showed that HRQoL was maintained from baseline after the tisagenlecleucel infusion. The mean EQ VAS score was 69.4 at baseline, which increased to 72.5 at 6 months, 75.9 at 12 months, and 71.9 at 24 months. However, these results should be interpreted with caution because the 24-month results were based on approximately half of the patients enrolled in the study.

Harms Results

At the data cut-off date of March 29, 2022, of the 97 patients evaluable for safety, 99% experienced at least 1 AE. The most commonly reported AEs for patients any time after infusion were CRS (49.5%), neutropenia (43.3%), anemia (25.8%), diarrhea (25.8%), headache (23.7%), decreased white blood cell count (22.7%), pyrexia (18.6%), thrombocytopenia (18.6%), fatigue (17.5%), nausea (17.5%), decreased neutrophil count (17.5%), constipation (16.5%), and hypogammaglobulinemia (15.5%). Serious adverse events (SAEs) were reported in 46.4% patients any time after infusion. The most commonly reported SAEs included CRS (19.6%), pneumonia (10.3%), and febrile neutropenia (8.2%). There were 13 deaths (13.4% of patients) reported after tisagenlecleucel infusion in the ELARA study: 7 patients died due to the study indication (after progression of the underlying disease) and 6 died for other reasons (AEs for 5 patients and euthanasia for 1 patient).

In terms of AEs of special interest, more than 40% of patients experienced any-grade CRS (49.5%), hematological disorders such as cytopenias (78.4%), and infections (55.7%). Across hematological disorders such as cytopenias, at least 25% of the patients reported neutropenia (43.3%) or anemia (25.8%). Overall, the majority of patients (74.2%) experienced hematological events of grade 3 or higher. Infections occurring any time after infusion were reported in 54 patients (55.7%), 16 of whom (16.5%) had infections suspected to be related to tisagenlecleucel. Most of the patients had either grade 1 or 2 infections, while 21.6% of patients had infections of grade 3 or higher (9.0% of whom had AEs suspected to be related to tisagenlecleucel). Death due to infection (pneumonia) was reported in 1 patient. Any-grade serious neurologic adverse reactions were reported in 12 patients (12.4%), 10 of whom experienced these events in the 8 weeks after tisagenlecleucel infusion. Grade 3 or 4 AEs considered to be serious were reported in 3 patients (3.1%), 1 of whom recovered.

According to the clinical experts consulted by CADTH, the safety profile of tisagenlecleucel is consistent with that of other CAR T-cell therapies, and no unexpected safety signals were observed in the ELARA trial.

Table 2: Summary of Key Results From the ELARA Trial

AE = adverse event; CI = confidence interval; CRR = complete remission rate; CRS = cytokine release syndrome; DoR = duration of response; EAS = efficacy analysis set; FU = follow-up; IRC = independent review committee; KM = Kaplan-Meier; ORR = overall response rate; OS = overall survival; PFS = progression-free survival; SAE = serious adverse event.

Source: ELARA Clinical Study Report.²⁹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Critical Appraisal

The single-arm, noncomparative design of the ELARA study is 1 of its key limitations. Interpreting the results of studies with this design is difficult because it may not be apparent whether the results are from the effect of the intervention, a placebo effect, or the effect of natural history. Although it is acknowledged that this study design has so far predominated in the evaluation of CAR T-cell therapies for advanced cancers across a variety of tumour types, and there may be practical limitations to conducting a randomized controlled trial (RCT) in patients with r/r FL, there is no clear rationale that makes a RCT infeasible. Subsequently, the lack of a comparator makes it difficult to determine whether the magnitude of the treatment effect would be replicated in a larger comparative trial or in the real world. Another limitation of the ELARA study is the relatively small sample size and selective study population.

Follow-up time was likely sufficient for assessing tumour response and safety outcomes associated with tisagenlecleucel in general. However, the follow-up duration was not long enough to fully capture the effects on OS and PFS, and thus these results are considered immature. In addition to the duration of the study and the noncomparative design, subsequent treatments make it difficult to interpret the OS and PFS results. The survival results (OS, PFS) should be considered in the context of subsequent treatments, because it may be difficult to tell which treatment has more impact on patient’s survival, especially when there is a lack of comparative data in the ELARA study.

The ELARA trial had an open-label design, which can result in a bias in the measurement of subjectively assessed outcomes, such as response, PFS, HRQoL, and AEs. In addition, the study presented patient-reported outcomes (PROs) and HRQoL data up to 24 months; however, there is a risk of attrition bias and drawing conclusions on a select population because the analyses at 24 months were based on half of the study population from baseline and the results could be biased to favour tisagenlecleucel.

According to the clinical experts consulted by CADTH, the ELARA study population generally represents the patients in Canada with r/r FL who would be receiving tisagenlecleucel. However, the clinical experts noted that patients seen in clinical practice would include those with a poorer performance status (the ELARA study only included patients with an ECOG PS of 0 or 1), those who had received prior CD19-targeted therapy, and those with more comorbidities.

Long-Term Extension Studies

There was no long-term extension study submitted by the sponsor.

Indirect Comparisons

Description of Studies

One indirect treatment comparison (ITC) was submitted by the sponsor and included in CADTH’s clinical review.³¹ Due to the lack of a common comparator, the sponsor conducted an unanchored matched-adjusted indirect comparison (MAIC) to estimate the comparative ORR, CRR, OS, and PFS between tisagenlecleucel and axicabtagene ciloleucel (2 × 10⁶ CAR T cells per kg) in patients with r/r FL after 2 or more lines of therapy. Although the comparator treatment used in the MAIC has not yet been reviewed by the CADTH pan-Canadian Oncology Drug Review Expert Review Committee for this patient population, the ITC was used to inform the sponsor's pharmacoeconomic model and, therefore, reviewed by the clinical team. The MAIC was based on individual data from patients who received tisagenlecleucel in the ELARA study and aggregate-level data from patients who received axicabtagene ciloleucel in the ZUMA-5 study.

Efficacy Results

The MAIC analysis compared 52 patients who received efficacy-evaluable, nonbridging chemotherapy in the ELARA study and 86 patients in the efficacy-evaluable set in the ZUMA-5 study who had at least 24 months of follow-up. Compared to axicabtagene ciloleucel, the MAIC estimated a response difference for tisagenlecleucel in ORR of –3.03% (95% CI, –13.67% to 7.61%) and in CRR of –5.03% (95% CI, –23.85% to 13.80%). Compared to axicabtagene ciloleucel, the MAIC analysis estimated the hazard of death and disease progression for tisagenlecleucel to be 0.49 (95% CI, 0.16 to 1.49) and 0.84 (95% CI, 0.37 to 1.90), respectively.

Harms Results

The 53 patients who received infused nonbridging chemotherapy in the ELARA study and the 124 patients in the infused set of the ZUMA-5 study were included in the MAIC of safety outcome. At least 1 AE of any grade was reported in 44.6% of patients in the ELARA study and 78.2% of patients in the ZUMA-5 study. AEs of grade 3 or higher were reported in no patients in the ELARA study and in 6.5% of patients in the ZUMA-5 study. Management of CRS with corticosteroids was documented in 3.0% and 15.3% of patients in the ELARA and ZUMA-5 studies, respectively. CRS management with tocilizumab was documented in 9.9% and 45.2% of patients in the ELARA and ZUMA-5 studies, respectively. Neurologic events of any grade were documented in 9.5% of patients in the ELARA study and in 56.5% of patients in the ZUMA-5 study. Neurologic events of grade 3 or higher were reported in 0.19% and 15.32% of patients in the ELARA and ZUMA-5 studies, respectively.

Critical Appraisal

For an unanchored MAIC to produce unbiased treatment-effect estimates, adjustment for all effect modifiers and prognostic variables in the analysis is needed. However, MAICs are rarely able to overcome the strict assumption and the bias resulting from missing covariates is very difficult to quantify. Key methodological differences between the ELARA and ZUMA-5 studies that could not be adjusted for and failure to match on key covariates may have confounded the study results. Furthermore, MAICs cannot account for unknown cross-trial differences; thus, MAIC estimates are susceptible to bias from unknown confounding. An evaluation of potential bias from residual confounding was not reported; therefore, the magnitude of this bias in the relative treatment-effect estimates is unclear. There is also concern about the loss of precision in the results, given the reduction in the effective sample size (ESS). Overall, the direction of bias could not be determined due to the abovementioned limitations, and the CADTH team could not draw any strong conclusions from the MAIC. Outcomes other than treatment response and survival that are important to patients, clinicians, and drug plans (e.g., HRQoL and symptoms) were not analyzed in the MAIC.

Studies Addressing Gaps in the Pivotal and RCT Evidence

Two studies provided additional context to the effectiveness and safety of tisagenlecleucel for the treatment of r/r FL. Tisagenlecleucel was compared to standard chemotherapy in 1 study (ELARA versus ReCORD-FL),³² and the second study was a single-arm, noncomparative trial.³³

ELARA Versus ReCORD-FL

Description of Studies

In the absence of a direct head-to-head comparison of tisagenlecleucel to SOC, the sponsor compared the treatment effect of tisagenlecleucel observed in the ELARA study to SOC, defined as standard chemotherapy as documented in the ReCORD-FL study. The ReCORD-FL study is a noninterventional, multicentre, retrospective chart review conducted by the sponsor to provide patient-level data that could be used to form a historical control group with which to compare standard chemotherapy in the ELARA study. Patient-level data were collected from patients treated for r/r FL between 1998 and 2020 from 10 sites across Europe and North America, including 1 Canadian site (n = 12). Where feasible, the ReCORD-FL study adopted the same inclusion and exclusion criteria as the ELARA study. Propensity score matching was used to achieve an approximate balance on the number of prior lines of therapy while balancing other key baseline prognostic variables between the ELARA and ReCORD-FL studies. The distribution of the weighted time-to-event end points of OS and PFS were estimated using a Kaplan-Meier (KM) analysis, whereas hazard ratios (HRs) were estimated using Cox proportional hazard regression. At the data cut-off, 97 patients with r/r FL in the ELARA study (March 29, 2022) and 143 patients in the ReCORD-FL study (December 31, 2021) were included. After weighting, patients in the ELARA (n = 97) and ReCORD-FL (ESS = 47.5) studies had a mean age of 55.4 years and 56.5 years, respectively, were mostly male (67% to 72%), and just more than a third had documented prior auto-SCT therapy. Approximately 68% and 70.2% of patients in the ELARA and ReCORD-FL studies, respectively, were documented as double refractory, and 77.3% and 63.9%, respectively, were documented as refractory to last prior therapy.

Efficacy Results

At the data cut-off, death events were observed in 13.4% of patients in the ELARA study and 45.2% of patients in the ReCORD-FL study. Median OS was not estimable (NE) in the ELARA study. For patients in the ReCORD-FL study, the median OS was 36.6 months (95% CI, 25.8 months to NE). The KM estimate for OS at 24 months was 90.8% (95% CI, 84.7% to 96.9%) and 64.8% (95% CI, 49.5% to 80.0%) for ELARA and ReCORD-FL, respectively. Compared to standard chemotherapy, tisagenlecleucel was associated with an estimated risk reduction in death of 72% (hazard ratio [HR] = 0.28; 95% CI, 0.07% to 0.49%).

At the data cut-off, disease progression was observed in 42.3% of patients in the ELARA study and 63.7% of patients in the ReCORD-FL study. Median PFS was NE in the ELARA study. For patients in the ReCORD-FL study, median PFS was 11.5 months (95% CI, 5.9 to 35.6 months). The KM estimate for PFS at 24 months was 58.6% (95% CI, 48.6% to 68.6%) and 38.3% (95% CI, 22.7% to 53.8%) in the ELARA and ReCORD-FL studies, respectively. Compared to standard chemotherapy, tisagenlecleucel was associated with an estimated risk reduction in death or starting a new anticancer therapy of 47% (HR = 0.53; 95% CI, 0.25% to 0.81%).

Harms Results

Harms outcomes were not compared between the ELARA and ReCORD-FL studies.

Critical Appraisal

The nonrandomized comparison of the ELARA and ReCORD-FL studies makes interpretation of the efficacy of tisagenlecleucel relative to standard chemotherapy challenging. To mitigate potential differences in baseline prognostic factors related to OS and PFS, inclusion and exclusion criteria from the ELARA study were applied to the ReCORD-FL study, and eligible patients in the ReCORD-FL study were systematically selected based on highest propensity scores. Moreover, a comparison of weighting by odds was conducted to assess the causal effects of prescribing tisagenlecleucel versus chemotherapy. However, several inclusion and exclusion criteria in the ELARA study could not be applied to the ReCORD-FL study. Moreover, prognostic factors considered important by the clinical experts consulted by CADTH for the purpose of this review, such as baseline ECOG PS and Follicular Lymphoma International Prognostic Index (FLIPI) scores, were not included in the propensity model. Consequently, there is uncertainty around the treatment effects of tisagenlecleucel relative to standard chemotherapy due to selection bias and unmeasured and residual confounding that cannot be entirely ruled out. Baseline characteristics after weighting were well balanced, as evidenced by absolute mean differences of less than 25%. However, the complete baseline demographic and disease characteristics of patients in the 2 studies were not reported after matching. Therefore, it is unclear what effect the weighting had on the balance of other relevant patient characteristics. The applications of weights resulted in a reduced ESS of 45.7, in which 52% of enrolled patients in the ReCORD-FL study were lost. The reduction in sample size may contribute to imprecision, leading to uncertainty of the results. Regarding the PFS efficacy outcome, date of disease progression was not available for most patients in the ReCORD-FL study. Moreover, radiographic assessment of disease progression tends to be less frequent in the real-world setting than in clinical trial protocols. Accordingly, the date of disease progression was considered at the time a new anticancer therapy was initiated in both the ReCORD and ELARA studies for comparative analysis of PFS, and censoring was redefined to occur at the last contact date, rather than the last assessment date, in the ELARA study to avoid bias due to timing of assessment. Uncertainty about outcome assessment is further compounded due to inconsistencies in assessment across patients included in the ReCORD-FL study. Because assessment of patients in the ReCORD-FL study was not planned according to a uniform protocol, physicians may have used subjective criteria to assess clinical response. Based on input from the clinical experts, the patients included in the comparison of the ELARA and ReCORD-FL studies appeared to be younger than what is typically seen in the clinical setting. The clinical experts noted that the selected anticancer treatments were appropriate SOC regimens. However, whether the change in PFS definition for the purpose of efficacy comparison between the ELARA and ReCORD-FL studies is an appropriate surrogate for the standard definition of PFS is uncertain.

Study by Schuster et al.

Description of Study

The study by Schuster et al. (2017)³³ was a single-centre, single-arm, phase IIa clinical trial conducted in the US. The objective of this study was to estimate the efficacy of a single infusion of tisagenlecleucel in patients with NHL and r/r CD19+ lymphomas after chemotherapy. Patients were eligible if they had CD19+ diffuse large B-cell lymphoma or FL with no curative treatment options, a limited prognosis (< 2 years of anticipated survival), and a partial response or stable disease after the most recent therapy. Patients with FL were eligible if they had measurable disease less than 2 years after the second line of immunochemotherapy (excluding single-drug monoclonal antibody therapy). A total of 15 patients with r/r FL were enrolled in this study, 14 of whom received the tisagenlecleucel treatment. Median age of the 14 patients was 59 years (range, 43 to 72 years). There was an equal distribution of male and female patients (50%). All patients (100%) had a baseline ECOG PS of 0 or 1. Overall, 64% of the patients had grade 1 to 2 FL and 86% had stage IV FL. Regarding prior therapy, patients received a median of 5.0 (range, 2 to 10) prior lines of treatment. The intervention of interest in this study is tisagenlecleucel, which was administered as a one-time, single infusion of CD19 CAR T cells by IV injection (total dose of 1 to 5 × 10⁸ CD19 CAR T cells). The median total dose of tisagenlecleucel was 5.00 × 10⁸ (range, 1.79 × 10⁸ to 5.00 × 10⁸), and the median dose of tisagenlecleucel per kilogram of body weight was 5.79 × 10⁶ (range = 3.08 × 10⁶ to 8.87 × 10⁶). The median number of days from apheresis to infusion was 39 (range, 27 to 145). All 14 patients with FL received lymphodepleting (LD) chemotherapy before the tisagenlecleucel infusion. The primary outcome in the study by Schuster et al. was ORR at 3 months in patients with NHL. The secondary outcomes were CRR, PRR, DoR, OS, PFS, and time to next treatment.

Efficacy Results

In the study by Schuster et al., the median OS was not reached at either 28.6 or 49.0 months of median follow-up, and the OS rate was 93% at 28.6 months median follow-up. The median PFS was also not reached at 28.6 months; however, a decline was observed at longer follow-ups (median PFS was 32.4 months and 26.2 months at median follow-ups of 49 months and 60 months, respectively). The estimated progression-free probabilities were 77%, 70%, and 43% at median follow-ups of 11.4 months, 28.6 months, and 60 months, respectively. The study by Schuster et al. assessed the response rate at shorter follow-ups (3 months and 6 months). This study reported an ORR of 79% at both assessment points. The CRR was 50% at 3 months and 71% at both 6 months and 49 months. The median DoR was not reached with tisagenlecleucel at the respective median follow-up times. PROs have not been included in the study by Schuster et al.

Harms Results

The main AEs were CRS of any grade and of grade 3 or 4, which were experienced by 42.9% and 14.3% of patients with FL, respectively. Tisagenlecleucel was administered as a single-time infusion; therefore, no patients discontinued treatment in the study by Schuster et al. One patient with FL who had encephalopathy had progressive neurologic deterioration that resulted in death. Data for other AEs of special interest were only reported for the overall NHL population in the study be Schuster et al. Eleven patients (39%) reported neurologic toxicities, including encephalopathy in 3 patients (27%), delirium in 2 patients (18%), and tremor in 2 patients (18%). In addition, cognitive disturbance, confusion, involuntary movements, and memory impairment were each reported in 1 patient (9%).

Critical Appraisal

The main limitations of the study be Schuster et al. are the single-arm design, the lack of a comparator, and the open-label nature, which limit the interpretation of effect. Moreover, the sample size calculation for this study was based on the overall NHL population and was not specific to the FL subgroup, which could limit detection of the magnitude of effect among the FL subgroup.

In terms of generalizability, this study was conducted in the US, which may have a different health system and different treatment conditions than Canada. The patient population in this study had a baseline ECOG PS of 0 or 1. It is not clear if the results are generalizable to patients with a poorer performance status.

Conclusions

Evidence from a single-arm study (ELARA) suggests that treatment with tisagenlecleucel is associated with clinically important tumour responses, including complete remission, in adults with r/r FL after 2 or more lines of systemic therapy. There is insufficient evidence — in part due to the limited follow-up duration of the trial — to determine the effects of tisagenlecleucel on OS and PFS. Results of PROs suggest that a patient’s QoL could be maintained; however, 24-month HRQoL data were based on only about 50% of the baseline population. It is unclear if treatment with tisagenlecleucel would improve HRQoL. The harms associated with the tisagenlecleucel infusion are consistent with its mechanism of action, and no unexpected safety signals observed.

The CADTH clinical assessment identified limitations of the sponsor’s comparison of the ELARA and ReCORD-FL studies and the sponsor-conducted MAIC (including small sample sizes, heterogeneity across study designs and populations, and the inability to adjust for all potential effect modifiers and prognostic variables), which substantially limited the ability to interpret the relative treatment effects observed for tisagenlecleucel, standard chemotherapy, and axicabtagene ciloleucel.

Introduction

The objective of this report is to review and critically appraise the evidence submitted by the sponsor on the beneficial and harmful effects of tisagenlecleucel (1.2 × 10⁶ to 6.0 × 10⁸ CAR-positive viable T cells for IV use) in the treatment of adults with r/r grade 1, 2, or 3a FL after 2 or more lines of systemic therapy.

Disease Background

The contents of this section have been informed by materials submitted by the sponsor and clinical expert input. The following information has been summarized and validated by the CADTH review team.

NHL accounts for approximately 90% of the cancers that affect the lymphocytes.³⁴ FL is the second most common subtype of NHL in Western countries and is characterized by follicular clusters or patterns of follicle centre B cells in a lymph node, bone marrow, or spleen.^1-3 Staging of FL describes the extent to which the disease has spread in the body: stage I FL involves 1 node or a group of adjacent nodes or single extranodal lesions without nodal involvement; stage II FL involves 2 or more nodal groups on the same side of the diaphragm and limited contiguous extranodal involvement; stage III FL involves nodes on both sides of the diaphragm and nodes above the diaphragm with spleen involvement; and stage IV FL has additional noncontiguous extralymphatic involvement.¹⁸ Grading of FL is determined by the number of large FL cells (also known as centroblasts) and reflects the aggressiveness of the tumour: grade 1 FL has 0 to 5 centroblasts per HPF; grade 2 FL has 6 to 15 centroblasts per HPF; and grade 3 FL has more than 15 centroblasts per HPF. Grade 3 FL has been subdivided into grade 3a (centrocytes are present) and grade 3b (solid sheets of centroblasts present).¹⁸ Grade 1, 2, and 3a FLs are generally considered to be low grade or slow growing, whereas grade 3b FL is faster growing. In the early stages, most patients experience few symptoms and slow disease progression.^1,35 The most common signs or symptoms in these stages are painless swelling or enlargement of 1 or more groups of lymph nodes in the neck, armpit, or groin that does not resolve with time.^1,35 Although usually painless, the swollen lymph nodes may cause compression of organs, restricted movement, and disfigurement.^11,19,36 In advanced stages, patients can experience fever, night sweats, and weight loss, as well as bone marrow involvement that results in bruising or bleeding, leucocytosis, thrombocytosis, thrombocytopenia, and leucopenia.^13,36-39 FL is a relapsing and remitting disease, characterized by recurrent disease progressions, shorter remission periods, and decreased survival (i.e., OS or PFS) with each treatment course.^4-7 Even though FL manifests as an indolent clinical course over many years in general, patients can still present with aggressive disease, which relapses more rapidly or is refractory to treatment, or the FL can undergo histologic transformation to a more aggressive type of lymphomas (e.g., 10% to 70% of FL transforms to diffuse large B-cell lymphoma over time, with a risk of 2% to 3% per year).^40-43 Most patients eventually develop increasingly resistant disease, resulting in reduced treatment options and poor prognosis for patients with r/r FL.^4,8 Many patients with indolent FL may survive for decades, although a significant proportion of patients experience aggressive disease, which results in approximately 20% of affected patients dying within 2 to 3 years of diagnosis.⁹ Patients who relapse within 12 months of SCT or progress within 24 months have an extremely poor prognosis.⁴⁴ Although Canada-specific mortality data for FL could not be identified, in 2022, it was projected that there would be 3,000 deaths due to NHL in Canada.¹⁰ Results of a retrospective analysis conducted in a single centre in the US showed that among patients with FL who had received multiple lines of treatment, median OS was 11.7 years, 8.8 years, and 5.3 years for patients who received the second-line, third-line, and fourth-line treatments, respectively.⁶ Compared with disease-free and newly diagnosed patients, those with r/r FL suffer from more significant HRQoL impairment.^11,12

In general, patients present with asymptomatic lymphadenopathy manifesting as painless superficial small to medium-sized lymph nodes that are typically present for many years before a FL diagnosis is established.^34,40 Thus, most patients are diagnosed at an advanced stage (i.e., 66% to 70% of patients are initially diagnosed at stage III or IV, which is considered incurable but treatable).⁴⁵ Most patients are usually diagnosed with advanced FL during the sixth decade of life, although approximately 25% of patients are diagnosed when they are younger than 40 years.⁴⁶ After diagnosis, FL is staged to determine the extent and distribution of the disease and to guide treatment decision-making. The most common method for staging lymphoma is the Ann Arbor staging system.^19,35,47 A grade is also assigned to the lymphoma: low grade is typically used to describe indolent or slow-growing tumours, and intermediate or high-grade is used to describe aggressive or fast-growing lymphomas.³⁵ One of the tools used most widely to estimate the prognosis of patients with FL is the FLIPI, which considers age (older than 60 years versus 60 years or younger), Ann Arbor stage (III to IV versus I to II), number of involved nodal areas (more than 4 versus no greater than 4), hemoglobin level (lower than 120 g/L versus 120 g/L or higher), and serum lactate dehydrogenase concentration (above normal versus normal or below).⁴⁸

The overall prevalence of FL is estimated to be 1 per 3,000 people, according to the Orphanet database.¹³ Despite being a rare disease, FL is considered to be among the most prevalent of the lymphomas due to chronicity and incurability in its advanced stages.^4,5,14,15 The overall incidence of FL is also low, with rates ranging from 2.2 to 3.5 per 100,000 new cases per year in Asia, Australia, Europe, and the US.¹⁶ According to a survey conducted in the US, the percentage of patients relapsing and receiving subsequent treatments is estimated to be 41%, 40%, and 30% for patients who have received first-line, second-line, and third-line treatments, respectively.¹⁷

Standards of Therapy

The contents of this section have been informed by materials submitted by the sponsor and clinical expert input. The following information has been summarized and validated by the CADTH review team.

The treatment goals for FL vary, depending upon the stage of the FL and individual patient factors. In general, available treatments for stage I to II FL have curative potential; however, for most patients with advanced stage III to IV FL, no curative therapies are available. Therefore, the main goals of treatment are to cure the lymphoma in patients with stage I to II FL and to extend remission in patients with stage III to IV FL.^19,20

Once a diagnosis of FL is confirmed, the gold standard for the management of asymptomatic patients with indolent FL is watchful waiting, also known as watch and wait.^23,49 On average, 49% of patients newly diagnosed with FL can be managed with a watchful waiting approach, based on the opinion of Canadian clinical experts consulted by the sponsor.¹⁷ According to the clinical experts consulted by CADTH, watch and wait is a common practice for many patients with FL, even after disease relapse.

First-Line Treatments

For small, localized symptomatic FL, radiotherapy is considered the SOC, according to North American and European guidelines.^19,20,50 This is supported by several provincial guidelines in Canada.^21,22 For grade 1, 2, and 3a FLs, the preferred chemoimmunotherapy regimen is BR, based on high-level evidence of efficacy and favourable tolerability in this population.^20,23,24 In frail and older adults, rituximab monotherapy, a chemotherapy-free approach, is the preferred first-line regimen, according to European and North American guidelines.^25,26 However, some Canadian centres do not have access to rituximab monotherapy. Instead, physicians keep patients on BR for several treatment cycles or for as long as possible. Beyond first-line treatment, there is currently no gold standard for the care of the r/r FL population.

Second-Line Treatments

Treatment options for second-line regimens for r/r FL depend on several factors, including level of fitness, prior treatment, and length of time to relapse.²⁰ Combined immunochemotherapy, such as O-CHOP, is the preferred treatment strategy in this population.²⁰ According to the clinicians consulted by the sponsor, SCT may be considered in young and fit patients with no comorbidities in the second-line setting. Auto-SCT is given more often than allo-SCT in this population. However, only a small subset of FL patients would be eligible for transplant as a second-line option. SCT is limited by highly selective eligibility criteria and typically reserved for younger, medically fit patients with chemotherapy-sensitive disease.^20,25,51,52 The clinical experts consulted by CADTH agreed with these strategies.

Third-Line Treatments

FL is a relapsing disease with continued unmet needs in adults with r/r FL after 2 or more lines of therapy, despite available established therapies. Patients with r/r FL in the third-line setting and beyond represent a heavily pretreated and advanced stage patient population. The SOC in Canada in the third-line setting for patients with r/r FL is heterogenous and varies across the regions. Based on Canadian clinician input, a heterogenous mix of immunochemotherapy (for most patients) and SCT (for a minority of patients) are the current treatment options in this hard-to-treat population.

Treatments in this setting may include SCT; however, there is controversy with regard to the clinical benefit of SCT in patients with r/r FL, and both allo-SCT and auto-SCT may be associated with significant mortality and morbidity.^20,25,51-53 Although SCT has been included in the treatment algorithm diagram for r/r FL in the third-line setting, it is expected that the number of patients with r/r FL eligible for SCT in the third-line setting will be small.

According to clinicians consulted by the sponsor, most patients in the third-line setting will keep recycling combined immunochemotherapy that might have been used in previous lines, including the following regimens (which are the most frequently used options in Canada): R-CVP, R-CHOP, O-CHOP, R-GDP (rituximab, gemcitabine, dexamethasone, and cisplatin), BR, and R-ICE (rituximab, ifosfamide, carboplatin, and etoposide). While used by some physicians in their clinical practice, the R2 immunomodulatory regimen is currently not officially indicated for the treatment of FL in Canada, per the latest Canadian product monograph for lenalidomide.⁵⁴ Furthermore, although no official Health Technology Assessment submission has been made by the sponsor to Canadian health technology assessment agencies, CADTH conducted a health technology review of R2 in r/r B-cell NHL and concluded that available evidence remains limited.⁵⁵ According to Canadian clinicians’ input, there is somewhat limited access to R2 in some provinces. In the third-line setting and beyond, idelalisib, a phosphoinositide 3-kinase (PI3K) inhibitor, is indicated for the treatment of patients with r/r FL in Canada.⁵⁶ However, idelalisib is not publicly reimbursed across Canada and is inaccessible at some sites, according to several clinicians.⁵⁷ Furthermore, it is generally at the bottom of the treatment list due to important side effects and is used as a palliative treatment. In recent years, CAR T-cell therapy has emerged as another form of immunotherapy for the treatment of blood cancers, including lymphomas. Although promising results have been reported for the CAR T-cell therapies in patients with advanced stage lymphomas, severe adverse effects related to CAR T-cell therapy, such as CRS and neurologic toxicities, are also reported.²⁷ One CAR T-cell product, axicabtagene ciloleucel, was recently approved by Health Canada (September 2022), with an NOC/c for the treatment of adults with r/r grade 1, 2, or 3a FL after 2 or more lines of systemic therapy.⁵⁸ Note that at the time of this report, axicabtagene ciloleucel has not been reviewed by CADTH for the current indication, and it has not been publicly funded for this indication.

In clinical practice, a patient’s response to treatment is commonly assessed using the Lugano classification,⁵⁹ in which fluorodeoxyglucose PET/CT was incorporated into the initial Ann Arbor staging system for fluorodeoxyglucose-avid lymphomas.

Drug Under Review

Tisagenlecleucel is a second-generation autologous CAR T-cell therapy directed at the cell-surface protein CD19, which is only expressed on B cells or their precursors, and not on hematopoietic stem cells or non-B-cell tissues. The CAR comprises a murine single-chain antibody fragment that recognizes CD19 and is fused to intracellular signalling domains from 4 to 1BB (CD137) and cluster of differentiation 3 (CD3) zeta. The CD3 zeta component is critical for initiating T-cell activation and antitumour activity, whereas 4 to 1BB enhances the expansion and persistence of tisagenlecleucel. CD19 is expressed throughout most stages of normal B-cell differentiation (early pre-B to mature B cells) and is present in a wide range of B-lymphoid malignancies that span different stages of B-cell differentiation. Targeting CD19 by genetically modifying a patient’s own T cells to express an anti-CD19 directed CAR has been shown to effectively treat various advanced B-cell malignancies. Upon binding to CD19-expressing cells, the CAR transmits a signal to promote T-cell expansion, activation, target-cell elimination, and persistence of tisagenlecleucel.²⁸

On December 19, 2022, tisagenlecleucel received an NOC/c from Health Canada for adults with r/r grade 1, 2, or 3a FL after 2 or more lines of systemic therapy.²⁸ For 1 of the conditions for the authorization of tisagenlecleucel, the sponsor committed to undertake a randomized phase III trial of patients with r/r FL. Patients will be randomized to tisagenlecleucel or the investigator’s choice of treatment in line with the SOC. The primary end point will be PFS, and OS and ORR will be the secondary end points.⁶⁰

Per the warnings and precautions in the tisagenlecleucel product monograph, “Treatment should only be administered in a treatment facility with personnel fully trained and approved for the care of patients receiving Kymriah infusion therapy. Fully trained staff will administer the Kymriah infusion using precautions for immunosuppressed patients.” The recommended dose of tisagenlecleucel for the indicated population is 0.6 to 6.0 × 10⁸ CAR-positive viable T cells (nonweight based) suspended in 1 or more patient-specific infusion bags for a single-dose, one-time, IV administration.²⁸

Key characteristics of tisagenlecleucel and another CAR T-cell therapy, axicabtagene ciloleucel, which is indicated for adults with r/r FL, are summarized in Table 3. Other comparators for tisagenlecleucel, such as immunochemotherapy and SCT, are not included in this table.

Table 3: Key Characteristics of Tisagenlecleucel and Axicabtagene Ciloleucel

AE = adverse event; CAR = chimeric antigen receptor; CRS = cytokine release syndrome: FL = follicular lymphoma.

^aHealth Canada indication (NOC/c).

Sources: Product monographs for tisagenlecleucel²⁸ and axicabtagene ciloleucel.⁵⁸

Stakeholder Perspectives

Patient Group Input

This section was prepared by CADTH staff based on the input provided by patient groups. The full original patient input(s) received by CADTH have been included in the stakeholder section at the end of this report.

One patient group, LC, provided input to the submission. LC is a national Canadian registered charity. It collected input from patients with r/r FL through an online anonymous survey from November 2022 to January 2023. Of the 44 respondents, only 1 had experience with tisagenlecleucel. Skipping logic was built into the survey so that respondents were asked questions only relevant to them. In addition, LC obtained patient feedback from a French patient organization called ELLyE. ELLyE collected feedback from 162 patients on CAR T-cell therapy, including 19 who had experience with tisagenlecleucel, from January 2019 to September 2020 in France.

LC Survey

Most of the respondents to the LC survey were female, living in Canada, aged 55 years to 74 years, and had been diagnosed with FL 9 to 10 years before the survey.

At the time of diagnosis of FL, the most common symptoms are enlarged lymph nodes and fatigue, followed by indigestion, abdominal pain, bloating, body aches, and pain. Patients mentioned that a low immunoglobulin level has a significant negative impact in everyday life. The most common psychosocial impacts reported were stress of cancer diagnosis, anxiety, and fear of progression or relapse. FL has a negative impact on the ability to work, visit family and friends, travel, and contribute financially to household expenses. Some patients mentioned that because of their compromised immune system, they are worried about getting COVID and must limit their social interactions.

In terms of experiences with currently available treatments, 44%, 24%, and 15% of the respondents (n = 34) received 1, 2, and 3 lines of treatment, respectively. The most common treatment received by respondents was chemotherapy. Important side effects of treatment reported were fatigue, low activity level, hair loss, and neuropathy. Seventy percent of patients had access to treatment locally, with various levels of difficulty based on the location and waiting time. Absence from work, drug costs, and travel costs are financial aspects of lymphoma treatment for patients.

Patients also had problems with long waiting times to be diagnosed, see a specialist, have a biopsy, and have an MRI or CT scan.

Patients indicated that there is a need for more therapeutic options that provide longer disease remission, a longer life span, and improve QoL. Respondents indicated that new therapies should be available to all patients in Canada, and that patients should not have to travel to another province to get the treatment. Some respondents think it would be better to get stem cell or CAR T-cell treatments as a first-line treatment or in the earlier stages of the disease, when QoL has not been affected very much and the chances of successful therapy is higher. One patient suggested a more integrative approach to cancer treatment that included natural medicine.

Only 1 respondent in the 75- to 84-year age group had experience with tisagenlecleucel. The respondent has been in remission for more than 2 years at the time of survey. Although this patient had side effects, such as neutropenia, thrombocytopenia, decreased appetite, and anemia, the patient would recommend this therapy to other patients with r/r FL.

ELLyE Survey

Of the 19 patients (5 female and 14 male) who had been treated with tisagenlecleucel, information on the number of prior lines of treatments was available for the 14 patients who received tisagenlecleucel as a second-line (n = 6) or later-line treatment (n = 8). The most common prior treatment, received by 11 patients, was chemotherapy with or without immunotherapy.

The main concern of this group of patients was the impact of FL on their own QoL and that of their loved ones, such as fatigue, nausea, weight loss, anxiety, disrupted social life, and being dependent. The patients were also concerned about the financial and distance implications of travelling to get treatment in France.

Some patients indicated that the benefits from CAR T-cell therapy included “complete remission,” “treatment of pain,” and “hope of recovery to follow.”

The patient group input suggested that in both Canada and France, CAR T-cell therapy should be provided at an earlier stage of disease or as an earlier line of therapy.

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., 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 tisagenlecleucel review, a panel of 4 clinical experts from across Canada was convened to characterize unmet therapeutic needs, assist in identifying and communicating situations in which there are 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 in the following.

Unmet Needs

The clinical panel indicated that for patients with FL, the most important goals of an ideal treatment are to prolong survival (both OS and PFS) and improve the patient’s QoL. However, patients with r/r FL relapse after frontline therapies or are refractory to available treatments, which subsequently affects their long-term PFS and QoL. In addition, some patients may not tolerate current treatments well due to the related AEs or complications associated with SCT. The clinical panel indicated that the greatest unmet needs for treatments such as tisagenlecleucel are in patients who progress in the 2 years after initial therapy, those who have already received SCT, and those who have been double refractory to chemotherapy and rituximab (implying limited treatment options).

Place in Therapy

The clinical panel noted that the treatment algorithm for adults with r/r FL is complicated. Many factors (e.g., patient characteristics, previous treatments, treatment effects and toxicity, whether a treatment is reimbursed by drug plans, disease progression and transformation, and patient preference) need to be considered before deciding which treatment to provide. Watch and wait is a common approach for many patients with FL, even after disease relapse. Patients who need active treatments typically receive bendamustine or rituximab-based therapies such as BR, R-CVP, R-CHOP, or R2. PI3K inhibitors are rarely used. Patients with relapsed disease after treatment with chemoimmunotherapy, particularly those who progress within 2 years, may receive auto-SCT if they are suitable candidates. After all these treatments, some patients maintain indolent status and some transform to large cell lymphoma; therefore, the proportion of patients who may be treated with CAR T-cell therapy is small. The clinical panel suggested that tisagenlecleucel be used as third-line or later-line treatments for patients with r/r FL. These patients usually have a treatment response that lasts less than 6 months from their previous treatment (medication or SCT). There are not many options available for the patients at this stage. CAR T-cell therapy would be considered due to its different mechanism of action.

The clinical panel noted that an auto-SCT is not mandatory before tisagenlecleucel can be given, because this is not the SOC in Canada.

For patients who have received previous CD19-targeted therapy, there is a lack of evidence to suggest whether the use of tisagenlecleucel is appropriate.

Patient Population

The clinical panel indicated that in practice, CAR T-cell therapy is used in a broader patient population than that enrolled in clinical trials, where a more selective population is recruited. For example, the panel noted that suitable patients for treatment with tisagenlecleucel would be those with acceptable organ function rather than those with excellent organ function, which is generally required for an auto-SCT. In clinical practice, suitable patients can be identified based on clinical judgment, which combines medical history, laboratory and imaging findings, and often a lymph node biopsy. When determining the suitability of tisagenlecleucel for a patient, the bulk of disease and rapid disease progression are among the factors that need to be taken into account.

The panel noted that patients who are likely to benefit from other available treatments or who have an ECOG PS of 3 or higher are least suitable for treatment with tisagenlecleucel.

The panel also noted that there is not a specific patient characteristic that can be used to predict who will respond to tisagenlecleucel and who will not.

Assessing the Response to Treatment

The panel indicated that in clinical practice, patients who receive treatments for FL, including CAR T-cell therapy, are evaluated and followed in a manner similar to that described the clinical trials of investigational treatments for FL. Remission and survival are measured. Physical exams and imaging exams are routinely conducted to assess the patient’s response to CAR T-cell therapy.

The panel suggested that meaningful responses to treatment with tisagenlecleucel would be not only a high CRR, but also the durability of treatment response and long-term PFS and OS. In addition, the clinicians are interested in knowing if the treatment is cost-effective.

The panel noted that after CAR T-cell therapy, the clinicians will assess the treatment response (e.g., with a CT scan) at 3 months, or sooner if needed.

Patients Who Go Through Pretreatment but Do Not Receive Tisagenlecleucel

The panel noted that it is rare for patients with FL to go through pretreatment but not receive tisagenlecleucel. If this happens, it can be due to rapid disease progression in the interim or to the emergence of major complications, such as a new myocardial infarction or stroke. Manufacturer failure is another reason for this situation.

When it happens, the cell therapy product can be manufactured again. If patients do not receive tisagenlecleucel after going through pretreatment, most can progress in the 6 months after their previous treatment, assuming these are high-risk patients. There are limited treatment options available for them. Palliative chemotherapy can be given. Other options may include radiation therapy, more chemotherapy, novel drugs, or a clinical trial, depending on a patient’s clinical status.

Subsequent Therapy After Tisagenlecleucel Failure

The panel indicated that in the case of a relapse after infusion with tisagenlecleucel, patients may participate in a clinical trial. In the absence of a clinical trial, they may try a chemoimmunotherapy that they haven’t been exposed to.

Prescribing Considerations

The panel emphasized that a multidisciplinary team, involving hematologists, infectious disease specialists, neurologists, an ICU team, and other specialists, is required to diagnose, treat, and monitor patients who receive tisagenlecleucel to ensure the safe and effective delivery of this treatment.

Clinician Group Input

This section was prepared by CADTH staff based on the input provided by clinician groups. The full original clinician group inputs received by CADTH have been included in the stakeholder section at the end of this report.

Two clinician groups provided input for the review of tisagenlecleucel: Cell Therapy Transplant Canada, and the Ontario Health (Cancer Care Ontario) Hematology Cancer Drug Advisory Committee.

Unmet Needs

The clinician group indicated that poor overall response rates and short remission intervals are the main challenges in the treatment of patients with FL in the third-line setting and beyond. The chemoimmunotherapeutic strategies, when used a second time around, will result in a lack of efficacy and can be toxic for older patients, which adversely affects QoL. Allo-SCT is an option with curative potential, but the number of eligible patients is minimal. Another option, auto-SCT, which has excellent responses, is often used as a second-line or third-line treatment, but it is not appropriate for patients who are refractory to chemotherapy. Because most patients with FL are not eligible for SCT as a third-line therapy or beyond, CD19-targeted CAR T-cell therapy may be an option for these patients.

Place in Therapy

The clinician groups suggested that tisagenlecleucel be used in patients with r/r FL who are not eligible to receive allo-SCT or auto-SCT in the third-line setting or beyond. In addition, CD19 CAR T-cell therapy, including tisagenlecleucel, will only be considered in patients without significant organ dysfunction in their opinion. The clinician groups believed that CAR T-cell therapy will shift the current treatment paradigm but will not change the current approach to the use of auto-SCT or allo-SCT.

Patient Population

The clinician groups considered the following patients to be eligible for tisagenlecleucel:

• adults with FL (grade 1, 2, or 3a) refractory to 2 or more lines of systemic therapy that includes an anti-CD20 monoclonal antibody and an alkylating drug

• adults with FL (grade 1, 2, or 3a) who relapse after the second line of therapy, including auto-SCT for those who are eligible for transplant.

The clinician groups considered the following patients to be ineligible for tisagenlecleucel:

• patients previously treated with allo-SCT or with active central nervous system (CNS) involvement

• patients with any suspicion of disease transformation (grade 3b FL or diffuse large B-cell lymphoma), based on a recent biopsy

• patients who previously received CD19-directed therapy.

The clinician groups indicated that significant organ-specific issues (liver, renal, pulmonary), as well as active and/or uncontrolled infection and uncontrolled autoimmune disease, may be reasons not to pursue CAR T-cell therapy. The use of CAR T-cell therapy might be preferred before auto-SCT in some patients.

Assessing Response to Treatment

The clinician groups noted that assessment of response to treatment should be based on standard lymphoma response criteria identified with clinical exams and imaging scans such as CT and PET. Outcomes such as remission rates, PFS, patient safety, and HRQoL should be measured.

Discontinuing Treatment

Not applicable.

Prescribing Conditions

The clinician groups indicated that CAR T-cell therapy in Canada is currently provided in cell therapy centres that are approved by Health Canada. Approval from the Foundation for the Accreditation of Cellular Therapy is also an important consideration. A foundation-approved transplant centre will have the necessary standard of programs and policies to safely provide CAR T-cell treatment. Institutional experience, trained CAR T-cell health care providers, tumour boards, and data collection capabilities are also required. Within an institution, infectious disease, ICU, or emergency department involvement, as well as specific expertise from the neurology service, need to be in place. To use tisagenlecleucel, a specific centre must have successfully completed an onboarding process that includes red cell microparticle dry runs, stem cell processing, and apheresis audits, as well as specific legal agreements.

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 4.

Table 4: Summary of Drug Plan Input and Clinical Expert Response

CAR = chimeric antigen receptor; CNS = central nervous system; CRS = cytokine release syndrome; DLBCL = diffuse large B-cell lymphoma; ECOG PS = Eastern Cooperative Oncology Group Performance Status; FL = follicular lymphoma; GDP = gemcitabine, dexamethasone, and cisplatin; ICE = ifosfamide, carboplatin, and etoposide; MALT = mucosa-associated lymphoid tissue; PAG = Provincial Advisory Group; pERC = CADTH pan-Canadian Oncology Drug Review Expert Review Committee; r/r = relapsed or refractory; R2 = lenalidomide plus rituximab; SCT = stem cell transplant.

Clinical Evidence

The objective of CADTH’s Clinical Review Report is to review and critically appraise the clinical evidence submitted by the sponsor on the beneficial and harmful effects of tisagenlecleucel in the treatment of adults with r/r grade 1, 2, or 3a FL after 2 or more lines of systemic therapy. The focus will be placed on comparing tisagenlecleucel to relevant comparators and identifying gaps in the current evidence.

A summary of the clinical evidence included by the sponsor in the review of tisagenlecleucel is presented in 3 sections, and CADTH’s critical appraisal of the evidence is included after each section. The first section, the Systematic Review, includes pivotal studies and RCTs that were selected according to the sponsor’s systematic review protocol. The second section includes indirect evidence from the sponsor. The third section includes additional studies that were considered by the sponsor to address important gaps in the pivotal and RCT evidence.

Included Studies

Clinical evidence from the following is included in the CADTH review and appraised in this document:

• 1 pivotal study (ELARA²⁹)

• 1 ITC (study comparing tisagenlecleucel and axicabtagene ciloleucel³¹)

• 2 additional studies addressing gaps in evidence (Schuster et al. [2017],³³ and a study comparing tisagenlecleucel with SOC³²).

Pivotal Studies and RCT Evidence

The contents of this section have been informed by materials submitted by the sponsor. The following information has been summarized and validated by the CADTH review team.

Description of Studies

Characteristics of the included studies are summarized in Table 5.

Table 5: Details of Evidence Identified by the Sponsor

Allo-SCT = allogeneic stem cell transplant; CAR = chimeric antigen receptor; CNS = central nervous system; CRR = complete response rate; DoR = duration of response; EAS = efficacy analysis set; ECOG PS = Eastern Cooperative Oncology Group Performance Status; FL = follicular lymphoma; IQR = interquartile range; IRC = independent review committee; NA = not applicable; ORR = overall response rate; OS = overall survival; PFS = progression-free survival; PRO = patient-reported outcome; PRR = partial response rate; ROA = route of administration; SD = standard deviation; SCT = hematopoietic stem cell transplant;.

^aDates reported are the study start and end dates from the Clinicaltrials.gov record.

^bEnrolment was defined as the point when the patient met all inclusion and/or exclusion criteria and the patient’s leukapheresis product was received and accepted by the manufacturing facility.

Source: ELARA Clinical Study Report.²⁹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

ELARA is a multicentre, international, open-label, single-arm phase II trial. The objective was to determine the efficacy, safety, pharmacokinetics (also referred to as cellular kinetics), and PROs of tisagenlecleucel in adults with r/r FL after 2 or more lines of systemic therapy. The ELARA study design included 3 sequential phases: screening, pretreatment, and treatment and follow-up, as illustrated in Figure 1. During the screening phase and before study enrolment, a patient’s blood mononuclear cells were collected via leukapheresis. In the pretreatment phase, patients could undergo optional bridging therapy to control their leukemia and LD chemotherapy. The treatment and follow-up phases involved a tisagenlecleucel one-time IV infusion and safety and efficacy follow-up that lasted at least 24 months. For all patients who received an infusion of tisagenlecleucel, additional survival follow-up was to be performed to determine survival status every 3 months. Efficacy was evaluated using PET, CT, or MRI, based on Lugano classification response criteria.⁵⁹ Efficacy was evaluated at months 3, 6, 9, 12, 18, and 24 after infusion, and every 6 months thereafter until end of study, which was defined as the point when all patients completed their month 24 evaluation or discontinued prematurely. Onsite assessments were to be performed any time disease progression or relapse was suspected, until disease progression or relapse, the start of new anticancer therapies, death, loss to follow-up, or withdrawal of consent. The primary efficacy end point was CRR per IRC through 24 months.

Data reported in this review are for the most recent data cut-off date of March 29, 2022 (corresponding with the date when 90 patients had completed follow-up of 24 months from infusion or discontinued earlier). A prior data cut-off of March 29, 2021 was conducted (corresponding with the date that 90 patients had completed follow-up of 12 months from infusion or discontinued earlier).

Figure 1: Design of the ELARA Study

Source: ELARA Clinical Study Report.²⁹

Populations

Inclusion and Exclusion Criteria

Patients eligible for inclusion in this study were required to be at least 18 years of age and to have histologically confirmed r/r FL (grade 1, 2, or 3a) by central pathology review. The FL had to meet at least 1 of the following criteria:

• refractory to a second or later line of systemic therapy (including an anti-CD20 antibody and an alkylating drug) or relapsed within 6 months of completion of a second or later line of systemic therapy

• relapsed during anti-CD20 antibody maintenance (following at least 2 lines of therapies) or within 6 months of maintenance completion

• relapsed after auto-SCT

• radiographically measurable disease at screening

• ECOG PS of 0 or 1 at screening

• a leukapheresis product of nonmobilized cells accepted for manufacturing.

Key exclusion criteria were evidence of histologic transformation and/or grade 3b FL and prohibited prior therapies, including prior anti-CD19 or anti-CD20 therapy, gene therapy, adoptive T-cell therapy, or allo-SCT. Patients who had active CNS abnormalities were excluded from the ELARA trial as well.

Interventions

The recommended dose of tisagenlecleucel in the ELARA trial was 0.6 to 6.0 × 10⁸ CAR-positive viable T cells administered as a one-time, single IV infusion. Prior to tisagenlecleucel infusion, all patients were required to receive LD chemotherapy. This step was to be omitted in cases of significant cytopenia (e.g., white blood cell count < 1,000 cells/μL, absolute lymphocyte count < 200 cells/μL) or any condition that, in the investigator’s opinion, precluded LD chemotherapy. LD chemotherapy was started 1 week before tisagenlecleucel infusion so that the CAR-positive viable T cells could be administered 2 to 6 days after completion of the LD chemotherapy. The chemotherapy start date varied based on the selected chemotherapy. For LD chemotherapy, cyclophosphamide-based regimens were the preferred drugs due to the vast experience with the use of these drugs in the facilitation of adoptive immunotherapy. The first option as a LD regimen was:

• fludarabine (25 mg/m² IV daily for 3 doses) plus cyclophosphamide (250 mg/m² IV daily for 3 doses starting with the first dose of fludarabine).

If there was previous grade 4 hemorrhagic cystitis with cyclophosphamide or if a patient had resistance to a previous cyclophosphamide-containing regimen, then the following regimen was allowed:

• bendamustine 90 mg/m² IV daily for 2 days.

No other regimen was allowed for LD chemotherapy.

Bridging Therapy

Of the 97 patients infused, 44 (45.4%) patients received optional antineoplastic bridging therapy before tisagenlecleucel infusion. The most common drugs (in ≥ 5% of patients) were rituximab (21.6%), dexamethasone (11.3%), gemcitabine (10.3%), oxaliplatin (7.2%), prednisolone (7.2%), etoposide (6.2%), cyclophosphamide (5.2%), and vincristine (5.2%). In 5 patients, only corticosteroids were administered as bridging therapy. Further, 2 patients received bridging radiotherapy: 1 patient received only radiotherapy and the other patient received radiotherapy and corticosteroids.

Lymphodepleting Chemotherapy

All 97 infused patients received LD chemotherapy before the tisagenlecleucel infusion. Most patients (92 [94.8%]) received fludarabine plus cyclophosphamide, whereas the remaining 5 (5.2%) patients received bendamustine.

Prior to tisagenlecleucel infusion, a minimum of 2 doses of tocilizumab per patient and emergency equipment had to be confirmed as available and had to be available for infusion within 2 hours for the management of CRS-related AEs.

All patients (n = 97) who received tisagenlecleucel also received concomitant medications (Table 6). As reported by Anatomic Therapeutic Chemical class, most patients received anilides (57.7%), sulfonamides plus trimethoprim, (54.6%), nucleosides plus nucleotides (49.5%), proton pump inhibitors (37.1%), and colony-stimulating factors (37.1%). In all, 17 (17.5%) patients in the infused set received anticytokine medication for CRS-related AEs.

Table 6: Concomitant Medication Use by ATC Class and Preferred Term in the ELARA Study

ATC = Anatomical Therapeutic Chemical.

Notes: In the ELARA study, concomitant treatments with n ≥ 20 are only included from a long list of therapies in the E2202 Clinical Study Report; the cut-off date was March 29, 2022.

Concomitant medications on or after the start of study treatment or starting before and continuing after the start of study treatment are summarized. A medication or therapy can be included in more than 1 ATC class.

Source: ELARA Clinical Study Report.²⁹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Outcomes

The efficacy end points assessed in this Clinical Review Report are provided in Table 7 and are subsequently summarized. Summarized end points are based on those included in the sponsor’s Summary of Clinical Evidence, as well as those identified as important to this review, according to stakeholders, such as clinical experts, clinician groups, or patient groups.

Table 7: Outcomes Summarized From the ELARA Study

AE = adverse event; CRR = complete response rate; DoR = duration of response, ORR = overall response rate, OS = overall survival; PFS = progression-free survival; PRO = patient-reported outcome.

Note: There was no key secondary outcome and secondary outcomes were not tested, according to a prespecified statistical hierarchy or adjusted for multiple comparisons. The analysis of all primary and all secondary outcomes were performed using the efficacy analysis set.

Source: ELARA Clinical Study Report.²⁹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Definitions of the various outcomes reported in the ELARA study are as follows:

• CRR was defined as the proportion of patients with a best overall response of complete response (CR) recorded from tisagenlecleucel infusion until progressive disease or the start of new anticancer therapy, whichever came first.

• ORR was defined as the proportion of patients with a best overall disease response of CR or partial response (PR).

• DoR was defined as the time from the date of first documented disease response (CR or PR) to the date of first documented progression or death due to FL. If a patient did not have an event before the earliest censoring event, DoR was censored at the date of the last adequate assessment on or before the earliest censoring event.

• PFS was defined as the time from the date of tisagenlecleucel infusion to the date of first documented progression or death due to any cause. If a patient did not have an event, PFS was censored at the date of the last adequate assessment. The censoring reasons could be any of the following:

  • ongoing without an event

  • lost to follow-up

  • withdrew consent

  • new anticancer therapy (including SCT if the patient proceeded to SCT after tisagenlecleucel infusion)

  • event documented after at least 2 missing tumour assessments

  • adequate assessments no longer available.

PFS was also analyzed as the time from enrolment to the date of progression or death due to any reason in the enrolled set.

• OS was defined as the time from the date of tisagenlecleucel infusion to the date of death due to any reason. If a death was not observed by the data cut-off date, OS was censored at the date of last contact. The distribution function of OS was estimated using the KM method. OS was also analyzed as the time from enrolment to the date of death due to any reason for in the enrolled set.

• Posttisagenlecleucel AE summaries included all AEs that started or worsened during the postinfusion period (i.e., within 8 weeks of the first tisagenlecleucel infusion, 8 weeks to 1 year after tisagenlecleucel infusion, more than1 year after the first tisagenlecleucel infusion, and any time after the tisagenlecleucel infusion. The postinfusion period was the main period of safety reporting.

• PROs included FACT-Lym, SF-36, and EQ-5D-3L, and were assessed at baseline and postbaseline scheduled assessments.

  • The FACT-Lym is a questionnaire designed to assess HRQoL in patients with lymphoma. It consists of a generic core HRQoL instrument (FACT-G) and a cancer-site-specific module (Lym-S). The FACT-G has 27 statements that patients need to rate on a 5-point scale (not at all, a little, somewhat, quite a bit, very much). The statements cover 5 subscales (physical well-being, social/family well-being, emotional well-being, functional well-being, and additional concerns). The Lym-S module consists of 15 statements patients need to rate on an identical 5-point scale specific for NHL. This scale is designed for patient self-administration. Higher scores demonstrate improvement in QoL. Minimally clinically important differences (MCIDs) were estimated to range from 5.5 to 11 for the FACT-Lym trial outcome index, 3 to 7 for the FACT-G total score, 2.9 to 5.4 for the FACT-Lym subscale score, and 6.5 to 11.2 for FACT-Lym total score.⁷⁶

  • The SF-36 is a widely used and extensively studied generic instrument to measure HRQoL among healthy patients and patients with acute and chronic conditions. It consists of 8 subscales that can be scored individually: physical functioning, role-physical, bodily pain, general health, vitality, social functioning, role-emotional, and mental health. Two overall summary scores, the physical component summary and the mental component summary, also can be computed. The SF-36 has proven useful in monitoring general and specific populations, comparing the relative burden of different diseases, differentiating the health benefits produced by different treatments, and screening individual patients.²⁹ There were no FL-specific MCIDs identified for the SF-36 in the literature; however, an absolute score increase of 3 to 5 points for both the physical component summary and mental component summary was reported as a MCID across various conditions, such as idiopathic pulmonary fibrosis and colitis.^77,78

  • The EQ-5D-3L is a widely used, self-administered questionnaire designed to assess health status in adults. The measure is divided into 2 distinct sections. The first section includes 1 item addressing each of 5 dimensions (mobility, self-care, usual activity, pain/discomfort, and anxiety/depression). Patients rate each of these items, from no problem, some problem, or extreme problem. A composite health index is then defined by combining the levels for each dimension. The second section of the questionnaire measures self-rated (global) health status using a vertically oriented VAS, where 100 represents the best possible health state and 0 represents the worst possible health state. Respondents are asked to rate their current health by placing a mark along this continuum. The recall period is “today,” and the questionnaire requires approximately 5 to 10 minutes to complete. In the ELARA study, only the change in VAS scores were reported. No disease-specific MCID for the EQ-5D-3L VAS score in patients with FL was reported; however, in the general population an increase in the VAS score of 10.9 was considered to indicate improved health.^29,79

Statistical Analysis

A prior study of idelalisib in patients with r/r FL after 2 lines of therapy reported an observed CRR of 14%.⁸⁰ This result was used as the null hypothesis for the ELARA study because idelalisib was the therapy with best CRR among the approved and widely used treatment options in this setting at the time the trial was being planned. Therefore, based on the null hypothesis of a CRR of 15% or less and assuming an underlying CRR of 30% for tisagenlecleucel, 90 patients in the primary analysis would provide at least 90% cumulative power to demonstrate statistical significance, using a 2-look Lan-DeMets group sequential design with O’Brien-Fleming type boundary and an exact CI at a 1-sided cumulative 0.025 level of significance, if the underlying CRR is 30%. In this setting, a CRR of 23.3% (21 of 90 patients) was needed to claim success. Therefore, assuming that 20% of enrolled patients will not be infused for reasons such as manufactory failure or worsening of the patient’s condition, at least 113 patients needed to be enrolled to ensure that 90 patients were treated and available for the primary analysis.

All statistical analyses were performed using data collected up to the cut-off date of March 29, 2022. The analysis of the primary outcome was performed by testing the null hypothesis of a CRR of 15% or less at a 1-sided cumulative 2.5% level of significance (i.e., P ≤ 0.15 for H₀ versus P > 0.15 for H_a). The 15% was derived from the CRR from approved PI3K inhibitors for the treatment of FL that relapsed after 2 prior systemic therapies based on single-arm, open-label, phase II studies. The CRR was analyzed at an interim analysis (using the interim efficacy analysis set [EAS]) and the primary analysis (EAS) of a group sequential design. The CRR was summarized along with 2-sided exact Clopper-Pearson CIs, with coverage level determined by the O’Brien-Fleming type alpha-spending approach, according to Lan and DeMets (1983).⁸¹ The study was considered successful if the lower bound of the 2-sided exact CI for ORR was greater than 15%, so that the null hypothesis that the CRR was less than or equal to 15% could be rejected. Patients who were of unknown clinical response were treated as nonresponders.

PFS, OS, and DoR were estimated using the KM method. The median OS and the proportion of patients alive at 3, 6, 12, 18, and 24 months, with 95% CIs, were presented. The median PFS and the proportion of patients without an event at 3, 6, 9, and 12 months, with 95% CIs, were presented. The ORR was summarized along with the 2-sided 95% exact Clopper-Pearson CIs. DoR was summarized for patients with a CR only, as well as for those with a CR or PR. The median response duration, as well as proportion of patients without event at 3, 6, 9, and 12 months, with 95% CIs, were presented.

Reporting of AEs (except for CRS) was based on MedDRA version 25.0 and Common Terminology Criteria for Adverse Events (CTCAE) version 4.03. There was no key secondary outcome and secondary outcomes were not tested, according to a prespecified statistical hierarchy, or adjusted for multiple comparisons. All analyses of secondary outcomes were performed using the EAS. For PROs, summary scores were calculated and descriptive statistics and change from baseline of the summary scores of each domain were derived from all available data at the time of the analysis, using the EAS.

Sensitivity Analyses

• The analysis of the primary outcome was performed on the enrolled set, the tisagenlecleucel-infused set, and the per-protocol set using the same methodology, and on the modified EAS and the EAS, which excluded patients who achieved a CR at the radiologic assessment at baseline per IRC. Consistent results were observed when CRR was analyzed across the different analysis sets, and the results were consistent with those of the EAS.

• The secondary outcomes included ORR, DoR, PFS, OS, safety, and PROs. IRC assessment was used in the main analysis of secondary outcomes that involved disease response. All analyses of the secondary efficacy outcomes were performed on the EAS. In addition, selected analyses were performed for the tisagenlecleucel-infused set and/or for the enrolled set.

Subgroup Analyses

The following prespecified subgroups of interest were used for the supporting efficacy analysis of the primary outcome:

• age (< 65 years, ≥ 65 years)

• sex (male, female)

• race (Asian, Black or African American, white, Native Hawaiian or Other Pacific Islander or American Indian, Alaska Native)

• ethnicity (Hispanic or Latino, not Hispanic or Latino)

• FLIPI (low or intermediate, high)

• histological grade (1, 2, 3a)

• number of prior lines of antineoplastic therapy (≤ 2 lines, 3 or 4 lines, > 4 lines)

• PI3K inhibitor use (naive, pretreated)

• prior SCT therapy (yes, no) and relapse (12 months or less from SCT, more than 12 months from SCT)

• disease status to last line of prior antineoplastic therapy (refractory, relapsed)

• progression of disease within 24 months of first-line treatment (POD24) with anti-CD20

• mAb-containing therapy (yes, no)

• bulky disease at baseline (defined per IRC as imaging showing any nodal or extranodal tumour mass that is > 7 cm in diameter or the involvement of at least 3 nodal sites, each with a diameter > 3 cm) (yes, no)

• bridging therapy (yes, no)

• lactate dehydrogenase (LDH) at study entry (≤ upper limit of normal, > upper limit of normal)

• R2 use lenalidomide plus rituximab in same regimen (naive, pretreated)

• US sites (yes, no)

• total metabolic tumour volume at baseline (low tumour burden [tumour volume ≤ 510 cm³], high tumour burden [tumour volume > 510 cm³]

• double refractory, defined as patients who failed to respond or relapsed within 6 months of therapy with anti-CD20 and alkylating drugs, any regimen (yes, no)

Of these subgroups, the ones relevant to the CADTH review (i.e., that may have important prognostic, confounding, effect modification influence on outcomes) based on clinician input are age, FLIPI score, number of prior lines of antineoplastic therapy, prior SCT, POD24, bulky disease, bridging therapy, and LDH at study entry.

Handling of missing values: Patients in this study who are of unknown clinical response were treated as nonresponders. Other missing data were simply noted as missing on appropriate tables and/or listings.

Table 8: Statistical Analysis of Efficacy End Points

CR = complete response; CRR = complete response rate; DoR = duration of response; NA = not applicable; OS = overall survival; PFS = progression-free survival; PPS = per-protocol set.

Source: ELARA Clinical Study Report.²⁹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Data Cut-off Dates

In the ELARA trial, analyses were conducted at various cut-off dates.

The data cut-off for the interim analysis was May 26, 2020, when 52 patients who received a tisagenlecleucel infusion had either been followed for 6 months or discontinued earlier.

The prespecified data cut-off for the primary analysis was September 28, 2020, when 94 patients had been followed for at least 6 months or discontinued earlier. This analysis was used for Health Authority interactions.

An extended follow-up analysis (data cut-off of March 29, 2021) was conducted after 90 patients who received a tisagenlecleucel were followed for at least 12 months or discontinued earlier, and was the basis for registration applications.

The 24-month follow-up analysis, with a data cut-off of March 29, 2022, was conducted after approximately 90 patients who received a tisagenlecleucel infusion and were followed for 24 months or discontinued earlier. Results of this analysis were presented in the current review.

Note that on the primary data cut-off date (September 28, 2020), the primary end point (CRR determined by IRC based on Lugano classification response criteria) was met, with a CRR of 66.0% (95% CI, 55.5% to 75.4%). The lower bound of the 95% CI was greater than 15%. The results were consistent across all data cut-off dates. The observed benefits were maintained at later data cut-off dates, including March 29, 2022.²⁹

Analysis Populations

The analysis populations in the ELARA trial are summarized in Table 9.

The EAS consisted of all patients who received tisagenlecleucel and had measurable disease at baseline per IRC. This analysis set was used for all efficacy analyses of the primary and secondary outcomes.

The safety set consisted of all patients who received tisagenlecleucel. This analysis set was used for all safety analyses.

Table 9: Analysis Populations of the ELARA Study

EAS = efficacy analysis set; IRC = independent review committee; mEAS = modified efficacy analysis set; PPS = per-protocol set.

Source: ELARA Clinical Study Report.²⁹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Results

Patient Disposition

A summary of patient disposition in the ELARA trial is provided in Table 10.

A total of 119 patients were screened by the data cut-off (March 29, 2022). Of these, 98 (82.4%) patients were enrolled and 21 (17.6%) patients were excluded. All but 1 enrolled patients (n = 97) received their tisagenlecleucel infusion. Of the 98 patients enrolled, 73 (74.5%) patients were in follow-up at the time of the data cut-off and 25 (25.5%) patients had discontinued the study. The most common reason for discontinuation was death (i.e., 13 [13.3%] patients).

Table 10: Summary of Patient Disposition From the ELARA Study

EAS = efficacy analysis set, PPS = per-protocol set.

Source: ELARA Clinical Study Report.²⁹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Baseline Characteristics

In total, 98 patients were enrolled in this study and 97 received treatment with tisagenlecleucel.

The baseline characteristics of the included study population in the ELARA trial are detailed in Table 11.

At the data cut-off date of March 29, 2022, a total of 94 patients with r/r FL had received a one-time, single infusion of tisagenlecleucel. The median age observed in the overall r/r FL population was 57 years (range, 29 to 73 years). More males (68.1%) were enrolled than females (31.9%), and most patients (84.0%) were white. Almost all patients (97%) had a baseline ECOG PS of 0 or 1. Most patients also had grade 1 to 2 (90.4%) and stage III to IV (80%) disease. In terms of prior therapies, enrolled patients had received a median of 4 (range, 2 to 13) prior lines of treatments. All of the included patients (n = 98; 100%) received prior therapy with an anti-CD20 mAb and alkylating drugs, and 36.7% of the patients received prior SCT. Of the 98 included patients, 77.6% were refractory to the last line of antineoplastic therapy. The proportion of patients who progressed within 24 months of first-line anti-CD20 mAb-containing therapy was 64.9%.

The previous therapies received by patients in the ELARA trial are summarized in Table 12.

Table 11: Summary of Patient Baseline Characteristics in the ELARA Trial

EAS = efficacy analysis set; ECOG PS = Eastern Cooperative Oncology Group Performance Status; FLIPI = Follicular Lymphoma International Prognostic Index; LDH = lactate dehydrogenase; NR = not reported; POD24 = progression of disease within 24 months of first-line treatment; Q1 = first quartile; Q3 = third quartile; SD = standard deviation; ULN = upper limit of normal.

^aThese 3 patients had an ECOG PS of 2 recorded just before receiving tisagenlecleucel infusion, and not at the time of signing the informed consent form.

Source: ELARA Clinical Study Report.²⁹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Table 12: Summary of FL Therapies Administered Prior to Tisagenlecleucel in the ELARA Trial

FL = follicular lymphoma; mAb = monoclonal antibody; PI3K = phosphoinositide 3-kinase; POD24 = progression of disease within 24 months of first-line treatment; SCT = hematopoietic stem cell transplant.

Source: ELARA Clinical Study Report.²⁹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Exposure to Study Treatments

As tisagenlecleucel is given as a single-dose, one-time IV infusion, all patients who were infused in the ELARA trial received 100% of the dose. The median time from enrolment to infusion is summarized in Table 13. In the ELARA trial, 97 patients received tisagenlecleucel treatment with a median duration of 46 days (range, 23 to 127 days) from enrolment to treatment infusion.

Table 13: Overview of Time to Infusion in the ELARA Trial

NA = not applicable; SD = standard deviation.

^aTime from enrolment to tisagenlecleucel infusion.

Source: ELARA Clinical Study Report.²⁹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

The recommended tisagenlecleucel dose range in the ELARA study was 0.6 to 6.0 × 10⁸ CAR-positive viable T cells. All except 4 patients received tisagenlecleucel within the targeted dose range. The median dose administered was 2.06 × 10⁸ CAR-positive viable T cells (range, 0.1 to 6.0 × 10⁸ cells). The median total viable cell count was 12 × 10⁸ cells (range, 0.4 to 34.0 × 10⁸ cells).

Overall, 22 (22.7%) patients received at least 1 new antineoplastic medication after the tisagenlecleucel infusion, mostly of whom had stable disease or progressive disease (Table 14). Regarding the types of subsequent therapy, most patients received lenalidomide (12.4%) and/or rituximab (11.3%) after the infusion.

Upon CADTH’s request, the sponsor specified that in the ELARA study, no patients received a subsequent auto-SCT, while 4.12% of the patients have received a subsequent allo-SCT.

Table 14: Antineoplastic Medication Received After Tisagenlecleucel Infusion in the ELARA Trial

Note: A medication can be included in more than 1 ATC class.

Source: ELARA Clinical Study Report.²⁹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Efficacy

Response Rates

Response rates with tisagenlecleucel used in the third-line setting for patients with r/r FL are provided in Table 15.

In the ELARA trial, among the 97 patients who were treated with tisagenlecleucel, the CRR was 68.1% (95% CI, 57.7% to 77.3%), the ORR was 86.2% (95% CI, 77.5% to 92.4%), and the PRR was 18.1% (95% CI, NR) at 24-month follow-up per IRC assessment. The median duration of follow-up was 28.85 months from the time of the infusion to the data cut-off.²⁹ The results from local assessment were consistent with the IRC assessment.

Table 15: Summary of Response Rates in the ELARA Trial

CI = confidence interval; CRR = complete response rate; EAS = efficacy analysis set; IRC = independent review committee; NR = not reported; ORR = overall response rate; PD = progressive disease; PRR = partial response rate.

Source: ELARA Clinical Study Report.²⁹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Results of the predefined sensitivity analyses were similar to those in the primary analysis of CRR per IRC assessment. The CRR per local assessment was 73.4% (95% CI, 63.3% to 82.0%), which was consistent with the IRC assessment. In addition, consistent results with the EAS were observed when CRR was analyzed using different analysis sets (i.e., enrolled set, tisagenlecleucel-infused set, and per-protocol set).

Results of the prespecified subgroup analyses were consistent with the primary analysis of CRR. The CRRs across the subgroups ranged from 40.0% to 84.8%, which were similar to the overall study population in the EAS. Results in the subgroups of age, FLIPI score, number of prior lines of therapies, prior SCT use, POD24 status, bulky disease, bridging therapy, and LDH level at baseline are presented in Table 36 in Appendix 1. These subgroups are considered clinically important by the experts consulted for this review.

Duration of Response

Median DoR was not reached with tisagenlecleucel at the median follow-up time in ELARA (Table 16). The median duration of follow-up was 20.7 months (range, 0 to 29.6 months).

Overall Survival

The ELARA study reported median OS and OS event rates with tisagenlecleucel at different time points (Table 17).

In the ELARA trial, median OS was not reached at the 24- month follow-up (cut-off date of March 29, 2022). Thirteen deaths occurred in the study, and the OS rate was 87.7% (95% CI, 78.3% to 93.2%) and 82.6% (95% CI, 70.2% to 90.2%) at 24 and 30 months, respectively.

Table 16: Summary of DoR in the ELARA Trial

CI = confidence interval; DoR = duration of response; EAS = efficacy analysis set; IRC = independent review committee.

Source: ELARA Clinical Study Report.²⁹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Table 17: Summary of OS in the ELARA Trial

CI = confidence interval; EAS = efficacy analysis set; FU = follow-up; NE = not estimable; OS = overall survival.

Source: ELARA Clinical Study Report.²⁹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Progression-Free Survival

Median PFS and PFS event rates with tisagenlecleucel are summarized in Table 18. The PFS rate with tisagenlecleucel was evaluated at multiple time points and data cut-off dates, ranging from 6 months to 24 months.

In the ELARA trial, the median PFS per IRC was not reached at the time of the data cut-off (March 29, 2022) and there were 38 PFS events in total (disease progression or death). The PFS rate was 77.8% (95% CI, 67.7% to 85.1%), 67.2% (95% CI, 56.3% to 75.9%), and 57.4% (95% CI, 46.2% to 67.0%) at 6, 12, and 24 months, respectively.

Fifty-six patients were censored from the analysis for the following reasons: 51 patients were ongoing without an event, 3 patients started new anticancer therapy other than SCT, 1 patient withdrew their consent, and 1 patient was lost to follow-up.

PFS results assessed by the local investigator were consistent with IRC assessment.

Table 18: Summary of PFS in the ELARA Trial

CI = confidence interval; EAS = efficacy analysis set; FU = follow-up; IRC = independent review committee; NE = not estimable; PFS = progression-free survival.

Source: ELARA Clinical Study Report.²⁹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Patient-Reported Outcomes

In the ELARA trial, PROs were assessed in the EAS (n = 94) to evaluate the impact of tisagenlecleucel on patients’ HRQoL. Three PRO instruments were used to capture impact on QoL: FACT-Lym, SF-36, and EQ-5D-3L. As previously reported, MCIDs were estimated to range from 5.5 to 11 for the FACT-Lym trial outcome index, 3 to 7 for FACT-G total score, 2.9 to 5.4 for the FACT-Lym subscale score, and 6.5 to 11.2 for the FACT-Lym total score.⁷⁶ For SF-36, MCIDs were estimated to be 3 for both the physical component and the mental component scores.⁷⁷

FACT-Lym and SF-36 Questionnaires

Over time, there were few notable changes in the proportion of patients reporting improved, stable, or deteriorated FACT-Lym or SF-36 scores after the tisagenlecleucel infusion (Figure 2). Overall, 70% to 88% of patients reported no deterioration in QoL based on the FACT-Lym and SF-36 scores at 12 months, with similar trends observed at 18 and 24 months.

Figure 2: Proportion of Patients Who Deteriorated, Improved, or Did Not Change on FACT-Lym and SF-36 Questionnaires (EAS)

EAS = efficacy analysis set; FACT = Functional Assessment of Cancer Therapy; Lym = lymphoma; MID = minimally important differences; SF-36 = 36-Item Short Form Survey.

Source: ELARA Clinical Study Report.²⁹

EQ-5D-3L Questionnaire

Results of the EQ-5D-3L VAS score showed that HRQoL was maintained from baseline after the tisagenlecleucel infusion. The mean EQ VAS score was 69.4 at baseline and increased to 72.5 at 6 months, 75.9 at 12 months, and 71.9 at 24 months (Table 19). The 24-month results were based on approximately half of the population at baseline.

Table 19: EQ VAS Score and Change From Baseline by Visit in the ELARA Trial (EAS)

EAS = efficacy analysis set; VAS = visual analogue scale; SD = standard deviation.

Source: ELARA Clinical Study Report.²⁹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Harms

In the ELARA study, all safety outcomes were reported in the safety analysis set (i.e., all patients who received an infusion of tisagenlecleucel). After the one-time infusion of tisagenlecleucel, patients were followed for up to 24 months in the ELARA study. Refer to Table 20 for harms data.

Adverse Events

Of the 97 patients evaluable for safety, 99% experienced any-grade AEs.

The most common AEs reported in at least 15% of patients any time after the infusion were CRS (49.5%), neutropenia (43.3%), anemia (25.8%), diarrhea (25.8%), headache (23.7%), decreased white blood cell count (22.7%), pyrexia (18.6%), thrombocytopenia (18.6%), fatigue (17.5%), nausea (17.5%), decreased neutrophil count (17.5%), constipation (16.5%), and hypogammaglobulinemia (15.5%).

Serious Adverse Events

A total of 46.4% of patients experienced an SAE at any time after the infusion. The most commonly reported SAEs included CRS (19.6%), pneumonia (10.3%), and febrile neutropenia (8.2%).

Withdrawal Due to Adverse Events

Not applicable.

Mortality

There were 13 deaths (13.4% of patients) reported after the tisagenlecleucel infusion in the ELARA study; 7 patients died from the study indication (i.e., progression of the underlying disease) and 6 from other causes (AEs for 5 patients and euthanasia for 1 patient).

Notable Harms

In terms of AEs of special interest, more than 40% of patients experienced any-grade CRS (49.5%), hematological disorders that included cytopenias (78.4%), and infections (55.7%). Most patients experienced CRS (n = 47) within 8 weeks of the tisagenlecleucel infusion. All CRS events except 1 (fatal, grade 5) were of low-grade severity (grade 1 or 2).

Across hematological disorders that included cytopenias, at least 25% of patients reported neutropenia (43.3%) or anemia (25.8%). Overall, the majority of patients (74.2%) experienced hematological events of grade 3 or higher.

Infections occurring any time after the infusion were reported in 54 patients (55.7%), 16 of whom (16.5%) had infections suspected to be related to tisagenlecleucel. Most of the patients had either grade 1 or 2 infections, although infections of grade 3 or higher were reported in 21.6% of patients (9.0% of whom had AEs suspected to be related to tisagenlecleucel). Death due to infection (pneumonia) was reported in 1 patient.

Any-grade serious neurologic adverse reactions were reported in 12 patients (12.4%), 10 of whom experienced these events within 8 weeks of tisagenlecleucel infusion. Grade 3 or 4 AEs considered serious were reported in 3 patients (3.1%), 1 of whom recovered.

Table 20: Summary of Harms in the ELARA Study (Safety Analysis Set)

AE = adverse event; CRS = cytokine release syndrome; NA = not applicable; SAE = serious adverse event.

Source: ELARA Clinical Study Report.²⁹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Critical Appraisal

Internal Validity

The single-arm, noncomparative study design of the ELARA study is 1 of its key limitations. Interpreting the results of studies with this design is difficult because it may not be apparent whether the results are from the effect of the intervention, a placebo effect, or the effect of natural history. It is acknowledged that this study design has so far predominated in the evaluation of CAR T-cell therapies for advanced cancers across a variety of tumour types. It is also acknowledged that there may be practical limitations to conducting an RCT in patients with r/r FL (beyond first-line treatment), such as decreasing population size with subsequent lines of therapy and lack of a gold-standard treatment in these later lines of treatment. However, there is no clear rationale that makes an RCT infeasible, and a search of clinicaltrials.gov showed that patients are being recruited into phase III RCTs comparing a CAR T-cell product with SOC therapy in patients with r/r FL (after first-line therapy).⁸² Some of the limitations of the single-arm design in determining the benefits of tisagenlecleucel are mitigated by the following. First, the hypothesis testing for the primary outcome of CRR was established against the response rate derived from studies of available PI3K inhibitors for the treatment of r/r FL. The clinical experts consulted by CADTH confirmed that a CRR of 15% was a clinically relevant threshold, and noted that any treatment that achieves a 30% CRR in the third or later line of treatment is considered a good therapy. Second, the primary assessment of CRR was done by IRC and confirmed by investigator assessment, providing internal validation of the tumour response results. And third, given that patients enrolled in the study primarily had failed 3 or more lines of prior therapies for FL, the achieved clinically important CRR observed in the ELARA study is likely valid. However, the lack of a comparator makes it difficult to determine whether the magnitude of the treatment effect would be replicated in a larger comparative trial or in the real-world.

Another limitation of the ELARA study is the relatively small sample size and the selective study population. For example, patients were excluded if they had an ECOG PS of 2 or had received prior anti-CD19 therapy or prior allo-SCT. Ninety-eight patients were enrolled, and 94 received treatment with tisagenlecleucel and were included in the primary analyses. The small sample size could restrict data interpretation, particularly for subgroup analyses, although the method of sample size and power calculation was justified. In addition, the statistical analyses for efficacy were based on the EAS population, which excluded patients who were enrolled and underwent apheresis but were subsequently unable to receive tisagenlecleucel. This approach is not aligned with the intention-to-treat principle, nor does it reflect the expected clinical practice population. This would potentially overestimate the benefit of tisagenlecleucel. However, because 94 of 98 enrolled patients were included in the EAS population, the impacts on the validity of the results from this analysis approach are minimal.

Follow-up time was likely sufficient for assessing tumour response and safety outcomes associated with tisagenlecleucel in general. However, median OS and median PFS were not reached at the 24-month follow-up, nor was the corresponding upper limit of the 95% CI, suggesting that the follow-up duration was not long enough to fully capture the effects on OS and PFS, so these results are considered immature. In addition to the duration of the study and the noncomparative design, subsequent treatments make it difficult to interpret the OS and PFS results. After the infusion of tisagenlecleucel, 23% of the patients received at least 1 subsequent antineoplastic medication and 4% received an allo-SCT. The survival results (OS, PFS) should be considered in the context of subsequent treatments because it may be difficult to tell which treatment had more of an impact on a patient’s survival, especially when there was a lack of comparative data in the ELARA study.

The ELARA trial was open-label, which can result in a bias in the measurement of subjectively assessed outcomes such as response, PFS, HRQoL, and AEs. As noted previously, response rates were assessed by IRC in addition to the investigators, which likely reduced the risk of bias in these outcome measures. As well, the tumour response results were consistent across the various assessment methods. PROs and HRQoL results are at risk of performance and detection bias related to the lack of patient and investigator blinding, although the extent and direction of the bias are uncertain.

The study presented PRO and HRQoL data up to 24 months; however, there is a risk of attrition bias and drawing conclusions on a select population because the analyses at 24 months were based on half of the study population from baseline and the results could be biased in favour of tisagenlecleucel.

External Validity

According to the clinical experts consulted by CADTH, the population of the ELARA study generally represents patients in Canada with r/r FL who would be receiving tisagenlecleucel. However, the clinical experts noted that patients seen in clinical practice would include those with poorer performance status (the ELARA study only included patients with an ECOG PS 0 or 1), those who received prior CD19-targeted therapy, and those who have more comorbidities. After screening, the procedures and cointerventions (including manufacturing process, depleting chemotherapy, bridging therapy, and posttisagenlecleucel interventions) were consistent with those adopted in the Canadian setting, although some discrepancies exist, per the clinical experts. The results of the ELARA study can be generalized to the patient population in Canada.

According to the clinical experts consulted by CADTH, the efficacy outcomes used in this study are clinically relevant and important for clinical trials of patients with r/r FL. Because ELARA was an open-label trial, all patients knew about the treatment that they received. This would have some impact on PROs such as HRQoL, but would be less likely to affect objective outcomes such as survival and response rates.

In addition, lack of long-term data on patient survival and response rates is another limitation, given that FL is an indolent and slowly progressive disease. Clinical benefits of the treatment, especially survival benefits and improvement in HRQoL, need to be evaluated in the longer term.

Long-Term Extension Studies

There are no long-term extension studies submitted by the sponsor for this review.

Other Relevant Evidence

This section includes additional relevant studies that were included in the sponsor’s submission to CADTH and were considered to address important gaps in the evidence included in the Systematic Review.

Results of 2 non-RCTs that may have addressed evidence gaps are presented in the Studies Addressing Gaps in the Pivotal and RCT Evidence section.

Indirect Evidence

The contents of this section have been informed by materials submitted by the sponsor. The following information has been summarized and validated by the CADTH review team.

Objectives for the Summary of Indirect Evidence

The objective of this section is to summarize and critically appraise available indirect evidence comparing tisagenlecleucel to other relevant treatments for r/r FL after 2 or more lines of therapy relevant to the Canadian clinical practice setting.

Description of Indirect Comparison

The efficacy and safety of tisagenlecleucel has been previously assessed in the ELARA study,²⁹ whereas its efficacy against standard chemotherapy was compared in the ReCORD-FL study.³² However, no head-to-head comparison of tisagenlecleucel against other advanced treatments for r/r FL after 2 or more lines of therapy was available for this review. Due to this gap in evidence, 1 sponsor-conducted ITC in the form of a MAIC of tisagenlecleucel versus axicabtagene ciloleucel was reviewed.³¹

Study Selection Method

Study selection criteria and key aspects of the methods used for the systematic review are summarized in Table 21.

Based on the prespecified eligibility criteria detailed in Table 21, the sponsor conducted a systematic literature search of clinical trials and observational studies to identify relevant comparators approved for the treatment of adults with r/r FL after 2 or more lines of therapy. The literature search included all studies published between 1998 and June 2022, based on the approval of rituximab in November 1997 (US) and June 1998 (EU) for FL. However, only studies published from 2014 onward, coinciding with the introduction of the new Lugano response criteria, were considered for inclusion.^59,83 Systematic literature searches were conducted of multiple electronic databases, including MEDLINE, Embase, the Cochrane Central Register of Controlled Trials, and the Cochrane Database of Systematic Reviews. In addition, abstracts from 5 relevant conference proceedings from 2019 to 2022 were hand-searched, and the reference lists of previously published systematic reviews and meta-analyses were screened.

The study screening, selection, and data extraction were conducted by 2 independent reviewers. Studies identified from the systematic literature search were first screened based on the title and abstract. A full-text screening was then carried out for studies selected at the title and abstract screening stage. A third reviewer was involved to resolve any discrepancies during the title, abstract, and full-text review. A list of excluded studies was reported, with reasons for exclusion. Two independent reviewers extracted data and assessed the risk of bias of the included studies using the National Institute for Health and Care Excellence and the Downs and Back checklist to determine the quality of reporting of RCTs and nonrandomized trials, respectively.^84,85

Table 21: Study Selection Criteria and Methods for ITCs Submitted by the Sponsor

Auto-SCT = autologous stem cell transplant; CDSR = Cochrane Database of Systematic Reviews; CENTRAL = Cochrane Central Register of Controlled Trials; CR = complete response; CRR = complete response rate; EU = European Union; FL = follicular lymphoma; FLIPI = Follicular Lymphoma International Prognostic Index; ITC = indirect treatment comparison; NHL = non-Hodgkin lymphoma; NICE = National Institute for Health and Care Excellence; ORR = overall response rate; OS = overall survival; PFS = progression-free survival; PR = partial response; RCT = randomized controlled trial; r/r = relapsed/refractory; SCT = stem cell transplant; SLR = systematic literature review.

Source: Clinical Study Report for sponsor-submitted MAIC.³¹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Objectives

The objective of the sponsor-submitted MAIC was to compare the efficacy and safety of tisagenlecleucel to axicabtagene ciloleucel in patients with r/r FL after 2 or more lines of therapy. Of note, axicabtagene ciloleucel received a NOC/c from Health Canada for the treatment of adults with r/r grade 1, 2, or 3a FL after 2 or more lines of systemic therapy on September 29, 2022,⁸⁶ and to date has not been reviewed by the CADTH pan-Canadian Oncology Drug Review Expert Review Committee for this patient population. The MAIC was reviewed by a CADTH clinical team because it was used to inform the sponsor's pharmacoeconomic model.

Study Selection Methods

A total of 4,865 records were screened by abstract, followed by a full-text screening of 1,133 articles. Of these, 132 unique studies published between January 1998 and June 2022 were extracted for clinical review. To be eligible for inclusion in the ITC, the population, outcomes, and the study design of studies identified from the systematic literature review were required to be aligned with those of the ELARA trial conducted by the sponsor. Moreover, comparators were only considered relevant if they had been reviewed to be approved or recommended interventions and met the following parameters:

• comparator treatments were regulatory approved regimens or recommended as treatments by evidence-based clinical guidelines

• studies included a sample size of more than 30 patients

• studies reported data for PFS (e.g., median or rate).

Using those parameters, 42 of the 132 studies were considered for inclusion in the ITC. Of these, 14 studies were conducted as clinical trials (9 single-arm trials, 3 RCTs, and 2 non-RCTs) and 28 were observational studies. The interventions assessed across these studies included Pi3K inhibitors (idelalisib, copanlisib, duvelisib), CAR T-cell therapies (axicabtagene ciloleucel; tisagenlecleucel), R2, mosunetuzumab, tazemetostat, rituximab and/or chemotherapy, SOC in general, and stem cells transplants (auto-SCT or allo-SCT).

These 42 studies were then assessed to determine if diagnosis, clinical management, and response criteria assessment were similar to those applied in the ELARA trial. As a result, 12 studies were excluded due to a publication date before the introduction of the new 2014 Lugano response criteria. After applying the study selection criteria and key aspects of the methods for the systematic review summarized in Table 21, 8 studies were considered to be potential comparator studies: 4 that assessed PI3K inhibitors (2 assessed idelalisib, 1 assessed copanlisib, and 1 assessed duvelisib),^80,87-89 and 1 each assessed CAR T-cell therapy (axicabtagene ciloleucel),⁹⁰ mosunetuzumab,⁹¹ salvage therapy after auto-SCT relapse,⁹² and SOC.⁹³ Interventions that were considered not to be relevant to the Canadian clinical setting were excluded. For example, idelalisib, although indicated for the treatment of patients with r/r FL in Canada, is not publicly reimbursed across Canada and is inaccessible at some sites, according to input the sponsor received from several clinicians.⁵⁷ Furthermore, copanlisib, duvelisib, and mosunetuzumab are not currently commercialized for treatment of FL in Canada.

The sponsor determined that axicabtagene ciloleucel was the only advanced therapy relevant to the Canadian clinical context. Accordingly, 1 study met the criteria as a comparator treatment to tisagenlecleucel for inclusion in the ITC: ZUMA-5.⁹⁴

MAIC Analysis Methods

A summary of the methods of analysis for the MAIC is presented in Table 23. An unanchored MAIC approach was selected for the indirect comparison between tisagenlecleucel and axicabtagene ciloleucel due to a lack of common comparator.³¹ Individual patient-level data from the ELARA study (data cut-off date of March 29, 2022) and aggregated data from the ZUMA-5 study (axicabtagene ciloleucel; data cut-off date of March 31, 2021) were used.^29,61,94

The covariates used for match-adjustment were informed by clinical experts consulted by the sponsor and against European Medicines Agency (EMA) recommendation based on importance. Relevant baseline prognostic factors were included in the MAIC if they were commonly and consistently reported in both the ELARA and ZUMA-5 studies. The prognostic factors considered and included in the propensity score model are detailed in Table 22. The covariates included in the MAIC were age (≥ 65 years versus < 65 years); sex (male versus female); race (white versus nonwhite [from original source]); ECOG PS (> 1 versus 1); Ann Arbor stage (IV versus III versus I to II); FLIPI score (≥ 3 versus 2 versus 0 to 1); high tumour bulk per Groupe d'Etude des Lymphomes Folliculaires (GELF) criteria (yes versus no); prior line of therapy (≥ 3 versus < 3); refractory to most recent regimen (yes versus no); POD24 from first anti-CD20 mAb-containing therapy; and prior auto-SCT.

Table 22: Prognostic Factors of Efficacy Outcomes in Patients With r/r FL Considered and Included in the MAIC of Tisagenlecleucel vs. Axicabtagene Ciloleucel

Auto-SCT = autologous stem cell transplant; CAR = chimeric antigen receptor; EMA = European Medicines Agency; FL = follicular lymphoma; FLIPI = follicular lymphoma international prognostic index; ITC = indirect treatment comparison; LDH = lactate dehydrogenase; MAIC = matching-adjusted indirect treatment comparison; POD24 = progression of disease within 24 months of first line of treatment; r/r = relapsed or refractory; vs. = versus.

^aCovariates included in the MAIC were age (≥ 65 years vs. < 65 years); sex (male vs female); race (white vs. nonwhite [from original source]); ECOG PS (> 1 vs. 1); Ann Arbor stage (IV vs. III vs. I to II); FLIPI score (≥ 3 vs. 2 vs. 0 to 1); high tumour bulk per GELF criteria (yes vs. no); prior lines of therapy (≥ 3 vs < 3); refractory to most recent regimen (yes vs. no); POD24 from first anti-CD20 mAb-containing therapy; and prior auto-SCT.

^bRegion was excluded because race was included in the MAIC.

^cInclusion criteria for both the ELARA and ZUMA-5 studies.

^dNo data available in the ZUMA-5 publication.

^eDefinition of double refractory unavailable in the ZUMA-5 publication.

Source: Clinical Study Report for sponsor-submitted MAIC.³¹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Baseline characteristics in the ELARA and ZUMA-5 studies were balanced by assigning weights to selected ELARA patients, such that:

• weighted mean baseline characteristics in the selected ELARA patients exactly matched those reported for patients in the ZUMA-5 study

• each individual patient weight was equal to the patient’s estimated odds (relative propensity) of being in the ZUMA-5 study versus the ELARA study, given the patient’s baseline characteristics.

Weights were determined from a logistic regression model for the propensity of inclusion in the ZUMA-5 versus ELARA study, with all matched-on-baseline characteristics included as independent variables in the model. Because only summary statistics for baseline characteristics were available for the ZUMA-5 study, the logistic model was estimated using the methods of moments approach as opposed to the maximum likelihood approach.^95-97 Baseline characteristics between the ELARA infused set nonbridging chemotherapy subgroup and the ZUMA-5 infused set after matching were compared using the Wald test.

Weighted ORR and CRR were calculated for the ELARA efficacy-evaluable set nonbridging chemotherapy subgroup using the same weights that were applied to match average baseline characteristics. Wald tests were used to compared ORR and CRR efficacy results between the ELARA and ZUMA-5 studies after matching.

Weighted median PFS, weighted median OS, and corresponding 95% Cis were calculated using the Nelson-Aalen estimator for the ELARA efficacy-evaluable set nonbridging subgroup, incorporating the same weights used to match baseline characteristics. The PFS and OS curves for the ZUMA-5 study were reconstructed using the Guyot method. A weighted long-rank test was performed for the comparison of the survival curves; the curve was unweighted for the ZUMA-5 study and weighted for the ELARA study. The HR and corresponding 95% CI were estimated using a weighted Cox proportional hazard model after matching.

The following safety outcomes were assessed in the MAIC: CRS, tocilizumab and corticosteroid use for CRS management, and neurologic events. Weighted CRS, tocilizumab and corticosteroid use for CRS management, and neurologic events were calculated for the ELARA infused set nonbridging chemotherapy subgroup, incorporating the same weights used to match baseline characteristics. Wald tests were used to compare safety outcomes in the ELARA and ZUMA-5 studies after matching.

Weights were only applied to ELARA patients for whom individual patient data were available; weighting did not impact the efficacy or safety outcomes reported for the ZUMA-5 efficacy-evaluable and ZUMA-5 infused sets, respectively, which were based on published aggregate data.

Table 23: MAIC Analysis Methods for Tisagenlecleucel vs. Axicabtagene Ciloleucel

Auto-SCT = autologous stem cell transplant; CRR = complete response rate; CRS = cytokine release syndrome; FL = follicular lymphoma; FLIPI = Follicular Lymphoma International Prognostic Index; LDH = lactate dehydrogenase; MAIC = matching-adjusted indirect treatment comparison; ORR = overall response rate; OS = overall survival; POD24 = progression of disease within 24 months of first-line of treatment; vs. = versus.

^aCovariates included in MAIC.

Source: Clinical Study Report for sponsor-submitted MAIC.³¹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Summary of Included Studies

The sponsor-submitted MAIC included 2 single-arm, open-label, multicentre phase II trials: ELARA and ZUMA-5. The ITC included individual patient-level data for nonbridging chemotherapy patients who received tisagenlecleucel in the ELARA study (n = 53) and aggregated data with patients who received axicabtagene ciloleucel in the ZUMA-5 study (n = 124).

A detailed description of the ELARA study is provided in the Pivotal Studies and RCT Evidence section of this clinical report. Briefly, patients were enrolled if they had r/r disease after at least 2 lines of systemic therapy or during or within 6 months of anti-CD20 mAb maintenance, or if they relapsed after auto-SCT. Optional bridging chemotherapy before tisagenlecleucel infusion was allowed. Patients received fludarabine-based or bendamustine-based LD chemotherapy before tisagenlecleucel infusion. The primary efficacy end point in the ELARA study was CRR assessed by a central IRC. ORR, PFS, OS, and safety outcomes were evaluated in the ELARA study as secondary end points. As of the March 29, 2022, data cut-off date, a total of 98 patients were enrolled and, among them, 97 patients were infused with tisagenlecleucel. The following patient sets from the ELARA study were used in the comparison of tisagenlecleucel with axicabtagene ciloleucel for the primary analysis:

• the ELARA infused set nonbridging chemotherapy subgroup (n = 53), which included patients who had received tisagenlecleucel infusion in the ELARA trial who did not receive bridging chemotherapy. This cohort was used for matching with axicabtagene ciloleucel on baseline characteristics and comparing safety outcomes.

• the ELARA efficacy-evaluable set nonbridging chemotherapy subgroup (n = 52), which included patients who had received tisagenlecleucel infusion in ELARA trial who did not receive bridging chemotherapy and who had measurable disease at baseline per IRC. This cohort was used for comparing efficacy outcomes.

ZUMA-5 is an ongoing pivotal single-arm, open-label, multicentre, phase II study aimed to assess the efficacy and safety of axicabtagene ciloleucel in adults with r/r indolent NHL, including FL and marginal zone lymphoma. Patients were enrolled if they had r/r disease after at least 2 lines of treatment with combination chemoimmunotherapy, which must have included an anti-CD20 mAb-containing therapy combined with an alkylating drug. Patients underwent leukapheresis and received conditioning chemotherapy (cyclophosphamide at 500 mg/m² per day and fludarabine at 30 mg/m² per day on days −5, −4, and −3) followed by a single infusion of axicabtagene ciloleucel (2 × 10⁶ CAR T cells per kg) on day 0. Bridging chemotherapy before axicabtagene ciloleucel infusion was used in 2 (1.6%) of 124 infused patients as of the March 12, 2020, data cut-off date. Patients received LD chemotherapy with fludarabine and cyclophosphamide before axicabtagene ciloleucel infusion. The primary efficacy end point in the ZUMA-5 study was ORR assessed by a central IRC. CRR, PFS, OS, and safety outcomes were evaluated in the ZUMA-5 study as secondary end points. Aggregate data for baseline characteristics were obtained directly from the ZUMA-5 publication,⁹⁴ and aggregate data for efficacy and safety outcomes were obtained directly from the American Society of Hematology 2021 ZUMA-5 publication.⁹⁸ As of the March 31, 2021, data cut-off, a total of 157 patients with r/r indolent NHL were enrolled. Among the 149 infused patients, 124 (83.2%) patients had r/r FL. The following patient sets from the ZUMA-5 study were used in the primary comparison with ELARA patients:

• the ZUMA-5 infused set (n = 124), which included axicabtagene ciloleucel–infused patients with r/r FL. This cohort was used as the reference for ELARA patients to match to, and for comparing safety outcomes.

• The ZUMA-5 efficacy-evaluable set (n = 86), which included axicabtagene ciloleucel–infused patients with r/r FL who had at least 24 months of follow-up. This cohort was used for comparing efficacy outcomes.

Fifty-three infused, nonchemotherapy bridging patients from the ELARA study were matched with 124 infused patients in the ZUMA study. Baseline characteristics of patients in the ELARA and ZUMA-5 studies before and after matching are summarized in Table 24. After matching, all matched-on-baseline characteristics were balanced in the ELARA and ZUMA- 5 studies. Patients included in the MAIC were mainly younger than 65 years (69.4%), male (59.9%), and white (92.7%). In regard to disease severity at baseline, 37.1% of patients had an ECOG PS of 1; 36.3% and 49.2% were assessed as having stage III and stage IV disease, respectively, on the Ann Arbor staging system; 38.7% and 43.6% had a FLIPI score of 2 and 3 or greater, respectively; and 51.6% were classified as having high tumour bulk based on the GELF criteria. Regarding prior therapies, 62.9% of a patients had 3 or more lines of prior therapy and 24.2% had prior auto-SCT. Overall, 67.7% of patients were refractory to their most recent regimen and 54.8% were POD24 from first anti-CD20 mAb-containing therapy.

Table 24: Patient Characteristics of the ELARA Nonbridging Chemotherapy Subgroup and the Zuma-5 Infused Set

Auto-SCT = autologous stem cell transplant; ECOG PS = Eastern Cooperative Oncology Group Performance Status; FLIPI = Follicular Lymphoma International Prognostic Index; GELF = Groupe d’Etude des Lymphomes Folliculaires; mAb = monoclonal antibody; NE = not estimable; POD24 = progression of disease within 24 months of first-line of treatment.

^aDenotes P value < 0.05. P values were calculated using the Wald test.

Source: Clinical Study Report for sponsor-submitted MAIC.³¹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Results

The MAIC analysis compared 52 ELARA efficacy-evaluable nonbridging chemotherapy patients to 86 patients in the ZUMA-5 efficacy-evaluable set. After-matching efficacy results are presented.

Treatment Response

Comparisons of ORR and CRR between the ELARA nonbridging chemotherapy subgroup and the ZUMA-5 study are summarized in Table 25.

The ORR was 91.2% (95% CI, 76.2% to 97.1%) and 94.2% (95% CI, 87.0% to 98.1%) for patients in the ELARA and ZUMA-5 studies, respectively. The ORR difference between the ELARA and ZUMA-5 studies was –3.03 (95% CI, –13.7% to 7.6%). The CRR was 74.0% (95% CI, 54.4% to 8.2%) and 79.1% (95% CI, 69.0% to 8.1%) for patients in the ELARA and ZUMA-5 studies, respectively. The CRR difference between the ELARA and ZUMA-5 studies was –5.0% (95% CI, –23.9% to 13.8%).

Sensitivity analyses comparing ORR and CRR between the ELARA infused set that included those with and without nonbridging chemotherapy and the ZUMA-5 study were consistent to the primary MAIC analysis.

Table 25: Comparisons of ORR and CRR Between the ELARA Efficacy-Evaluable Nonbridging Chemotherapy Subgroup and the ZUMA-5 Efficacy-Evaluable Set After Matching

CI = confidence interval; CRR = complete response rate; ORR = overall response rate.

^aP values were calculated using the Wald test.

Source: Clinical Study Report for sponsor-submitted MAIC.³¹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Overall Survival

KM curves comparing PFS in the ELARA to ZUMA-5 studies are presented in Figure 3.

Median OS was not estimated for either the ELARA study or the ZUMA-5 study. The hazard of death for tisagenlecleucel compared to axicabtagene ciloleucel was 0.49 (95% CI, 0.16 to 1.49; P = 0.21).

Sensitivity analysis of OS between the ELARA infused set that included those with and without nonbridging chemotherapy and the ZUMA-5 study documented an HR of 0.34 (95% CI, 0.14 to 0.84).

Progression-Free Survival

KM curves comparing PFS in the ELARA to ZUMA-5 studies are presented in Figure 4.

Median time of PFS was 36.6 months (95% CI, 25.7 to 29.6 months) for in the ZUMA-5 study. Median PFS was not estimated for tisagenlecleucel. The hazard of disease progression or death for tisagenlecleucel compared to axicabtagene ciloleucel was 0.84 (95% CI, 0.37 to 1.90).

Sensitivity analysis of PFS between the ELARA infused set that included patients with and without nonbridging and the ZUMA-5 study noted an HR of 1.17 (95% CI, 0.63 to 2.18).

Figure 3: KM Curves and HRs for OS Comparing the ELARA Efficacy-Evaluable Set Nonbridging Chemotherapy Subgroup and the ZUMA-5 Efficacy-Evaluable Set After Matching

Axi-cel = axicabtagene ciloleucel; CI = confidence interval; HR = hazard ratio; KM = Kaplan-Meier; NR = not reported; OS = overall survival; Tisa-cel = tisagenlecleucel.

Note: The OS curve for the ZUMA-5 trial was reconstructed using the Guyot method. The median OS and 95% CI were estimated using the Nelson-Aalen estimator and log transformation.

Source: Clinical Study Report for sponsor-submitted MAIC.³¹

Figure 4: KM Curves and HRs of PFS Comparing the ELARA Efficacy-Evaluable Set Nonbridging Chemotherapy Subgroup and the ZUMA-5 Efficacy-Evaluable Set After Matching

Axi-cel = axicabtagene ciloleucel; CI = confidence interval; HR = hazard ratio; KM = Kaplan-Meier; NR = not reported; PFS = progression-free survival; Tisa-cel = tisagenlecleucel.

Note: The PFS curve for the ZUMA-5 trial was reconstructed using the Guyot method. The median PFS and 95% CI were estimated using the Nelson-Aalen estimator and log transformation.

Source: Clinical Study Report for sponsor-submitted MAIC.³¹

Harms

Fifty-three ELARA infused nonbridging chemotherapy patients and 124 patients in the ZUMA-5 infused set were included in the MAIC of safety outcome. Harms outcomes in the ELARA infused set nonbridging chemotherapy subgroup and the ZUMA-5 study after matching are summarized in Table 26.

At least 1 AE of any grade was reported in 44.6% of patients in the ELARA study and 78.2% of patients in the ZUMA-5 study, for a response difference of –33.7% (95% CI, –54.0% to –13.3%). AEs of grade 3 or above were reported in no patients in the ELARA study and in 6.5% of patients in the ZUMA-5 study, for a response difference of –6.4% (95% CI, –10.8% to –2.1%).

The management of CRS with corticosteroids was documented in 3.0% of patients in the ELARA study and 15.3% in the ZUMA-5 study, for a response difference of –12.3% (95% CI, –21.0% to –3.6%). CRS management with tocilizumab was documented in 9.9% of patients in the ELARA study and 45.2% of patients in the ZUMA-5 study, for a response difference of –35.3% (95% CI, –47.8% to –22.7%).

Neurologic events of any grade were documented in 9.5% of patients in the ELARA study and 56.5% of patients in the ZUMA-5 study, for a response difference of –47.0% (95% CI, –61.9 to –32.1). Neurologic events of grade 3 and above were reported among 0.19% of patients in the ELARA study and 15.3% of patients in the ZUMA-5 study, for a response difference of –15.1% (95% CI, –21.5% to –8.8%).

Sensitivity analysis comparing safety outcomes between the ELARA infused set that included patients with and without nonbridging chemotherapy and the ZUMA-5 study were consistent with the primary MAIC analysis.

Table 26: Comparison of Safety Outcomes Between the ELARA Infused Set Nonbridging Chemotherapy Subgroup and the ZUMA-5 Infused Set After Matching

AE = adverse event; CI = confidence interval; CRS = cytokine release syndrome.

Source: Clinical Study Report for sponsor-submitted MAIC.³¹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Critical Appraisal

The MAIC was based on studies identified from a sponsor-conducted systematic review of treatments for FL in adults.⁹⁹ The systematic review was comprehensive, involving multiple electronic databases, clinical registries, and supplementary manual searches. By following the PRISMA guidelines, the systematic review minimized error and bias in the study selection process. The reasons for study exclusion were reported, and selection and data extraction process were well defined. Although the risk of bias of the comparator trial was assessed, this rating was not detailed in the systematic review.

The MAIC was limited by important methodological differences between the ELARA and ZUMA-5 studies that could not be matched in the analyses. These differences are summarized in Table 27. As previously described, eligibility criteria in the ELARA study required patients to be r/r FL after 2 or more lines of systemic therapy, during or within 6 months of anti-CD20 mAb maintenance, or relapsed after auto-SCT. However, the eligibility criteria for the ZUMA-5 study did not include the 6-month caveat. Moreover, the definition of refractory differed; in the ELARA study, refractory was defined as failure to respond to previous treatment or progressive disease within 6 months of prior therapy completion, and in the ZUMA-5 study, patients were considered refractory on progression within 6 months of completion of the most recent prior treatment. Of note, although re-treatment was not permitted in the ELARA study, it was allowed in the ZUMA-5 study. Overall, 11 (8.9%) patients in the ZUMA-5 infused set received re-treatment. Due to a lack of access to ZUMA-5 individual patient data, the MAIC was unable to account for these factors. Finally, the definition of neurologic event in the ELARA study was more detailed than that used in the ZUMA-5 study. According to the clinical experts consulted by CADTH for this review, safety outcomes would bias neurologic event results toward the ZUMA-5 study due to difference in definitions.

The choice to conduct an unanchored MAIC was appropriately justified by the lack of a common comparator between the ELARA and ZUMA-5 studies. However, for an unanchored MAIC to produce unbiased treatment-effect estimates, all effect modifiers and prognostic variables need to be adjusted in the analysis. The covariates used for match-adjustment in the sponsor-submitted MAIC were informed by consultation with clinical experts experienced in the treatment of patients with r/r FL and EMA recommendations. A total of 31 covariates were identified as having some importance by the clinical experts or were recommended by the EMA. However, the MAIC was unable to adjust for all baseline characteristics of interest; only observed factors commonly and consistently reported in both studies were included for adjustment. In the end, the final MAIC matched on 11 variables (age, sex, race, ECOG PS, Ann Arbor stage, FLIPI score, high tumour bulk, prior lines of therapy, refractoriness to the most recent regimen, progression of disease within 12 months of first-line treatment from first anti-CD20 mAb-containing therapy, and prior auto-SCT). Of concern was the absence of matching on a number of prognostic factors considered important by the clinical experts and EMA, including nodal site involvement, LDH, lymphoma symptoms, chemoresistance, and time since diagnosis, because no data were available in the ZUMA-5 publication. Further compounding the issue was the differences in samples used for adjustment among the infused set and those used for efficacy comparison in the efficacy-evaluable set. The limitation would likely be more apparent in the ZUMA-5 study, given the larger sample size difference between the infused set (n = 124) and the efficacy-evaluable set (n = 86). Although the analysis assumed that patient characteristics in the ELARA infused set were comparable to those in the efficacy-evaluable set, the assumption was not validated. This difference has the likely potential to bias the results in favour of tisagenlecleucel. As the MAIC failed to meet the core assumption that all effect-modifier and prognostic factors were accounted for, conclusions about the effect of tisagenlecleucel relative to axicabtagene ciloleucel could not be drawn, owing to unknown amounts of residual bias.^100,101 As noted by the National Institute for Health and Care Excellence (NICE) Decision Support Unit technical guidance, the failure of this assumption leads to an unknown amount of bias in the unanchored estimate.⁹⁵

A sensitivity analysis was conducted to compare patients with and without bridging chemotherapy from the ELARA with ZUMA-5 studies. However, due to the lack of details about how the sensitivity analysis was conducted and how the 2 populations compared, the CADTH review team was unable to comment on the results with any certainty.

Unadjusted and matched-adjusted baseline covariates were reported. Baseline characteristics after matching were well balanced, with almost perfect matching for the covariates included in the MAIC. However, the complete baseline demographic and disease characteristics for patients in both trials were not reported after matching; only the balance of patient characteristics relevant to the covariates used in the matching were reported. Therefore, it is unclear what effect the matching had on the balance of other relevant patient characteristics.

The application of weight resulted in a reduced ESS of 26 patients, in which 51.9% of enrolled patients in the ELARA study were lost. The reduction in ESS suggests poor population overlap, with the results being heavily influenced by a subset of patients from the ELARA study, which may not be representative of the entire sample. The reduction in sample size may contribute to imprecision, increasing the uncertainty of the results.

Baseline characteristics and clinical results from the comparator study were derived from aggregate patient data, further limiting the interpretability of the MAIC results. Published KM curves pooled data over different covariates that may affect survival. A major limitation of the KM extraction method used in MAICs is the inability to derive separate KM curves for different subgroups or to model the joint effects of covariates and treatment.¹⁰² As a result, the treatment effects estimated may be impacted by aggregation bias.¹⁰² Furthermore, the precision of data extraction from KM curves is often dependent on the quality of the initial input (e.g., clear versus blurry figures) and depth of information provided in the original publications.

Of note, the MAIC assessed the end points of CRR, ORR, OS and PFS, as well as harms. Other efficacy end points of interest to patients and clinical experts, such as HRQoL, were not investigated.

Table 27: Assessment of Homogeneity for MAIC

Auto-SCT = autologous stem cell transplant; CRS = cytokine release syndrome; FL = follicular lymphoma; FLIPI = Follicular Lymphoma International Prognostic Index; IRC = independent review committee; LD = lymphodepleting; mAb = monoclonal antibody; MAIC = matching-adjusted indirect treatment comparison; ORR = overall response rate; OS = overall survival; POD24 = progression of disease within 24 months of first-line treatment; r/r = relapsed or refractory.

Source: Sponsor-submitted MAIC of ELARA vs. ZUMA-5.³¹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Studies Addressing Gaps in the Pivotal and RCT Evidence

The contents of this section have been informed by materials submitted by the sponsor. The information has been summarized and validated by the CADTH review team.

The sponsor indicated that for ethical and practical reasons, novel therapies such as CAR T-cell therapy for rare diseases have typically been evaluated in single-arm trials, which results in a lack of direct comparative evidence with a control group (e.g., SOC or usual care).^{29,33,61,80,94,105} Further, a single-arm design is supported by the rarity of patients with r/r FL, the lack of an established SOC for r/r FL, and low response rates for currently available therapies in this high-risk population. These complicating factors make the identification of a relevant comparator difficult. Consultations with both the FDA and the EMA on the clinical development program and registration strategy for tisagenlecleucel in r/r FL resulted in agreement on the study design, primary and secondary end points, and on the statistical analysis methods undertaken in the ELARA study. For the tisagenlecleucel submission, the pivotal study, ELARA, was a single-arm trial; therefore, the lack of a control group constitutes an important evidence gap. To address this evidence gap, a retrospective cohort study, ReCORD-FL, was undertaken in adults with grade 1, 2, or 3a r/r FL after 2 or more lines of therapy to construct a historic cohort of control patients to augment current and future single-arm trials in the r/r FL setting.⁹³

In addition, the pivotal ELARA study provided efficacy outcomes with up to 30 months of follow-up. Longer-term efficacy data and safety data are needed to explore the treatment effect of tisagenlecleucel in the study population. Therefore, results of the study be Schuster et al. (2017)³³ are summarized in this review.

Table 28: Summary of Gaps in the Evidence

CRR = complete remission rate; DoR = duration of response; EFS = event-free survival; FL = follicular lymphoma; ORR = overall response rate; OS = overall survival; PFS = progression-free survival; r/r = relapsed or refractory; RCT = randomized controlled trial; TNT-D = time to next treatment or death.

Description of the ELARA and ReCORD-FL Trials

In the absence of direct head-to-head comparisons of tisagenlecleucel to SOC, the sponsor compared the treatment effect of tisagenlecleucel as observed in the ELARA study to SOC, defined as standard chemotherapy, as documented in the retrospective ReCORD-FL cohort study.

The ELARA study (n = 98) is an ongoing, single-arm, open-level, multicentre, phase II study assessing the efficacy and safety of tisagenlecleucel in adults with r/r FL after 2 or more lines of therapy. A detailed description of the ELARA study is provided in the Studies Addressing Gaps in the Pivotal and RCT Evidence section of this clinical report. The ReCORD-FL study (n = 187) is a noninterventional, multicentre, retrospective chart review conducted by the sponsor with the purpose of providing patient-level data for a historical control group to compare with standard chemotherapy data in the ELARA trial. The study scheme for the ReCORD-FL study is illustrated in Figure 5. Patient-level data were collected from patients treated for r/r FL between 1998 and 2020 from 10 sites in Europe and North America, including 1 Canadian site (n = 12). The medical records of patients with r/r FL were abstracted and provided by physicians or their designates who have been or are treating patients with FL. Clinical outcomes were measured from the start of each observed line of therapy until the earliest of death, the last available follow-up, or the data cut-off date. Anonymized patient data were entered into an electronic data collection form using a secure web-based data collection portal. Data were collected or derived at 3 time points:

• from the initial diagnosis of FL (the earliest was in 1998) through the study index date

• at the time of the study index date (between January 2, 2000, and December 31, 2018)

• from the study index date until the earliest of the following: death, last available record and/or follow-up, or December 31, 2020.

Figure 5: The ReCORD-FL Study Scheme

FL = follicular lymphoma.

Source: Clinical Study Report for ELARA versus ReCORD.³²

Populations

Where feasible, the ReCORD-FL study adopted the same inclusion and exclusion criteria as the ELARA study. For inclusion into the ReCORD-FL study, patients 18 years and older were required to meet 1 of the following criteria defining r/r FL:

• refractory to a second line or later line of systemic therapy (including an anti-C20 antibody and an alkylating drug) or relapse within 6 months of completion of a second line or later line of systemic therapy

• relapse during anti-C20 antibody maintenance (after at least 2 lines of therapies) or within 6 months of maintenance completion

• relapse after auto-SCT.

Moreover, patients must have initiated at least 1 additional line of systemic therapy after first meeting the criteria for r/r FL; have at least 3 months of available follow-up data after the index date, unless the patient died in the 3 months after the index date; were confirmed histologically as having grade 1, 2, or 3a disease at the time of the index date; have an index date during the period between January 1, 2000, and December 31, 2018, with an initial diagnosis of FL occurring before the index date but not earlier than January 1, 1998; and have an ECOG PS of 0 or 1 at the time of the index date. Of note, eligible patients may have been living or deceased at the time of data abstraction. Patients were excluded if the following criteria were noted at or before the study index date: evidence of histological transformation; grade 3b FL; prior anti-CD19 therapy; prior gene therapy; prior adoptive T-cell therapy; prior allo-SCT; active CNS involvement by malignancy; active neurologic autoimmune or inflammatory disorders; and any enrolment in a CAR T-cell trial.

Of note, not all inclusion and exclusion criteria in the ELARA study were adopted in the ReCORD-FL study. Notable inclusion and exclusion criteria that were not feasible to apply in the ReCORD-FL study are summarized in Table 29.

Table 29: ELARA Inclusion and Exclusion Criteria Not Feasible to Apply in the ReCORD-FL Study

AV = atrioventricular; ECHO = echocardiogram; FL = follicular lymphoma; LVEF = left ventricular ejection fraction; MI = myocardial infarction; MRA = coronary magnetic resonance angiography; MUGA = multigated acquisition; NYHA = New York Hearts Association.

Source: Clinical Study Report for ELARA vs. ReCORD.³²

Interventions

Tisagenlecleucel

Tisagenlecleucel as administered in the ELARA study is described in detail in the Pivotal Studies and RCT Evidence section.

Standard of Care

SOC included treatment with chemotherapy, anti-CD20 antibody monotherapy, a PI3K inhibitor, high-dose chemotherapy as salvage therapy for auto-SCT or allo-SCT, or radioimmunotherapy. A chemotherapy mix that included auto-SCT at the index date or later was also accepted as SOC. In addition, other cancer-directed systemic therapies, including investigational drugs, except for CAR T-cell therapy, were considered. Watch and wait was not considered a relevant therapy option for the purpose of this study. Previous treatment with other FL-targeting medications, such as PI3K inhibitors, was permitted if patients recovered from all treatment-related AEs.

Outcomes

Overall Survival

OS was defined as time to death from any cause. OS was censored at the last date the patient was known to be alive if no death had been observed at the time of analysis.

Progression-Free Survival

Dates of disease progression were not available for most patients in the ReCORD-FL study. Accordingly, a comparison analysis of PFS was performed by considering death or initiation of new anticancer therapy for patients in both the ReCORD-FL and ELARA studies. Patients were censored at the date of last contact if no progression, death, or new anticancer therapy initiation occurred.

Statistical Analysis

Selection of Eligible Line of Chemotherapy

It was noted that patients in the ReCORD-FL study may meet inclusion and exclusion criteria at the start of more than 1 line of therapy. A propensity score, operationalized as the probably of being enrolled in the ELARA study conditional on observed variables measured at the start of the line of therapy, was estimated for each patient at each eligible line of therapy in the ReCORD-FL study. This approach was used to achieve an approximate balance of the number of prior lines of therapy while also balancing other key baseline prognostic variables between the ELARA and ReCORD-FL studies. Covariates were considered relevant for adjustment if deemed to have some level of importance (less important to very important) by clinical experts consulted by the sponsor or if recommended by the EMA and aligned with health authorities. Relevant baseline prognostic factors were included in the propensity score model if they were commonly and consistently reported in both the ELARA and ReCORD-FL studies. Prognostic factors considered and included in the propensity score model are detailed in Table 30. The covariates included in the final propensity model were age at treatment initiation (continuous variable); region (Europe or North America); sex (female or male); history of auto-SCT (yes or no); number of previous lines of systematic treatment (> 4 or 2 to 4); disease stage at FL diagnosis and initiation of treatment (continuous variable); sites of nodal involvement at treatment initiation (> 4 or ≤ 4); double refractory (yes or no); and progression of disease within 24 months of first-line treatment (yes or no).

Table 30: Prognostic Factors of Efficacy Outcomes in Patients With r/r FL Considered and Included in the Propensity Score Model for the ITC of Tisagenlecleucel vs. SOC

Auto-SCT = autologous stem cell transplant; EMA = European Medicines Agency; FL = follicular lymphoma; FLIPI = Follicular Lymphoma International Prognostic Index; ITC = indirect treatment comparison; LDH = lactate dehydrogenase; POD24 = progression of disease within 24 months of first-line of treatment; r/r/ = relapsed or refractory; SOC = standard of care.

^aCovariates included in the propensity score model were age at treatment initiation (continuous variable); region (Europe or North America); sex (female or male); history of auto-SCT (yes or no); number of previous lines of systematic treatment (> 4 or 2 to 4); disease stage at initial FL diagnosis and initiation of treatment (continuous variable); sites of nodal involvement at treatment initiation (> 4 or ≤ 4); double refractory (yes or no); and POD24 (yes of no).

^bNo data available in the ReCORD-FL study.

^cExclusion criteria.

^dMissing data in the ReCORD-FL study.

^eThree of the 5 risk factors of the FLIPI score were included in the propensity score model (age, number of nodal sites, and disease stage).

^fExcluded because double refractory already captures refractoriness status, and included in the propensity score model.

Source: Clinical Study Report for comparison of ELARA vs. ReCORD-FL.³² (Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

For each line of therapy, a patient’s propensity score was conditional on the values of their covariates at baseline for that line of therapy. One eligible line of therapy per patient was systematically selected to correspond with the patient’s highest propensity score. Accordingly, for each patient in the ReCORD-FL study, the eligible line of therapy was selected, at which time key baseline prognostic factors were mostly closely aligned with those in the ELARA study. After the selection of 1 line of therapy for each patient in the ReCORD-FL study, patients with a selected line of any chemotherapy, including monotherapy or combination therapy with or without anti-CD20 body, were included in the comparison of SOC in the ReCORD-FL study versus tisagenlecleucel in the ELARA study.

Estimates of Comparison

Estimates of the treatment effect were based on the comparison of patients from the ELARA study and patients from the ReCORD-FL study. Propensity scores were recalculated to ensure that baseline prognostic values were balanced in the 2 groups. The point estimates of propensity scores were based on a logistic regression adjusted for the included covariates listed in Table 30. The model was fitted using maximum likelihood estimation. Using the propensity score estimates, the odds of a patient in the ReCORD-FL study being enrolled in the ELARA study was calculated, given their baseline variables at the selected line of chemotherapy.

The time-to-event end point of OS and PFS were evaluated from the start of enrolment in the ELARA study and from the start date of the selected line of therapy in the ReCORD-FL study. The weighting by odds method was used in the comparison analysis.¹⁰⁶ The distribution of the weighted time-to-event end points were estimated using KM analysis, whereas HRs were estimated using Cox proportional hazard regression. Patients from the ELARA study were assigned a weight of 1, and patients from the ReCORD-FL study were assigned a weight equal to their odds of being enrolled in the ELARA study. The median survival time proportion of patients without an event at 6, 12, 18, and 24 months, as well as corresponding HRs, were also calculated. Bootstrap 95% CIs were provided.

Results

Patient Disposition

Baseline Characteristics

Baseline characteristics in the ELARA and ReCORD-FL studies before and after weighting are summarized in Table 31.

After weighting, included patients from the ELARA (N = 97) and ReCORD-FL (ESS = 45.7) studies appeared balanced on age, sex, prior auto-SCT, number of previous lines of therapies, disease stage at initial diagnosis, time between diagnosis and treatment, sites of nodal involvement, double refractoriness, and POD24. Patients were, on average, middle aged (mean age, 55.4 to 56.5 years), mainly male (67% to 72%), and just more than a third were documented to have had prior auto-SCT therapy. Approximately 68.0% and 70.2% of patients in the ELARA and ReCORD-FL studies, respectively, were documented as double refractory. Approximately 82% of patients in the ReCORD-FL study were refractory to the last prior therapy compared to 73% of patients in the ELARA study, whereas approximately 11% to 40% of patients in the ReCORD-FL study had intermediate or low FLIPI scores, compared to 21% to 19% of patients in the ELARA study.

Table 31: Comparison of Baseline Characteristics in the ELARA Study and the ReCORD-FL Study Before and After Weighting

Auto-SCT = autologous stem cell transplant; FL = follicular lymphoma; FLIPI = Follicular Lymphoma International Prognostic Index; POD24 = progression of disease within 24 months of first-line treatment; PS = propensity score; RoW = rest of world; SD = standard deviation; SMD = standard mean difference.

^aSample size after weighting (i.e., sum of weights) was 99.8 for the ReCORD-FL study and the effective sample size was 45.7.

^bBecause double refractoriness and last prior therapy refractory status both capture refractoriness status, only 1 prognostic factor (double refractoriness) was included in the propensity model. Furthermore, refractoriness to last prior therapy was already very well balanced (SMD < 0.25) before weighting. FLIPI score was excluded from the model on the basis of missingness (12 [15%] additional chemotherapy-treated patients would be excluded from the analysis if FLIPI were included in the model). Considering that 3 of the 5 risk factors of the FLIPI score are already included in the model (age, number of nodal sites, disease stage), which achieved excellent balance with absolute SMDs < 0.25, FLIPI (without missing values) was also well balanced in the cohorts before weighting and, thus, to conserve sample size, FLIPI score was excluded.

Source: Clinical Study Report for comparison of ELARA vs. ReCORD-FL.³² (Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Exposure to Study Treatments

Median follow-up time, defined as time to death or last follow-up date, was 26.2 months for patients who received tisagenlecleucel in the ELARA study and 26.2% in patients who received chemotherapy in the ReCORD-FL study.

Among the patients included from the ReCORD-FL study, the most frequent (> 5%) regimens were BR (14.1%), bendamustine monotherapy (7.7%), cyclophosphamide plus fludarabine plus rituximab (6.4%), R-ICE (6.4%), and R-CHOP (5.1%).

Efficacy

Overall Survival

Comparison of OS between the ELARA and ReCORD-FL studies is summarized in Table 32. KM plots for OS are illustrated in Figure 6.

After weighting, death events were observed in 13.4% of patients in ELARA study and 45.2% of patients in ReCORD-FL study at the time of data cut-off. Median OS was not estimable (NE) for ELARA study. Among patients in ReCORD-FL study, the median OS was 36.6 months (95% CI, 25.8 to NE) after weighting. The KM estimate for OS at 24 months was 90.8% (95% CI, 84.7 to 96.9) and 64.8% (95% CI, 49.5 to 80.0) for ELARA and ReCORD-FL studies, respectively. Compared to standard chemotherapy, tisagenlecleucel was associated with an estimated risk reduction in death of 72% (HR = 0.28; 95% CI, 0.07 to 0.49).

Table 32: Comparison of OS in the ELARA Study and the ReCORD-FL Study Before and After Weighting

CI = confidence interval; HR = hazard ratio; NA = not available; NE = not estimable; OS = overall survival.

Note: Data cut-off dates were March 29, 2022, for the ELARA study and December 31, 2021, for the ReCORD-FL study.

^aSample size after weighting (i.e., sum of weights) was 99.8 for the ReCORD-FL study and the effective sample size was 45.7.

Source: Clinical Study Report for comparison of ELARA vs. ReCORD-FL.³² (Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Progression-Free Survival

The PFS comparison of the ELARA and ReCORD-FL studies is summarized in Table 33. KM plots for PFS after weighting are illustrated in Figure 7.

After weighting, disease progression events (i.e., death or the start of a new anticancer therapy) were observed in 42.3% of patients in the ELARA study and 63.7% of patients in the ReCORD-FL study at the time of data cut-off. Median PFS was NE for the ELARA study. Among patients in the ReCORD-FL study, median PFS was 11.5 months (95% CI, 5.9 to 35.6 months). The KM estimate for PFS at 24 months was 58.6% (95% CI, 48.6% to 68.6%) and 38.3% (95% CI, 22.7% to 53.8%) for the ELARA and ReCORD-FL studies, respectively. Compared to standard chemotherapy, tisagenlecleucel was associated with an estimated risk reduction in disease progression of 47% (HR = 0.53; 95% CI, 0.25 to 0.81).

Figure 6: KM Curves of OS in the ELARA and ReCORD-FL Studies After Weighting

E2202 = the ELARA study; KM = Kaplan-Meier; NE = not estimable; OS = overall survival.

Note: Data cut-off dates were March 29, 2022, for the ELARA study and December 31, 2021, for the ReCORD-FL study. Sample size after weighting (i.e., sum of weights) was 99.8 for the ReCORD-FL study and the effective sample size was 45.7.

Source: Clinical Study Report for comparison of ELARA versus ReCORD-FL.³² (Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.)

Table 33: Comparison of PFS in the ELARA Study and the ReCORD-FL Study Before and After Weighting

CI = confidence interval; HR = hazard ratio; PFS = progression-free survival.

Notes: Data cut-off dates were March 29, 2022, for the ELARA study and December 31, 2021, for the ReCORD-FL study.

^aSample size after weighting (i.e., sum of weights) was 99.8 for the ReCORD-FL study and the effective sample size was 45.7.

Source: Clinical Study Report for comparison of ELARA vs. ReCORD-FL.³² (Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Figure 7: KM Curves of PFS in the ELARA and ReCORD-FL Studies After Weighting

E2202 = the ELARA study; KM = Kaplan-Meier; PFS = progression-free survival.

Notes: Data cut-off dates were March 29, 2022, for the ELARA study and December 31, 2021, for the ReCORD-FL study. Sample size after weighting (i.e., sum of weights) was 99.8 for the ReCORD-FL study and the effective sample size was 45.7.

Source: Clinical Study Report for comparison of ELARA versus ReCORD-FL.³² (Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Harms

Harms outcomes were not compared between the ELARA and ReCORD-FL studies.

Critical Appraisal

Internal Validity

The nonrandomized comparison of the ELARA and ReCORD-FL studies makes interpretation of the efficacy and safety of tisagenlecleucel challenging. To mitigate potential differences in baseline prognostic factors related to OS and PFS, the inclusion and exclusion criteria of the ELARA study were applied to the ReCORD-FL study, and eligible patients from the ReCORD-FL study were systematically selected based on the highest propensity scores. Moreover, comparisons by weighting by odds were conducted to assess the causal effects of prescribing tisagenlecleucel versus chemotherapy. However, important prognostic factors such as baseline ECOG PS and FLIPI scores were not included in the propensity model. Also, important exclusion criteria — as noted by the clinical experts consulted by CADTH for the purpose of this review — specific to the ELARA study, such as the exclusion of patients with a cardiac or cardiac repolarization abnormality and a previous or concurrent malignancy, could not be applied to the ReCORD-FL study. Consequently, there is uncertainty around the treatment effects of tisagenlecleucel relative to standard chemotherapy due to selection bias and unmeasured and residual confounding that cannot be entirely ruled out.

Unadjusted and match-adjusted baseline covariates were reported. Baseline characteristics after matching were well balanced, as evidenced by absolute mean differences of less than 25%. However, the complete baseline demographic and disease characteristics for patients in both trials were not reported after matching; only the balance of patient characteristics relevant to the covariates used in the matching was reported. It was noted that there were differences in the reported characteristics between the cohorts before the weights were applied, indicating heterogeneity. Therefore, it is unclear what effect the matching had on the balance of other relevant patient characteristics. The application of weights resulted in a reduced ESS of 45.7; the reduction in ESS in the primary analysis contributes to imprecision in the estimates, leading to uncertainty of the results.

Standard procedures define PFS as the time to first documented disease progression or death from any cause. However, the date of disease progression was not available for most patients in the ReCORD-FL study. Moreover, radiographic assessment of disease progression is less frequent (or not performed at all) in the real-world setting than in clinical trial protocols. Accordingly, for a comparative analysis of PFS, the date of disease progression was considered at the time a new anticancer therapy was started in both the ReCORD-FL and ELARA studies. Accordingly, censoring was redefined to occur at the last contact date, rather than the last assessment date, in ELARA study to avoid bias related to the timing of assessment. However, due to the inherent limitations of retrospective studies, the accuracy of the date of last contact is uncertain. The validity of using the date of a new anticancer therapy as a surrogate for disease progression is uncertain. Uncertainty about outcome assessment is further compounded by inconsistencies in the assessment of patients included in the ReCORD-FL study. Because the assessment of patients included in the ReCORD-FL study was not planned in the uniform protocol, physicians may have used subjective criteria to assess clinical response or disease progression.

External Validity

Based on input from the clinical experts, the patients included in the comparison of the ELARA and ReCORD-FL studies appeared to be younger than what is typically seen in the clinical setting. The clinical experts noted that the selected anticancer treatments were appropriate SOC regimens. However, whether the change in PFS for the purpose of efficacy comparison between the ELARA and ReCORD-FL studies is appropriate is uncertain.

Description of the Schuster Study

The study by Schuster et al. (2017)³³ was conducted at a single centre in the US. This study reported data from a cohort of patients with FL (n = 14) enrolled in a single-arm, phase IIa clinical trial conducted in patients with NHL and r/r CD19+ lymphomas after chemotherapy. The objective of this study was to estimate the efficacy of tisagenlecleucel in patients with NHL. As a part of the FL subgroup, a total of 15 patients with r/r FL were enrolled, of whom 14 received tisagenlecleucel. This was a pilot study designed to test the central hypothesis of antitumour responses of CD19 CAR T-cell infusions in patients with advanced B-cell NHL. At entry, the suitability of each patient’s T cells for CD19 CAR T-cell manufacturing was determined, and those who had adequate T cells went for leukapheresis. Purified T cells were transduced with CD19-BB-zeta. After leukapheresis at the discretion of physician, patients received bridging therapy followed by staging and LD chemotherapy.

Details of the Schuster et al. study are summarized in Table 34.

Table 34: Details of the Schuster et al. Study

CAR = chimeric antigen receptor; CRR = complete response rate; DoR = duration of response; FL = follicular lymphoma; mFU = median follow-up; NA = not applicable; NR = not reported; ORR = overall response rate; OS = overall survival; PFS = progression-free survival; PRR = partial response rate; ROA = route of administration; TTNT = time to next treatment.

^aDates reported are the study start and end dates from the clinicaltrials.gov record for each trial.

^bEnrolment was defined as the point when the patient met all inclusion and/or exclusion criteria and the patient’s leukapheresis product was received and accepted by the manufacturing facility.

Sources: Schuster et al.³³ and NCT02030834.¹¹⁶ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Populations

Patients eligible for inclusion in this study were required to have CD19+ diffuse large B-cell lymphoma or FL with no curative treatment options (e.g., allo-SCT or auto-SCT), with a limited prognosis (several months to < 2 years of anticipated survival) and a PR or stable disease after the most recent therapy. Patients with FL were eligible if they had measurable progression of disease less than 2 years after the second line of immunochemotherapy (excluding single-drug monoclonal antibody).

Interventions

Tisagenlecleucel was administered as a one-time, single infusion of CD19 CAR T cells by IV injection (total dose of 1 to 5 × 10⁸ cells).³³

Bridging Therapy

In the overall study population, 10 of 28 patients received bridging therapy after apheresis and before LD chemotherapy; however, the specific number of patients with FL who received bridging therapy was not provided.

Lymphodepleting Chemotherapy

The LD chemotherapy received by 14 patients included bendamustine in 6 patients, cyclophosphamide plus fludarabine in 1 patient, radiation plus cyclophosphamide in 3 patients, cyclophosphamide alone in 2 patients, carboplatin plus gemcitabine in 1 patient, and modified EPOCH (doxorubicin, etoposide, cyclophosphamide, but no prednisone or vincristine) in 1 patient.

Outcomes

The primary outcome in the Schuster et al. (2017)³³ study was ORR at 3 months in patients with NHL, based on the 1999 International Working Group response criteria. The secondary outcomes included CRR, PRR, DoR, OS, PFS, safety, and time to next treatment.

Statistical Analysis

The primary outcome of ORR and the associated 95% CIs were calculated based on binomial distribution. PFS and DoR were estimated using the KM methodology. The median survival time and rate of survival with 95% CIs were calculated at different time points. The handling of missing data and sensitivity analyses have not been reported in this study.

Sample Size and Power Calculation

Sample size of the Shuster et al. study was estimated based on a response rate of 50% (15 of 30 patient responses), with 95% CIs of the true response rate being more than 30%. Even though the population of interest in this study was patients with FL, the sample size calculation was based on the overall NHL population.

Results

Patient Disposition

A total of 15 patients with r/r FL were enrolled, of whom 14 received tisagenlecleucel treatment. One patient did not receive the treatment because of rapid disease progression.

Baseline Characteristics

Among the 14 patients who received treatment with tisagenlecleucel, the median age was 59 years (range, 43 to 72 years). There was equal distribution of male and female patients (50% each). The median lines of prior therapy was 5.0 (range, 2 to 10 lines). Overall, 12 patients (86%) were at the advanced stage of disease, which is defined as stage III or IV, according to the modified Ann Arbor staging system. In 4 patients (28%), bone marrow was involved. Elevated LDH was recorded in 9 patients (64%). All patients (100%) had a baseline ECOG PS of 0 or 1. Eight of 14 patients (57%) had double refractory FL, which was defined as disease in which progressive or stable disease was considered to be the best response to chemotherapy (stable disease was defined as disease that is less than 12 months in duration after the patient has undergone at least 4 cycles of first-line therapy or 2 cycles of second-line, third-line, or later line therapies) or relapsed in the 12 months after auto-SCT. Patients must have received an anti-CD20 mAb (unless they had negative test results for CD20) and an anthracycline as 1 of their previous treatment regimens. Three patients (21%) had received prior auto-SCT and 1 patient (7%) had received prior allo-SCT.³³

Exposure to Interventions

As tisagenlecleucel is given as a single-dose, one-time IV infusion, all patients who were infused in the included studies received 100% of the dose. The median number of days from apheresis to tisagenlecleucel infusion was 39 (range, 27 to 145 days).³³ The median total dose of tisagenlecleucel in the overall study population was 5.00 × 10⁸ cells (range, 1.79 × 10⁸ to 5.00 × 10⁸ cells) and the median dose of tisagenlecleucel per kilogram of body weight was 5.79 × 10⁶ cells (range, 3.08 × 10⁶ to 8.87 × 10⁶ cells). No information pertaining to concomitant medications, cointerventions, or subsequent treatments was reported.

Efficacy

In the Schuster et al. study, the response rates were analyzed at 3 months and 6 months. This study reported an ORR of 79% at both assessment points. The CRR was 50% at 3 months and 71% at both 6 months and 49 months.

The median OS was not reached at either 28.6 months or 49 months of median follow-up, and the rate was 93% at 28.6 months of median follow-up. Median PFS was also not reached at 28.6 months; however, a decline was observed at longer follow-up (i.e., median PFS was 32.4 months and 26.2 months at a median follow-up of 49 months and 60 months, respectively). The estimated progression-free probabilities were 77%, 70%, and 43% at median follow-up of 11.4 months, 28.6 months, and 60 months, respectively. The median DoR was not reached at the respective median follow-up time.^33,107-109

Key efficacy results are summarized in Table 35.

Table 35: Summary of Key Efficacy Outcomes in the Schuster et al. Study

CI = confidence interval; CRR = complete response rate; DoR = duration of response; mFU = median follow-up; NE = not estimable; ORR = overall response rate.

Sources: Schuster et al. (2017),³³ Chong et al. (2019),¹⁰⁷ Chong et al. (2021),¹⁰⁸ and Chong et al. (2016).¹⁰⁹ (Note: Details from table have been taken from the sponsor’s Summary of Clinical Evidence.³⁰)

Harms

Adverse Events

Overall, 6 patients (42.9%) with FL experienced any-grade CRS, 2 (14.3%) of whom reported grade 3 or 4 CRS.¹⁰⁹

Serious Adverse Events

Refer to the Adverse Events and Notable Harms sections for CRS and neurologic toxicities.

Withdrawal Due to Adverse Events

Tisagenlecleucel was administered as a one-time infusion; therefore, no patients discontinued treatment in the Schuster et al. study.

Mortality

One death was reported during the study in a patient with FL and progressive neurologic deterioration who had encephalopathy.

Notable Harms

CRS was reported in 6 patients: 4 patients had grade 2 CRS; 1 patient had grade 3 CRS, and 1 patient had grade 4 CRS.

Data for other AEs of special interest were only reported for the overall NHL population in the Schuster et al. study. Eleven patients (39%) reported neurologic toxicities, including encephalopathy in 3 patients (27%), delirium in 2 patients (18%), and tremor in 2 patients (18%). In addition, cognitive disturbance, confusion, involuntary movements, and memory impairment were each reported in 1 patient (9%).

Critical Appraisal

Internal Validity

The main limitations of the Schuster et al. study are the single-arm design, lack of comparator, and open-label nature of the study, which limit interpretation of the treatment effect attributable to tisagenlecleucel. Moreover, the sample size calculation for this study was based on the overall NHL population and was not specific to the FL subgroup, which could limit the detection of magnitude of the effect among the FL subgroup.

External Validity

In terms of generalizability, this study was conducted in the US, which may have a different health system and treatment conditions than Canada.

The patient population in this study had a baseline ECOG PS of 0 or 1. It is not clear if the results are generalizable to patients with a poorer performance status.

Discussion

Summary of Available Evidence

One clinical study, ELARA, was included in the systematic review. The ELARA study is a phase II, open-label, single-arm study that evaluated the efficacy and safety of tisagenlecleucel in patients with r/r grade 1, 2, or 3a FL after 2 or more lines of systemic therapy. The primary end point was CRR per IRC through 24 months. Secondary end points included ORR, OS, PFS, DoR, and HRQoL. A total of 94 of the 98 patients enrolled in the ELARA study received tisagenlecleucel and were included in the primary efficacy analyses. Data with up to 30 months of follow-up were available at the time of this review (data cut-off date of March 29, 2022). The median age observed in the overall r/r FL population was 57 years (range, 29 to 73 years). More males (68.1%) were enrolled than females (31.9%), and most patients (84.0%) were white. Almost all patients (97.0%) had a baseline ECOG PS of 0 or 1. Most patients also had grade 1 or 2 (90.4%) and stage III or IV (80.0%) disease. Enrolled patients had received a median of 4 (range, 2 to 13) prior lines of treatments. Of the 98 included patients, 77.6% were refractory to the last line of antineoplastic therapy. The proportion of patients who progressed within 24 months of first-line anti-CD20 mAb-containing therapy was 64.9%.

One sponsor-submitted ITC was summarized and critically appraised. Although the comparator treatment used in the MAIC has not yet been reviewed by the CADTH pan-Canadian Oncology Drug Review Expert Review Committee for this indication, the ITC was used to inform the sponsor’s pharmacoeconomic model and therefore was reviewed by the clinical team. The sponsor performed a MAIC to estimate the ORR, CRR, OS, and PFS of tisagenlecleucel relative to axicabtagene ciloleucel in patients with r/r FL after 2 more line of therapies. The MAIC included 2 single-arm, open-label, multicentre phase II trials, and was based on individual data from patients who received tisagenlecleucel in the ELARA study and on aggregate data from patients who received axicabtagene ciloleucel in the ZUMA-5 study, which was identified during a systematic review of r/r FL anticancer therapies.

The sponsor also provided an additional study in which patients in the ELARA study were matched with patients included in the retrospective cohort ReCORD-FL study to provide comparator efficacy data for tisagenlecleucel versus standard chemotherapy.

In addition, a single-centre, single-arm study (by Schuster et al.) evaluating the effectiveness and safety of tisagenlecleucel in patients with NHL (including 14 patients with r/r FL) was reviewed. OS and PFS data with up to 60 months of follow-up were reported.

Interpretation of Trial Results

Efficacy

CRR was the primary outcome in the ELARA study. The CRR per IRC was 68.1% (95% CI, 57.7% to 77.3%) in the treated population. The results of sensitivity analyses based on various analysis sets and subgroup analyses based on various patient characteristics at baseline supported the primary analysis of CRR. The clinical experts consulted by CADTH indicated that the results of the CRR was clinically important based on their clinical experience in treating patients with r/r FL.

Survival outcomes were identified by CADTH with input from patient groups and clinicians as some of the most important efficacy outcomes to assess treatment effect in patients with r/r FL. Prolonged survival may be correlated with high response rates (e.g., ORR and CRR); patients who achieve a CR after CAR T-cell therapy have a better prognosis (e.g., more favourable survival) than those who do not respond well but this is not always the case, per the clinical experts. In a meta-analysis evaluating the relationship between response rates and median PFS in patients with NHL (including FL, which accounted for 23% of the study population), a strong correlation between response rates and PFS was found (the coefficient of determination [r²] was 0.78 for ORR versus median PFS and 0.74 for CRR versus median PFS). The results were similar in the subgroup of patients with r/r FL and treatment-naive FL.¹¹⁷ In another meta-analysis examining the correlation between response and survival outcomes, a moderate correlation was observed between CRR and median PFS in patients with FL (r² = 0.69). In this study, the authors noted that because the median OS was usually not reached in clinical trials of NHL, limiting the data, none of the median OS-related correlation analysis results were evaluable.¹¹⁸

At the data cut-off of March 29, 2022, based on the 30-month data in the ELARA study, the survival rate was 82.6% (95% CI, 70.2% to 90.2%) in the EAS, whereas the median OS was not reached at the time of analysis. Another survival outcome, PFS, measures the length of time from tisagenlecleucel infusion to the date of disease progression or death from any cause. The proportion of patients who remained progression-free at 24 months was 57.4% (95% CI, 46.2% to 67.0%). According to the clinical experts consulted by CADTH, the survival data look promising in patients with r/r FL. However, given that the OS data and the PFS data are immature (30 months of data thus far), it is still too early to determine the impact of tisagenlecleucel on a patient’s longer-term survival. In addition, the survival benefit gained from treatment with tisagenlecleucel compared to currently available treatments for r/r FL could not be claimed due to the noncomparative design of the ELARA study. It is also worth considering the OS in patients with FL who do not receive CAR T-cell therapy when interpreting these results. Results from a retrospective analysis conducted in a single centre in the US showed that among patients with FL who had received multiple lines of treatment, median OS was 11.7 years, 8.8 years, and 5.3 years for patients who received the second-line, third-line, and fourth-line treatments, respectively. In this study, recurrent uses of single-drug rituximab (9% to 31%), alkylator-based chemotherapy (22% to 26%), and radiotherapy alone or radioimmunotherapy (10% to 18%) were common in second-line to sixth-line therapy. Ten percent of treated patients received SCT during the course of their therapy (auto-SCT, 6%; allo-SCT, 4%). Investigational therapies (not specified and unclear whether a CAR T-cell product was used) ranged from 8% to 22% when second-line or later therapy was required.⁶

After tisagenlecleucel infusion, some patients received subsequent antineoplastic medications (23.0%) or allo-SCT (4.1%) when they were considered to have stable disease or progressive disease. It is unknown how subsequent therapy with allo-SCT and other cancer treatments would impact a patient’s survival or response to tisagenlecleucel treatment.

HRQoL was an outcome indicated as important by patients with r/r FL, as well as clinicians. HRQoL is also useful in contextualizing the impacts of other efficacy outcomes and harms. Although the findings related to HRQoL based on the 3 tools from the ELARA study suggested that a patient’s HRQoL at baseline was maintained at 24 months, firm conclusion regarding the effect of tisagenlecleucel on HRQoL could not be made due to the large amount of missing data (at 24 months, HRQoL data were available in only half of the baseline population). It is unclear if treatment with tisagenlecleucel is associated with improved HRQoL based on the available evidence.

Due to lack of direct comparative evidence, results from the sponsor-submitted MAIC suggest that ORR, CRR, OS, and PFS associated with tisagenlecleucel treatment are similar to those outcomes associated with axicabtagene ciloleucel treatment. However, definitive conclusions related to the response and survival benefits of tisagenlecleucel, compared to axicabtagene ciloleucel, cannot be drawn from the MAIC analysis due to methodological limitations, including trial design differences that could not be adjusted for in the analysis; adjustment on only a limited number of potential prognostic factors and effect modifiers; and small sample sizes.

Results of a comparison between tisagenlecleucel and standard chemotherapy (from the ReCORD-FL study, which was a noninterventional, multicentre, retrospective chart review used as an external comparator) suggested that, compared to standard chemotherapy, tisagenlecleucel was associated with improved OS and PFS. However, interpretation of the comparative results is limited by the potential for selection bias and residual confounding despite propensity score matching. Therefore, the statistical inference from the comparison of the ELARA to ReCORD-FL studies has low reliability and validity and the results are highly uncertain.

Results of a small single-arm study by Schuster et al. reported favourable CRR per IRC assessment (71% at 49 months), a 93% OS rate at 29-month follow-up, and a 43% PFS rate at 60-month follow-up. Interpretation of the study findings is also limited due to the study design.

Harms

At the data cut-off date of March 29, 2022, all 97 patients in the safety analysis set of the ELARA study reported at least 1 AE. The most common AEs reported were CRS, neutropenia, anemia, diarrhea, headache, decreased white blood cell count, pyrexia, thrombocytopenia, fatigue, nausea, decreased neutrophil count, constipation, and hypogammaglobulinemia. Most of these AEs are symptoms of CRS, which is a usual but severe AE associated with CAR T-cell therapy. A total of 46.4% of patients experienced any SAE. The most commonly reported SAEs included CRS, pneumonia, and febrile neutropenia. There were 13 deaths reported after tisagenlecleucel infusion in the ELARA study (13.4% of patients; 7 patients died from the study indication and 6 from other causes). In terms of AEs of special interest, more than 40% of patients experienced any-grade CRS (49.5%), hematological disorders including cytopenias (78.4%), and infections (55.7%). According to the clinical experts consulted by CADTH, the safety profile of tisagenlecleucel is consistent with other CAR T-cell therapies, and no unexpected safety signals were observed in the ELARA study.

Fifty-three ELARA patients who received infused nonbridging chemotherapy and 124 patients in the ZUMA-5 infused set were included in the MAIC of safety outcomes. At least 1 AE of any grade was reported in 44.6% of patients in the ELARA study and in 78.2% of patients in the ZUMA-5 study. AEs of grade 3 or above were reported in no patients in the ELARA study and in 6.5% of patients in the ZUMA-5 study. Management of CRS with corticosteroids was documented in 3.0% and 15.3% of patients in the ELARA and ZUMA-5 studies, respectively. CRS management with tocilizumab was documented in 9.9% and 45.2% of patients in the ELARA and ZUMA-5 studies, respectively. Neurologic events of any grade were documented in 9.5% of patients in the ELARA study and in 56.5% of patients in the ZUMA-5 study, respectively. Neurologic events of grade 3 or above were reported among 0.19% and 15.32% of patients in the ELARA and ZUMA-5 studies, respectively. Based on the MAIC, tisagenlecleucel appears to have a better toxicity profile than axicabtagene ciloleucel. However, due the previously described limitations of the MAIC analysis, including the application of different definitions of neurologic events in the ELARA and ZUMA-5 studies, the results from the comparative harms analyses between tisagenlecleucel and axicabtagene ciloleucel were associated with too much uncertainty to draw a definitive conclusion.

Harms outcomes were not compared between the ELARA and ReCORD-FL studies.

In the Schuster et al. study, CRS of any grade and of grade 3 or 4 was recorded in 42.9% and 14.3% of patients with r/r FL, respectively. Other harm outcomes were reported for the overall population with NHL in this study, and results specific for patients with r/r FL were not reported.

Conclusion

Evidence from a single-arm study (ELARA) suggests that treatment with tisagenlecleucel is associated with clinically important tumour responses, including complete remission, in adults with r/r FL after 2 or more lines of systemic therapy. There is insufficient evidence — in part due to the limited follow-up duration of the trial — to determine the effects of tisagenlecleucel on OS and PFS. PROs suggest that a patient’s QoL could be maintained; however, 24-month HRQoL data were based on only about 50% of the baseline population. It is unclear if treatment with tisagenlecleucel would improve HRQoL. The harms associated with the tisagenlecleucel infusion are consistent with its mechanism of action, and no unexpected safety signals observed.

The CADTH clinical assessment identified limitations of the sponsor’s comparison of the ELARA and ReCORD-FL studies and the sponsor-conducted MAIC (including small sample sizes, heterogeneity across study designs and populations, and the inability to adjust for all potential effect modifiers and prognostic variables), which substantially limited the ability to interpret the relative treatment effects observed for tisagenlecleucel, standard chemotherapy, and axicabtagene ciloleucel.

References

1.Canadian Cancer Society. Follicular lymphoma. 2022; https://cancer.ca/en/cancer-information/cancer-types/non-hodgkin-lymphoma/what-is-non-hodgkin-lymphoma/follicular-lymphoma. Accessed 2022 Oct 03.

2.American Cancer Society. Types of B-cell lymphomas. About non-Hodgkin lymphoma 2022; https://www.cancer.org/cancer/non-hodgkin-lymphoma/about/b-cell-lymphoma.html. Accessed 2022 Oct 03.

3.Freedman A, Friedberg JW. Histologic transformation of follicular lymphoma. Waltham (MA): UpToDate; 2022: www.uptodate.com. Accessed 2023 Feb 17.

4.Link BK, Day BM, Zhou X, et al. Second-line and subsequent therapy and outcomes for follicular lymphoma in the United States: data from the observational National LymphoCare Study. Br J Haematol. 2019;184(4):660-663. PubMed

5.Hubel K, Salles G, Marcus R, Zinzani PL, Dreyling M. New treatment options in advanced stage follicular lymphoma. Hemasphere. 2018;2(6):e156. PubMed

6.Batlevi CL, Sha F, Alperovich A, et al. Follicular lymphoma in the modern era: survival, treatment outcomes, and identification of high-risk subgroups. Blood Cancer J. 2020;10(7):74. PubMed

7.Rivas-Delgado A, Magnano L, Moreno-Velázquez M, et al. Response duration and survival shorten after each relapse in patients with follicular lymphoma treated in the rituximab era. Br J Haematol. 2019;184(5):753-759. PubMed

8.Johnson PW, Rohatiner AZ, Whelan JS, et al. Patterns of survival in patients with recurrent follicular lymphoma: a 20-year study from a single center. J Clin Oncol. 1995;13(1):140-147. PubMed

9.Klyuchnikov E, Bacher U, Kroger NM, et al. Reduced-intensity allografting as first transplantation approach in relapsed/refractory grades one and two follicular lymphoma provides improved outcomes in long-term survivors. Biol Blood Marrow Transplant. 2015;21(12):2091-2099. PubMed

10.Brenner DR, Poirier A, Woods RR, et al. Projected estimates of cancer in Canada in 2022. CMAJ. 2022;194(17):E601-607. PubMed

11.Pettengell R, Donatti C, Hoskin P, et al. The impact of follicular lymphoma on health-related quality of life. Ann Oncol. 2008;19(3):570-576. PubMed

12.Price M, Ravelo A, Sae-Hau M, et al. Patient-reported disease burden in chronic lymphocytic leukemia, diffuse large B-cell lymphoma, and follicular lymphoma: results from a national patient advocacy survey. J Clin Oncol. 2019;37(15_suppl):e18198.

13.Orphanet. Follicular lymphoma (ORPHA:545). 2010; https://www.orpha.net/consor/cgi-bin/Disease_Search.php?lng=EN&data_id=8750&Disease_Disease_Search_diseaseGroup=Follicular-lymphoma&Disease_Disease_Search_diseaseType=Pat&Disease(s)/group%20of%20diseases=Follicular-lymphoma&title=Follicular%20lymphoma&search=Disease_Search_Simple. Accessed 2022 Oct 03.

14.Morton LM, Wang SS, Devesa SS, Hartge P, Weisenburger DD, Linet MS. Lymphoma incidence patterns by WHO subtype in the United States, 1992-2001. Blood. 2006;107(1):265-276. PubMed

15.Kahl BS, Yang DT. Follicular lymphoma: evolving therapeutic strategies. Blood. 2016;127(17):2055-2063. PubMed

16.Cerhan JR. Epidemiology of follicular lymphoma. Hematol Oncol Clin North Am. 2020;34(4):631-646. PubMed

17.Drug Reimbursement Review sponsor submission: Kymriah (tisagenlecleucel), target of 0.6 to 6.0*108 CAR-positive viable T-cells (non-weight based) and cell suspension in patient-specific single infusion bag for intravenous infusion [internal sponsor's package]. Dorval (QC): Novartis Pharmaceuticals Canada Inc.; 2023 Feb 08.

18.Freedman A, Aster JC. Clinical manifestations, pathologic features, diagnosis,and prognosis of follicular lymphoma. In: Post TW, ed. UpToDate. Waltham (MA): UpToDate; 2022: www.uptodate.com. Accessed 2023 Feb 17.

19.Dreyling M, Ghielmini M, Rule S, et al. Newly diagnosed and relapsed follicular lymphoma: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2021;32(3):298-308. PubMed

20.NCCN Guidelines. B-cell lymphomas, version 5.2022. Plymouth Meeting (PA): National Comprehensive Cancer Network; 2022: https://www.nccn.org/. Accessed 2022 Oct 04.

21.Guideline Resource Unit. Clinical practice guideline on lymphoma, version 16. Edmonton (AB): Cancer Care Alberta, Alberta Health Services; 2021: https://www.albertahealthservices.ca/assets/info/hp/cancer/if-hp-cancer-guide-lyhe002-lymphoma.pdf. Accessed 2022 Oct 05.

22.BC Cancer. Hodgkin lymphoma. (Cancer management manual: lymphoma, chronic leukemia, myeloma). Vancouver (BC): Provincial Health Services Authority; 2013: http://www.bccancer.bc.ca/health-professionals/clinical-resources/cancer-management-manual/lymphoma-chronic-leukemia-myeloma/hodgkin-lymphoma#Treatment. Accessed 2022 Oct 05.

23.Kuruvilla J, Assouline S, Hodgson D, et al. A Canadian evidence-based guideline for the first-line treatment of follicular lymphoma: joint consensus of the Lymphoma Canada Scientific Advisory Board. Clin Lymphoma Myeloma Leuk. 2015;15(2):59-74. PubMed

24.Lymphoma Canada. Follicular lymphoma guidelines. 2015; http://www.lymphoma.ca/medpath/#md:cid=1&md:viewname=dashboardTreeView&ca:viewname=CADefault&ca:id=404&car:viewname=CourseActivityResourcesEmpty&cat:viewname=CourseActivityTreeActivityByType. Accessed 2022 Oct 05.

25.Robinson SP, Canals C, Luang JJ, et al. The outcome of reduced intensity allogeneic stem cell transplantation and autologous stem cell transplantation when performed as a first transplant strategy in relapsed follicular lymphoma: an analysis from the Lymphoma Working Party of the EBMT. Bone Marrow Transplant. 2013;48(11):1409-1414. PubMed

26.Buske C, Hutchings M, Ladetto M, et al. ESMO Consensus Conference on malignant lymphoma: general perspectives and recommendations for the clinical management of the elderly patient with malignant lymphoma. Ann Oncol. 2018;29(3):544-562. PubMed

27.National Cancer Institute. CAR T cells: enginneering patients' immune cells to treat their cancers. 2022; https://www.cancer.gov/about-cancer/treatment/research/car-t-cells. Accessed 2023 Apr 8.

28.Kymriah (tisagenlecleucel): 1.2 x 106 to 6.0 x 108 CAR-positive viable T cells, cell suspension in infusion bag, for intravenous use [product monograph]. Dorval (QC): Novartis Pharmaceuticals Canada Inc.; 2022 Dec 19.

29.Clinical Study Report: CCTL019E2202. A phase II, single arm, multicenter open label trial to determine the efficacy and safety of tisagenlecleucel (CTL019) in adult patients with refractory or relapsed follicular lymphoma, version 1.0 [internal sponsor's report]. Dorval (QC): Novartis Pharmaceuticals Canada Inc.; 2022 Aug 12.

30.Kymriah (tisagenlecleucel) for treatment of adult patients with refractory or relapsed follicular lymphoma after two or more lines of therapy [internal sponsor's report]. In: Drug Reimbursement Review sponsor submission: Kymriah (tisagenlecleucel) cell suspension in infusion bag, 1.2 x 106 to 6.0 x 108 CAR-positive viable T cells for intravenous use. Dorval (Quebec): Novartis Pharmaceuticals Canada Inc.; 2023 Feb 08.

31.Matching-adjusted indirect treatment comparisons of efficacy and safety of tisagenlecleucel versus axicabtagene ciloleucel among adult patients with relapsed/refractory follicular lymphoma [internal sponsor's report] In: Drug Reimbursement Review sponsor submission: Kymriah (tisagenlecleucel), target of 0.6 to 6.0*108 CAR-positive viable T-cells (non-weight based) and cell suspension in patient-specific single infusion bag for intravenous infusion. Dorval (QC): Novartis Pharmaceuticals Canada Inc.; 2022 Aug.

32.Indirect comparison of study E2202 (data cut-off: March 2022) versus ReCORD standard of care / chemotherapy-treated patients in relapsed/refractory follicular lymphoma [internal sponsor's report] In: Drug Reimbursement Review sponsor submission: Kymriah (tisagenlecleucel), target of 0.6 to 6.0*108 CAR-positive viable T-cells (non-weight based) and cell suspension in patient-specific single infusion bag for intravenous infusion. Dorval (QC): Novartis Pharmaceuticals Canada Inc.; 2022 Sep.

33.Schuster SJ, Svoboda J, Chong EA, et al. Chimeric Antigen Receptor T Cells in Refractory B-Cell Lymphomas. N Engl J Med. 2017;377(26):2545-2554. PubMed

34.Shankland KR, Armitage JO, Hancock BW. Non-Hodgkin lymphoma. Lancet. 2012;380(9844):848-857. PubMed

35.Understanding NHL: a patient's guide to non-Hodgkin lymphoma. Mississauga (ON): Lymphoma Canada; 2020: https://z9a2f9c2.rocketcdn.me/wp-content/uploads/2020/05/2020430-NHL-Interior-WEB.pdf. Accessed 2022 Oct 04.

36.Kohrt HEK, Ugarte A. Follicular Lymphoma: a guide for patients - information based on ESMO clinical practice guidelines, v.2014.1 Lugano (CH): ESMO; 2014: https://www.esmo.org/content/download/52236/963497/1/EN-Follicular-Lymphoma-Guide-for-Patients.pdf. Accessed 2022 Oct 03.

37.Salles GA. Clinical features, prognosis and treatment of follicular lymphoma. Hematology Am Soc Hematol Educ Program. 2007;2007(1):216-225. PubMed

38.Conlan MG, Armitage JO, Bast M, Weisenburger DD. Clinical significance of hematologic parameters in non-Hodgkin's lymphoma at diagnosis. Cancer. 1991;67(5):1389-1395. PubMed

39.Freedman A. Follicular lymphoma: 2018 update on diagnosis and management. Am J Hematol. 2018;93(2):296-305. PubMed

40.Freedman A. Follicular lymphoma: 2014 update on diagnosis and management. Am J Hematol. 2014;89(4):429-436. PubMed

41.Casulo C, Byrtek M, Dawson KL, et al. Early relapse of follicular lymphoma after rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone defines patients at high risk for death: an analysis from the National LymphoCare Study. J Clin Oncol. 2015;33(23):2516-2522. PubMed

42.Casulo C, Burack WR, Friedberg JW. Transformed follicular non-Hodgkin lymphoma. Blood. 2015;125(1):40-47. PubMed

43.Seymour JF, Marcus R, Davies A, et al. Association of early disease progression and very poor survival in the GALLIUM study in follicular lymphoma: benefit of obinutuzumab in reducing the rate of early progression. Haematologica. 2019;104(6):1202-1208. PubMed

44.Singh D, Singh A, Mukkamalla SKR. Relapsed and refractory follicular lymphoma. In: [Internet] S, ed. Treasure Island (FL): StatPearls Publishing; 2023: https://www.ncbi.nlm.nih.gov/books/NBK589677/. Accessed 2023 Apr 21.

45.Sarkozy C, Maurer MJ, Link BK, et al. Cause of death in follicular lymphoma in the first decade of the rituximab era: a pooled analysis of French and US cohorts. J Clin Oncol. 2019;37(2):144-152. PubMed

46.Jaglowski SM, Linden E, Termuhlen AM, Flynn JM. Lymphoma in adolescents and young adults. Semin Oncol. 2009;36(5):381-418. PubMed

47.Carbone PP, Kaplan HS, Musshoff K, Smithers DW, Tubiana M. Report of the Committee on Hodgkin's Disease Staging Classification. Cancer Res. 1971;31(11):1860-1861. PubMed

48.Solal-Céligny P, Roy P, Colombat P, et al. Follicular lymphoma international prognostic index. Blood. 2004;104(5):1258-1265. PubMed

49.NCCN Guidelines for Patients. Follicular lymphoma. Plymouth Meeting (PA): National Comprehensive Cancer Network; 2022: https://www.nccn.org/patients/guidelines/content/PDF/nhl-follicular-patient.pdf. Accessed 2022 Oct 04.

50.Lopez-Guillermo A, Caballero D, Canales M, et al. Clinical practice guidelines for first-line/after-relapse treatment of patients with follicular lymphoma. Leuk Lymphoma. 2011;52 Suppl 3:1-14. PubMed

51.Klyuchnikov E, Bacher U, Woo Ahn K, et al. Long-term survival outcomes of reduced-intensity allogeneic or autologous transplantation in relapsed grade 3 follicular lymphoma. Bone Marrow Transplant. 2016;51(1):58-66. PubMed

52.Smith SM, Godfrey J, Ahn KW, et al. Autologous transplantation versus allogeneic transplantation in patients with follicular lymphoma experiencing early treatment failure. Cancer. 2018;124(12):2541-2551. PubMed

53.Sarkozy C, Trneny M, Xerri L, et al. Risk factors and outcomes for patients with follicular lymphoma who had histologic transformation after response to first-line immunochemotherapy in the PRIMA trial. J Clin Oncol. 2016;34(22):2575-2582. PubMed

54.Teva-Lenalidomide (lenalidomide as lenalidomide hydrochloride monohydrate) 2.5 mg, 5 mg, 10 mg, 15 mg, 20 mg, and 25 mg oral capsules [product monograph]. Toronto (ON): Teva Canada Limited; 2022 Sep 13: https://pdf.hres.ca/dpd_pm/00067563.PDF. Accessed 2023 Aug 28.

55.Khangura SD, Ryce A. Lenalidomide plus rituximab chemotherapy for relapsed or refractory indolent B-cell Non-Hodgkin lymphomas. Can J Health Technol. 2021;1(11). 10.51731/cjht.2021.203. Accessed 2022 Nov 01.

56.Zydelig (idelalisib): 150 mg and 100 mg, oral tablet [product monograph]. Mississauga (ON): Gilead Sciences Canada, Inc.; 2020 Aug 18: https://pdf.hres.ca/dpd_pm/00057731.PDF. Accessed 2023 Aug 28.

57.pCODR Expert Review Committee (pERC) final recommendation: idelalisib (Zydelig - Gilead Sciences, Inc). Ottawa (ON): CADTH; 2016 Sep 29: https://www.cadth.ca/sites/default/files/pcodr/pcodr_idelalisib_zydelig_fl_fn_rec.pdf. Accessed 2022 Oct 31.

58.Yescarta (axicabtagene ciloleucel): target of 2 × 106 chimeric antigen receptor (CAR)-positive viable T cells per kg body weight with a maximum of 2 × 108 CAR-positive viable T cells, csuspension in patient-specific single infusion bag, for intravenous infusion [product monograph]. Mississauga (ON): Gilead Sciences Canada, Inc.; 2022 Oct 13: https://pdf.hres.ca/dpd_pm/00067754.PDF. Accessed 2023 Aug 28.

59.Cheson BD, Fisher RI, Barrington SF, et al. Recommendations for initial evaluation, staging, and response assessment of Hodgkin and non-Hodgkin lymphoma: the Lugano classification. J Clin Oncol. 2014;32(27):3059-3068. PubMed

60.Qualifying notice for Kymriah. Ottawa (ON): Health Canada; 2022: https://www.canada.ca/en/health-canada/services/drugs-health-products/drug-products/notice-compliance/conditions/qualifying-notice-kymriah-263500.html. Accessed 2023 May 25.

61.Fowler NH, Dickinson M, Dreyling M, et al. Tisagenlecleucel in adult relapsed or refractory follicular lymphoma: the phase 2 ELARA trial. Nat Med. 2022;28(2):325-332. PubMed

62.Fowler NH, Dickinson M, Ghosh M, et al. Assessment of healthcare resource utilization and costs in patients with relapsed or refractory follicular lymphoma undergoing CAR-T cell therapy with tisagenlecleucel: results from the Elara study. Blood. 2021;138(Supplement 1):3533-3533.

63.Fowler NH, Dickinson M, Martinez‐Lopez J, et al. Patient‐reported quality of life (Qol) following tisagenlecleucel (Tisa‐Cel) infusion in adult patients (Pts) with relapsed/refractory follicular lymphoma (R/R Fl). Hematol Oncol. 2021;39(S2):358-361.

64.Schuster SJ, Dickinson M, Dreyling M, et al. Efficacy and Safety of Tisagenlecleucel (Tisa‐Cel) in Adult Patients (Pts) with Relapsed/Refractory Follicular Lymphoma (R/R Fl): Primary Analysis of the Phase 2 Elara Trial. Hematol Oncol. 2021;39(S2):139-141.

65.Fowler NH, Dickinson M, Dreyling M, et al. Efficacy and safety of tisagenlecleucel in adult patients with relapsed/refractory follicular lymphoma: interim analysis of the phase 2 Elara trial. Blood. 2020;136(Supplement 1):1-3.

66.Hao Y, Hsu W-C, Parzynski CS, et al. Comparison of Clinical Outcomes Among Patients with Relapsed/Refractory Follicular Lymphoma Treated with Tisagenlecleucel in the Elara Trial Versus a Real-World External Control Arm of Patients Treated with Standard of Care. Blood. 2021;138(Supplement 1):2419.

67.Schuster SJ, Dickinson M, Dreyling M, et al. V484: Efficacy and safety of Tisagenlecleucel (Tisa-cel) in adult patients (Pts) with relapsed/refractory follicular lymphoma (r/r FL): primary analysis of the phase 2 ELARA trial. Oncol Res Treat. 2021;44(Suppl 2):104-105.

68.Schuster SJ, Dickinson M, Dreyling M, et al. IBCL-195: primary analysis of the phase 2 ELARA trial: tisagenlecleucel efficacy and safety in adult patients with relapsed/refractory follicular lymphoma (r/r FL). Clin Lymphoma Myeloma Leuk. 2021;21(Suppl 1):S403-S404.

69.Schuster SJ, Dickinson MJ, Dreyling MH, et al. Efficacy and safety of tisagenlecleucel (Tisa-cel) in adult patients pPts) with relapsed/refractory follicular lymphoma (r/r FL): primary analysis of the phase 2 Elara trial. J Clin Oncol. 2021;39(15_suppl):7508.

70.Schuster SJ, Dickinson M, Dreyling M, et al. S210: Efficacy and safety of tisagenlecleucel in adult patients with relapsed/refractory follicular lymphoma: primary analysis of the phase 2 Elara trial. Hemasphere. 2021;5(S2):62.

71.Thieblemont C. Efficacy of tisagenlecleucel in adult patients with high-risk relapsed/refractory follicular lymphoma: subgroup analysis of the phase II ELARA study. Presented at: ASH 2021 Meeting, 2021 Dec 11-14; Atlanta (GA). 2021.

72.Thieblemont C, Dickinson M, Martinez-Lopez J, et al. Efficacy of tisagenlecleucel in adult patients (pts) with high-risk relapsed/refractory follicular lymphoma (r/r FL): subgroup analysis of the phase II ELARA study. Blood. 2021;138(Supplement 1):131.

73.Efficacy and safety of tisagenlecleucel in adult patients with relapsed/refractory follicular lymphoma: Interim analysis of the phase 2 ELARA trial. Clin Adv Hematol Oncol. 2021;19(1 Suppl 2):12.