CADTH Reimbursement Review

Ibrutinib (Imbruvica)

Sponsor: Janssen Inc.

Therapeutic area: Chronic lymphocytic leukemia

This multi-part report includes:

Clinical Review

Pharmacoeconomic 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 of Application Submitted for Review

CLL = chronic lymphocytic leukemia; NOC = Notice of Compliance.

Introduction

Chronic lymphocytic leukemia (CLL) is a lymphoproliferative B-cell malignancy characterized by the progressive expansion of monoclonal B lymphocytes in the blood, bone marrow (BM), lymph nodes, or other lymphoid tissue.^1,2 CLL is a rare disease with low prevalence and incidence worldwide.^3,4 However, it is the most common adult leukemia in Canada.⁵ In 2018, 1,725 patients were diagnosed with CLL (1,095 men and 630 women).⁵ Patients are usually diagnosed with CLL between the ages of 65 and 70 years;⁶ however, more than 10% of patients are diagnosed with CLL when they are younger than 55 years.⁷ Other than factors related to patient, disease, and treatment,⁸ several genetic alterations can influence prognosis, including the deletion of 17p (del[17p]) resulting in the loss of tumour protein 53 (TP53), 1 of the poorest prognostic factors for CLL. Other genetic alterations, including TP53 mutation without del(17p), unmutated immunoglobulin heavy chain variable region (IGHV) gene, deletion of 11q (del[11q]), and complex karyotype (i.e., more than 3 cytogenetic aberrations), are associated with a poor prognosis in CLL. For many patients with CLL, disease burden is increased by the presence of major comorbidities and frailty, as well as by toxicities associated with standard-of-care chemotherapy-based regimens. CLL is generally considered incurable. The 5-year net survival for CLL is 83%; in 2020, 554 patients in Canada died from CLL.⁵ Median life expectancy for patients with del(17p) or TP53 mutation is less than 2 to 3 years from the time of initial diagnosis.⁹

For patients deemed ineligible for fludarabine, cyclophosphamide, and rituximab (FCR), treatment with chemoimmunotherapy, such as chlorambucil plus obinutuzumab, venetoclax plus obinutuzumab (VO), and continuous Bruton tyrosine kinase (BTK) inhibitors (e.g., ibrutinib or acalabrutinib) may be used.^10,11 For patients with high-risk features (i.e., del[17p] and/or TP53 mutation), regardless of age or fitness, BTK inhibitors are the preferred treatment option, even though VO is still available to them.^10,11 For young patients (aged 18 to 64 years) without del(17p) or TP53 mutations, FCR is recommended as first-line treatment for those with mutated IGHV. For young and fit patients with unmutated IGHV, BTK inhibitors are preferred over FCR. VO is also a treatment option for these patients, albeit with less durable remission compared with BTK inhibitors, and is not reimbursed publicly.^10,12 Based on clinician input collected by CADTH, continuous BTK inhibitors are most commonly used in Canada for younger patients with higher-risk mutations, such as TP53 mutations, 11q mutations, or unmutated IGHV.

Ibrutinib monotherapy is indicated for CLL and small lymphocytic lymphoma (SLL), mantle cell lymphoma, and Waldenström macroglobulinemia, all of which have been reviewed by CADTH.^13-15 Venetoclax monotherapy is indicated for CLL and acute myeloid leukemia,^16-18 both of which have been reviewed by CADTH. Venetoclax in combination with obinutuzumab is indicated for CLL, and venetoclax in combination with rituximab is indicated for CLL, both of which have been reviewed by CADTH.^19,20 The requested indication for reimbursement is ibrutinib in combination with venetoclax for the treatment of adult patients with previously untreated CLL, including those with del(17p). The requested indication is the same as the Health Canada–approved indication (post–Notice of Compliance). Ibrutinib in combination with venetoclax is indicated for the treatment of CLL in Europe.²¹ The FDA has not approved the combination of ibrutinib and venetoclax as a treatment for any disease.²²

The objective of this report is to review and critically appraise the evidence submitted by the sponsor on the beneficial and harmful effects of ibrutinib (Imbruvica) (140 mg oral capsule) in combination with venetoclax (Venclexta) in the treatment of CLL in previously untreated adult patients, including those with del(17p).

Stakeholder Perspectives

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

Patient Input

Lymphoma Canada (LC) is a national charity that engages in education, support, advocacy, and research activities for patients and the lymphoma community. LC conducted an online anonymous patient survey between March 22 and May 2, 2023. A total of 87 patients, including 49 patients from Canada, responded to the survey. As most patients with CLL experience no or minor symptoms, many respondents indicated that their daily activities were not strongly impacted by CLL at diagnosis. A total of 64 respondents rated fatigue (47%), high white blood cell counts (26%), and body aches and pains (25%) as having a highly negative impact (3 to 5 out of 5) at diagnosis. Among the 71 respondents who reported psychosocial impacts of their CLL diagnosis, anxiety or worry (61%), stress of diagnosis (59%), and difficulty sleeping (28%) were the most common. The most highly rated negative physical symptoms (3 to 5 out of 5) among 70 respondents included fatigue (44%), body aches and pains (27%), and indigestion, abdominal pain, or bloating (17%). The most negatively rated impacts on quality of life (QoL) among the 87 respondents included anxiety or worry (42%), difficulty sleeping (31%), and stress of diagnosis (28%). When considering a novel CLL treatment, respondents cited living longer (81%), controlling symptoms (75%), longer remission (71%), better QoL (66%), and fewer side effects (35%) as extremely important.

Of 10 patients with CLL who had specific experience with ibrutinib-venetoclax regimen, 5 patients were in remission for 2 to 5 years. In 10 patients treated with ibrutinib-venetoclax regimen, there was improvement in high white blood cell counts (80%), enlarged lymph nodes (70%), low platelet and red blood cell counts (60%), and weight loss (30%).

The input highlighted that a time-limited, oral ibrutinib-venetoclax therapy option would be especially beneficial for those living in rural areas and cost-saving for the health care system. Of note, 24% of respondents reported preference for the fixed duration (FD) of treatment. Also, 55% of patients reported having more treatment options is very important.

Clinician Input

Input From the Clinical Expert Consulted by CADTH

The clinical expert consulted by CADTH indicated that the most important goals of treatment for patients with CLL is to reverse symptoms and control the disease for as long as possible with treatments that have minimal toxicity and do not have a significant negative impact on QoL. The clinical expert stated that the biggest limitation of current treatments for patients with CLL is that tumour cell resistance usually occurs, and patients stop responding or relapse on therapy. Other limitations include toxicity and drug interactions, the requirement for continuous ongoing treatment, and that there are no curative treatments for patients with CLL. The clinical expert stated that other considerations might be addressed, including the opportunity for FD treatment versus continuous treatment and achieving undetectable minimal residual disease (MRD). The clinical expert stated that this combination therapy could be used in a first-line setting and possibly in recurrent disease and in patients with CLL who have some level of resistance or at least have been exposed to the drugs in the combination. In addition, the clinical expert speculated that it is possible that this combination therapy could be used in patients who are resistant to 1 of the drugs in the combination, although this would have to be shown in the evidence. Overall, the clinical expert stated that ibrutinib-venetoclax could be used in patients who have been exposed to, but are not necessarily resistant to, either of these drugs. The clinical expert stated that there is a possible shift in the current treatment paradigm with the combination therapy because the drugs are stopped after a fix period, which is different than the paradigm of continuous therapy that is used for the different BTK inhibitors. The clinical expert commented that older patients or younger patients with a high Cumulative Illness Rating Scale (CIRS) score or a reduced creatinine clearance (CrCl), regardless of del(17p) or TP53 mutations status, should be suited to ibrutinib-venetoclax. In addition, patients with bulky disease or high lymphocytes counts, or with impaired CrCl, who are considered at higher risk for tumour lysis syndrome (TLS), would also be suitable for ibrutinib-venetoclax because ibrutinib was given before venetoclax to reduce the risk of tumour lysis. The clinical expert mentioned that there were no factors identified in the subgroup analysis indicating any subset of patients who would be most likely to respond or, conversely, less likely to respond to the combination therapy. All symptomatic patients with CLL need treatment. The clinical expert indicated that response to treatment is assessed by changes in peripheral blood (PB) counts, which can easily be documented by clinicians looking after patients. In addition, per feedback from the clinical expert, MRD assessments were also used from PB and BM using 2 technologies: next-generation sequencing (NGS) and multicolour flow cytometry. The clinical expert stated that death and disease progression, measured by increasing lymphocyte count or enlarging lymph nodes or spleen, are major reasons for discontinuing the treatment of ibrutinib in combination with venetoclax. Similar to other BTK inhibitors, the clinical expert stated that ibrutinib in combination with venetoclax treatment should be managed by a specialist (i.e., hematologist or medical oncologist) who is familiar with this class of drug to optimally manage toxicities and dosing.

Clinician Group Input

Seven clinicians from the Ontario Health (Cancer Care Ontario) Hematology Cancer Drug Advisory Committee (OH-CCO Hem DAC) and 2 clinicians who treat patients with CLL and SLL in Canada submitted 2 separate clinician group inputs. The 2 clinician groups and a clinical expert consulted by CADTH agree that ibrutinib-venetoclax would be the first-line option for patients with CLL and also suitable for later settings in certain clinical situations. OH-CCO Hem DAC added that ibrutinib-venetoclax may be limited to a healthy (i.e., younger) population, due to safety concerns related to cardiovascular adverse effects. All clinicians agreed that having a time-limited treatment option is an unmet need, that discontinuation of therapy would be considered in the case of disease progression, and that hematologists and/or medical oncologists should be involved in the management of patients with CLL being treated with ibrutinib-venetoclax. Additionally, the clinical expert consulted by CADTH stated that the development of tumour cell resistance, drug interactions, and a lack of curative therapy are also unmet needs. Both clinician groups and the clinical expert consulted by CADTH agree that symptom control, disease control (undetectable MRD), and time off treatment are important goals of therapy with ibrutinib-venetoclax treatment. The 2 clinician groups added that improvement in progression-free survival (PFS) and QoL, along with minimal toxicity from ibrutinib-venetoclax treatment as added by the clinical expert, are the desirable goals of therapy. None of the clinicians had experience treating patients with CLL with the ibrutinib-venetoclax combination.

Drug Program Input

The drug programs that participate in the CADTH reimbursement review process identified potential implementation issues regarding the potential comparators for ibrutinib-venetoclax; the re-treatment of patients whose disease recurs after the completion of therapy; the dosing regimen for ibrutinib-venetoclax; the continuation of venetoclax monotherapy in patients who are intolerant to ibrutinib; access to ibrutinib-venetoclax for patients who are currently receiving ibrutinib monotherapy and have not experienced disease progression; the place in therapy for ibrutinib-venetoclax; care provision issues related to storage requirements; the potential of ibrutinib-venetoclax for drug-drug, drug-food, and drug-herb interactions, requiring assessment and/or intervention; and concerns regarding the feasibility of adoption.

Clinical Evidence

Pivotal Studies and Randomized Controlled Trial Evidence

Description of Studies

The GLOW trial is a multicentre, randomized, open-label phase III study that compared the efficacy and safety of the combination of ibrutinib-venetoclax to chlorambucil-obinutuzumab for the first-line treatment of patients with CLL. GLOW enrolled older patients (aged 65 years and older) with previously untreated CLL who were unfit for treatment with a fludarabine-based regimen. Participants with del(17p) or known TP53 mutation were excluded because these aberrations are associated with inferior outcomes with chemoimmunotherapy (i.e., chlorambucil-obinutuzumab). Participants (N = 211) were randomized in a 1:1 ratio to receive ibrutinib-venetoclax (n = 106) or chlorambucil-obinutuzumab (n = 105). The CAPTIVATE trial is a multicentre, phase II, single-arm study assessing time-limited treatment with the combination of ibrutinib-venetoclax in sequentially enrolled participants with treatment-naive CLL or SLL in either an MRD-guided discontinuation cohort (n = 164) or a FD cohort (n = 159). The MRD cohort will not be further discussed because the sponsor is not proceeding with this treatment regimen. This review focuses on the CAPTIVATE FD cohort, which enrolled 159 patients eligible for a fludarabine-based regimen. Eligible patients in the CAPTIVATE FD cohort were aged 18 to 70 years with previously untreated CLL or SLL requiring treatment, per International Workshop on Chronic Lymphocytic Leukemia (iwCLL) criteria, and with measurable nodal disease by CT, with an Eastern Cooperative Oncology Group Performance Status (ECOG PS) of 0 to 2, and with adequate hepatic, renal, and hematologic function. There were 2 study Canadian sites included in the GLOW trial and no sites in Canada for the CAPTIVATE FD cohort.

The primary end points were PFS per independent review committee (IRC) in the GLOW trial and complete response (CR) rate per investigator assessment (IA) in the CAPTIVATE FD cohort. Other secondary or exploratory outcomes of interest included PFS per IA, overall survival (OS), overall response rate (ORR) per IRC and IA, CR rate per IRC, improvement in hematological parameters (secondary outcome in the GLOW trial; exploratory outcome in the CAPTIVATE FD cohort), duration of response (DOR) per IRC and IA, time to next treatment (TTNT) (GLOW study only), MRD negativity rate, TLS risk reduction, and health-related quality of life (HRQoL) (GLOW study only).

For the fludarabine-ineligible patients included in the GLOW trial, the majority of patients enrolled were men (57.8%) and white (95.7%); 42.2% of patients were female and other races included Asian (0.5%), multiple (0.5%), and 3.3% of patients did not report their race. Median age was 71 years (range, 47 to 93 years), with 87.2% of patients 65 years and older and 34.1% of patients 75 years and older. Advanced stage disease at baseline was reported for 54.8% of patients based on Rai stage III or IV disease and 42.1% of patients based on Binet stage C disease. About half of patients had a baseline ECOG PS of 1 (53.1%). Overall, the proportion of patients with high-risk disease, defined by the presence of del(11q), unmutated IGHV, or TP53 mutation, was similar between treatment arms (59.4% versus 57.1% for ibrutinib-venetoclax versus chlorambucil-obinutuzumab). For fludarabine-eligible patients in the CAPTIVATE FD cohort, the median age at baseline was 60.0 years (range, 33 to 71 years), and 28.3% of patients were 65 years and older. More patients were male (66.7%), and the majority of patients were white (92.5%); 33.3% of patients were female and other races included Asian (1.9%), Black or African American (0.6%), and native Hawaiian or other Pacific Islander (0.6%), and 4.4% of patients did not report their race. At baseline, more patients (69.2%) had an ECOG PS score of 0. Cytogenetic characteristics indicative of poor prognosis (per hierarchical classification) were del(17p) (12.6%) and del(11q) (17.6%). Other poor prognostic characteristics included mutated TP53 (10.1%), mutated del(17p) or TP53 (17.0%), unmutated IGHV (56.0%), and complex karyotype (19.5%).

Efficacy Results

Unless otherwise specified, the key efficacy results of the GLOW and CAPTIVATE trials are summarized in Table 2. Refer to Appendix 1 for the hierarchical testing order of secondary end points for the GLOW study and detailed efficacy data for the GLOW study and the CAPTIVATE FD cohort.

Progression-Free Survival

In the GLOW study, PFS per IRC was the primary end point. At the time of primary analysis (data cut-off: February 26, 2021), the data recorded a median follow-up time of 27.7 months (95% confidence interval [CI], 27.50 to 27.83 months) in the ibrutinib-venetoclax arm and 27.89 months (95% CI, 27.53 to 28.58 months) in the chlorambucil-obinutuzumab arm. Median PFS per IRC was not reached in the ibrutinib-venetoclax arm and was 21.0 months in the chlorambucil-obinutuzumab arm. The hazard ratio (HR) for PFS events was 0.216 (95% CI, 0.131 to 0.357; P < 0.0001) based on IRC assessment. Events of disease progression or death were reported for 20.8% of patients in the ibrutinib-venetoclax arm (13 disease progression events and 9 deaths) and 63.8% of patients in the chlorambucil-obinutuzumab arm (65 disease progression events and 2 deaths). Subgroup analyses and prespecified sensitivity analyses of PFS per IA in the primary analysis were generally consistent with the primary analysis across all prespecified subgroups except for race and disease diagnosis at baseline. In general, results were similar at the time of the extended follow-up (data cut-off: August 25, 2022).

In the CAPTIVATE FD cohort, in the primary analysis (data cut-off: November 12, 2020), median PFS by IRC assessment was not reached in the all-treated FD cohort, based on an overall median follow-up of 27.9 months. The PFS event rate probabilities in the all-treated FD cohort were 14.5% (21 disease progression events and 2 deaths) at the time of primary analysis. Of note, results for median PFS per IRC for all patients and for patients with del(17p) were based on an immature time point after the median follow-up and were therefore considered to be unreliable by the sponsor. Generally, similar results were observed for PFS per IA and in the extended follow-up analysis (data cut-off: August 4, 2021).

Overall Survival

In the GLOW study, in the primary analysis (data cut-off: February 26, 2021), median OS was not reached in either arm. With a median follow-up of 27.7 months for the ibrutinib-venetoclax arm and 27.89 months for the chlorambucil-obinutuzumab arm, there were 11 (10.4%) deaths observed in the ibrutinib-venetoclax arm and 12 (11.4%) deaths observed in the chlorambucil-obinutuzumab arm (HR = 1.048; 95% CI, 0.454 to 2.419; nominal P = 0.9121). Similarly, median OS was not reached in either arm in the extended follow-up analysis (data cut-off: August 25, 2022).

In the CAPTIVATE FD cohort, in the primary analysis (data cut-off: November 12, 2020), median OS was not reached in the all-treated FD cohort based on an overall median follow-up of 27.9 months; there were 3 deaths (1.9%) reported in the all-treated FD cohort and no deaths reported in patients with del(17p). Most patients were alive and on study with OS probabilities of 98.1% at 24 months at the primary data cut-off date and at 36 months at the extended follow-up analysis data cut-off date (August 4, 2021) for the all-treated FD cohort. All patients with del(17p) were alive and on study at the primary data cut-off date and at the extended follow-up analysis data cut-off date.

Overall Response Rate

At the GLOW primary analysis (data cut-off: February 26, 2021), the IRC-assessed ORR (of partial response [PR] or better) was similar in the ibrutinib-venetoclax and chlorambucil-obinutuzumab arms (86.6% and 84.8%, respectively; relative response = 1.02; 95% CI, 0.92 to 1.14; P = 0.6991). Similar results were observed in the ORR based on IA. Similar results were observed in the extended follow-up analysis (data cut-off: August 25, 2022). Of note, as the difference in ORR based on IRC assessment between treatment arms was not statistically significant (P = 0.6991) in the primary analysis, the hierarchical statistical testing strategy ended at ORR per IRC. The remaining key secondary end points (i.e., OS, sustained hematological improvements, and time to improvement on the Functional Assessment of Chronic Illness Therapy [FACIT]-Fatigue Scale) and ORR per IA were considered not statistically significant.

In the CAPTIVATE FD cohort, in the primary analysis (data cut-off: November 12, 2020), the ORR per IRC assessment was 96.2% (95% CI, 93.3% to 99.2%) for all treated patients and 100.0% (95% CI, 100.0% to 100.0%) for patients with del(17p). Generally, similar results were observed with ORR per IA and in the extended follow-up analysis (data cut-off: August 4, 2021).

CR (CR and CR With Incomplete Bone Marrow Recovery) Rate

In the GLOW study, in the primary analysis (data cut-off: February 26, 2021), the IRC-assessed CR rate was higher in the ibrutinib-venetoclax arm compared with the chlorambucil-obinutuzumab arm (38.7% and 11.4%, respectively; relative response = 3.43; 95% CI, 1.91 to 6.15; P < 0.0001). Similar results were observed at the time of the 18-month extended follow-up (data cut-off: August 25, 2022) and with CR rate per IA.

In the CAPTIVATE FD cohort, CR rate by IA was assessed as the primary end point. In patients without del(17p) in the FD cohort, the CR rate per IA was 55.9% (95% CI, 47.5% to 64.2%), which exceeded the prespecified minimum CR rate of 37% (1-sided P < 0.0001) in the primary analysis (data cut-off: November 12, 2020). Similar results were observed in the extended follow-up analysis (data cut-off: August 4, 2021) with a CR rate of 58.1% (95% CI, 49.8% to 66.4%) based on IA. The investigator-assessed CR rate in the all-treated patients was 55.3% (95% CI, 47.6% to 63.1%) in the primary analysis and 57.2% (95% CI, 49.5% to 64.9%) in the extended follow-up analysis. For patients with del(17p), the CR rates per IA were identical, at 50.0% (95% CI, 28.1% to 71.9%), in the primary analysis and in the extended follow-up analysis.

Subgroup analyses of CR rate per IA, sensitivity analyses using CR rate per IRC and duration of CR, and supportive analysis of the duration of CR were generally consistent with the primary analysis. The subgroup analysis results were consistent across all prespecified subgroups except for bulky disease. Refer to Appendix 1 for detailed subgroup analyses data.

Sustained Hematologic Improvement

In the GLOW study, in the primary analysis (data cut-off: February 26, 2021), the proportion of patients with sustained improvement in hemoglobin was similar for the ibrutinib-venetoclax and chlorambucil-obinutuzumab arms (44.3% and 50.5%, respectively; nominal P = 0.3854). The proportion of patients with sustained improvement in platelets was similar for the ibrutinib-venetoclax and chlorambucil-obinutuzumab arms (24.5% and 29.5%, respectively; nominal P = 0.4346). Similar results were observed at the time of the 18-month extended follow-up (data cut-off: August 25, 2022).

In the CAPTIVATE FD cohort, in the primary analysis (data cut-off: November 12, 2020), the proportion of patients achieving a sustained improvement in hemoglobin was 41.5% (95% CI, 33.9% to 49.2%) in all treated patients and 60.0% (95% CI, 38.5% to 81.5%) in patients with del(17p). The proportion of patients with sustained improvement in platelets was 17.6% (95% CI, 11.7% to 23.5%) in all treated patients and 15.0% (95% CI, 0 to 30.6%) in patients with del(17p). Similar results were observed in the extended follow-up analysis (data cut-off: August 4, 2021).

Duration of Response

In the GLOW study, as of the data cut-off for the primary analysis (February 26, 2021), with an overall median follow-up of 27.7 months, the median DOR for patients who achieved an IRC-assessed PR or better was 28.9 months (95% CI, 28.7 to not estimable [NE]) in the ibrutinib-venetoclax arm and 21.1 months (95% CI,15.9 to 25.1) in the chlorambucil-obinutuzumab arm. Similar results were observed at the time of the 18-month extended follow-up (data cut-off: August 25, 2022) and for DOR per IA.

In the CAPTIVATE FD cohort, in the primary analysis (data cut-off: November 12, 2020), with a median follow-up of 27.9 months, the median DOR per IRC assessment for the FD cohort was not reached for all patients or for patients with del(17p) (with the lower end of the 95% CI of 18.9 months). The median DOR per IRC for patients with del(17p) in the primary analysis was considered not reliable due to the limited number (| | |) of patients at risk at 36 months after initial response. Similar results were observed for DOR per IA and for the extended follow-up analysis (data cut-off: August 4, 2021).

Time to Next Treatment

At the GLOW primary analysis (data cut-off: February 26, 2021), fewer patients in the ibrutinib-venetoclax arm received subsequent anticancer therapy compared to that in the chlorambucil-obinutuzumab arm (3.8% and 25.7%, respectively; HR = 0.143, 95% CI, 0.050 to 0.410; nominal P < 0.0001). Similar results were observed in the extended follow-up analysis (data cut-off: August 25, 2022).

TTNT was not reported in the CAPTIVATE FD cohort.

MRD Negativity

In the GLOW trial, a higher proportion of patients reported negative overall MRD by NGS in the ibrutinib-venetoclax arm than in the chlorambucil-obinutuzumab arm in BM (55.7% and 21.0%, respectively; relative response = 2.65; 95% CI, 1.75 to 3.99; P < 0.0001) and in PB (59.4% and 40.0%, respectively; nominal P = 0.0055) in the primary analysis (data cut-off: February 26, 2021). The MRD negativity rate was not assessed in the extended follow-up analysis (data cut-off: August 25, 2022).

In the all-treated FD cohort of the CAPTIVATE trial, overall MRD negativity rates by flow cytometry were as follows: 59.7% (95% CI, 52.1% to 67.4%) for all patients and 45.0% (95% CI, 23.2% to 66.8%) for patients with del(17p) in the BM; and 76.7% (95% CI, 70.2% to 83.3%) for all patients and 80.0% (95% CI, 62.5% to 97.5%) for patients with del(17p) in the PB. Identical results were reported for the overall MRD negativity rate in the extended follow-up analysis (data cut-off: August 4, 2021).

TLS Risk Reduction

In the ibrutinib-venetoclax arm of the GLOW study, 26 (24.5%) patients were at high risk for TLS based on tumour burden at baseline. After the ibrutinib lead-in, 22 (20.8%) patients shifted to medium or low risk in the primary analysis (data cut-off: February 26, 2021). TLS risk reduction was not assessed in the extended follow-up analysis (data cut-off: August 25, 2022).

In the FD cohort of the CAPTIVATE study, high tumour burden was observed in 34 of 159(21.4%) all-treated patients; among them, 1 (5%) patient had del(17p) at baseline. After 3 cycles of single-drug ibrutinib lead-in therapy, 33 (20.8%) patients shifted to medium or low risk in the primary analysis (data cut-off: November 12, 2020); among them, 1 (5%) patient had del(17p). TLS risk reduction was not assessed in the extended follow-up analysis (data cut-off: August 4, 2021).

Health-Related Quality of Life

In the GLOW trial, HRQoL was a secondary outcome. Generally, in the primary analysis of the GLOW trial (data cut-off: February 26, 2021), patients in the ibrutinib-venetoclax arm had early deteriorations and later improvements in HRQoL compared to patients in the chlorambucil-obinutuzumab arm, as measured by European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 30 (EORTC QLQ-C30) global health status, 5-Level EQ-5D questionnaire visual analogue scale (EQ-5D-5L VAS), EQ-5D-5L utility score, and FACIT-Fatigue Scale. All HRQoL results were considered not statistically significant according to the prespecified hierarchical statistical testing strategy because ORR per IRC failed to demonstrate statistical significance. The time to worsening or improvement in EORTC global health status, EQ-5D-5L VAS, EQ-5D-5L utility score, and FACIT-Fatigue Scale score were not assessed in the extended follow-up analysis (data cut-off: August 25, 2022).

HRQoL was not measured in the CAPTIVATE FD cohort.

Harms Results

Harms results for the GLOW trial and the CAPTIVATE FD cohort are summarized in Table 3. In fludarabine-ineligible patients reported in the GLOW trial, at least 1 adverse event (AE) was reported for a similar proportion of patients in the ibrutinib-venetoclax arm and the chlorambucil-obinutuzumab arm (99.1% versus 94.3% for ibrutinib-venetoclax versus chlorambucil-obinutuzumab) in GLOW. AEs of any grade were reported more frequently in the ibrutinib-venetoclax arm than in the chlorambucil-obinutuzumab arm, including diarrhea (50.9% versus 12.4%); nausea (26.4% versus 25.7%), infections and infestations (60.4% versus 48.6%); metabolism and nutrition disorders (42.5% versus 23.8%); respiratory, thoracic, and mediastinal disorders (35.8% versus 28.6%); musculoskeletal and connective tissue disorders (34.0% versus 25.7%); and nervous system disorders (30.2% versus 20.0%).

More patients in ibrutinib-venetoclax arm experienced at least 1 serious adverse event (SAE) of any grade than the chlorambucil-obinutuzumab arm (46.2% versus 27.6% for ibrutinib-venetoclax versus chlorambucil-obinutuzumab). More patients reported AEs leading to discontinuation in the ibrutinib-venetoclax arm (20.8% versus 7.6%). There were 11 (10.4%) patients in the ibrutinib-venetoclax arm and 12 (11.4%) patients in the chlorambucil-obinutuzumab arm who died during the study period and AE was the most frequent cause of death in the ibrutinib-venetoclax arm (6.6% versus 1.9%). In the GLOW study, more patients in the ibrutinib-venetoclax arm than in the chlorambucil-obinutuzumab arm reported atrial fibrillation (14.2% versus 1.9%). Grade 3 or 4 atrial fibrillation was reported in 2 (1.9%) patients in the ibrutinib-venetoclax arm while no patients in the chlorambucil-obinutuzumab arm. Similarly, a higher proportion of patients reported major hemorrhage in the ibrutinib-venetoclax arm than in the chlorambucil-obinutuzumab arm (3.8% versus 1.0%).

In fludarabine-eligible patients in the CAPTIVATE FD cohort, at least 1 AE was reported in 158 (99.5%) patients and AEs of grade 3 or 4 were reported in 98 (61.6%) patients. The most commonly reported AEs were diarrhea (62.3%), nausea (42.8%), neutropenia (41.5%), and arthralgia (33.3%). Grade 3 or 4 neutropenia was reported in 52 (32.7%) patients. There were 36 (22.6%) patients who experienced at least 1 SAE of any grade and 30 (18.9%) patients who experienced at least 1 SAE of grade 3 or 4. Eight (5.0%) patients had AEs leading to ibrutinib discontinuation, while 3 (1.9%) patients had AEs leading to venetoclax discontinuation. There were 3 (1.9%) patients who died during the study period and 7 (4.4%) patients who had atrial fibrillation, 2 (1.3%) of which were of grade 3 or 4. Major hemorrhage was reported in 3 (1.9%) patients.

Table 2: Summary of Key Results From Pivotal Studies and RCT Evidence (ITT Analysis Set for GLOW; All-Treated Analysis for CAPTIVATE)

C+O = chlorambucil plus obinutuzumab; CI = confidence interval; CR = complete response; CRi = complete response with incomplete bone marrow recovery; del(17p) = deletion of 17p; FD = fixed duration; HR = hazard ratio; I+V = ibrutinib plus venetoclax; IA = investigator assessment; IRC = independent review committee; ITT = intention to treat; MRD = minimal residual disease; NA = not applicable; NE = not estimable; NR = not reported; ORR = overall response rate; OS = overall survival; PFS = progression-free survival; RCT = randomized controlled trial.

^aEstimate is not reliable because only 3 patients were at risk at 39 months.

^bNext-generation sequencing was used as the primary method of MRD analysis for the GLOW trial; flow cytometry method was used for the MRD analysis in the CAPTIVATE FD cohort.

Sources: GLOW primary analysis CSR;²³ GLOW extended follow-up analysis CSR;²⁴ GLOW PRO CSR;²⁵ CAPTIVATE CSR.²⁶ (Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.)

Table 3: Summary of Harms — Pivotal and RCT Evidence (Safety Analysis Set for GLOW; All-Treated Analysis for CAPTIVATE)

FD = fixed duration; NR = not reported (when incidence rates fall below the specified reporting threshold in either treatment arm within the specified adverse event severity); RCT = randomized controlled trial; SAE = serious adverse event; TEAE = treatment-emergent adverse event.

Note: At the GLOW extended 18-month follow-up from the primary analysis (data cut-off: August 25, 2022), no changes in safety data were noted, except for 1 additional patient in the chlorambucil-obinutuzumab treatment arm who was diagnosed with 2 new treatment-emergent SAEs, (myelodysplastic syndrome and myeloproliferative neoplasm) after primary analysis and 7 additional patients who were reported to have developed a nontreatment-emergent secondary malignancy (4 patients from the ibrutinib-venetoclax arm and 3 patients from the chlorambucil-obinutuzumab arm).

Sources: GLOW primary analysis CSR;²³ CAPTIVATE CSR.²⁶ (Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.)

Critical Appraisal

Fludarabine-Ineligible Population

The open-label design of the GLOW trial had the potential to introduce reporting bias into the assessment of subjective outcomes reported by patients, such as HRQoL and AEs. Disease response outcomes (PFS, ORR, DOR, CR rate) were assessed by investigator assessments and an IRC to help mitigate the biases associated with the open-label study design for the GLOW trial. In the GLOW trial, the overall median duration of exposure was substantially longer for the ibrutinib-venetoclax arm (13.8 months) than for the chlorambucil-obinutuzumab arm (5.1 months), which may bias the results in favour of ibrutinib-venetoclax. In addition, fewer patients in the chlorambucil-obinutuzumab arm discontinued treatment due to AEs compared to the ibrutinib-venetoclax arm, and a higher proportion of patients completed the study treatment in the chlorambucil-obinutuzumab arm than the ibrutinib-venetoclax arm, which indicates that patients in the chlorambucil-obinutuzumab arm had better treatment compliance than those in the ibrutinib-venetoclax arm in the GLOW trial, which may bias the results against the ibrutinib-venetoclax arm. A higher proportion of patients in the chlorambucil-obinutuzumab arm received subsequent anticancer therapy compared to the ibrutinib-venetoclax arm. The clinical expert consulted by CADTH indicated that the use of subsequent therapies would influence OS. The CADTH review team agreed with the clinical expert and noted that the use of subsequent anticancer therapy results in an indirectness of the estimated OS effect. It is difficult in this setting to isolate the direct effect of ibrutinib-venetoclax treatment on OS due to the intercurrent use of subsequent anticancer therapies. There were critical protocol amendments in the GLOW trial impacting the conduct of the trial after patients had first been randomized, which may have biased the results and increased uncertainty because of increased heterogeneity in the patient population for the ibrutinib-venetoclax arm. For example, the last 3 cycles (cycles 16, 17, and 18) of ibrutinib monotherapy were removed from the treatment in the ibrutinib-venetoclax arm. This may increase uncertainty in the estimate of treatment effect due to patient heterogeneity because there may be a mix of patients who received and did not receive 3 cycles of ibrutinib monotherapy. Due to limited data, the magnitude and direction of the bias is uncertain. In the GLOW trial, median OS was not reached in the ibrutinib-venetoclax arm in the primary analysis, so the OS data were considered to be immature.

Fludarabine-Eligible Population

The FD cohort in the CAPTIVATE trial was designed as a single-arm study and, given the lack of a comparator arm, the ability to make definitive conclusions about the comparative efficacy of ibrutinib-venetoclax in fludarabine-eligible patients with CLL is limited. In addition, the open-label design had the potential to introduce reporting bias in the assessment of subjective outcomes reported by patients (i.e., AEs). Disease response outcomes (PFS, ORR, DOR, CR rate) were assessed by investigators and by an IRC to help mitigate the biases associated with the open-label study design for the FD cohort in the CAPTIVATE trial. The Median of OS and PFS were not reached in the CAPTIVATE FD cohort in the primary analysis, so the OS data were considered immature. In addition, the CADTH review team noted that although the CAPTIVATE FD cohort included a subgroup of patients with del(17p), no formal statistical testing was performed between subgroups and the sample size for patients with del(17p) was small (n = 20); thus, no conclusions can be drawn from the study results for patients with del(17p). Although the subgroup analyses were prespecified, there is no evidence that the studies were powered to detect subgroup differences. HRQoL is considered a relevant outcome by patients with CLL and by clinicians. However, there was no assessment for HRQoL in the CAPTIVATE FD cohort; thus, it is uncertain whether treatment with ibrutinib-venetoclax would improve HRQoL in fludarabine-eligible patients with CLL.

External Validity

Both the GLOW trial and CAPTIVATE FD cohort required eligible patients to have measurable nodal disease. However, according to the clinical expert consulted by CADTH, there is a small proportion of patients who only have elevated white cell counts and cytopenia and may not have an enlarged lymph node in clinical practice; these patients are important and would fit in the patient population for the ibrutinib-venetoclax regimen. The clinical expert stated that there is a more diversified patient population that includes patients from Asia and other parts of the world in their clinical practice compared to the patient population in the GLOW trial and CAPTIVATE FD cohort. The baseline characteristics of the 2 studies may be indicative of the overrepresentation of white patients (≥ 92%) with CLL in both the fludarabine-eligible and fludarabine-ineligible populations and thus present an evidence gap in patients’ generalizability. Although the inclusion criteria mandated that eligible patients in the GLOW trial have a CrCl of less than 70 mL/min and/or a CRIS score higher than 6, the baseline CrCl for both treatment arms is higher than what the clinical expert would expect in high-risk patients with CLL who are ineligible to FCR in clinical practice, which may indicate that patients in the GLOW trial have better kidney functions than patients in clinical practice. This may compromise the study results for general fludarabine-ineligible patients with CLL. Generally, the risk profile of patients in the CAPTIVATE FD cohort is what the clinical expert would expect in fludarabine-eligible patients with CLL in clinical practice.

Long-Term Extension Studies

No long-term extension studies were identified for this review.

Indirect Comparisons and Observational Studies

Description of Studies

In patients who were ineligible to received fludarabine, the sponsor provided 2 indirect treatment comparisons (ITCs) and 2 individual patient data (IPD) observational studies that evaluated the efficacy and safety of ibrutinib-venetoclax versus bendamustine plus rituximab (BR),²⁷ ibrutinib,²⁸ VO,²⁹ and acalabrutinib.³⁰ In fludarabine-eligible patients, an IPD analysis was conducted comparing ibrutinib-venetoclax versus FCR.³¹ The ITC analyses were based on matching-adjusted indirect treatment comparison (MAIC) methods (comparison with acalabrutinib and VO) and the IPD analyses were based on propensity score methods (comparison with BR, ibrutinib, and FCR). All comparisons included patients with untreated CLL but varied in terms of age (all adults or ≥ 65 years only) and the presence of comorbidities and high-risk mutations, such as del(17p) and TP53. The median follow-up duration ranged from 38 months to 54.4 months, depending on the treatment group and the analysis. ||| |||| |||| |||| |||||| |||| |||||||| ||| | |||||||||| ||||| |||| ||||||||||| |||||| |||| ||||| ||| |||| |||||||| ||||||| ||| ||| |||||||| |||||||||| ||| || ||||||||| |||||||||||| ||||||||| |||| |||| |||||| ||||| ||||||||| ||| ||||||||| ||||||||| |||||| |||||||| ||| ||||||.

Efficacy Results

Fludarabine-Ineligible Population

For ibrutinib-venetoclax versus acalabrutinib, the time-varying analysis of PFS estimated an HR of |||| |||| ||| |||| || ||||| | | ||||| ||| | || |||||| ||| |||| |||| ||| |||| || ||||| | | ||||| for more than 12 months for the base-case model. The comparison with VO reported a base-case PFS HR of |||| |||| ||| |||| || ||||| | | ||||) for the first 12 months of follow-up; after 12 months, the estimated HR was |||| |||| ||| |||| || ||||| | | |||||).

The observational study comparing ibrutinib-venetoclax versus BR estimated a PFS HR of |||| |||| || |||| || ||||| | | ||||||) for the base-case analysis. For the comparison of ibrutinib-venetoclax versus ibrutinib, ||||| ||| ||||| of patients in the ibrutinib-venetoclax and ibrutinib groups, respectively (| | ||||), reported a progression event (|| ||| ||| || ||| ||||||||).

For all comparisons in the fludarabine-ineligible population, the comparative efficacy of ibrutinib-venetoclax on OS was unclear, as the results had high uncertainty due to the low number of events, the limited sample size, and, in many cases, ||| || |||| ||||||| ||| |||| ||| |||||| |||||||||.

Fludarabine-Eligible Population

In the base-case analysis for PFS, the model estimated an HR of |||| |||| ||| |||| || ||||| | | |||||), favouring ibrutinib-venetoclax over FCR. The analysis of OS also suggested a benefit favouring ibrutinib-venetoclax but was limited by small sample size and low event rates, and thus should be interpreted with caution.

Harms Results

No safety data were reported in the fludarabine-ineligible population.

Among fludarabine-eligible patients, grade 3 or 4 treatment-emergent adverse events (TEAEs) were reported by ||| || ||| of patients in the FCR group compared with ||| || ||| of those in the ibrutinib-venetoclax group, across the base-case and sensitivity analyses conducted. Comparative odds ratios (ORs) of grade 3 or 4 AEs were not reported. The incidence of treatment discontinuation ranged from || || || for ibrutinib-venetoclax to ||| || ||| for FCR across the analyses.

Critical Appraisal

Fludarabine-Ineligible Population

For the MAICs comparing ibrutinib-venetoclax with acalabrutinib and VO, there was poor overlap between trial populations in the GLOW study and comparator trials, which was evident, given the low effective sample size after weighting (||||| || || ||| ||||||||). The comparator trials included patients with high-risk mutations who were excluded from the GLOW study, and the presence of these mutations could not be controlled for in the adjusted analyses. The ability to achieve balance in effect modifiers was also limited due to missing covariate data, and imbalances between groups were noted for important patient characteristics in the base-case analysis. ||||||||| ||| |||||||||||| |||||| |||||||||| ||| ||| ||| ||| ||| ||| ||| ||| |||||||||||| ||| |||||||| |||||| ||||||||| ||||||| |||| |||||| |||||||||. Given these limitations, the findings of the MAIC were considered highly uncertain.

The selection of covariates for inclusion in the propensity score model is important for inference to be valid in observational studies. In these analyses, it is unclear if all known confounders and prognostic factors were included. Moreover, there were issues with missing covariate data which may have impacted the specification of the propensity scores. ||| |||||||||||| || |||||||||| |||||| |||||| | |||||||||||| |||||||||| || |||||||| ||| |||| ||| ||||||||||| || ||||||||| ||| ||||||||| |||||| |||||||||| ||||| |||| ||| ||||| |||||||| |||||| || |||| |||| |||||||| ||||||| |||||||| ||| |||| |||| | |||||||||||||||| ||||||||| || ||| ||||||||| ||||||| ||| |||||| ||||| |||| ||||||||| |||||| |||| |||| || |||||||| || ||| ||| |||||| ||| ||| ||| ||| |||||||| || ||| |||||||||| ||||||.

In the fludarabine-ineligible population, none of the pairwise comparisons reported safety outcomes, thus the comparative safety is unknown.

Regarding external validity, there were differences in the patients included in each pairwise comparison that should be considered when interpreting the results. The comparison with BR and ibrutinib included only patients 65 years and older (with or without comorbidities), whereas the comparison with VO included any adult (≥ 18 years) with comorbidities (CIRS score > 6 and CrCl ≤ 70 mL/min), and the comparison with acalabrutinib included all patients 65 years and older but only younger adults who had comorbidities. Thus, it appears that the patients may not be comparable across analyses and the external validity of each pairwise comparison should be assessed separately.

Fludarabine-Eligible Population

As there was no randomized control group in the CAPTIVATE study, the sponsor conducted an observational study based on propensity score methods. For the IPD analysis, it was unclear if all known confounders and prognostic factors were included in the model and, in addition, there were issues with missing covariate data. Therefore, there may be residual confounding from measured, as well as unmeasured, covariates that could bias the effect estimates. Also, the data were based on a small sample size (||| ||| ||||||||| ||| ||| |||||||| || ||| |||| |||| ||||||||), and for OS, there were low event rates. Limited safety data were reported; thus, the comparative safety is unclear. The external validity of the base-case findings may be limited by the inclusion of patients with comorbidities and high-risk mutations that would not be eligible to receive FCR in clinical practice.

Conclusions

Patients and clinicians highlighted the need for new effective treatments for CLL that prolong life, control disease and symptoms, maintain QoL, and reduce side effects compared to current treatments. According to 1 pivotal trial, ibrutinib-venetoclax demonstrated a clinically meaningful improvement in PFS compared with chlorambucil-obinutuzumab in fludarabine-ineligible patients with CLL without del(17p). In fludarabine-eligible patients with CLL without del(17p), ibrutinib-venetoclax demonstrated a meaningful improvement in the CR rate compared with the prespecified minimum CR rate of 37%. Analyses of secondary outcomes supported the efficacy of ibrutinib-venetoclax in prolonging PFS and delaying disease progression. Data on OS were considered immature and not interpretable at the time of the analysis. Due to limitations in the statistical analysis, a lack of identified minimally important difference (MID) estimates in fludarabine-ineligible patients with CLL, and the lack of assessment in fludarabine-eligible patients, no definitive conclusions can be drawn from the HRQoL analyses in the GLOW trial. The pivotal study results were subjected to key limitations, such as the exclusion of patients without measurable disease, a major amendment in protocol, and the use of a comparator treatment for fludarabine-ineligible patients that may not be currently relevant in Canada. No new safety signals were identified in either fludarabine-ineligible or fludarabine-eligible patients with CLL.

The supplementary comparative evidence from the ITCs and observational studies submitted by the sponsor had significant limitations that impact the internal validity of the findings. For each of these analyses, there was potential residual confounding from measured and unmeasured confounders, as the statistical methods were unable to control for all sources of heterogeneity in patient and study characteristics. In addition, most analyses were based on a small sample size, and there were low event rates for the OS analysis, with some results showing a wide 95% CI and poor precision. In the fludarabine-ineligible patient population, the comparative efficacy of ibrutinib-venetoclax versus ibrutinib, BR, VO, or acalabrutinib is unclear. The observational data comparing ibrutinib-venetoclax to FCR in fludarabine-eligible patients shared the same limitations, and thus definitive conclusions on comparative efficacy cannot be drawn from this study. In addition, the lack of comparative efficacy for patients with del(17p) were not sufficiently addressed by the supporting evidence.

No comparative safety data were available in the fludarabine-ineligible population and limited data in the fludarabine-eligible population, thus comparative safety of ibrutinib-venetoclax is unknown.

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 ibrutinib (Imbruvica), (140 mg oral capsule), in combination with venetoclax (Venclexta) in the treatment of CLL in previously untreated adult patients, including those with del(17p).

Disease Background

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

CLL is a lymphoproliferative B-cell malignancy characterized by the progressive expansion of monoclonal B lymphocytes in the blood, BM, lymph nodes, or other lymphoid tissue.^1,2 The accumulation of genetic lesions and the interactions of leukemic cells with antigen through the B-cell receptor, as well as the microenvironment, are believed to play a key role in the survival and proliferation of CLL cells.³²

Patients are usually diagnosed with CLL between the ages of 65 and 70 years;⁶ however, more than 10% of patients are diagnosed with CLL when they are younger than 55 years.⁷ The prognosis of patients with CLL is dependent on a variety of patient-related (age, sex, comorbidities, performance status), disease-related (disease stage, cytogenetics, marrow failure, immunodeficiency, lymphomatous transformation, biomarkers), and treatment-related (type of treatment, response, toxicity, MRD status) factors.⁸ Several genetic alterations can influence prognosis, including del(17p), which can result in the loss of TP53 and 1 of the poorest prognostic factors for CLL. Other genetic alterations, including TP53 mutation without del(17p), unmutated IGHV, del(11q), and complex karyotype (i.e., more than 3 cytogenetic aberrations) are associated with a poor prognosis in CLL.

For many patients with CLL, the disease burden is increased by the presence of major comorbidities and frailty, as well as by toxicities associated with standard-of-care chemotherapy-based regimens. Patients with CLL rarely experience symptoms at the time of diagnosis, but symptoms that may arise include lymphadenopathy, weight loss, recurrent disease-related infections, anemia, night sweats, and bleeding.^33-35 Patients with CLL experiencing symptoms can have a substantial impact on their QoL. Untreated patients with CLL have been shown to have a reduced QoL compared to healthy individuals.³⁶ For patients receiving treatment, conventional therapy is thought to compound this impact on QoL, as these patients report poorer QoL compared with those receiving no therapy or healthy individuals.^37-39 CLL is generally considered incurable. The 5-year net survival rate for CLL is 83% ; in 2020, 554 people died from CLL in Canada.⁵ Median life expectancy for patients with del(17p) or TP53 mutation is less than 2 to 3 years from the time of initial diagnosis.⁹

CLL is a rare disease with low prevalence and incidence worldwide.^3,4 However, it is the most common adult leukemia in Canada.⁵ In 2018, 1,725 patients were diagnosed with CLL (1,095 men and 630 women).⁵ Disease prevalence by region is summarized in Table 4.

Table 4: Estimated Prevalence of CLL in Canada

CLL = chronic lymphocytic leukemia; NR = not reported.

^aLower and upper estimates have not been reported by the sponsor.

^bPrevalence of CLL in the Non-Insured Health Benefits population is assumed to be the same as the pan-Canadian prevalence.

Source: Statistics Canada.⁴⁰

To distinguish a diagnosis of CLL from other lymphoproliferative disease, a blood smear, immunophenotype, and sometimes the genetics of a patient’s circulating lymphoid cells must be evaluated.² Diagnosis of CLL requires presence of 5 × 10⁹/L or more B lymphocytes in PB for at least 3 months, with clonality confirmed with immunoglobulin light chain restriction using flow cytometry. Regarding immunophenotype, a panel of CD19, CD 5, CD 20, CD 23, kappa, and lambda is usually sufficient to establish the diagnosis. Molecular genetics may help predict prognosis.

Standards of Therapy

The content within this section has been informed by materials submitted by the sponsor and clinical expert input. The following has been summarized and validated by the CADTH review team.

Because CLL is incurable, the goal of treatment is to achieve effective and durable disease control, with minimal toxicity and an acceptable HRQoL. The clinical expert consulted by CADTH added that the most important goal is to reverse symptoms and control the disease for as long as possible with treatments that have minimal toxicity and no negative impact on QoL. Also, achieving undetectable MRD is also considered a goal, according to the clinician consulted by CADTH.

There are a number of treatment options in the first-line therapy setting based on individual patient and disease characteristics, such as fitness for fludarabine-based therapy and/or risk status defined by IGHV mutation, del(17p) or TP53 mutation, and age.¹⁰ Eligibility for fludarabine-based therapy is also based on patient fitness; patients who are fit have a CIRS score of less than 6, an ECOG PS of 0 to 2, an absence of cardiac and renal diseases (CrCl equal or greater than 60 mL/min), favourable cytogenic and/or mutational status, or have clinical rationale from their physician (Figure 1).

Among patients deemed ineligible for FCR due to age (adults older than 65 years), a lack of overall fitness, cytogenetic and/or mutational status, or other clinical rationale from their physicians, treatment options must account for toxicity and infectious complications. In these older patients or in patients with comorbidities without del(17p) and/or TP53 mutation, treatment with chemoimmunotherapy, such as chlorambucil-obinutuzumab, VO, or continuous BTK inhibitors (e.g., ibrutinib or acalabrutinib), may be used.^10,11 Chemoimmunotherapy has been shown to be the least effective treatment and is only recommended in limited scenarios.¹⁰ VO is an option in patients with mutated and unmutated IGHV; however, it has shorter expected PFS compared to continuous BTK inhibitors in those with unmutated IGHV. According to the clinical expert consulted by CADTH, VO has good activity in CLL and could be used for those who prefer FD therapy or for patients with chronic comorbidities (e.g., hypertension, atrial fibrillation, or an indication for blood thinners) that may be exacerbated by BTK inhibitors. However, the clinical expert noted that VO could be less desirable if patients with CLL are at high risk of TLS (i.e., elevated creatinine, nodal bulk, or very high lymphocyte counts).

For patients with high-risk features (i.e., del[17p] and/or TP53 mutation), regardless of age or fitness, BTK inhibitors are the preferred treatment option even if VO is still available to them.^10,11 The clinical expert mentioned that cross-trial comparisons of the SEQUOIA, ELEVATE-TN, and CLL 14 trials suggest longer PFS with BTK inhibitors for patients with mutated TP53. According to the clinical expert, ibrutinib is being used less because other BTK inhibitors may have less toxicity and FD treatments have become an option. For patients with unmutated IGHV regions, ibrutinib, acalabrutinib, or venetoclax and obinutuzumab are options. For patients who are much older and very frail, FD treatment with mild alkylating drugs, such as chlorambucil-obinutuzumab, are often offered to these patients (even though they would likely have better results with BTK inhibitors with access permitted).

Among young (ages 18 to 64) patients without del(17p) or TP53 mutations, FCR is recommended as a first-line treatment for those with mutated IGHV. According to the clinical expert, FCR with chemoimmunotherapy can induce very long remissions and perhaps even cures for younger patients with good CIRS scores without high-risk mutations. For young and fit patients with unmutated IGHV, BTK inhibitors are preferred over FCR. VO is also a treatment option for these patients albeit with less durable remission compared to BTK inhibitors and is not reimbursed publicly.^10,12 Based on clinician input collected by CADTH, continuous BTK inhibitors are most commonly used in Canada for younger patients with higher-risk mutations, such as TP53 mutations, 11q mutations, or unmutated IGHV genes.

Figure 1: Current Treatment Paradigm of CLL

BR = bendamustine-rituximab; BTKI = Bruton’s tyrosine kinase inhibitor; CIT = chemoimmunotherapy; Chl-R = chlorambucil plus rituximab; CLL = chronic lymphocytic leukemia; del(17p) = deletion of 17p; FCR = fludarabine, cyclophosphamide, and rituximab; VO = venetoclax plus obinutuzumab.

^a First-line treatment options are based on the 2021 CADTH provisional funding algorithm and publicly available provincial criteria, supplemented with the 2022 update to the Canadian evidence-based guideline for front-line treatment of CLL.

^b Relapsed or refractory CLL treatment options are based on the 2021 CADTH provisional funding algorithm and publicly available provincial criteria. Canadian evidence-based guideline in the relapsed or refractory setting is currently in development.

^c Idelalisib-rituximab is available only in cases of intolerance to a BTK inhibitor or for bridging to cellular therapy.

^d Venetoclax monotherapy is only funded after failure of a BTK inhibitor.

^e Zanubrutinib for the treatment of relapsed or refractory CLL is currently under CADTH review. Sequencing of zanubrutinib in the relapsed or refractory setting is not specified, as it is a pre–Notice of Compliance submission to CADTH.

Source: CADTH provisional funding algorithm for CLL.^10,12

Drug Under Review

Key characteristics of ibrutinib in combination with venetoclax are summarized in Table 5, along with other treatments available for CLL.

For treatment of untreated CLL with an ibrutinib-venetoclax combination, the recommended dosing schedule is ibrutinib 420 mg (three 140 mg capsules) administered orally once daily for 3 28-day cycles, followed by ibrutinib 420 mg plus venetoclax 400 mg administered orally daily for 12 28-day cycles. Venetoclax dosing should be initiated in cycle 4, with dose ramp-up over 5 weeks (20 mg, 50 mg, 100 mg, 200 mg, and 400 mg daily), and continued at 400 mg daily from cycle 5 onward.

Ibrutinib monotherapy is indicated for CLL and SLL, mantle cell lymphoma, and Waldenström macroglobulinemia, all of which have been reviewed by CADTH.^13-15 Venetoclax monotherapy is indicated for CLL and acute myeloid leukemia, both of which have been reviewed by CADTH.^16-18 Venetoclax in combination with obinutuzumab is indicated for CLL and venetoclax in combination with rituximab is indicated for CLL, both of which have been reviewed by CADTH.^19,20

B-cell receptor signalling is a key mechanism of disease progression in B-cell malignancy, wherein BTK plays a role in the signalling cascade.^41,42 Ibrutinib is an oral, first-in-class, targeted BTK inhibitor.⁴³ Specifically, the target of ibrutinib and its active metabolite, PCI-45227, is a cysteine residue located on site 481 within the adenosine triphosphate binding domain of BTK. Both molecules bind covalently and irreversibly to this residue, providing potent and sustained inhibition of BTK enzymatic activity.^41,42 Venetoclax is an oral inhibitor of B-cell lymphoma 2.⁴⁴ Given their distinct and complementary mechanism of action, ibrutinib and venetoclax work synergistically to eradicate CLL by eliminating both dividing and resting leukemic subpopulations.⁴⁴ Ibrutinib effectively inhibits tumour cell proliferation while mobilizing leukemic cells from protective lymphoid niches.⁴⁴ Further, ibrutinib increases the sensitivity of CLL cells to B-cell lymphoma 2 inhibition, thereby accelerating apoptotic cell killing by venetoclax.⁴⁵

The requested indication for reimbursement is ibrutinib in combination with venetoclax for the treatment of adult patients with previously untreated CLL, including those with del(17p). The requested indication is the same as the Health Canada–approved indication (post-NOC). Ibrutinib in combination with venetoclax is indicated for the treatment of CLL in Europe.²¹ The FDA has not approved the combination of ibrutinib and venetoclax as a treatment for any disease.²²

Table 5: Key Characteristics of Comparator Regimens for Ibrutinib-Venetoclax in Previously Untreated Patients With CLL

BCL-2 = B-cell lymphoma; BR = bendamustine plus rituximab; BTK = Bruton tyrosine kinase; CLL = chronic lymphocytic leukemia; del(17p) = deletion of 17p; FCR = fludarabine, cyclophosphamide, and rituximab; IgG1 = immunoglobulin G, subclass 1; PML = progressive multifocal leukoencephalopathy; PRES = posterior reversible leukoencephalopathy syndrome; SJS = Stevens-Johnson syndrome; TEN = toxic epidermal necrolysis; TLS = tumour lysis syndrome; VO = venetoclax plus obinutuzumab.

Sources: Sponsor’s clinical summary,⁴⁶ Cancer Care Ontario – Drug Formulary,⁴⁷ BC Cancer.⁴⁸

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.

LC is a national charity that engages in education, support, advocacy, and research activities for patients and the lymphoma community. LC conducted an online anonymous patient survey between March 22 and May 2, 2023. A total of 87 patients, including 49 patients from Canada, responded to the survey. As most patients with CLL experience no or minor symptoms, many respondents indicated that their daily activities were not strongly impacted by CLL at diagnosis. A total of 64 respondents rated fatigue (47%), high white blood cell counts (26%), and body aches and pains (25%) as having a highly negative impact (3 to 5 out of 5) at diagnosis. Among the 71 respondents who reported psychosocial impacts of their CLL diagnosis, anxiety or worry (61%), stress of diagnosis (59%), and difficulty sleeping (28%) were the most common. The most highly rated negative physical symptoms (3 to 5 out of 5) among 70 respondents included fatigue (44%), body aches and pains (27%), and indigestion, abdominal pain, or bloating (17%). The most negatively rated impacts on QoL among the 87 respondents included anxiety/worry (42%), difficulty sleeping (31%), and stress of diagnosis (28%). When considering a novel CLL treatment, respondents cited living longer (81%), controlling symptoms (75%), extending remission (71%), improving QoL (66%), and reducing side effects (35%) as extremely important.

Of the10 patients with CLL who had specific experience with ibrutinib-venetoclax regimen, 5 patients have been in remission for 2 to 5 years. In 10 patients treated with ibrutinib-venetoclax regimen, there was improvement in high white blood cell counts (80%), enlarged lymph nodes (70%), low platelet and red blood cell counts (60%), and weight loss (30%).

The input highlighted that a time-limited, oral ibrutinib-venetoclax therapy option would be especially beneficial for those living in rural areas and cost-saving for health care system savings. Of note, 24% of respondents reported preference for the FD of treatment. Also, 55% of patients reported that having more treatment options is very important.

Clinician Input

Input From the Clinical Expert Consulted by CADTH

All CADTH review teams include at least 1 clinical specialist with expertise regarding the diagnosis and management of the condition for which the drug is indicated. Clinical experts are a critical part of the review team and are involved in all phases of the review process (e.g., providing guidance on the development of the review protocol, assisting in the critical appraisal of clinical evidence, interpreting the clinical relevance of the results, and providing guidance on the potential place in therapy). The following input was provided by 1 clinical specialist with expertise in the diagnosis and management of adult patients with previously untreated CLL.

Unmet Needs

According to the clinical expert consulted by CADTH, the first-line treatment options for adult patients with CLL depend upon the patient’s age, CIRS score, and the features of the CLL, including IGHV mutational status and cytogenetic profile. For younger patients with a good CIRS score and no high-risk mutations, chemoimmunotherapy with FCR can induce very long remissions, and perhaps even cure the disease. For younger patients with higher-risk mutations, such as TP53 mutations, 11q mutations, or unmutated IGHV, continuous BTK inhibitors are most commonly used in Canada. In addition, FD treatments with VO are now funded and being used for this patient population. For older patients, FD treatments with mild alkylating drugs, such as chlorambucil-obinutuzumab, are occasionally offered. The clinical expert also stated that many of these patients will have better results with less toxicity with BTK inhibitors, which are now accessible in the first-line setting.

The clinical expert consulted by CADTH indicated that the most important goals of treatment for patients with CLL is to reverse symptoms and control the disease for as long as possible with minimal toxicity and no significant negative impact on the QoL. The clinical expert stated that the biggest limitation to current treatments for patients with CLL is that tumour cell resistance usually occurs, and patients stop responding or relapse on therapy. Other limitations include toxicity and drug interactions, the requirement for continuous ongoing treatment, and the fact that there are no curative treatments for patients with CLL. The clinical expert stated that other considerations to be addressed include the opportunity for FD treatment versus continuous treatment and achieving undetectable MRD.

Place in Therapy

The clinical expert consulted by CADTH indicated that the addition of venetoclax to BTK inhibitor therapy is directed at the underlying disease process, as the mechanism of the combination therapy addresses a cell resistance mechanism that is biologically activated in many cancers. The clinical expert stated that this combination therapy could be used in the first-line setting where CLL cells have not acquired resistance to either of the drugs or mechanisms of cytotoxicity used by this combination. The clinical expert also indicated that the combination therapy could possibly be used in recurrent disease and in patients with CLL who have some level of resistance, or at least have been exposed to the drugs in the combination. In addition, the clinical expert speculated that it is possible that this combination therapy could be used in patients who are resistant to 1 of the drugs in the combination, although this would have to be shown in evidence. Overall, the clinical expert stated that ibrutinib-venetoclax could be used in patients who have been exposed but are not necessarily resistant to either of these drugs. This is relevant to venetoclax, which is used as FD therapy, and thus patients will eventually relapse not because of resistance but because the treatment has been stopped.

The clinical expert stated that there is a possible shift in the current treatment paradigm with the combination therapy because the drugs are stopped after a fix perioded, which is different than the paradigm of continuous therapy that is used for the different BTK inhibitors.

Patient Population

The clinical expert consulted by CADTH commented on the inclusion and exclusion criteria of the GLOW trial that older patients or younger patients with a high CIRS score or a reduced CrCl were all included. Although patients with del(17p) or TP53 mutations were excluded, the study did not exclude patients based on other disease characteristics, cytogenetics (except for del[17p] or TP53 mutation), or bulk of disease. The study also included patients with bulky disease, high lymphocyte counts, or impaired CrCl, which are considered at higher risk for TLS, according to the clinical expert. The clinical expert indicated that venetoclax therapy can result in TLS; therefore, ibrutinib was given before the venetoclax to reduce the risk of tumour lysis. The clinical expert mentioned that there were no factors identified in the subgroup analysis indicating that any subset of patients who would be most likely to respond or, conversely, less likely to respond to the combination therapy. All symptomatic patients with CLL need treatment. The clinical expert commented that patients with del(17p)/TP53 mutations were included in the CAPTIVATE trial and similar MRD negativity rates were observed between those patients with and without del(17p)/TP53 mutations; therefore, the presence of del(17p)/TP53 mutations should not be an exclusion criterion for patients with CLL.

According to the clinical expert consulted by CADTH, patients best suited for treatment with the drug under review can be identified by the treating physician, without the need for a companion diagnostic. In addition, the clinical expert stated that it is very unusual to misdiagnose patients with CLL.

Assessing the Response to Treatment

The clinical expert consulted by CADTH indicated that response to treatment is assessed by changes in PB counts, which can easily be documented by clinicians looking after patients. The clinical expert stated that repeat BM biopsies are not often performed in clinical practice but were required as part of the clinical trial formal response criteria. The clinical expert stated that DOR or response to next treatment are important end points used by clinicians to choose appropriate treatments and to inform prognosis. The clinical expert stated that objective responses often correlate with improvements in cytopenia, which may result in decreased transfusion requirements or decrease risk of infection. In addition, MRD assessments were also used from PB and BM using 2 technologies: NGS and multicolour flow cytometry.

Discontinuing Treatment

According to the clinical expert consulted by CADTH, death and disease progression (measured by increasing lymphocyte count or enlarging lymph nodes or spleen) are major reasons for discontinuing the treatment of ibrutinib in combination with venetoclax. The clinical expert indicated that increasing lymphocyte count is usually accompanied by anemia or thrombocytopenia.

Prescribing Considerations

The clinical expert stated that ibrutinib in combination with venetoclax treatment should be managed by a specialist (i.e., hematologist or medical oncologist) who is familiar with this class of drugs to optimally manage toxicities and dosing.

Additional Considerations

The clinical expert mentioned that, with FD treatment, if a patient has a good response (i.e., PFS of 3 years or more), they may be offered the same treatment again, as they are not necessarily resistant. This could be an advantage of the combination therapy over continuous BTK inhibitor treatment, as chronic BTK inhibitor therapy often results in accumulation of side effects over time. In addition, the undetectable MRD rates with the combination therapy are significantly higher than with BTK inhibitor monotherapy, which could indicate deeper responses.

Clinician Group Input

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

OH-CCO Hem DAC provides evidence-based clinical and health system guidance. Seven clinicians from OH-CCO Hem DAC submitted input. In addition, 2 clinicians who treat CLL and SLL in Canada submitted input based on literature and consultation with other clinicians, including those from LC.

Unmet Needs

Both clinician groups noted that time-limited, all-oral combination ibrutinib-venetoclax regimen is the important unmet need. The ibrutinib-venetoclax option would neither require IV access nor continuous treatment. This unmet need would reduce travel to clinics, caregiver cost, chronic side effects, and associated ongoing costs. The 2 clinicians said that ibrutinib-venetoclax can be used for any previously untreated patients with CLL and might be better tolerated than venetoclax-obinutuzumab (e.g., very low risks of TLS). Also, they added that patients should be able to be re-treated with ibrutinib or venetoclax in the second line if they are not refractory to either drug. In contrast, OH-CCO Hem DAC stated that ibrutinib-venetoclax replaces BTK inhibitor monotherapy and practical onboarding of venetoclax-obinutuzumab can be considered.

Place in Therapy

The 2 clinicians stated that ibrutinib-venetoclax will be an alternate first-line option among other funded therapies. In contrast, OH-CCO Hem DAC suggested that ibrutinib-venetoclax would be an option for time-limited therapy for the highest-risk groups (e.g., those with del[17p] or TP53 mutations) and/or would increase accessibility of a time-limited treatment option for those unable to easily access venetoclax-obinutuzumab.

Patient Population

The 2 clinicians mentioned that previously untreated patients are eligible for ibrutinib-venetoclax regimen, with access available for those who already take either drug as the first-line treatment. However, OH-CCO Hem DAC suggested that ibrutinib-venetoclax therapy be limited to patients without cardiovascular comorbidities and likely younger population due to safety concerns regarding sudden deaths reported in a study by Lampson et al. (2017).⁴⁹ According to OH-CCO Hem DAC, molecular testing for genetic mutations is standard of care for CLL and SLL. If jurisdictions do not provide access to molecular testing, genetic testing should be required for TP53, IGVH, and del(17p) mutations to select the eligible population for ibrutinib-venetoclax therapy. OH-CCO Hem DAC also added that the population that may benefit the most are those taking continuous standard-of-care therapies or having access issues.

Assessing Response to Treatment

The 2 groups said that improved PFS and reduced symptoms would be the meaningful response and standard of outcomes for CLL studies. Additionally, the 2 clinicians and OH-CCO Hem DAC noted that time off treatment and QoL are measures to determine responses to treatment, respectively.

Discontinuing Treatment

Both clinician groups emphasized that ibrutinib-venetoclax regimen is time-limited therapy. The 2 groups mentioned that early disease progression while on therapy and significant intolerance are necessary factors to be considered for discontinuation of therapy.

Prescribing Conditions

Both groups agreed that clinicians familiar with the treatment of CLL (e.g., hematologists and medical oncologists) should be prescribers of the ibrutinib-venetoclax combination (as outpatient therapy). The 2 clinicians added that a cardiologist may also be required due to safety concern. Both groups emphasized that additional lab monitoring (e.g., weekly ramp-up intensive lab monitoring with venetoclax) is likely required.

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 expert consulted by CADTH are summarized in Table 6.

Table 6: Summary of Drug Plan Input and Clinical Expert Response

BTK = Bruton tyrosine kinase; CLL = chronic lymphocytic leukemia; DOR = duration of response; MRD = minimal residual disease; pERC = CADTH pan-Canadian Oncology Review Expert Review Committee; PFS = progression-free survival; SLL = small lymphocytic lymphoma.

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 ibrutinib (140 mg capsule, oral) in combination with venetoclax (10 mg, 50 mg, 100 mg tablet, oral) for the treatment of previously untreated adults with CLL, including those with del(17p). The focus was placed on comparing ibrutinib in combination with venetoclax to relevant comparators and identifying gaps in the current evidence.

A summary of the clinical evidence included by the sponsor in the review of ibrutinib in combination with venetoclax 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 randomized controlled trials (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:

• 2 pivotal studies: 1 phase III, open-label, multicentre RCT (GLOW) and a phase II, multicohort, international trial (CAPTIVATE)

• 5 nonrandomized comparisons: 2 ITCs using MAICs and 3 IPD observational studies.

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

Two pivotal trials (GLOW and CAPTIVATE) met in the inclusion criteria for the systematic review conducted by the sponsor. Their characteristics are summarized in Table 7.

GLOW Study (Fludarabine-Ineligible Population)

The objective of the GLOW study was to evaluate FD ibrutinib-venetoclax versus the standard chemoimmunotherapy combination of chlorambucil-obinutuzumab in older patients and/or those with comorbidities with previously untreated CLL. In this phase III, open-label trial, treatment was randomized and administered through a parallel assignment model. Patients were centrally randomized in a 1:1 ratio to either treatment arm, balanced by using permuted blocks and stratified by IGHV mutational status (mutated versus unmutated versus not available) and the presence of del(11q) (yes versus no). The trial enrolled patients aged 65 years or older and patients 18 to 64 years with a CIRS score greater than 6 and/or a CrCl less than 70 mL/min. All patients had active CLL or SLL requiring treatment, per iwCLL criteria. Patients were excluded if they had del(17p) or known TP53 mutations, bleeding disorders, central nervous system involvement, Richter syndrome, or uncontrolled autoimmune hemolytic anemia or thrombocytopenia. Between May 2018 and April 2019, 211 patients were enrolled and randomly assigned to treatment with either ibrutinib-venetoclax (n = 106) or chlorambucil-obinutuzumab (n = 105). The trial was conducted at 86 sites in 14 countries, including 2 sites in Canada. Patients in the ibrutinib-venetoclax arm received 3 cycles of ibrutinib lead-in at 420 mg once daily followed by 12 cycles of ibrutinib-venetoclax, with a venetoclax ramp-up from 20 to 400 mg/day over 5 weeks and 400 mg/day from cycle 5 onward. Patients in the chlorambucil-obinutuzumab arm received 6 cycles of obinutuzumab. In cycle 1, patients received 100 mg of obinutuzumab on day 1 and 900 mg on day 2, followed by 1,000 mg on days 8 and 15. In cycles 2 through 6, patients received 1,000 mg of obinutuzumab on days 1, 8, and 15. Patients received chlorambucil per the label (0.5 mg/kg body weight on days 1 and 15 of each cycle). A diagrammatic representation of the GLOW study design is presented in Figure 2.

Figure 2: Schema for the GLOW Study

Source: GLOW primary analysis CSR.²³

CAPTIVATE Study (Fludarabine-Eligible Population)

CAPTIVATE was a 2-cohort phase II study assessing both MRD-guided discontinuation (MRD cohort) and FD therapy (FD cohort) with the combination of ibrutinib-venetoclax in patients with treatment-naive CLL or SLL. The FD cohort was enrolled sequentially after the MRD cohort and is an open-label, single-arm cohort. For this review, only the FD cohort is discussed. The MRD cohort is not further discussed because the sponsor is not proceeding with this treatment regimen. The objective of the CAPTIVATE FD cohort was to evaluate the depth of response with the combination of ibrutinib-venetoclax administered for a FD of therapy by assessment of CR (CR and CR with incomplete BM recovery [CRi]) rate. Eligible patients for the CAPTIVATE FD cohort were aged 18 to 70 years with previously untreated CLL or SLL requiring treatment, per iwCLL criteria, and had measurable nodal disease by CT, an ECOG PS of 0 to 2, and adequate hepatic, renal, and hematologic function. Patients with known allergy to xanthine oxidase inhibitors and/or rasburicase were excluded because of the requirement for TLS prophylaxis per venetoclax prescribing information. At the start of the study, 159 patients enrolled in the FD cohort received 420 mg ibrutinib daily for 3 28-day cycles, followed by 12 cycles (28 days) of ibrutinib-venetoclax, with a target dose of 400 mg once daily after a 5-week ramp-up period, for a total of 15 cycles of open-label treatment.

Table 7: Details of Pivotal Studies and RCT Evidence Identified by the Sponsor

AE = adverse event; ANC = absolute neutrophil count; CIRS = Cumulative Illness Rating Scale; CLL = chronic lymphocytic leukemia; CNS = central nervous system; CR = complete response; CrCl = creatinine clearance; CRi = complete response with incomplete bone marrow recovery; del(17p) = deletion of 17p; DOR = duration of response; ECOG PS = Eastern Cooperativity Oncology Score Performance Status; EORTC QLQ-C30 = European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 30; EQ-5D-5L = 5-Level EQ-5D; FACIT = Functional Assessment of Chronic Illness Therapy; FD = fixed duration; Hgb = hemoglobin; IRC = independent review committee; iwCLL = International Workshop on CLL; MRD = minimal residual disease; NA = not applicable; NR = not reported; ORR = overall response rate; OS = overall survival; PFS = progression-free survival; RCT = randomized controlled trial; SD = standard deviation; SLL = small lymphocytic leukemia; TLS = tumour lysis syndrome; TTNT = time to next treatment.

Note: One additional report was included: Health Canada report.⁵¹

Sources: GLOW primary analysis CSR;²³ CAPTIVATE CSR.²⁶ (Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.)

Populations

Inclusion and Exclusion Criteria

Both the GLOW and CAPTIVATE (FD cohort) studies required patients to be treatment naive and to be 18 years or older, with GLOW requiring patients younger than 65 years to have a CIRS score greater than 6 and/or a CrCl of less than 70 mL/min. Both studies had an inclusion criterion of an ECOG PS of 0 to 2. The GLOW study excluded patients with the presence of del(17p) and/or known TP53 mutations detected at a variable allele frequency of more than 10%, whereas the CAPTIVATE study allowed patients with del(17p) and/or TP53 mutations to enrol.

Interventions

Patients enrolled in the GLOW trial were randomized in a 1:1 ratio to treatment with either ibrutinib-venetoclax or chlorambucil-obinutuzumab, balanced by using permuted blocks and stratified by IGHV mutational status (mutated versus unmutated versus not available) and the presence or absence of del(11q). For patients in the ibrutinib-venetoclax treatment arm, ibrutinib was initiated with a lead-in phase consisting of 420 mg taken orally once daily for 3 cycles (28 days per cycle), then continued for 12 additional cycles in combination with venetoclax. Venetoclax was initiated in cycle 4 with a 5-week dose ramp-up phase in the sequence of 20 mg, 50 mg, 100 mg, 200 mg, then 400 mg daily administered orally, followed by continuous 400 mg per day treatment in combination with ibrutinib until the end of treatment. For the chlorambucil-obinutuzumab arm, patients received 6 cycles (28 days per cycle) of 1,000 mg daily of IV obinutuzumab on days 1, 8, and 15 of cycle 1, then day 1 for cycles 2 through 6. Chlorambucil 0.5 mg/kg on was given on day 1 and day 15 of each cycle. The first administration of obinutuzumab could be split into 100 mg on day 1 and 900 mg on day 2 for cycle 1 only. Patients assessed by IRC as having confirmed disease progression were eligible to receive subsequent therapy with single-drug ibrutinib, administered at 420 mg orally once daily until further disease progression or unacceptable toxicity.

Patients enrolled in the CAPTIVATE FD cohort received ibrutinib-venetoclax for 15 cycles of 28 days each, or until confirmed disease progression. Ibrutinib was initiated with a lead-in phase consisting of 420 mg taken orally once daily (three 140 mg capsules) for 3 cycles then continued for 12 cycles in combination with venetoclax. Venetoclax was given starting in cycle 4 with a target of 400 mg once daily taken orally after a standard 5-week ramp-up of 20 mg, 50 mg, 100 mg, 200 mg, and 400 mg doses.

Outcomes

A list of efficacy end points assessed in this report is provided in Table 8, and summarized subsequently. The summarized end points are based on those included in the sponsor’s Summary of Clinical Evidence as well as any identified as important to this review according to stakeholders, such as the clinical expert, clinician groups, or patient groups.

Table 8: Outcomes Summarized From Pivotal Studies Identified by the Sponsor

CR = complete response; DOR = duration of response; EORTC QLQ-C30 = European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 30; EQ-5D-5L = 5-Level EQ-5D; FACIT = Functional Assessment of Chronic Illness Therapy; FD = fixed duration; MRD = minimal residual disease; NR = not reported; ORR = overall response rate; PD = progressive disease; PFS = progression-free survival; OS = overall survival; TLS = tumour lysis syndrome; TTNT = time to next treatment.

^aFor the GLOW trial, the primary analysis data cut-off date was February 26, 2021, and the extended follow-up analysis data cut-off date was August 25, 2022.

^bFor the CAPTIVATE FD cohort, the primary analysis data cut-off date was November 12, 2020, and the extended follow-up analysis data cut-off date was October 4, 2021.

^cStatistical testing for these end points was adjusted for multiple comparisons (e.g., hierarchal testing).

Sources: GLOW primary analysis CSR;²³ CAPTIVATE CSR.²⁶ (Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.)

Progression-Free Survival

In the GLOW trial, PFS was the primary end point. PFS was defined as the time between the date of randomization and the date of progressive disease (PD) (as assessed by the IRC), or date of death due to any cause, whichever occurred first, regardless of the use of subsequent antileukemic therapy or treatment discontinuation before documented PD or death. Patients who were progression-free and alive were censored at the date of last disease assessment.

In the CAPTIVATE FD cohort, PFS was the primary end point. The definition of PFS was the time from the date of first study treatment to the date of disease progression, per IA, or date of death from any cause, whichever occurs first, regardless of the use of subsequent antineoplastic therapy before documented PD or death. For patients without documented PD or death, PFS was censored at the date of last adequate disease assessment.

Overall Survival

In both the GLOW and CAPTIVATE FD cohorts, OS was a secondary end point. In the GLOW study, OS was defined as the time from the date of randomization to the date of death from any cause, whereas in the CAPTIVATE FD cohort, OS was defined as the time from the date of first dose until the date of death due to any cause. If patients were alive or of unknown status by the end of study, their survival end point was censored at the date the patient was last known to be alive. In the GLOW trial, death due to COVID-19 was considered a potential competing risk to disease-related OS, and a composite variable strategy was adopted (i.e., considering death due to COVID-19 as an OS event) in the primary analysis of OS. If the number of patient deaths due to COVID-19 was more than 5% of total OS events, then a hypothetical outcomes strategy was adopted (censoring at the date of death due to COVID-19) as a supplementary analysis.

Overall Response Rate

ORR was a secondary end point in both the GLOW and CAPTIVATE trials. ORR was defined as the proportion of patients achieving a best overall response of either CR, CRi, nodular partial response (nPR), or PR, per iwCLL criteria, as evaluated by IRC or IA on or before initiation of subsequent antileukemic therapy (including subsequent single-drug ibrutinib). Disease response was assessed by investigators using CT or MRI as well as by IRC in the GLOW trial. Patients with missing postrandomization data were considered nonresponders.

CR Rate

In the GLOW study, CR rate was assessed as a secondary end point. CR rate was defined as the proportion of patients who achieved CR or CRi on or before initiation of subsequent antileukemic therapy (including subsequent single-drug ibrutinib).

In the CAPTIVATE FD cohort, CR rate for patients without del(17p), per IA, was the primary end point. CR rate was defined as the proportion of patients achieving a best overall response of CR or CRi on or before initiation of subsequent antineoplastic therapy or, if applicable, reintroduction of study treatment, whichever occurred first.

Improvement in Hematologic Parameters

Improvement in hematologic parameters was assessed as a secondary end point in the GLOW study and as an exploratory end point in the CAPTIVATE FD cohort. In both trials, sustained hemoglobin improvement was defined as the percentage of patients who achieved an increase in hemoglobin levels from baseline by 2 g/dL or more that lasted at least 56 days without blood transfusion or growth factors. In both trials, sustained platelet improvement was defined as the percentage of patients who achieved an increase in platelet levels from baseline by 50% or more that lasted for at least 56 days without blood transfusion or growth factors.

Duration of Response

In both trials, DOR was a secondary end point and was defined as the time elapsed between the date of initial documentation of a response, including PR with lymphocytosis, and the date of the first documented evidence of PD or death. Censoring of DOR data followed the same rule as for PFS. In the GLOW study, DOR was per IRC assessment, whereas in the CAPTIVATE FD cohort, it was per IA.

Time to Next Treatment

In the GLOW study, TTNT was assessed as a secondary end point. TTNT was measured from the date of randomization to the start date of any subsequent anticancer therapy. Patients without subsequent therapy were censored at the last known date to not have received subsequent therapy. TTNT was not assessed in the CAPTIVATE FD cohort.

MRD Negativity Rate

In both trials, the MRD negativity rate was assessed as a secondary end point and was defined as patients who achieved MRD negativity (< 1 CLL cell/10⁴ leukocytes) in BM and PB. NGS was used as the primary method of MRD analysis, and flow cytometry was used for supplementary analyses in the GLOW study. In the CAPTIVATE FD cohort, only the flow cytometry method was used for the MRD analysis.

TLS Risk Reduction

TLS risk reduction was a secondary end point in both the GLOW and CAPTIVATE trials. In the GLOW trial, TLS risk reduction was defined as the proportion of patients whose TLS risk was reduced from high at baseline to medium or low after ibrutinib lead-in treatment; thus, it was only assessed for the ibrutinib-venetoclax arm. The CAPTIVATE FD cohort compared the proportion of patients who were at high risk of TLS after ibrutinib lead-in phase with the proportion at baseline. In both trials, TLS risk categories were defined per the venetoclax US package insert based on lymph node size and absolute lymphocyte count tumour burden measurements: Low risk was defined as all lymph nodes smaller than 5 cm and absolute lymphocyte counts less than 25 × 10⁹/L; medium risk was defined as any lymph node 5 cm to less than 10 cm or absolute lymphocyte count 25 × 10⁹/L or greater; and high risk was defined as any lymph node of at least 10 cm or absolute lymphocyte count 25 × 10⁹/L or greater and any lymph node 5 cm or greater.

Health-Related Quality of Life

The GLOW trial also explored HRQoL by assessing time to worsening using the EQ-5D-5L, EORTC QLQ-C30, and FACIT-Fatigue instruments. The collection of patient-reported outcome assessments was to be stopped after the primary PFS analysis was completed.

EORTC QLQ-C30

EORTC QLQ-C30 includes 30 separate items resulting in 5 functional scales (physical functioning, role functioning, emotional functioning, cognitive functioning, and social functioning), 1 global health status scale, 3 symptom scales (fatigue, nausea and vomiting, and pain), and 6 single items (dyspnea, insomnia, appetite loss, constipation, diarrhea, and financial difficulties). Scores were derived using validated scoring algorithms, according to EORTC QLQ-C30 scoring manual. Scores range from 0 to 100; for functional and global QoL scales, higher scores indicate a better level of functioning. EORTC QLQ-C30 improvement or worsening is defined as a change of 10 points or more at each postbaseline assessment.⁵²

EQ-5D-5L

The EQ-5D-5L consists of a 5-item descriptive system and the EQ-5D VAS of self-rated health, with scores ranging from 0 (worst imaginable health state) to 100 (best imaginable health state). Responses for the 5 dimensions are combined into a 5-digit number describing a respondents’ health state that can be converted into a single index value or utility score (using the UK weights), ranging from –1 to 1, where lower scores indicate a worse health status. A minimum difference of 0.07 points or more change in utility score is considered clinically important; for the VAS health rating, MID is a change of 7 points or more.⁵³

FACIT-Fatigue

The FACIT-Fatigue Scale measures fatigue severity and its impact on daily activities. It includes 13 items that assess tiredness, weakness, and difficulty conducting usual activities due to fatigue. Scores range from 0 to 52, with high scores indicating less fatigue. FACIT improvement or worsening is defined as a change of 3 points or more at each postbaseline assessment.⁵⁴

Harms

In the GLOW and CAPTIVATE trials, the proportion of patients who experienced AEs up until 30 days after the last dose of study treatment had their experiences recorded and reported using the National Cancer Institute Common Terminology Criteria for Adverse Events (version 4.03.26) coded with Medical Dictionary for Regulatory Activities (version 23.0). Specifically, hematological AEs were graded using iwCLL criteria, nonhematological AEs were graded using National Cancer Institute Common Terminology Criteria for Adverse Events (version 4.03.26), and TLS was assessed using Howard criteria.⁵⁵ In both the GLOW and CAPTIVATE trials, AEs of special interest included atrial fibrillation and major hemorrhages.

Table 9: Summary of HRQoL Outcome Measures and Their Measurement Properties

CLL = chronic lymphocytic leukemia; EORTC QLQ-C30 = European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 30; EQ-5D-5L = 5-Level EQ-5D; FACIT = Functional Assessment of Chronic Illness Therapy; ICC = intraclass correlation coefficient; MID = minimally important difference; QoL = quality of life; SD = standard deviation; SEM = standard error of the mean; VAS = visual analogue scale.

Statistical Analysis

Sample Size and Power Calculation

A summary of statistical analysis of efficacy end points for the GLOW and CAPTIVATE trials is provided in Table 10.

With a 1:1 randomization, the GLOW study planned to enrol 200 patients (100 patients into ibrutinib-venetoclax and 100 patients into the chlorambucil-obinutuzumab treatment arm), expecting to observe 71 PFS events under an assumed alternative of HR = 0.5 and baseline exponential. Under the assumed alternative, the sample size would provide 80% power of rejecting a null of no difference in PFS using a log-rank test. Actual enrolment was 211 patients, with 106 and 105 randomly assigned to the ibrutinib-venetoclax and chlorambucil-obinutuzumab treatment arms, respectively.

The CAPTIVATE trial conducted separate sample size and power calculations for the FD cohorts. Sample sizes were estimated based on the primary end point of CR rate in patients without the del(17p) mutation (irrespective of TP53 mutation). A sample size of 125 patients was determined to be sufficient to exclude a minimum CR rate of 37%, with a 1-sided significance level of 0.025, with 83% power, assuming an alternative CR probability for ibrutinib-venetoclax of 50%.⁵⁹

The choice of a minimum CR rate of 37% was based on the results of the CLL10 study, which evaluated first-line chemoimmunotherapy with BR versus FCR in patients with advanced CLL excluding patients with del(17p). In this study, a CR rate of 31% was observed in the BR arm, whereas treatment with FCR resulted in a CR rate of 40%.⁵⁹ Based on that, a minimum CR rate of 37% was determined to ensure that the lower acceptable CR boundary for ibrutinib-venetoclax is comparable to the CR rate observed with FCR and BR in the CLL10 study, the standard of care for patients without del(17p) at the time the CAPTIVATE trial was designed.

Statistical Test or Model

In the GLOW study, all interval estimations were reported using 2-sided 95% CIs. No formal interim analysis was planned for efficacy outcomes due to the small sample size and short accrual period. Comparison tests across treatment groups were conducted using the Cochran-Mantel-Haenszel (CMH) chi-square test for discrete variables and a 2-sided log-rank test for time-to-event variables. The PFS analysis was performed on the intention-to-treat (ITT) population, defined as all randomized patients. The Kaplan-Meier method was used to estimate the distribution of PFS for each treatment arm. A difference in treatment effect of ibrutinib-venetoclax compared with chlorambucil-obinutuzumab on PFS was tested using a stratified log-rank test. The HR for ibrutinib-venetoclax relative to chlorambucil-obinutuzumab and its corresponding 95% CI were calculated using a stratified Cox regression model. In both cases, the stratification variables were those used for the block randomization: IGHV mutational status (mutated versus unmutated versus not available) and del(11q) (yes versus no). The GLOW study used a nonstratified Cox regression model to analyze the treatment effect on PFS as a sensitivity analysis. The model adjusted for select prespecified baseline factors as explanatory variables. The baseline factors are del(11q) (no, yes), IGHV (mutated, unmutated, not available), age group (< 65 years, > 65 years), sex (female, male), Rai stage (0 to II, III to IV), ECOG PS (0, 1 to 2), CIRS total score (< 6, > 6), elevated lactate dehydrogenase (LDH) (no, yes), bulky disease (< 5 cm, > 5 cm), cytopenia (no, yes), and serum beta-2 microglobulin (beta-2M) (< 3.5mg/L, > 3.5 mg/L).

In the CAPTIVATE trial, the estimate of CR rate per IA and the corresponding 95% CI based on normal approximation to the binomial distribution were provided. In the FD cohort non-del(17p) population, the P value for testing the CR rate (≤ 37% versus > 37%) was calculated using an asymptotic test for the binomial proportion. Planned analyses included a primary analysis and a final analysis. The primary analysis for the FD cohort was conducted after the last patient enrolled had the opportunity to be followed for at least 30 cycles (15 cycles of treatment plus 15 cycles of posttreatment follow-up), and all safety and efficacy outcomes were analyzed. The Kaplan-Meier method was used to estimate the median and the distribution of time-to-event variables (e.g., PFS, OS). The primary analysis was reported in the included publications for the FD cohort; as indicated in the statical analysis plan, a final analysis will be conducted when the study closes.

Multiple Testing Procedure

In the GLOW study, a serial gatekeeping procedure was used to account for multiplicity incurred from testing primary and secondary end points. Each hypothesis for secondary outcomes was only tested if the null hypotheses for the primary outcome and all preceding secondary outcomes were rejected. The secondary efficacy end points were tested in a hierarchal manner, at the nominal 0.05 significance level (2-sided), in the following order:

1. MRD negativity rate by NGS in BM

2. CR rate (per IRC)

3. ORR (ORR per IRC followed by ORR per IA)

4. OS

5. Rate of sustained platelet improvement

6. Rate of sustained hemoglobin improvement

7. Time to improvement in FACIT-Fatigue score.

In the CAPTIVATE FD cohort, hypothesis testing was performed independently (without multiplicity adjustment) for the primary end point of CR rate, which was tested in non-del(17) patients at a 1-sided alpha level of 0.025. Other end points were summarized descriptively, with 95% CIs whenever applicable.

Data Imputation Methods

There were no data imputation methods reported for the GLOW and CAPTIVATE trials.

Subgroup Analyses

The GLOW trial conducted predefined subgroup analyses for PFS per IRC for the following select demographic and baseline variables: age (< 65 years, ≥ 65 years), sex (male, female), race (white, nonwhite), diagnosis (CLL, SLL), Rai stage at screening for CLL patients only (0 to II, II to IV), Binet stage at screening for CLL patients only (A, B, C), ECOG PS (0, 1 to 2), CIRS total score (≤ 6, > 6), bulky disease (yes [longest diameter ≥ 5 cm], no [longest diameter < 5 cm]), IGHV mutation status (mutated, unmutated, not available), del(11q) mutation (yes, no), high-risk population (yes, no; defined as yes if patients presented with TP53 mutation or del[11q] or unmutated IGHV status at baseline), elevated LDH at baseline (yes [> upper limit of normal], no [≤ upper limit of normal]), cytopenia at baseline (yes, no; defined as yes if platelet count ≤ 100,000/μL, or Hgb ≤ 11 g/dL, or absolute neutrophil count ≤ 1,500/μL is observed), and serum beta-2M (≤ 3.5 mg/L, > 3.5 mg/L). The nonstratified log-rank test and Cox regression model were used for each of the predefined subgroup analyses to investigate whether the treatment effects on PFS persisted across populations. Estimated HRs across the 2 treatment arms within each subgroup and their 95% CIs were also provided.

The CAPTIVATE trial conducted predefined subgroup analyses for outcomes of CR rate, MRD negativity rate, and AEs for select variables. For CR rate and MRD negativity rate, select variables for subgroups included the following: age (< 65 years, ≥ 65 years), sex (male, female), race (white, nonwhite), ECOG PS (0, 1 to 2), Rai stage (0 to II, III to IV), bulky disease (< 5 cm, ≥ 5 cm), del(17p) mutated (yes, no), del(17p) or TP53 mutated (yes, no), fluorescence in situ hybridization (del[17p], del[11q], and others), and IGHV per central lab (unmutated, mutated). For AEs, select variables were age, sex, and race stratification, along with CrCl rate (< 60 mL/min, ≥ 60 mL/min) and National Cancer Institute Organ Dysfunction Working Group liver function classification (normal, abnormal).

Sensitivity Analyses

In the GLOW study, sensitivity and supplementary analyses were performed for PFS, as follows:

1. Nonstratified log-rank test and nonstratified Cox regression model were conducted for PFS based on IRC assessment in the ITT population.

2. Use of subsequent antileukemic therapy before documented PD or death: Patients were censored at the last disease assessment if there was no evidence of PD before the use of subsequent antileukemic therapy.

3. Disease assessment follow-up: Patients were censored at the last disease evaluation if they progressed or died after missing at least 2 consecutive planned disease evaluation visits.

4. PFS based on PD assessed by investigators: The same censoring rules and analysis methods used for primary analysis of IRC-reported PFS were used. The concordance rate between the IRC-assessed PD and investigator-assessed PD was evaluated. The number and percentage of PD events and non-PD cases assessed by investigator and by IRC were cross-tabulated.

The sponsor also planned to conduct a supplementary analysis censoring patients at the last disease assessment before pre-PD death due to COVID-19. However, the sponsor did not conduct this supplementary analysis, as patients with pre-PD death due to COVID-19 was less than the prespecified threshold of 5% of total PFS events.

In the CAPTIVATE trial, a sensitivity analysis for the primary end point, CR rate, was conducted, per IRC assessment. A supportive analysis included duration of complete response (duration of CR was defined as the interval between the date of initial CR or CRi until PD or death from any cause, whichever occurred first) for patients who achieved a CR or CRi, and durable CR rate (defined as the proportion of patients with a duration of CR of ≥ 336 days [12 cycles]). For discrete events (e.g., ORR), CIs were estimated using the normal approximation of the binomial distribution.

Secondary Outcomes of the Studies

Progression-Free Survival

In the CAPTIVATE trial, PFS was a secondary end point, and the PFS distribution was estimated using the Kaplan-Meier method; median PFS and 24-month and 36-month survival probability estimates with 2-sided 95% CIs were provided.

Overall Survival

In the GLOW study, OS was analyzed using similar analysis methods as used for the primary analysis of PFS. The sponsor also performed a supplementary analysis censoring patients at the date of death due to COVID-19.

In the CAPTIVATE trial, Kaplan-Meier estimates of median OS time and landmark time (i.e., 6-month, 12-month, 18-month, 24-month, 30-month, and 36-month) estimates with 2-sided 95% CIs were provided.

Overall Response Rate

ORR was estimated according to the crude proportion of confirmed responders (CR, CRi, nPR, or PR), based on the best overall response, and was summarized by treatment arm in the GLOW study. ORR was compared between treatment arms using the CMH chi-square test, adjusted for the randomization stratification factors. The relative response and its associated 95% CI were estimated according to CMH. Both ORR based on IRC and IA were analyzed based on the same analysis methods.

In the CAPTIVATE trial, the point estimate of ORR (per IRC and per IA) and the corresponding 95% CI based on normal approximation to the binomial distribution were provided.

Improvement in Hematologic Parameters

In both the GLOW and CAPTIVATE trials, the proportion of patients achieving sustained hemoglobin and platelet improvement were summarized by treatment arm, respectively. In the GLOW trial, these proportions were compared using the CMH chi-square test.

Duration of Response

DOR was analyzed using the same method as used for PFS in the GLOW study. Only patients who achieved a PR or better were included in the analysis of DOR. DOR was summarized descriptively using the Kaplan-Meier method, and no inferential comparison was made between treatment arms.

In the CAPTIVATE trial, Kaplan-Meier estimates of median DOR time (per IRC and per IA) and 6-month,12-month,18-month, 24-month, 30-month, and 36-month landmark estimates with 95% CIs were provided.

Time to Next Treatment

TTNT was analyzed using the same methods to those used for PFS outcome in the GLOW study. The number of patients who received subsequent antileukemic therapy were summarized by therapy type. TTNT was not assess in the CAPTIVATE trial.

MRD Negativity Rate

In the GLOW study, all randomized patients were included in this analysis; patients with missing MRD data were considered to be MRD positive. The overall MRD negativity rate in the BM, assessed by NGS, was the primary MRD analysis used for hierarchical testing. The MRD negativity rate by NGS in the PB was considered as the supportive analysis for this end point. The MRD negativity rate among patients who achieved a best overall response of CR or CRi per IRC assessment was conducted as a supplementary analysis. The CMH chi-square test adjusted for the randomization stratification factors was used to test for differences in MRD negativity rates across treatment groups.

In the CAPTIVATE trial, the point estimate of the MRD negativity rate and the corresponding 95% CI based on normal approximation to the binomial distribution were provided.

TLS Risk Reduction

In the GLOW study, descriptive statistics were provided. The proportion of patients with hospitalization indicated due to TLS risk (i.e., patients with TLS risk based on high tumour burden, patients with TLS risk based on medium tumour burden and CrCl < 80 mL/min) was also summarized at baseline and after the ibrutinib lead-in.

In the CAPTIVATE trial, the estimated difference in proportions of patients with high risk of TLS (at baseline after ibrutinib lead-in) and the corresponding 95% CI (adjusting correlation of paired data) were provided as primary analysis. Reduction of the proportion of patients with hospitalization indicated due to TLS risk (i.e., patients with high TLS risk and patients with medium risk of TLS and CrCl < 80 mL/min) were provided as a supportive analysis.

Health-Related Quality of Life

In the GLOW study, descriptive statistics were provided. Mixed-effects models for repeated measures analysis, time to deterioration (first clinically meaningful deterioration relative to baseline), and time to improvement (first clinically meaningful improvement relative to baseline) were provided for EORTC QLQ-C30 global health status, FACIT-Fatigue score, EQ-5D-5L utility score, and EQ-5D-5L VAS. HRQoL outcomes were not reported for the CAPTIVATE trial.

Table 10: Statistical Analysis of Efficacy End Points

AE = adverse event; BM = bone marrow; CLL = chronic lymphocytic leukemia; CR = complete response; CrCl = creatine clearance; CRi = complete response with incomplete bone marrow recovery; del(11q) = deletion of 11q; DOR = duration of response; EORTC QLQ-C30 = European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 30; EQ-5D-5L = 5-Level EQ-5D; FACIT = Functional Assessment of Chronic Illness Therapy; FD = fixed duration; IRC = independent review committee; ITT = intention to treat; MMRM = mixed-effects models for repeated measures; MRD = minimal residual disease; NA = not applicable; NR = not reported; ORR = overall response rate; OS = overall survival; PB = peripheral blood; PD = progressed disease; PFS = progression-free survival; TLS = tumour lysis syndrome; TTNT = time to next treatment.

Sources: GLOW primary analysis CSR;²³ CAPTIVATE CSR.²⁶ (Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.)

Analysis Populations

A summary of analysis populations in the GLOW and CAPTIVATE trials is provided in Table 11.

The ITT analysis set was used for efficacy analyses in the GLOW study. The ITT population was defined as all patients randomized in the study, and it was analyzed according to assigned treatment group, regardless of the actual treatment received. The ITT population was used for all analyses of primary and secondary efficacy end points and for patient-reported outcomes, analyses of disposition, and demographic and baseline disease characteristics.

The safety population was defined as all randomized patients who receive at least 1 dose of any of the 4 study drugs (ibrutinib, venetoclax, chlorambucil, or obinutuzumab). The safety population is used for all safety analyses and analyses of exposure. All patients were analyzed according to the treatment they were assigned.

In the CAPTIVATE trial, the all-treated population was used in all analyses, including baseline, disposition, efficacy, and safety. The non-del(17p) population was used in the primary analysis of the primary and main secondary end points for the FD cohort.

Table 11: Analysis Populations of the GLOW and CAPTIVATE Studies

CIRS = Cumulative Illness Rating Scale; del(11q) = deletion of 11q; del(17p) = deletion of 17p; ECOG PS = Eastern Cooperativity Oncology Group Performance Status; FD = fixed duration; FISH = fluorescence in situ hybridization; ITT = intention to treat; LDH = lactate dehydrogenase; PFS = progression-free survival; PRO = patient-reported outcome.

Sources: GLOW primary analysis CSR;²³ CAPTIVATE CSR.²⁶ (Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.)

Results

Patient Disposition

Patient disposition for the GLOW trial and the CAPTIVATE FD cohort are summarized in Table 12.

GLOW

At the time of data cut-off for the primary analysis (February 26, 2021) in the GLOW trial, more patients in the ibrutinib-venetoclax arm had discontinued study treatment than in the chlorambucil-obinutuzumab arm (22.6% versus 4.8% for ibrutinib-venetoclax versus chlorambucil-obinutuzumab). The most common reason for treatment discontinuation was AE (10.4% versus 1.9%). Other reasons for study treatment discontinuation ibrutinib-venetoclax and chlorambucil-obinutuzumab arms included patient refusal of further study treatment (3.8% versus 1.0%), PD (2.8% versus 1.0%), and physician decision (1.9% versus 1.0%). A number of patients died during the study period (11 [10.4%] patients in the ibrutinib-venetoclax arm and 11 [10.5%] patients in the chlorambucil-obinutuzumab arm). Among them, 4 patients in the ibrutinib-venetoclax arm discontinued study treatment due to death, and the remaining 18 patients either completed the study treatment (2 patients in the ibrutinib-venetoclax arm and 9 patients in the chlorambucil-obinutuzumab arm) or discontinued study treatment due to AEs (4 patients in the ibrutinib-venetoclax arm), physician’s decision (1 patient in the ibrutinib-venetoclax arm), PD (1 patient in the chlorambucil-obinutuzumab arm), or patient refusal (1 patient in the chlorambucil-obinutuzumab arm) before they died. Five (2.4%) patients elected to terminate study participation prematurely (2 [1.9%] patients in the ibrutinib-venetoclax arm and 3 [2.9%] patients in the chlorambucil-obinutuzumab arm).

CAPTIVATE

In the FD cohort (n = 159), 147 patients (92.5%) completed the planned ibrutinib treatment and 149 patients (93.7%) completed the planned venetoclax treatment. The most common reason for treatment discontinuation was AEs (ibrutinib: 4.4%;venetoclax: 1.9%). The most common reason for study discontinuation was withdrawal by patient (3.8%).

Table 12: Summary of Patient Disposition From Pivotal Studies and RCT Evidence Submitted by the Sponsor (ITT Analysis Set for GLOW; All-Treated Analysis Set for CAPTIVATE)

AE = adverse event; FAS = full analysis set; FD = fixed duration; ITT = intention to treat; NA = not applicable; NR = not reported; progressive disease; RCT = randomized controlled trial.

Sources: GLOW primary analysis CSR;²³ CAPTIVATE CSR.²⁶ (Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.)

Impact of COVID-19 on Patient Disposition

In the GLOW trial, in the primary analysis (data cut-off: February 26, 2021), no patient discontinued study treatment due to the COVID-19 pandemic. Six (2.8%) patients had not completed the FD treatment phase when the COVID-19 outbreak was declared a pandemic by WHO in March 2020. The sponsor considered the COVID-19 pandemic to have had a limited impact on patient disposition.

In the CAPTIVATE trial, in the primary analysis (data cut-off: November 12, 2020) and with extended follow-up (data cut-off: August 4, 2021), none of the patients in the FD cohort discontinued treatment or exited the study due to COVID-19.

Baseline Characteristics

The GLOW Study

The majority of patients in the trial were men (57.8%) and white (95.7%). Median age at enrolment was 71 years (range, 47 to 93 years), with 87.2% of patients aged 65 years or older and 34.1% of patients aged 75 years or older. Patient characteristics were generally balanced across treatment arms at baseline. However, a difference of more than 10% was observed for CIRS score greater than 6 (69.8% versus 58.1% for ibrutinib-venetoclax versus chlorambucil-obinutuzumab) and LDH elevation (33.0% versus 48.6%). CLL was the initial diagnosis in 93.4% of patients; 6.6% of patients had SLL. Advanced stage disease at baseline was reported for 54.8% of patients based on Rai stage III or IV disease and for 42.1% of patients based on Binet stage C disease. In total, 37.6% of patients had bulky disease based on the presence of a tumour lesion at least 5 cm in longest diameter and 58.3% of patients had cytopenia at baseline. About half the patients had a baseline ECOG PS of 1 (53.1%). Overall, the proportion of patients with high-risk disease, defined as the presence of del(11q), unmutated IGHV, or TP53 mutation, was similar in treatment arms (59.4% versus 57.1%). Although patients with del(17p) or known TP53 mutation at baseline were excluded from the study, patients with unknown TP53 mutation status were allowed to participate. After randomization, central laboratory testing identified 9 (4.3%) patients with a TP53 mutation, 7 (6.6%) in the ibrutinib-venetoclax arm and 2 (1.9%) in the chlorambucil-obinutuzumab arm.

The CAPTIVATE Study

In the FD cohort, the median age at baseline was 60.0 years (range, 33 to 71 years), with 28.3% of patients aged 65 years or older (Table 13). More patients were male (66.7%), and the majority of patients were white (92.5%). At baseline, more patients (69.2%) had an ECOG PS score of 0. Fewer patient had progressed to stage III or IV (27.7%), and 30.2% of patients had bulky disease of at least 5 cm. Cytogenetic characteristics indicative of poor prognosis (per hierarchical classification) in the study population were del(17p) (12.6%) and del 11q (17.6%). Other poor prognostic characteristics included mutated TP53 (10.1%), mutated del(17p) or TP53 (17.0%), unmutated IGHV (56.0%), and complex karyotype (19.5%).

The baseline characteristics outlined in Table 13 are limited to those that are most relevant to this review or were felt to affect outcomes or interpretation of the study results.

Table 13: Summary of Baseline Characteristics of Pivotal Studies and RCT Evidence Submitted by the Sponsor

CIRS = Cumulative Illness Rating Scale; CLL = chronic lymphocytic leukemia; CrCl = creatinine clearance; del(11q) = deletion of 11q; del(13q) = deletion of 13q; del(17p) = deletion of 17p; ECOG PS = Eastern Cooperativity Oncology Group Performance Status; FD = fixed duration; LDH = lactate dehydrogenase; NR = not reported; RCT = randomized controlled trial; SD = standard deviation; SLL = small lymphocytic leukemia.

Sources: GLOW primary analysis CSR;²³ CAPTIVATE CSR.²⁶ (Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.)

Exposure to Study Treatments

GLOW

A summary of exposure to study treatments is provided in Table 14.

At the time of the primary analysis (data cut-off: February 26, 2021), patients in the ibrutinib-venetoclax arm had a longer median duration of treatment than patients in the chlorambucil-obinutuzumab arm (13.8 months versus 5.1 months for ibrutinib-venetoclax versus chlorambucil-obinutuzumab). Ibrutinib and venetoclax dose reductions due to an AE were reported for 19.8% and 17% of patients, respectively. Dose interruptions (defined as doses missed for ≥ 7 consecutive days) for ibrutinib and venetoclax were noted in 52.8% and 36.8% of patients, respectively. For both study drugs, the majority of the dose interruptions were due to an AE. In the chlorambucil-obinutuzumab arm, the majority of patients received chlorambucil (79%) at the dose intended, per protocol, and without any prescribed dose reductions. Chlorambucil dosing was reduced or delayed because of an AE in 21.0% and 34.3% of patients, respectively. For obinutuzumab, which is administered intravenously, dose reductions were not permitted. Obinutuzumab dosing was skipped or delayed because of an AE in 2.9% and 39% of patients, respectively. Additionally, obinutuzumab infusion was temporarily interrupted, the infusion rate reduced, and/or the infusion aborted in 57.1% of patients due to an AE.

CAPTIVATE

In the FD cohort, the overall median treatment duration was 13.8 months at the primary analysis data cut-off date of November 12, 2020 (range, 0.5 to 24.9 months). The individual median treatment durations for ibrutinib and venetoclax were 13.8 months (range, 0.5 to 24.9 months) and 11.0 months (range, 1.3 to 22.1 months), respectively. One patient had a 2-day dose hold, but there were no other dose modifications.

Table 14: Summary of Patient Exposure From Pivotal Studies and RCT Evidence Submitted by the Sponsor (Safety Analysis Set for GLOW; All-Treated Analysis Set for CAPTIVATE)

FD = fixed duration; NR = not reported; SD = standard deviation.

^aDose interruption is defined as missing a dose for ≥ 7 consecutive days.

^bPatients with any prescribed dose reduction and/or dose delay due to adverse event.

Sources: GLOW primary analysis CSR;²³ CAPTIVATE CSR.²⁶ (Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.)

Concomitant Medications and Co-Interventions

A summary of concomitant medications and co-interventions in the GLOW and CAPTIVATE trials is presented in Table 15.

All patients in the primary analysis safety analysis set (data cut-off: February 26, 2021) received concomitant medication in both treatment arms in the GLOW trial. Concomitant medications reported with a frequency at least 15% higher in the ibrutinib-venetoclax arm compared to the chlorambucil-obinutuzumab arm include drugs acting on the renin-angiotensin system (61.3% versus 45.7% for ibrutinib-venetoclax versus chlorambucil-obinutuzumab); antithrombotic drugs (38.7% versus 13.3%); antidiarrheals, intestinal anti-inflammatory, and/or anti-infective drugs (34.9% versus 3.8%); and all other nontherapeutic products (27.4% versus 9.5%). Concomitant medications with a frequency at least 15% higher in the chlorambucil-obinutuzumab arm compared with the ibrutinib-venetoclax arm include analgesics (61.3% versus 100.0%), antihistamines for systemic use (21.7% versus 100.0%), corticosteroids for systemic use (17.0% versus 100.0%), and antiemetics and antinauseants (27.6% versus 9.4%). Similar results were observed in the extended follow-up analysis (data cut-off: August 25, 2022).

In CAPTIVATE FD cohort, in the primary analysis (data cut-off: November 12, 2020), the most common concomitant medications were antigout preparations (96.9%), antibacterials for systemic use (66.7%), and analgesics (54.7%). Similar results were observed in the extended follow-up analysis (data cut-off: August 4, 2021).

Table 15: Summary of Concomitant Treatment^a (Safety Analysis Set GLOW; All-Treated Analysis Set for CAPTIVATE)

FD = fixed duration.

^aConcomitant treatment was defined as any medication administered while the study treatment was ongoing (i.e., during the treatment phase from first dose to last dose of the study treatment).

Sources: GLOW primary analysis CSR;²³ CAPTIVATE CSR.²⁶

Subsequent Treatment

Both the GLOW and CAPTIVATE trials reported on patients who received subsequent treatments. A summary of subsequent treatment is presented in Table 16.

In the GLOW study, at primary analysis (data cut-off: February 26, 2021), fewer patients received anticancer therapy in the ibrutinib-venetoclax arm than in the chlorambucil-obinutuzumab arm (3.8% versus 25.7% for ibrutinib-venetoclax versus chlorambucil-obinutuzumab). The commonly reported subsequent therapies include single-drug ibrutinib (0 versus 5.7% for ibrutinib-venetoclax versus chlorambucil-obinutuzumab), acalabrutinib (0 versus 1.0% for ibrutinib-venetoclax versus chlorambucil-obinutuzumab), and venetoclax (0 versus 1.0% for ibrutinib-venetoclax versus chlorambucil-obinutuzumab). Similar results were observed in the extended follow-up analysis (data cut-off: August 25, 2022).

In the CAPTIVATE FD cohort, at primary analysis (data cut-off: November 12, 2020), 9 (5.7%) patients of the FD cohort received subsequent therapy, 9 (5.7%) and 2 (1.3%) of whom received single-drug ibrutinib and venetoclax, respectively. Similar results were observed in the extended follow-up analysis (data cut-off: August 4, 2021).

Table 16: Summary of Subsequent Treatment From Pivotal Studies and RCT Evidence Submitted by the Sponsor (ITT Analysis Set for GLOW; All-Treated Analysis Set for CAPTIVATE)

FD = fixed duration; ITT = intention to treat; NR = not reported; RCT = randomized controlled trial.

^aAny subsequent therapy other than reintroduced treatment, either ibrutinib or ibrutinib-venetoclax.

Sources: GLOW primary analysis CSR;²³ CAPTIVATE CSR.²⁶ (Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.)

Protocol Amendments

In the GLOW study, there were a total of 4 protocol amendments reported. Of these 4, the following was particularly of note and impactful. Protocol amendment 1, made June 6, 2018, removed eligibility for patients with galactose intolerance, Lapp lactase deficiency, or glucose-galactose malabsorption from GLOW study enrolment.

There were a total of 3 protocol amendments reported for the CATIVATE trial. No impactful amendment was identified by the CADTH review team.

Protocol Deviation

Both the GLOW and CAPTIVATE trials reported on patients who had major or important protocol deviations (Table 17).

In the GLOW study, at the time of primary analysis (data cut-off: February 26, 2021), a similar proportion of patients reported major protocol deviations in the ibrutinib-venetoclax and chlorambucil-obinutuzumab arms (5.7% versus 5.7% for ibrutinib-venetoclax versus chlorambucil-obinutuzumab). Five (4.7%) patients in the ibrutinib-venetoclax arm and 3 (2.9%) patients in the chlorambucil-obinutuzumab arm received the wrong treatment or an incorrect dose, including 1 patient in the ibrutinib-venetoclax arm who received expired venetoclax and 3 patients in the chlorambucil-obinutuzumab arm who received more than 175% of the prescribed chlorambucil daily dose. One (0.9%) patient in the ibrutinib-venetoclax arm and 3 (2.9%) patients in the chlorambucil-obinutuzumab arm missed a disease evaluation visit due to the COVID-19 pandemic; these were considered to be major protocol deviations because they potentially delayed the detection of PD. One (0.9%) patient in the ibrutinib-venetoclax arm did not meet an inclusion criterion (adequate organ function: absolute neutrophil count ≥ 750 cells/μL) for study eligibility.

In the CAPTIVATE trial, at the time of primary analysis (data cut-off: November 12, 2020), 2 patients had important protocol deviations related to not meeting all eligibility criteria, 1 patient was enrolled despite the need to be treated with 20 mg of prednisone during the screening period to control autoimmune hemolytic anemia, and 1 patient did not have an activated partial thromboplastin time coagulation test performed at screening. One patient refused to have CT scans performed at multiple time points, which was considered to have an impact on efficacy. On 3 separate occasions, 1 patient experienced grade 3 to 4 neutropenia related to venetoclax treatment, and the dose of venetoclax was not reduced or withheld per protocol requirements which was considered relating to safety.

Table 17: Summary of Patients With Major Protocol Deviations (ITT Analysis Set for GLOW; All-Treated Analysis Set for CAPTIVATE)

FD = fixed duration; ITT = intention to treat; NR = not reported.

Sources: GLOW primary analysis CSR;²³ CAPTIVATE CSR.²⁶

Treatment Compliance

In the GLOW study, at the time of the primary analysis (data cut-off: February 26, 2021), for all 4 study drugs, around 95% of patients received the investigator-prescribed dose (96.2%, 96.9%, 94.3%, and 100% for ibrutinib, venetoclax, obinutuzumab, and chlorambucil, respectively). Three patients in the chlorambucil-obinutuzumab arm received a dose of chlorambucil that was more than 175% of the protocol-specified dose. These account for the chlorambucil mean compliance rate greater than 100%. All 3 overdose cases were reported as major protocol deviations.

In the CAPTIVATE trial, at the time of primary analysis (data cut-off: November 12, 2020), 147 patients (92.5%) completed the planned ibrutinib treatment and 149 patients (93.7%) completed the planned venetoclax treatment.

Efficacy

Unless otherwise specified, the key efficacy results of the GLOW and CAPTIVATE trials are summarized in Table 18. Refer to Appendix 1 for the hierarchical testing order of secondary end points for the GLOW trial and detailed efficacy data for the GLOW trial and CAPTIVATE FD cohort.

Progression-Free Survival

In the GLOW study, PFS per IRC was the primary end point. In the primary analysis (data cut-off: February 26, 2021), there were 89 IRC-assessed PFS events, 22 patients (20.8%) in the ibrutinib-venetoclax arm and 67 patients (63.8%) in the chlorambucil-obinutuzumab arm had experienced a PFS event. Of these, 13 (12.3%) had disease progression and 9 (8.5%) died in the ibrutinib-venetoclax arm, 65 (61.9%) had disease progression and 2 (1.9%) died in the chlorambucil-obinutuzumab arm. The median time on study was 27.7 months (95% CI, 27.50 to 27.83 months) in the ibrutinib-venetoclax arm and 27.89 months (95% CI, 27.53 to 28.58 months) in the chlorambucil-obinutuzumab arm. The primary analysis showed an improvement in PFS assessed by IRC for patients receiving FD ibrutinib-venetoclax compared with chlorambucil-obinutuzumab (HR = 0.216; 95% CI, 0.131 to 0.357; P < 0.0001). This represents a 78% reduction in the risk of PD or death for patients treated with ibrutinib-venetoclax compared with chlorambucil-obinutuzumab. Median PFS was not reached in the ibrutinib-venetoclax arm and was 21.0 months in the chlorambucil-obinutuzumab arm. The probability of not experiencing a PFS event by 24 months was 84.4% (95% CI, 75.8% to 90.1%) in the ibrutinib-venetoclax arm and 44.1% (95% CI, 34.2% to 53.6%) in the chlorambucil-obinutuzumab arm in the primary analysis. The Kaplan-Meier plot of PFS per IRC in the primary analysis is depicted in Figure 3.

Subgroup analyses of PFS per IRC in the primary analysis were general consistent with the primary analysis across all prespecified subgroups, except for race and disease diagnosis at baseline. Refer to Appendix 1 for detailed subgroup analyses data.

In addition, several prespecified sensitivity analyses based on the IRC assessment of PFS were included in the statistical analysis plan, including unstratified analysis, censoring patients who initiated subsequent anticancer therapy before PD, censoring patients at the last disease evaluation if they progressed or died after missing at least 2 consecutive planned disease evaluation visits, and using PFS per IA. The results were generally consistent with the results of the primary analysis and showed HR values ranging from 0.207 (95% CI, 0.120 to 0.357) to 0.233 (95% CI, 0.143 to 0.379). The overall concordance between IA and IRC assessments of PD and non-PD events was robust in both arms (93.4% versus 81.0% for ibrutinib-venetoclax versus chlorambucil-obinutuzumab). Refer to Appendix 1 for the Kaplan-Meier plot of PFS per IA. A planned supplementary PFS analysis censoring patients who died pre-PD due to COVID-19 was not conducted because only 1 pre-PD death due to COVID-19 was reported and the threshold specified in the statistical analysis plan (i.e., the number of pre-PD deaths > 5% of the total PFS events) was not met.

With a median time on study for all patients of 46.1 months (95% CI, 45.86 to 46.26 months) at the time of the extended follow-up (data cut-off: August 25, 2022), PFS analysis by IRC continued to show an improvement with FD ibrutinib-venetoclax compared to chlorambucil-obinutuzumab (HR = 0.214; 95% CI, 0.138 to 0.334; nominal P < 0.0001). This improvement represents a 79% reduction in the risk of PD or death with ibrutinib-venetoclax treatment compared with chlorambucil-obinutuzumab treatment. The Kaplan-Meier plot of PFS per IRC in the extended follow-up analysis is depicted in Figure 4.

In the CAPTIVATE FD cohort, in the primary analysis (data cut-off: November 12, 2020), median PFS by IRC assessment was not reached in the all-treated FD cohort based on an overall median follow-up of 27.9 months. PFS event probabilities in the all-treated FD cohort were 14.5% (21 disease progression events and 2 deaths) at the time of primary analysis and 25.2% (38 disease progression events and 2 deaths) in the follow-up analysis (data cut-off: August 4, 2021). The probability of not experiencing a PFS event based on IRC assessment was 88.9% (95% CI, 82.7% to 92.9%) at 24 months in the primary analysis and 85.5% (95% CI, 78.8% to 90.2%) at 36 months in the extended follow-up analysis for all treated patients. Among patients with del(17p), 5 (25.0%) reported a PFS event, the median PFS by IRC was not reached, and the PFS event-free probability was 75.0% (95% CI, 50.0% to 88.7%) at 24 months in the primary analysis and 80.0% (95% CI, 55.1% to 92.0%) at 36 months in the extended follow-up analysis. The Kaplan-Meier plot of PFS per IRC in the primary analysis is depicted in Figure 5; refer to Appendix 1 for the Kaplan-Meier plot of PFS per IRC in the extended follow-up analysis.

Generally, similar results were observed for PFS per IA, and median PFS by IA was not reached in the all-treated FD cohort. PFS event probabilities were 12.6% (18 disease progression events and 2 deaths) at the time of the primary analysis (data cut-off: November 12, 2020) and 17.6% (26 disease progression events and 2 deaths) in the extended follow-up analysis (data cut-off: August 4, 2021) for all treated patients. The probability of not experiencing a PFS event based on IA were 94.8% (95% CI, 89.8% to 97.3%) at 24 months in the primary analysis and 88.1% (95% CI, 81.7% to 92.3%) at 36 months in the extended follow-up analysis. In patients with del(17p), 4 (20.0%) reported a PFS event, median PFS by IA was not reached (the lower end of the 95% CI was 28.5 months), PFS event-free rates were 84.2% (95% CI, 58.7% to 94.6%) at 24 months in the primary analysis and 78.6% (95% CI, 52.5% to 91.4%) at 36 months in the extended follow-up analysis. The Kaplan-Meier plot of PFS per IA in the primary analysis was depicted in Figure 6; refer to Appendix 1 for the Kaplan-Meier plot of PFS per IA in the extended follow-up analysis.

Figure 3: Kaplan-Meier Plot of PFS per IRC at Primary Analysis for GLOW (ITT Analysis Set)

Clb+Ob = obinutuzumab plus chlorambucil; Ibr+Ven = ibrutinib plus venetoclax; IRC = independent review committee; ITT = intention to treat; PFS = progression-free survival.

Note: Data cut-off was February 26, 2021.

Source: GLOW Primary Analysis CSR.²³

Figure 4: Kaplan-Meier Plot of PFS per IRC at Extended Follow-Up Analysis for GLOW (ITT Analysis Set)

Clb+Ob = obinutuzumab plus chlorambucil; Ibr+Ven = ibrutinib plus venetoclax; IRC = independent review committee; ITT = intention to treat; PFS = progression-free survival.

Note: Data cut-off was August 25, 2022.

Source: GLOW extended follow-up analysis CSR.²⁴

Figure 5: Kaplan-Meier Curves for PFS per IRC Assessment at Primary Analysis for CAPTIVATE (All-Treated Analysis Set)

Del17P = deletion of 17p; FD = fixed duration; IRC = independent review committee; MRD = minimal residual disease; PFS = progression-free survival.

Note: Data cut-off was November 12, 2020.

Source: CAPTIVATE primary analysis CSR.

Figure 6: Kaplan-Meier Curves for PFS per IA at Primary Analysis for CAPTIVATE (All-Treated Analysis Set)

Del17P = deletion of 17p; FD = fixed duration; IA = investigator assessment; MRD = minimal residual disease; PFS = progression-free survival.

Note: Data cut-off was November 12, 2020.

Source: CAPTIVATE primary analysis CSR

Overall Survival

In the GLOW study, at the time of primary analysis (data cut-off: February 26, 2021), median OS was not reached in either arm. With a median follow-up of 27.7 months for the ibrutinib-venetoclax arm and 27.89 months for the chlorambucil-obinutuzumab arm, there were 11 (10.4%) deaths observed in the ibrutinib-venetoclax arm and 12 (11.4%) deaths observed in the chlorambucil-obinutuzumab arm (HR = 1.048; 95% CI, 0.454 to 2.419; nominal P = 0.9121). OS probabilities decreased from 91.4% at 12 months to 90.4% at 24 months in the ibrutinib-venetoclax arm and from 98.1% to 91.3%, respectively, in the chlorambucil-obinutuzumab arm. A supplementary OS analysis that censored patients who died due to COVID-19 showed results consistent with the primary OS analysis, with a HR of 1.298 (95% CI, 0.512 to 3.290). In the extended follow-up analysis (data cut-off: August 25, 2022), similar to the primary analysis, median OS was not reached in either treatment arm. With a median follow-up of 46.1 months, 15 (14.2%) death events were observed in the ibrutinib-venetoclax arm and 30 (28.6%) were observed in the chlorambucil-obinutuzumab arm. This corresponds to an HR of 0.487 (95% CI, 0.262 to 0.907; nominal P = 0.0205). OS survival probabilities at 42 months were 87.5% and 77.6% in the ibrutinib-venetoclax and chlorambucil-obinutuzumab arms, respectively. Results of a supplementary OS analysis that censored patients who died due to COVID-19 were consistent with the primary OS analysis, with an HR of 0.527 (95% CI, 0.268 to 1.035). The Kaplan-Meier plot of OS in the primary analysis is depicted in Figure 7.

In the all-treated CAPTIVATE FD cohort, at the time of primary analysis (data cut-off: November 12, 2020), median OS was not reached. Based on an overall median follow-up of 27.9 months, there were 3 deaths (1.9%) reported in the all-treated FD cohort, 3 deaths (2.2%) reported in patients without del(17p), and no deaths reported in patients with del(17p). Most patients were alive and on study, with OS probabilities of 98.1% at 24 months at the primary data cut-off date and 98.1% at 36 months at the extended follow-up analysis data cut-off date (August 4, 2021) for the all-treated FD cohort. All patients with del(17p) were alive and on study at the primary data cut-off date and at the extended follow-up analysis data cut-off date. The Kaplan-Meier plot of OS in the primary analysis is depicted in Figure 8.

Figure 7: Kaplan-Meier Plot of OS at Primary Analysis for GLOW (ITT Analysis Set)

Clb+Ob = obinutuzumab plus chlorambucil; Ibr+Ven = ibrutinib plus venetoclax; ITT = intention to treat; OS = overall survival.

Note: Data cut-off was February 26, 2021.

Source: GLOW primary analysis CSR.²³

Figure 8: Kaplan-Meier Curves for OS at Primary Analysis for CAPTIVATE (All-Treated Analysis Set)

Del17P = deletion of 17p; FD = fixed duration; MRD = minimal residual disease; OS = overall survival.

Note: Data cut-off was November 12, 2020.

Source: CAPTIVATE primary analysis CSR.

Overall Response Rate

In the GLOW primary analysis (data cut-off: February 26, 2021), the IRC-assessed ORR (of PR or better) was similar in the ibrutinib-venetoclax and chlorambucil-obinutuzumab treatment arms (86.6% versus 84.8% for ibrutinib-venetoclax versus chlorambucil-obinutuzumab). The majority of patients had PR (44.3% versus 70.5), followed by CR (35.8% versus 11.4%), nPR (3.8% versus 2.9%), and CRi (2.8% versus 0%). In this analysis, the relative response of ORR for ibrutinib-venetoclax compared to chlorambucil-obinutuzumab was 1.02 (95% CI, 0.92 to 1.14; P = 0.6991). Similar results were observed in the ORR based on IA (92.5% versus 87.6%; relative response = 1.05; 95% CI, 0.96 to 1.15; nominal P = 0.2585). Generally, similar results were observed in the extended follow-up analysis (data cut-off: August 25, 2022).

Of note, as the difference in ORR based on IRC assessment between treatment arms was not statistically significant (P = 0.6991) in the primary analysis, the hierarchical statistical testing strategy ended at ORR per IRC. The remaining key secondary end points (i.e., OS, sustained hematological improvements, and time to improvement in the FACIT-Fatigue score) and ORR per IA were considered not statistically significant. Refer to Appendix 1 for the hierarchical testing order of secondary end points for the GLOW study.

In the CAPTIVATE FD cohort, in the primary analysis (data cut-off: November 12, 2020), the ORR per IRC assessment was 96.2% (95% CI, 93.3% to 99.2%) for all treated patients, 95.6% (95% CI, 92.1% to 99.0%) for patients without del(17p), and 100.0% (95% CI, NE) for patients with del(17p). The majority of all treated patients had CR (57.9%), followed by PR (35.2%), CRi (1.9%), and nPR (1.3%). Similar results were observed for patients with and without del(17p) and for ORR based on IA. Similar results were observed in the extended follow-up analysis (data cut-off: August 4, 2021).

CR (CR and CRi) Rate

In the GLOW study, in the primary analysis (data cut-off: February 26, 2021), the IRC-assessed CR rate was higher in the ibrutinib-venetoclax arm than in the chlorambucil-obinutuzumab arm (38.7% versus 11.4% for ibrutinib-venetoclax versus chlorambucil-obinutuzumab; relative response = 3.43; 95% CI, 1.91 to 6.15; P < 0.0001). Similar results were observed for the investigator-assessed CR probability in the primary analysis (45.3% versus 13.3%; relative response = 3.42; 95% CI, 2.01 to 5.82; nominal P < 0.0001). Generally, similar results were observed in the extended follow-up analysis (data cut-off: August 25, 2022).

In the CAPTIVATE FD cohort, investigator-assessed CR probability was assessed as the primary end point. In patients without del(17p) in the FD cohort, CR rate per IA was met, at 55.9% (95% CI, 47.5% to 64.2%), which exceeded the prespecified minimum CR probability of 37% (1-sided P < 0.0001) and the historical FCR comparator probability (40%) in the primary analysis (data cut-off: November 12, 2020). The investigator-assessed CR rate was 55.3% (95% CI, 47.6% to 63.1%) in all treated patients and 50.0% (95% CI, 28.1% to 71.9%) in patients with del(17p) in the primary analysis. Generally, similar results were observed in the extended follow-up analysis (data cut-off: August 4, 2021).

Subgroup analyses in the primary analysis of the CR probability based on IA for all patients in the FD cohort were, in general, consistent with the primary analysis across all prespecified subgroups except for bulky disease. Refer to Appendix 1 for detailed subgroup analyses data.

In sensitivity analysis, IRC-assessed results were similar to the results based on IA. The concordance of CR and CRi per investigator and IRC assessments was 78.0% at the primary analysis, and no appreciable change in concordance was identified with extended follow-up (76.1%). In the supportive analysis of durable CR in the primary analysis, around 50% of patients had a durable CR — 49.1% (95% CI, 41.3% to 56.8%) for all treated patients and 48.5% (95% CI, 40.1% to 56.9%) for patients without del(17p) — and similar results were seen in the extended follow-up analysis. The analysis of duration of CR in patients who achieved a CR was performed as a supportive analysis in the primary analysis, with a median follow-up of 27.9 months. The median duration of CR was not reached for all patients or for patients without del(17p), and similar results were observed in the extended follow-up analysis.

Sustained Hematologic Improvement

In the GLOW study, at the time of primary analysis (data cut-off: February 26, 2021), the proportion of patients with sustained improvement in hemoglobin was similar the ibrutinib-venetoclax arm compared with the chlorambucil-obinutuzumab arm (44.3% and 50.5%, respectively; nominal P = 0.3854). The proportion of patients with sustained improvement in platelets was also similar in the ibrutinib-venetoclax and chlorambucil-obinutuzumab arms (24.5% and 29.5%, respectively; nominal P = 0.4346). In the extended follow analysis (data cut-off: August 25, 2022), the proportions of patients with sustained improvement in hemoglobin and platelet counts in the ibrutinib-venetoclax and chlorambucil-obinutuzumab arms were similar to those in the primary analysis (hemoglobin: 52.8% versus 51.4% for ibrutinib-venetoclax versus chlorambucil-obinutuzumab; platelet count: 32.1% versus 32.4%).

In the CAPTIVATE FD cohort, in the primary analysis (data cut-off: November 12, 2020), the proportion of patients achieving a sustained improvement in hemoglobin was 41.5% (95% CI, 33.9% to 49.2%) in all treated patients and 60.0% (95% CI, 38.5% to 81.5%) in patients with del(17p). The proportion of patients with sustained improvement in platelets was 17.6% (95% CI, 11.7% to 23.5%) in all treated patients and 15.0% (95% CI, 0 to 30.6%) in patients with del(17p). Generally, similar results were observed in the extended follow-up analysis (data cut-off: August 4, 2021).

Duration of Response

In the GLOW study, as of the data cut-off for the primary analysis (February 26, 2021), with an overall median follow-up of 27.7 months, the median DOR for patients who achieved an IRC-assessed PR or better was 28.9 months (95% CI, 28.7 months to NE) in the ibrutinib-venetoclax arm and 21.1 months (95% CI,15.9 to 25.1 months) in the chlorambucil-obinutuzumab arm; based on investigator-assessed responses, median DOR was not reached in the ibrutinib-venetoclax arm and was 21.5 months (95% CI, 15.7 to 28.5 months) in the chlorambucil-obinutuzumab arm. In the extended follow-up analysis (data cut-off: August 25, 2022), with an overall median follow-up of 46.1 months, median DOR for patients who achieved an IRC-assessed PR or better was not reached in the ibrutinib-venetoclax arm and was 21.2 months (95% CI, 17.7, 28.6 months) in the chlorambucil-obinutuzumab arm; similarly, median DOR per IA was not reached in the ibrutinib-venetoclax arm and was 21.6 months (95% CI, 15.7 to 29.1 months) in the chlorambucil-obinutuzumab arm.

In the CAPTIVATE FD cohort, in the primary analysis (data cut-off: November 12, 2020), with a median follow-up of 27.9 months, the median DOR per IRC assessment in the FD cohort was not reached for all patients, patients without del(17p) (with the lower end of the 95% CI of 27.6 months), and patients with del(17p) (with the lower end of the 95% CI of 18.9 months). Similar results were observed for investigator-assessed DOR at the same analysis. In the extended follow-up analysis (data cut-off: August 4, 2021), with a median follow-up of 38.7 months, the median DOR per IRC was 36.4 months for all treated patients, not reached for patients without del(17p) (with the lower end of the 95% CI of 36.3 months), and 35.9 months (95% CI, 35.6 months to NE) for patients with del(17p). Of note, the median DOR for patients with del(17p) in the primary analysis was considered not reliable due to the limited number (n = 2) of patients at risk at 36 months after initial response. At the same analysis, similar results were observed for investigator-assessed DOR.

Time to Next Treatment

At the GLOW primary analysis (data cut-off: February 26, 2021), the proportion of patients who received subsequent anticancer therapy was lower in the ibrutinib-venetoclax arm (3.8%) compared with the chlorambucil-obinutuzumab arm (25.7%). The median time to treatment failure was not reached in either arm. The HR of subsequent anticancer therapy comparing the ibrutinib-venetoclax arm to the chlorambucil-obinutuzumab arm was 0.143 (95% CI, 0.050 to 0.410; nominal P < 0.0001). Similar results were observed in the extended follow-up analysis (data cut-off: August 25, 2022).

TTNT was not reported in the CAPTIVATE FD cohort.

MRD Negativity

In the GLOW trial, a higher proportion of patients reported negative overall MRD by NGS in the ibrutinib-venetoclax arm than in the chlorambucil-obinutuzumab arm in BM (55.7% versus 21.0% for ibrutinib-venetoclax versus chlorambucil-obinutuzumab; relative response = 2.65; 95% CI, 1.75 to 3.99; P < 0.0001) and in PB (59.4% versus 40.0% for ibrutinib-venetoclax versus chlorambucil-obinutuzumab; nominal P = 0.0055) in the primary analysis (data cut-off: February 26, 2021). The 12-month MRD negativity rate in PB was 49.1% in the ibrutinib-venetoclax arm and 12.4% in the chlorambucil-obinutuzumab arm. MRD negativity rate by NGS was not assessed in the extended follow-up analysis (data cut-off: August 25, 2022).

In the all-treated FD cohort of the CAPTIVATE trial, overall MRD negativity rates by flow cytometry were as follows: 59.7% (95% CI, 52.1% to 67.4%) for all patients, 61.8% (95% CI, 53.6% to 69.9%) for patients without del(17p), and 45.0% (95% CI, 23.2% to 66.8%) for patients with del(17p) in BM; and 76.7% (95% CI, 70.2% to 83.3%) for all patients, 76.5% (95% CI, 69.3% to 83.6%) for patients without del(17p), and 80.0% (95% CI, 62.5% to 97.5%) for patients with del(17p) in PB. Identical results were reported for overall MRD negativity rates in the extended follow-up analysis (data cut-off: August 4, 2021).

TLS Risk Reduction

In the ibrutinib-venetoclax arm of the GLOW study, 26 (24.5%) patients had a high TLS risk by tumour burden at baseline. After ibrutinib lead-in, 22 (20.8%) patients shifted to medium and/or low risk in the primary analysis (data cut-off: February 26, 2021). TLS risk reduction was not assessed in the extended follow-up analysis (data cut-off: August 25, 2022).

In the FD cohort of the CAPTIVATE study, a high tumour burden was observed for 34 (21.4%) all treated patients; among them, 1 (5%) patient had del(17p) at baseline. After 3 cycles of single-drug ibrutinib lead-in therapy, 33 (20.8%) patients shifted to medium or low risk in the primary analysis (data cut-off: November 12, 2020); among them, 1 (5%) patient had del(17p). TLS risk reduction was not assessed in the extended follow-up analysis (data cut-off: August 4, 2021).

EORTC QLQ-C30 Global Health Status

In the primary analysis of the GLOW trial (data cut-off: February 26, 2021), 47 (44.3%) patients in the ibrutinib-venetoclax arm and 43 (41.0%) patients in the chlorambucil-obinutuzumab arm reported a meaningful deterioration (decreased ≥ 10 points) in the EORTC QLQ-C30 global health status score. The median time to first meaningful deterioration was 14.95 months (95% CI, 8.38 months to NE) in the ibrutinib-venetoclax group and 24.18 months (95% CI, 13.86 months to NE) in the chlorambucil-obinutuzumab group, with an HR of 1.149 (95% CI, 0.756 to1.746; nominal P = 0.5130). At the same analysis, 65 (61.3%) patients in the ibrutinib-venetoclax arm and 58 (55.2%) patients in the chlorambucil-obinutuzumab arm reported a meaningful improvement (increased ≥ 10 points) in EORTC QLQ-C30 global health status. The median time to first meaningful improvement in EORTC QLQ-C30 global health status score was 5.72 months (95% CI, 3.88 to 13.86 months) in the ibrutinib-venetoclax group and 6.11 months (95% CI, 4.14 to 11.56 months) in the chlorambucil-obinutuzumab group, with an HR of 0.874 (95% CI, 0.610 to 1.253; nominal P = 0.4576). The time to worsening or improvement in EORTC QLQ-C30 global health status score was not assessed in the extended follow-up analysis (data cut-off: August 25, 2022).

EORTC QLQ-C30 global health status was not measured in the CAPTIVATE FD cohort.

EQ-5D-5L VAS

In the primary analysis of the GLOW trial (data cut-off: February 26, 2021), 52 (49.1%) patients in the ibrutinib-venetoclax arm and 43 (41.0%) patients in the chlorambucil-obinutuzumab arm reported a meaningful decrease (≥ 7 points) in the EQ-5D-5L VAS score. The median time to first meaningful deterioration was 8.34 months (95% CI, 5.65 months to NE) in the ibrutinib-venetoclax group and 24.18 months (95% CI, 11.27 months to NE) in the chlorambucil-obinutuzumab group, with an HR of 1.382 (95% CI, 0.916 to 2.084; nominal P = 0.1188). At the same analysis, 60 (56.6%) patients in the ibrutinib-venetoclax arm and 75 (71.4%) patients in the chlorambucil-obinutuzumab arm reported a meaningful improvement (an increase of ≥ 7 points) in the EQ-5D-5L VAS score. The median time to first meaningful improvement was 4.76 months (95% CI, 3.78 to 8.31 months) in the ibrutinib-venetoclax group and 3.42 months (95% CI, 2.14 to 3.81 months) in the chlorambucil-obinutuzumab group, with an HR of 1.436 (95% CI, 1.015 to 2.031; nominal P = 0.0359). The time to worsening or improvement in EQ-5D-5L VAS score was not assessed in the extended follow-up analysis (data cut-off: August 25, 2022).

EQ-5D-5L VAS was not measured in the CAPTIVATE FD cohort.

EQ-5D-5L Utility Score

In the primary analysis of the GLOW trial (data cut-off: February 26, 2021), 51 (48.1%) patients in the ibrutinib-venetoclax arm and 46 (43.8%) patients in the chlorambucil-obinutuzumab arm reported a meaningful decrease (≥ 0.07 points) in the EQ-5D-5L utility score. The median time to first meaningful deterioration was 14.29 months (95% CI, 8.15 months to NE) in the ibrutinib-venetoclax group and 24.11 months (95% CI, 8.34 months to NE) in the chlorambucil-obinutuzumab group, with an HR of 1.188 (95% CI, 0.793 to 1.779; nominal P = 0.3997). At the same analysis, 59 (55.7%) patients in the ibrutinib-venetoclax arm and 65 (61.9%) patients in the chlorambucil-obinutuzumab arm reported a meaningful improvement (increased ≥ 0.07 points) in the EQ-5D-5L utility score. The median time to first meaningful improvement was 5.59 months (95% CI, 3.81 to 11.30 months) in the ibrutinib-venetoclax group and 4.67 months (95% CI, 3.75 to 6.14 months) in the chlorambucil-obinutuzumab group, with an HR of 1.177 (95% CI, 0.817 to 1.695; nominal P = 0.3757). The time to worsening or improvement in the EQ-5D-5L utility score was not assessed in the extended follow-up analysis (data cut-off: August 25, 2022).

EQ-5D-5L utility score was not measured in the CAPTIVATE FD cohort.

FACIT-Fatigue Scale

In the primary analysis of the GLOW trial (data cut-off: February 26, 2021), 62 (58.5%) patients in the ibrutinib-venetoclax arm and 51 (48.6%) patients in the chlorambucil-obinutuzumab arm reported a meaningful decrease (≥ 3 points) in the FACIT-Fatigue Scale score. The median time to first clinically meaningful deterioration in the FACIT-Fatigue score was 8.15 months (95% CI, 3.98 to 10.94 months) in the ibrutinib-venetoclax group and 14.03 months (95% CI, 8.61 months to NE) in the chlorambucil-obinutuzumab group, with an HR of 1.5 (95% CI, 1.054 to 2.237; nominal P = 0.0235). At the same analysis, 61 (57.5%) patients in the ibrutinib-venetoclax arm and 73 (69.5%) patients in the chlorambucil-obinutuzumab arm reported a meaningful improvement (increased ≥ 3 points) in the FACIT-Fatigue Scale score. The median time to first meaningful improvement in FACIT-Fatigue score was 5.59 months (95% CI, 3.81 to 11.20 months) in the ibrutinib-venetoclax group and 3.75 months (95% CI, 2.20 to 5.75 months) in the chlorambucil-obinutuzumab group, with a HR of 1.369 (95% CI, 0.959 to 1.954; nominal P = 0.0776). The time to worsening or improvement in the FACIT-Fatigue score was not assessed in the extended follow-up analysis (data cut-off: August 25, 2022).

FACIT-Fatigue Scale was not measured in the CAPTIVATE FD cohort.

Table 18: Summary of Key Efficacy Results From Pivotal and RCT Evidence (ITT Analysis Set for the GLOW Study; All-Treated Analysis Set for the CAPTIVATE Study)

BM = bone marrow; C+O = chlorambucil plus obinutuzumab; CI = confidence interval; CR = complete response; CRi = complete response with incomplete bone marrow recovery; del(17p) = deletion of 17p; DOR = duration of response; EORTC QLQ-C30 = European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 30; EQ-5D-5L = 5-Level EQ-5D; FACIT = Functional Assessment of Chronic Illness Therapy; FD = fixed duration; HR = hazard ratio; I+V = ibrutinib plus venetoclax; IA = investigator assessment; IRC = independent review committee; ITT = intention to treat; MRD = minimal residual disease; NA = not applicable; NE = not evaluable; NR = not reported; nPR = nodular partial response; ORR = overall response rate; OS = overall survival; PB = peripheral blood; PFS = progression-free survival; PR = partial response; TLS = tumour lysis syndrome; TTNT = time to next treatment.

^aEstimate is not reliable because only 3 patients were at risk at 39 months.

^bEstimate is not reliable because only 2 patients were at risk at 36 months after initial response.

^cNext-generation sequencing was used as the primary method of MRD analysis for the GLOW trial; the flow cytometry method was used for the MRD analysis in the CAPTIVATE FD cohort.

^dDeterioration or improvement is defined as a change of ≥ 10 points in the EORTC QLQ-C30 global health status score for each assessment after baseline.

^eMID of ≥ 7 points is considered clinically important for the EQ-5D-5L VAS health rating.

^fA minimum difference of ≥ 0.07 points in EQ-5D-5L utility score is considered clinically important.

^gFACIT-Fatigue improvement or worsening is defined as ≥ 3-point change for each assessment post baseline.

Sources: GLOW Primary Analysis CSR;²³ GLOW Extended Follow-up Analysis CSR;²⁴ GLOW PRO CSR;²⁵ CAPTIVATE CSR.²⁶ (Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.)

Harms

A summary of harms is provided in Table 19.

Adverse Events

In the GLOW study, at least 1 AE was reported in a similar proportion of patients in the ibrutinib-venetoclax arm compared with the chlorambucil-obinutuzumab arm (99.1% versus 94.3% for ibrutinib-venetoclax versus chlorambucil-obinutuzumab). AEs of any grade were reported more frequently in the ibrutinib-venetoclax arm, including diarrhea (50.9% versus 12.4%), nausea (26.4% versus 25.7%), infections and infestations (60.4% versus 48.6%), metabolism and nutrition disorders (42.5% versus 23.8%), respiratory, thoracic and mediastinal disorders (35.8% versus 28.6%), musculoskeletal and connective tissue disorders (34.0% versus 25.7% for), and nervous system disorders (30.2% versus 20.0% obinutuzumab). AEs of grade 3 or 4 were also reported in similar proportions in the ibrutinib-venetoclax arm and chlorambucil-obinutuzumab arm (68.9% versus 67.1%). The following AEs of grades 3 or 4 were reported more frequently (≥ 5% difference) in the ibrutinib-venetoclax arm than in the chlorambucil-obinutuzumab arm: diarrhea (10.4% versus 1.0% for obinutuzumab), vascular disorders (8.5% versus 1.9% for), and cardiac disorders (12.3% versus 2.9%).

In the CAPTIVATE FD cohort, at least 1 AE was reported in 158 (99.5%) patients, and AEs of grade 3 or 4 were reported in 98 (61.6%) patients. The most commonly reported AEs were diarrhea (62.3%), nausea (42.8%), neutropenia (41.5%), and arthralgia (33.3%). Grade 3 or 4 neutropenia was reported in 52 (32.7%) patients.

Serious Adverse Events

In the GLOW study, more patients in the ibrutinib-venetoclax arm experienced at least 1 SAE of any grade (46.2% versus 27.6% for ibrutinib-venetoclax versus chlorambucil-obinutuzumab). Of these SAEs, the most reported SAEs that were more frequently reported in the ibrutinib-venetoclax arm were cardiac disorders (13.2% versus 2.9%), infections (12.3% versus 8.6%), general disorders and administration-site conditions (5.7% versus 1.9%), skin and subcutaneous tissue disorders (4.7% versus 0%), and gastrointestinal disorders (3.8% versus 1.9%). Atrial fibrillation was reported more frequently in the ibrutinib-venetoclax arm (6.6% versus 0%) whereas TLS was more common in the chlorambucil-obinutuzumab arm (0% versus 2.9%).

In the CAPTIVATE FD cohort, 36 (22.6%) patients experienced at least 1 SAE of any grade and 30 (18.9%) patients experienced at least 1 SAE of grade 3 or 4.

Withdrawal Due to Adverse Events

In the GLOW trial, more patients reported AEs leading to discontinuation in the ibrutinib-venetoclax arm than in the chlorambucil-obinutuzumab arm (20.8% versus 7.6% for ibrutinib-venetoclax versus chlorambucil-obinutuzumab).

In the CAPTIVATE FD cohort, 8 (5.0%) patients had AEs leading to ibrutinib discontinuation and 3 (1.9%) had AEs leading to venetoclax discontinuation.

Mortality

In the GLOW trial, all patients had completed the study treatment at the time of the primary analysis (data cut-off: February 26, 2021). Overall, 11 (10.4%) patients in the ibrutinib-venetoclax arm and 12 (11.4%) patients in the chlorambucil-obinutuzumab arm died during the study period. AEs were the most frequent cause of death in the ibrutinib-venetoclax arm (6.6% versus 1.9% for ibrutinib-venetoclax versus chlorambucil-obinutuzumab), whereas other causes were the most frequent cause of death in the chlorambucil-obinutuzumab arm (2.8% versus 9.5%).

In the CAPTIVATE FD cohort, 3 (1.9%) patients died during the study period.

Notable Harms

AEs of special interest included atrial fibrillation and major hemorrhages. In the GLOW study, more patients reported atrial fibrillation (14.2% versus 1.9% for ibrutinib-venetoclax versus chlorambucil-obinutuzumab). Grade 3 or 4 atrial fibrillation was reported in 2 (1.9%) patients in the ibrutinib-venetoclax arm and in no patients in the chlorambucil-obinutuzumab arm. Similarly, a higher proportion of patients reported major hemorrhage in the ibrutinib-venetoclax arm than the chlorambucil-obinutuzumab arm (3.8% versus1.0% for ibrutinib-venetoclax versus chlorambucil-obinutuzumab).

In the CAPTIVATE FD cohort, 7 (4.4%) patients had atrial fibrillation, and 2 (1.3%) of these events were grade 3 or 4. Major hemorrhage was reported in 3 (1.9%) patients.

Table 19: Summary of Harms — Pivotal and RCT Evidence (Safety Analysis Set for GLOW; All-Treated Analysis Set for CAPTIVATE)

AE = adverse events; FD = fixed duration; NR = not reported (when incidence rates fall below the specified reporting threshold in either treatment arm within the specified AE severity); RCT = randomized controlled trial; SAE = serious adverse event; TEAE = treatment-emergent adverse event; TLS = tumour lysis syndrome.

Note: At the GLOW extended 18-month follow-up from primary analysis (data cut-off: August 25, 2022), no changes in safety data were noted, except for 1 additional patient in the chlorambucil-obinutuzumab treatment arm who was diagnosed with 2 new treatment-emergent SAEs (myelodysplastic syndrome and myeloproliferative neoplasm) after primary analysis and 7 additional patients who were reported to have developed a nontreatment-emergent secondary malignancy (4 patients from the ibrutinib-venetoclax arm and 3 patients from the chlorambucil-obinutuzumab arm).

^aFor the GLOW study, any-grade TEAEs occurring in ≥ 20% of patients in any treatment arm by system organ class and preferred term, and grade 3 or 4 TEAEs occurring in ≥ 2% of patients in either arm by system organ class and preferred term are reported.

^bFor the CAPTIVATE study, any-grade TEAEs occurring in ≥ 20% of patients by MedDRA preferred term and maximum severity, and grades 3 or 4 TEAEs occurring in ≥ 2% of patients by MedDRA preferred term and maximum severity are reported.

^cFor the GLOW study, any-grade and grade 3 or 4 treatment-emergent SAEs occurring in ≥ 2% of patients in any treatment arm by system organ class and preferred term are reported.

^dFor the CAPTIVATE study, any-grade and grade 3 or 4 serious TEAEs by MedDRA preferred term and maximum severity are reported if they occurred in ≥ 2% patients.

Sources: GLOW Primary Analysis CSR;²³ CAPTIVATE CSR.²⁶ (Note: details from the table have been taken from the sponsor’s Summary of Clinical Evidence.)

Critical Appraisal

Internal Validity

Fludarabine-Ineligible Population

The GLOW trial was a phase III, randomized study evaluating the efficacy and safety of the combination ibrutinib-venetoclax compared with chlorambucil-obinutuzumab in the first-line treatment of patients with CLL who are not eligible to fludarabine-based therapy. In the GLOW trial, patients were assigned to a treatment group via a computer-generated randomization schedule, which involved using IGHV mutation status and the presence of del(11q) as stratification factors were considered appropriate. There were imbalances reported in CIRS scores and LDH elevations, which are considered prognostic factors by the clinical expert consulted by CADTH. More patients in the ibrutinib-venetoclax arm than in the chlorambucil-obinutuzumab arm had a CIRS score higher than 6, and TP53 mutations were reported more frequently in the ibrutinib-venetoclax arm, which may bias the study against the ibrutinib-venetoclax arm, and more patients in the chlorambucil-obinutuzumab arm had LDH elevation, which may bias the study results in favour of the ibrutinib-venetoclax arm.

For the GLOW trial, the CADTH review team considered the open-label design to be reasonable, given the distinct dosing regimens of ibrutinib-venetoclax and chlorambucil-obinutuzumab, which would likely allow investigators and patients to make inferences about treatment assignment regardless of blinding. The CADTH review team noted that the open-label design of the GLOW trial had the potential to introduce reporting bias in the assessment of subjective outcomes reported by patients, such as HRQoL and AEs. Disease response outcomes (i.e., PFS, ORR, DOR, and CR rate) were assessed by IA and by IRC to help mitigate the biases associated with the open-label study design for the GLOW trial.

In the GLOW trial, the overall median duration of exposure was substantially longer in the ibrutinib-venetoclax arm (13.8 months) than for the chlorambucil-obinutuzumab arm (5.1 months). In addition, fewer patients in the chlorambucil-obinutuzumab arm discontinued treatment due to AEs compared to the ibrutinib-venetoclax arm, and a higher proportion of patients completed the study treatment in the chlorambucil-obinutuzumab arm than the ibrutinib-venetoclax arm, which indicates that patients in the chlorambucil-obinutuzumab arm had better treatment compliance than those in the ibrutinib-venetoclax arm in the GLOW trial, this may bias the results against the ibrutinib-venetoclax arm. A higher proportion of patients received concomitant analgesics, antihistamines, and corticosteroids in the chlorambucil-obinutuzumab arm than the ibrutinib-venetoclax arm in the GLOW trial. Although the clinical expert did not think the imbalances in concomitant treatments between treatment arms would bias the efficacy results, as these concomitant medications do not have any anti-CLL activity, except for corticosteroids, which were given at relatively low doses, the CADTH review team noted that these imbalances may bias the safety and HRQoL results against the ibrutinib-venetoclax arm. According to the clinical expert,, the use of concomitant medications is possibly a reflection of a patient population with comorbidities. Regarding subsequent anticancer therapies, in the GLOW trial, a higher proportion of patients in the chlorambucil-obinutuzumab arm received subsequent anticancer therapy compared to the ibrutinib-venetoclax arm. The clinical expert consulted by CADTH indicated that the use of subsequent therapies would influence OS. The CADTH review team agreed with the clinical expert and noted that the use of subsequent anticancer therapy results in an indirectness of the estimated OS effect. It is difficult in this setting to isolate the direct effect of ibrutinib-venetoclax treatment on OS due to the intercurrent use of subsequent anticancer therapies.

In the GLOW study, patients in both treatment arms had major protocol deviations. Most frequently, the wrong treatment or an incorrect dose was administered, which may increase uncertainty in the estimate of the treatment effect due to increased treatment heterogeneity and may be attributed to the difference between the treatment of interest and the treatment implemented.

Many of the outcomes used in the GLOW trial (PFS, OS, ORR, CR rate, DOR) are standard in oncology trials and in alignment with the end points used for similar indications for ibrutinib and other CLL therapies.⁵¹ In the GLOW trial (fludarabine-ineligible patients) the primary efficacy end point investigated was PFS as measured by IRC; PFS per IA was reported as a sensitivity analysis. Both IA and IRC used the iwCLL criteria for CLL, which were considered adequate for response assessment by CADTH and Health Canada.⁵¹ The primary end point was met at the February 26, 2021, cut-off for the GLOW trial as PFS per IRC demonstrated superiority in the ibrutinib-venetoclax arm compared to the chlorambucil-obinutuzumab arm.

The median of OS was not reached in the ibrutinib-venetoclax arm in the GLOW trial in the primary analysis, and OS data were considered immature. Although the sponsor conducted subsequent final PFS analyses in the extended follow-up analyses, there is uncertainty about the effect of ibrutinib-venetoclax on long-term OS for fludarabine-ineligible patients with CLL. In addition, in the GLOW trial, the sponsor conducted a sensitivity analysis that used death due to COVID as a censor in the analysis of OS; the CADTH review team considers censoring patients who died due to COVID is not sufficient to estimate the direct effect of treatment on OS in a COVID-free population, as the patients who died of COVID may have been different than the patients who stayed on trial in terms of other variables and baseline characteristics.

A serial gatekeeping procedure was employed to control the overall type I error for the primary end point and selected secondary end points in the GLOW trial for the primary analysis. Of note, the hierarchical statistical testing strategy ended at ORR per IRC. The remaining key secondary end points (i.e., OS, sustained hematological improvements, and time to improvement in FACIT-Fatigue score) and ORR per IA were considered not statistically significant due to the failure of ORR to demonstrate statistical significance, per statistical testing hierarchy. Therefore, any results with a P value less than the prespecified significance level should be interpreted with caution, considering the potentially inflated type I error rate. Although the subgroup analyses were prespecified, there is no evidence that the studies were powered to detect subgroup differences. In addition, no data imputation was performed in the GLOW trial.

HRQoL is considered a relevant outcome by patients with CLL and by clinicians. The EORTC QLQ-C30, EQ-5D-5L utility, EQ-5D VAS, and FACIT-Fatigue measurements used for HRQoL assessment in the GLOW study are commonly used in oncology trials; however, the validity, reliability, and responsiveness of these measurements have not been studied in patients with CLL. The minimal clinically important differences proposed by the sponsor were based on patients with breast cancer (EORTC QLQ-C30) or multiple types of cancer (EQ-5D-5L utility, and VAS, and FACIT-Fatigue). Given that AEs were the most common cause of discontinuation in the GLOW trial, there is a risk of reporting bias results from patients who remained in the trial, impacting the interpretability of HRQoL trends over time. Moreover, there was no multiplicity adjustment for the analysis of HRQoL, which may have further added uncertainty to the study results.

Fludarabine-Eligible Population

The CAPTIVATE trial was a phase II, multicohort, international trial with 2 cohorts: the MRD cohort and the FD cohort. This review focused on the CAPTIVATE FD cohort, which assessed the efficacy and safety of ibrutinib-venetoclax in fludarabine-eligible patients with treatment-naive CLL or SLL. Because the FD cohort of the CAPTIVATE trial was designed as a single-arm study, given the lack of a comparator arm, the ability to make definitive conclusions about the comparative efficacy of ibrutinib-venetoclax in fludarabine-eligible patients with CLL is limited. In addition, the open-label design had the potential to introduce reporting bias in the assessment of subjective outcomes reported by patients (i.e., AEs). Disease response outcomes (PFS, ORR, DOR, CR rate) were assessed by IA and by IRC to help mitigate the biases associated with the open-label study design for the FD cohort in the CAPTIVATE trial. Many of the outcomes used in the CAPTIVATE FD cohort (PFS, OS, ORR, CR rate, DOR) are standard in oncology trials and in alignment with the end points used for similar indications for ibrutinib and other CLL therapies. The primary efficacy end point was CR rate per IA; CR per IRC was also reported as a sensitivity analysis. Both IA and IRC used the iwCLL criteria for CLL, which were considered adequate for response assessment by CADTH and Health Canada.⁵¹

The primary end point for the CAPTIVATE FD cohort was met at the time of primary analysis (data cut-off: November 12, 2020) because the CR rate per IA in patients without del(17p) exceeded the prespecified minimum CR rate of 37%. The medians of OS and PFS were not reached in the CAPTIVATE FD cohort at the time of primary analysis, and OS data were considered immature. Therefore, there is uncertainty in the interpretation of long-term OS and PFS results for FCR-eligible patients with CLL in the CAPTIVATE FD cohort. In addition, the CADTH review team noted that the CAPTIVATE FD cohort included a subgroup of patients with del(17p); however, there was no formal statistical testing performed between subgroups and the sample size for patients with del(17p) is small (n = 20), thus no conclusions can be drawn from the study results for patients with del(17p). Although the subgroup analyses were prespecified, there is no evidence that the study were powered to detect subgroups differences. HRQoL is considered a relevant outcome by patients with CLL and by clinicians, but there was no assessment of HRQoL in the CAPTIVATE FD cohort, thus it is uncertain whether treatment with ibrutinib-venetoclax would improve HRQoL in fludarabine-eligible patients with CLL.

External Validity

Both the GLOW trial and CAPTIVATE FD cohort required eligible patients to have measurable nodal disease; however, according to the clinical expert consulted by CADTH, there is a small proportion of patients who only have elevated white blood cell counts and cytopenia and may not have an enlarged lymph node in clinical practice. These patients are important and would fit in the patient population for the ibrutinib-venetoclax regimen. In addition, the GLOW trial excluded patients with galactose intolerance, Lapp lactase deficiency, or glucose-galactose malabsorption from study enrolment at Protocol Amendment 1 made on June 6, 2018, which is not reflective of real-world clinical practice, as there are patients with lactose intolerance who take ibrutinib based on the feedback provided by the clinical expert consulted by CADTH. Furthermore, the exclusion of patients who are intolerant to galactose may underestimate the risk of abdominal pain and diarrhea in patients with CLL, as the formulation of ibrutinib contains lactose, and symptoms of lactose intolerance (i.e., abdominal pain and diarrhea) can occur in these patients, according to the clinical expert. The clinical expert stated that there is a more diversified patient population, including patients from Asia and other parts of the world, in heir clinical practice compared to the patient population in the GLOW trial and CAPTIVATE FD cohort. The baseline characteristics of the 2 studies may be indicative of the overrepresentation of white patients (≥ 92%) with CLL in both fludarabine-eligible and fludarabine-ineligible populations; this presents an evidence gap in the generalizability of results to all patients. Although the inclusion criteria mandated that eligible patients in the GLOW trial have a CrCl of less than 70 mL/min and/or a CRIS score greater than 6, the baseline CrCl for both treatment arms was higher than what the clinical expert would expect in high-risk patients with CLL who are ineligible to fludarabine-based therapy in clinical practice, which may indicate that patients in the GLOW trial have better kidney function than patients in clinical practice. This may compromise the generalizability of the study results to the general FCR-ineligible patients with CLL. Generally, the risk profile of patients in the CAPTIVATE FD cohort is what the clinical expert would expect in fludarabine-eligible patients with CLL in clinical practice.

The GLOW trial used chlorambucil-obinutuzumab as the treatment for the comparator arm. The clinical expert comment that chlorambucil-obinutuzumab may not be the best choice of comparator for ibrutinib-venetoclax, as it is an old historic treatment being used as a standard therapy in an older population of patients with CLL, which is about 5% of the patient population in the clinical expert’s clinical practice. Moreover, according to the clinical expert, the dosing regimen for the chlorambucil-obinutuzumab arm used in the GLOW trial (i.e., 6 cycles of 28 days) is a standard regimen commonly used in clinical practice; however, chlorambucil can be given in many different ways based on physician decision and patient preference in clinical practice in Canada (i.e., having a pause for 4 days every month or receiving daily low-dose chlorambucil), and there are variations in treatment length (i.e., 12-month duration). These factors may introduce uncertainty in the estimate of the treatment effect in practice in Canada due to limitations in the generalizability of the study results.

Long-Term Extension Studies

No long-term extension studies were identified for this review.

Indirect Evidence and Observational 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

No comparative evidence was available in the pivotal trials for patients with CLL who were eligible for fludarabine nor direct comparative data to treatment regimens other than chlorambucil-obinutuzumab in the fludarabine-ineligible population. This section provides a summary and appraisal of the indirect and observational evidence submitted by the sponsor to address these knowledge gaps. These data were used to inform the pharmacoeconomic model.

Description of Indirect Comparisons and Observational Evidence

For patients who were ineligible to received fludarabine, the sponsor provided 2 ITCs and 2 IPD observational studies that evaluated the efficacy and safety of ibrutinib-venetoclax versus BR,²⁷ ibrutinib,²⁸ VO,²⁹ and acalabrutinib.³⁰ In fludarabine-eligible patients, an IPD analysis was conducted comparing ibrutinib-venetoclax versus fludarabine in combination with cyclophosphamide and rituximab.³¹ The ITC analyses were based on MAIC methods (comparison with acalabrutinib and VO) and the IPD analyses were based on propensity score methods (comparison with BR, ibrutinib, and FCR).

Indirect Treatment Comparison Design

Objectives

The aims of the sponsor-submitted analyses were to estimate the relative efficacy and safety of ibrutinib-venetoclax versus other treatment options in Canada for patients with previously untreated CLL.

Study Selection Methods

A systematic literature review was conducted to identify clinical trials evaluating first-line treatment of CLL in fludarabine-ineligible and fludarabine-eligible patients. The date of the last search update was October 27, 2022. The study selection criteria were broader than in the ITC criteria with respect to population, comparators, and outcomes. A summary of the selection criteria is presented in Table 20.

Although the sponsor reported having conducted a systematic review, no further information was provided on the number of trials that met the inclusion criteria, or their characteristics. According to the sponsor’s reports,^27-31 the comparator studies used in the ITCs were selected based on the availability of IPD, had inclusion criteria that were similar to the ibrutinib-venetoclax trials (GLOW or CAPTIVATE FD cohort), or were used for regulatory approval of the drug.

Table 20: Study Selection Criteria and Methods for ITCs Submitted by the Sponsor

AE = adverse event, BR = bendamustine plus rituximab, CLL = chronic lymphocytic leukemia, CR = complete response, FCR = fludarabine plus cyclophosphamide plus rituximab, ITC = indirect treatment comparison, ORR = overall response rate, OS = overall survival, PFS = progression-free survival, RCT = randomized controlled trial, TTNT = time to next treatment, VO = venetoclax plus obinutuzumab.

Source: Sponsor’s Summary of Clinical Evidence.⁶⁰

ITC Analysis Methods

In fludarabine-ineligible patients, 2 pairwise comparisons were conducted using MAIC methods to compare the relative efficacy and safety of ibrutinib-venetoclax with VO²⁹ and acalabrutinib.³⁰

Ibrutinib-Venetoclax Versus Acalabrutinib

The sponsor identified 2 RCTs that used a common comparator (chlorambucil-obinutuzumab) that could be used for the ITC comparing ibrutinib-venetoclax to acalabrutinib; however, there were differences in the recruited populations in these 2 trials and the sponsor concluded that a Bucher or network meta-analysis approach would yield biased results. IPD were available for ibrutinib-venetoclax from the GLOW trial (median follow-up 46 months) and aggregate data were available for acalabrutinib from the ELEVATE-TN study (median follow-up 46.9 months). Therefore, the sponsor opted to use anchored MAIC methods to compare the efficacy of ibrutinib-venetoclax versus acalabrutinib. Detailed methods, including outcomes, model estimation, covariates, and sensitivity analyses, are outlined in Table 21.

To conduct the MAIC, first the inclusion and exclusion criteria of the 2 trials were reviewed to determine the degree of overlap between the studies. The ELEVATE-TN study included patients who were not eligible to enrol in the GLOW study; specifically, patients with del(17p) or known TP53 mutation were included in the ELEVATE-TN study but not in the GLOW study. Given that the inclusion criteria for the ELEVATE-TN study were broader than the GLOW study, there were no criteria identified that could be used to exclude patients from the GLOW study that would otherwise have been excluded from the ELEVATE-TN study. In the next step of the analysis, patients from the GLOW study were weighted to match the mean values of relevant baseline characteristics for the comparator trial. Weighting was based on a propensity score model, in which patients in the GLOW study were weighted by the inverse odds of being in the GLOW trial rather than in the ELEVATE-TN study. The weighting used a generalized method of moments model to estimate the propensity scores, according to Signorovitch et al. (2012),⁵¹ and followed guidance from the NICE Decision Support Unit.^61,62 Four variables were selected for the base-case propensity score model (age, ECOG PS, CIRS score, TP53 status) that were deemed most important by a UK clinical advisory board. The fully matched analysis included 17 variables that were reported for both studies. Region was not included in the fully matched model due to poor overlap between studies. Patients with missing covariate data were excluded from the propensity score model, under the assumption that they were missing completely at random.⁶³ The outcomes planned for analysis were PFS, OS, and TTNT. No safety end points were analyzed. The Engauge Digitizer software was used to digitize Kaplan-Meier curves from the ELELVATE-TN study, which were then used to generate IPD (time and censoring status) for each curve based on the method by Guyot et al. (2012).⁶⁴ Weighted patient data from the GLOW trial and simulated patient data from the ELEVATE-TN trial were analyzed using a weighted Cox proportional hazard model.

A time-varying Cox model was used for sensitivity analyses under a piecewise constant hazard assumption. The rationale for these analyses was that the proportional hazards assumption was violated for PFS and OS, and these violations persisted after matching. These analyses were run with a cut point of 12 months for PFS and 15.5 months for OS, based on the observed change in shape for the survival curves at these time points. A second sensitivity analysis was run that matched for all 17 variables plus region. In either case, the patient data from the GLOW trial were weighted using the propensity scores from the MAIC analysis.

Ibrutinib-Venetoclax Versus VO

The sponsor identified 2 RCTs that used a common comparator (chlorambucil-obinutuzumab) that could be used for the ITC comparing ibrutinib-venetoclax to VO. The sponsor concluded that a Bucher or network meta-analysis approach would yield biased results, as there were significant differences in the patient populations enrolled in the trials and the distribution of effect modifiers. Anchored MAIC methods were used to compare the efficacy of ibrutinib-venetoclax versus VO using IPD for ibrutinib-venetoclax from the GLOW trial (median follow-up 46 months) and aggregate data for VO from Study CLL14 (median follow-up 39.6 and 54.4 months). The methods used were largely the same as those used for the comparison with acalabrutinib; additional details are outlined in Table 21.

Based on the sponsor’s review of the inclusion and exclusion criteria of the 2 trials, patients who would not have qualified for entry in the CLL14 study were excluded from the GLOW dataset (CIRS score less than 6 and CrCl greater than 70 mL/min). Patients in the GLOW dataset were then weighted using the methods described for the acalabrutinib ITC. The selection and ranking of covariates used in the weighting were based on the availability of data and ranking by a UK clinical advisory board. The base-case analysis was matched for 4 variables (age, ECOG PS, CIRS score, TP53 status), and the fully weighted model included 5 additional variables (IGHV status, CrCl, sex, beta-2M level, and median time since initial diagnosis). The sponsor stated that the Binet stage was not included in the fully adjusted model due to poor overlap between trials on this variable. Patients with missing covariate data were excluded from the propensity score model under the assumption that they were missing completely at random.⁶³

The outcomes planned for analysis were PFS, OS, and TTNT. No safety end points were analyzed. The primary analyses used 39.6-month median follow-up data from the CLL14 study, with a sensitivity analysis using 54.4-month data. Time-varying Cox models were run as sensitivity analyses with piecewise constant hazards for PFS (≤ 12 months, > 12 months) and OS (≤ 15.5 months, > 15.5 months), as noted for the comparison with acalabrutinib. A third sensitivity analysis was run that matched on 10 variables, including Binet stage.

Observational Studies Analysis Methods

In patients who were fludarabine-ineligible, 2 IPD analyses were conducted comparing ibrutinib-venetoclax to BR²⁷ and to ibrutinib monotherapy.²⁸ One IPD analysis was conducted in fludarabine-eligible patients to compare the relative efficacy and safety of ibrutinib-venetoclax with FCR.³¹

Ibrutinib-Venetoclax Versus BR

The sponsor had access to IPD from the ALLIANCE study, which enrolled a patient population similar to that in the GLOW study. However, the sponsor concluded that there were sufficient differences in the populations enrolled that a naive comparison, or a Bucher method analysis, would likely be biased. Instead, an analysis was conducted using IPD for ibrutinib-venetoclax from the GLOW study (median follow-up 46 months) and for BR from the ALLIANCE study (median follow-up 38 months). Common inclusion criteria from the trials were applied and patients were included in the analyses if they met the following criteria: did not have del(17p), were at least 65 years of age, had a Rai stage of I to IV, absolute neutrophil count of 1,000 cells/μL or greater, alanine aminotransferase and aspartate aminotransferase less than or equal to 3.0 times the upper limit of normal, and a platelet count of 50,000 cells/μL or greater.

The treatment propensity score weights were estimated using logistic regression for use in inverse probability treatment weighing estimation methods to compare the efficacy of ibrutinib-venetoclax versus BR. Average treatment among the treated (ATT) weighting was the primary analysis. Kaplan-Meier curves and weighted log-rank tests were also used to further present comparisons for PFS and OS. Analyses of CR and ORR were based on a weighted logistic regression model. Secondary weighting schemes included the average treatment among the controls (ATC) and the average treatment among the overlap (ATO). Patients with missing covariate data were assigned to a category “missing,” under the assumption that the reason of missingness is similar in both cohorts.⁶³

Details of the analyses conducted, including methods, outcomes, covariates, and sensitivity analyses, are outlined in Table 21. The selection of variables for inclusion in the propensity model was based on a review of the variables adjusted for in prior reports of CLL, as well as the availability of data from both trials. Sensitivity analyses were conducted that excluded patients with missing covariates in the ALLIANCE study, and for the population that had no TP53 mutation as well as no del(17p). The sponsor reported balance diagnostic data in the form of standardized mean difference plots for the covariates included in the propensity model and in the distribution plots of propensity scores and weights.

Efficacy analyses included PFS and OS. Although safety outcomes were planned for analysis (TEAEs, grade 3 or 4 TEAEs, treatment discontinuation, and time to treatment discontinuation [TTD]), the sponsor stated that due to the limitations with the data available from the ALLIANCE trial, safety end points were not analyzed.⁶³

Ibrutinib-Venetoclax Versus Ibrutinib

The sponsor stated that there were no RCTs comparing chlorambucil-obinutuzumab to ibrutinib, thus a network meta-analysis was not possible for this comparison. IPD were available for ibrutinib-venetoclax from the GLOW study and for ibrutinib from the ALLIANCE study. For this comparison, the same approach was taken as for the comparison of ibrutinib-venetoclax versus BR (Table 21). One additional sensitivity analysis was conducted based on pooled data from the ibrutinib and the ibrutinib plus rituximab groups in the ALLIANCE study.

Ibrutinib-Venetoclax Versus FCR

For patients who were eligible for fludarabine, IPD were available for ibrutinib-venetoclax from the FD cohort of the CAPTIVATE study (median follow-up 38.7 months and 49.7 months) and from the FCR group in the E1912 study (median follow-up of 48 months). For this comparison, the same approach was taken as for the ibrutinib-venetoclax comparisons versus BR and ibrutinib, with propensity score matching and inverse probability treatment weighting methods (ATT, ATC, and ATO weights). The outcomes analyzed included PFS, OS, ORR, CR, TEAEs, grade 3 or 4 TEAEs, treatment discontinuations, and TTD. Detailed methods, including outcomes, model estimation, covariates, and sensitivity analyses, are outlined in Table 21.

Based on the inclusion criteria that were common to both cohorts, the base-case population included adults aged 18 to 70 years with previously untreated CLL who were eligible to receive fludarabine, did not have the del(17p), and who had received at least 1 dose of the study drug. The selection of variables for inclusion in the propensity model was based on a review of the variables adjusted for in prior reports of CLL, as well as on the availability of data from both trials.

Three sensitivity analyses were planned that excluded patients with the TP53 mutation (3% and 5% in the CAPTIVATE and E1912 studies, respectively). The sponsor stated that TP53 mutation is of equal adverse prognostic value as the deletion of 17p, and therefore these patients are not suitable for FCR therapy. A second sensitivity analysis excluded patients with known TP53 mutation or missing TP53 status (18% in the E1912 study). A third analysis was limited to patients with a very good prognosis for treatment with FCR and included patients younger than 65 years with good renal function (CrCl ≥ 70 mL/min) and no TP53 mutation.

Table 21: Indirect Comparison and Observational Study Analysis Methods