CADTH Reimbursement Review

Darolutamide (Nubeqa)

Sponsor: Bayer Inc.

Therapeutic area: Metastatic castration-sensitive prostate cancer

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.

Introduction

Prostate cancer is the most common cancer and the third leading cause of death from cancer in men living in Canada.¹ It is estimated that 24,600 men will be diagnosed with prostate cancer in 2022.¹ Metastatic castration-sensitive prostate cancer (mCSPC) refers to cancer that responds to androgen deprivation therapy (ADT). Bone metastasis (84%) is common and often accompanied by bone pain.² Disease-related skeletal complications such as pathological fracture and spinal cord compression can occur.³ Progression to metastatic castration-resistant prostate cancer (mCRPC), a disease state associated with morbidity, poor health-related quality of life (HRQoL), and poor prognosis, occurs despite standard of care (SoC) treatment. The median survival for mCSPC is approximately 5 years, with a large variability in survival among patients, according to the clinical experts consulted by CADTH.

Treatment intensification with docetaxel or androgen receptor-axis-targeted therapy (ARAT) (i.e., enzalutamide, apalutamide, or abiraterone plus prednisone) in addition to ADT is the current SoC in men with mCSPC and has been shown to improve survival and delay disease progression. Recently, “triplet therapy” with abiraterone and prednisone plus docetaxel and ADT is also observed in clinical practice in light of new clinical trial evidence.⁴ This triplet therapy is currently under review at CADTH for mCSPC. Treatment intensification is the SoC; however, there is currently no consensus on the optimal treatment intensification strategy for mCSPC. The choice of therapy, in general, depends on functional assessment, comorbidities, patient preference, drug toxicity profiles, and drug accessibility, as per the clinical experts.

Darolutamide is an ARAT that is indicated for the treatment of patients with mCSPC in combination with docetaxel. The product monograph recommends darolutamide 600 mg (2 tablets of 300 mg) to be given twice daily orally with food until disease progression or unacceptable toxicity. Patients receiving this drug should receive a gonadotropin-releasing hormone (GnRH) analogue concurrently or have had a bilateral orchiectomy.⁵ Darolutamide was previously reviewed by CADTH for the treatment of non-mCSPC.

The objective of this review is to perform a systematic review of the beneficial and harmful effects of darolutamide 300 mg oral tablet in combination with docetaxel for the treatment of patients with mCSPC who are chemotherapy-eligible.

Stakeholder Perspectives

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

Patient Input

Two patient groups, the Canadian Cancer Survivor Network (CCSN) and the Canadian Cancer Society, submitted patient group input for this review. The CCSN gathered data through an online survey with responses from 24 patients with prostate cancer (6 of whom were diagnosed with metastatic disease) and from 2 caregivers. The Canadian Cancer Society conducted surveys and interviews with 39 patients with mCSPC and with 2 caregivers. In both submissions, all patients were living in Canada and a total of 8 patients had experience with darolutamide.

The Canadian Cancer Society respondents indicated that symptoms associated with mCSPC had a moderate-to-severe negative impact on their ability to engage in sexual activity, work, exercise, and travel; fulfill family obligations; and maintain their mental health. Common side effects following currently available treatments by patients in both submissions included changes in libido and sexual function, hot flashes, fatigue, loss of muscle mass, incontinence, and weight gain. Five patients from the CCSN submission rated how their experience with darolutamide compared to other treatments, with 4 of these patients indicating that darolutamide was easier to use and better addressed disease progression, 3 patients stating that they experienced a reduction in side effects compared to current treatments, and 1 indicating that it was better able to control their symptoms. Adverse events (AEs) identified among respondents with experience with darolutamide were consistent with those of currently available treatments. Across both submissions, respondents reported that they would like to see future treatments that delay the onset of symptoms and the need for chemotherapy, have fewer side effects, improve survival, are easy to use, and allow them to maintain their HRQoL.

Clinician Input

Input From Clinical Experts Consulted by CADTH

The clinical experts noted that although there are currently 2 available treatment-intensification strategies using chemotherapy or ARAT that can improve long-term outcomes beyond ADT alone, they are not curative. There remains a compelling need to extend survival longer while improving and maintaining quality of life overall. The clinical experts expected triplet therapy with darolutamide plus docetaxel and ADT to be considered as a first-line treatment option for men with mCSPC who are eligible for cytotoxic chemotherapy. The clinical experts noted that a major shift in the prescribing pattern is unlikely in the absence of direct evidence between darolutamide triplet therapy versus ARAT plus ADT, the most common regimen currently prescribed in Canada.

The clinical experts noted that there is no consensus among clinicians in Canada on who should be offered triplet therapy, and that the decision usually involves a case-by-case discussion between the patient and the treating physician. The clinical experts agreed that triplet therapy with darolutamide, docetaxel, and ADT should be available to all men with mCSPC who are candidates for cytotoxic chemotherapy.

The clinical experts noted that treatment response is, in general, evaluated based on clinical status, radiologic response, and prostate-specific antigen (PSA) response, and that the frequency of assessment is highly variable in clinical practice. The clinical experts noted that treatment discontinuation is considered in patients who have unacceptably toxicities or disease progression (clinical, PSA, and/or radiologic progression). In addition, the clinical experts noted that rapid progression and the absence of ongoing clinical benefit may warrant the initiation of a subsequent line of therapy.

Clinician Group Input

Clinician group input was received from 7 groups: the BC Cancer Agency (12 clinicians); the Canadian Cancer Society (12 clinicians); genitourinary oncologists from the Maritime provinces (5 clinicians); the Allan Blair Cancer Centre (5 clinicians); the Ottawa Hospital Cancer Centre — Genitourinary Oncology Group (3 clinicians); the Ontario Health — Cancer Care Ontario Genitourinary Cancer Drug Advisory Committee (4 clinicians); and the Genitourinary Disease Site Group of the Cancer Centre of Southeastern Ontario (2 clinicians). The various submissions noted that current treatment goals are to reduce symptom burden, prolong survival, and delay disease progression. The clinician groups noted that current treatment for mCSPC includes either chemotherapy (docetaxel) or an ARAT (i.e., abiraterone acetate plus prednisone, enzalutamide, or apalutamide) in combination with ADT. The clinician groups noted that mCSPC is incurable and that many patients fail systemic treatment shortly after treatment initiation. The groups emphasized a significant unmet need for treatments that further improve survival, increase HRQoL, and increase duration of treatment response while providing less toxicity burden. In terms of place in therapy, the submissions stated that darolutamide would be used as a first-line treatment for mCSPC in combination with ADT and docetaxel in patients who are fit for chemotherapy. The submissions stated that response to treatment would be assessed using PSA response, radiographic response, and clinical assessment (i.e., worsening pain or symptoms). Treatment would typically be discontinued upon disease progression (PSA, radiological, or symptomatic progression) or unacceptable toxicities.

Drug Program Input

The drug plans noted interest in understanding the treatment eligibility of patients with regional lymph node metastases, Eastern Cooperative Oncology Group (ECOG) performance status greater than or equal to 2, prior ADT in the adjuvant setting, and recent initiation of ADT in the metastatic setting, as well as identifying the appropriate criteria for treatment discontinuation and switch in therapy.

The clinical experts noted that, to receive triplet therapy, patients should have hormone sensitivity at the onset of treatment. Therefore, eligible patients would include those who are naive to ADT or who had received ADT within 6 months of starting the intensification, or metastatic patients who had completed adjuvant ADT more than 1 year prior. Patients who develop nodal disease alone should also be considered for triplet therapy. The clinical experts noted that the requirement for ECOG performance status is less stringent in regular practice than the inclusion criteria of ECOG performance status greater than or equal to 2 in the ARASENS trial. Patients with a poor baseline performance status should not be necessarily excluded from treatment and their overall medical status should be thoroughly assessed, as per the clinical experts.

The clinical experts noted that patients who are unable to tolerate all 6 cycles of docetaxel should be allowed to continue darolutamide plus ADT and that it is reasonable for patients who are unable to tolerate darolutamide to switch to an alternative ARAT (i.e., abiraterone) plus docetaxel plus ADT. The clinical experts commented that it would be reasonable to align discontinuation criteria for darolutamide with other ARATs. The clinical experts noted that the treatment should be discontinued upon unacceptable toxicity or disease progression as determined by the composite of clinical, radiological, and PSA determinants.

The clinical experts commented that the addition of darolutamide to docetaxel plus ADT would be reasonable if done at the onset of treatment intensification before the completion of docetaxel cycles. They noted that a switch in treatment regimen in patients who have well-controlled disease on an established regimen is unlikely in clinical practice. It would be uncommon for a treating physician to consider switching from ARAT plus ADT to triplet therapy except in the early phase of treatment (i.e., within 3 months of ADT initiation).

Clinical Evidence

Pivotal Studies and Protocol-Selected Studies

Description of Studies

The CADTH systematic review identified 1 relevant study, ARASENS,⁶ which was a phase III, double-blind, randomized controlled trial (RCT) comparing darolutamide (600 mg twice daily) to placebo, in combination with docetaxel (75 mg/m² via IV infusion every 21 days for 6 cycles) and ADT in patients with de novo or metachronous mCSPC (N = 1,306) in a first-line metastatic setting. Patients were randomized to the treatment arms on a 1:1 ratio and continued treatment until disease progression (symptomatic disease progression or a change of therapy) or unacceptable toxicity. The primary end point was overall survival (OS), and the secondary end points were time to castration-resistant prostate cancer (CRPC), time to initiation of subsequent systemic antineoplastic therapy, time to pain progression, time to first symptomatic skeletal event (SSE), and symptomatic skeletal event-free survival (SSE-FS).

In the final efficacy analysis (data cut-off on October 25, 2021), the median age of patients at baseline was 67.0 (range = 41 to 89) years. The majority of patients were white (52.0%) or Asian (36.4%), and had stage IV disease at initial diagnosis (87.6%), an ECOG performance status of 0 (71.1%), and bone metastases (82.8%) at baseline. Most patients did not receive prior local therapy and no patients had prior systemic antineoplastic therapy for prostate cancer other than ADTs.

Efficacy Results

The key efficacy results from the ARASENS trial are summarized in Table 2.

Overall Survival

In the final OS analysis (primary outcome), the median duration of follow-up in the darolutamide plus docetaxel and ADT arm, and placebo plus docetaxel and ADT arm, was 43.7 months and 42.4 months (standard deviation [SD] not reported in both), respectively. The median OS was not reached in the darolutamide plus docetaxel and ADT arm and was 48.9 (95% confidence interval [CI], 44.4 to not reached [NR]) months in the placebo plus docetaxel and ADT arm, which corresponded to a hazard ratio (HR) of 0.68 (95% CI, 0.57 to 0.80; P < 0.0001), in favour of darolutamide.

Time to CRPC

The median time to CRPC (secondary outcome) was not reached in the darolutamide plus docetaxel and ADT arm, and was 19.1 (95% CI, 16.5 to 21.8) months in the placebo plus docetaxel and ADT arm, with an HR of 0.36 (95% CI, 0.30 to 0.42; P < 0.0001) in favour of darolutamide.

Time to Initiation of Subsequent Systemic Antineoplastic Therapy

The median time to initiation of subsequent systemic antineoplastic therapy (secondary outcome) was not reached in the darolutamide plus docetaxel and ADT arm, and was 25.3 (95% CI, 23.1 to 28.8) months in the placebo plus docetaxel and ADT arm, with a HR of 0.39 (95% CI, 0.33 to 0.46; P < 0.0001), in favour of darolutamide.

Time to Pain Progression

The median time to pain progression (secondary outcome) was not reached in the darolutamide plus docetaxel and ADT arm, and was 27.5 (95% CI, 22.0 to 36.1) months in the placebo plus docetaxel and ADT arm, with a HR of 0.79 (95% CI, 0.66 to 0.95; P = 0.0058), in favour of darolutamide.

Health-Related Quality of Life

HRQoL (exploratory outcome) was measured using the National Comprehensive Cancer Network prostate cancer symptom index 17 item questionnaire/Functional assessment of cancer therapy (NCCN-FACT FPSI-17) questionnaire. The mean total score and subscale scores were similar between treatment arms at baseline and remained stable at most assessment time points until near the end of the treatment, when the scores trended toward deterioration in both arms. There was no notable difference in the mean change in score from baseline between treatment arms at most time points; however, the difference between treatments was not statistically tested.

Objective Response Rate

This outcome was not measured in the study.

Time to First SSE

The median time to first SSE (secondary outcome) was not reached in both treatment arms and the HR was 0.71 (95% CI, 0.54 to 0.94; P = 0.0081), in favour of darolutamide.

Symptomatic Skeletal Event-Free Survival

The median SSE-FS (secondary outcome) was 51.2 (95% CI, 47.2 to NR) months in the darolutamide plus docetaxel and ADT arm, and 39.7 (95% CI, 36.0 to 42.3) months in the placebo plus docetaxel and ADT arm, with a HR of 0.61 (95% CI, 0.52 to 0.72; P < 0.0001), in favour of darolutamide.

PSA Outcomes

The following analyses were exploratory and the difference between treatment arms was not adjusted for multiplicity.

The median time to PSA progression was not reached in the darolutamide plus docetaxel and ADT arm, and was 22.4 (95% CI, 22.1 to 27.6) months in the placebo plus docetaxel and ADT arm, with an HR of 0.26 (95% CI, 0.21 to 0.31).

The risk difference in absolute PSA response rate (the proportion of patients with PSA level below 0.2 ng/mL) between treatment arms was 25.0% (95% CI, 20.0% to 29.9%) at 6 months, and 34.2% (95% CI, 29.2% to 39.1%) at 12 months.

The relative 90% PSA response rate (proportion of patients with at least 90% PSA reduction) was numerically higher in the darolutamide plus docetaxel and ADT arm than the placebo plus docetaxel and ADT arm at 3, 6, and 12 months. Analyses of relative 50% and 30% PSA response rate showed similar findings.

Harms Results

The key harms results from the ARASENS trial are summarized in Table 2.

Table 2: Summary of Key Results From the Pivotal Study

ADT = androgen deprivation therapy; CI = confidence interval; CRPC = castration-resistant prostate cancer; FAS = full analysis set; NR = not reached; SAF = safety analysis set; TEAE = treatment-emergent adverse event.

^aBased on Kaplan-Meier estimates.

^bThe hazard ratio and 95% CI were based on a Cox regression model, stratified by extent of disease (M1a vs. M1b vs. M1c) and alkaline phosphatase (< upper limit of normal vs. ≥ upper limit of normal).

^cP value was derived from the 1-sided log-rank test, stratified by extent of disease (M1a vs. M1b vs. M1c) and alkaline phosphatase (< upper limit of normal vs. ≥ upper limit of normal).

^dPatients with multiple events were only counted for the category in which the first event occurred.

^eTEAEs were defined as any event(s) arising or worsening after the first dose of darolutamide or placebo, until 30 days after the last dose of darolutamide or placebo administration.

Data cut-off: October 25, 2021.

Source: ARASENS Clinical Study Report.⁷

Almost all patients reported at least 1 treatment-emergent adverse event (TEAE) in both treatment arms (99.5% of patients in the darolutamide plus docetaxel and ADT arm, and 99.8% of patients in the placebo plus docetaxel and ADT arm). There was no notable difference in the TEAEs between treatment arms, except for the incidence of decreased appetite of any grade and hypertension of grade 3 or higher, both of which were numerically higher in the darolutamide plus docetaxel and ADT arm than the placebo plus docetaxel and ADT arm (decreased appetite of any grade: 18.6% versus 13.1%; hypertension of grade 3 or higher: 6.6% versus 3.2%). At least 1 serious TEAE was reported in 44.8% of patients in the darolutamide plus docetaxel and ADT arm, and 42.3% of patients in the placebo plus docetaxel and ADT arm, with the most common serious TEAE in both arms being febrile neutropenia (6.1% and 6.0%, respectively).

The proportion of patients who discontinued darolutamide or placebo due to TEAE was 13.5% in the darolutamide plus docetaxel and ADT arm, and 10.6% in the placebo plus docetaxel and ADT arm, while the proportion of patients who discontinued docetaxel due to TEAE was 8.0% in the darolutamide plus docetaxel and ADT arm, and 10.3% in the placebo plus docetaxel and ADT arm. Death events were reported in 35.1% of patients in the darolutamide plus docetaxel and ADT arm, and 46.8% of patients in the placebo plus docetaxel and ADT arm. The majority of deaths were attributed to disease progression in both treatment arms.

Critical Appraisal

Appropriate methods of randomization were used in the study. Reporting bias in favour of the darolutamide arm might have been involved for subjective efficacy outcomes (i.e., time to pain progression, time to first SSE, SSE-FS, and HRQoL) as a result of unblinding by error, although the extent of bias was likely to be small given the small number of unblinded patients (|||||| in the darolutamide triplet arm, |||||| in the control arm). A high number of important protocol deviations (||||||) was reported, which, according to the sponsor, was due to the use of broad and conservative definitions for important deviations and was not expected to compromise study data. The sponsor noted that none of the important protocol deviations was considered to be major according to the old International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH E3) classification. The statistical analyses were generally well designed, with adequate sample size and power considerations and multiplicity adjustments for all secondary outcomes using a hierarchal gatekeeping approach. HRQoL, PSA outcomes, and subgroup analyses, however, were not adjusted for multiplicity and were considered exploratory due to increased risk of type I error. There is also uncertainty in the HRQoL outcomes due to high risk of bias in measurement of the outcome and large losses to follow-up. The clinical experts noted the duration of follow-up for OS (median approximately 3.5 years) was adequate for assessing the efficacy and safety of systemic treatments for mCSPC, although longer follow-up would increase certainty of the OS results.

In consultation with the clinical experts, the inclusion and exclusion criteria of the trial were generally reflective of the treatment eligibility criteria in clinical practice; however, the exclusion of patients with significant comorbidities (e.g., cardiovascular diseases) and poor performance status may limit the generalizability of study results because this patient population is commonly seen in clinical practice. ARATs were the most commonly used subsequent antineoplastic therapy in the darolutamide arm. However, the clinical experts noted that the use of a second-line ARAT is unlikely to be adopted in clinical practice because second- or later-line re-treatment with an alternate ARAT is not funded by most jurisdictions. While the comparator regimen, docetaxel plus ADT, is an appropriate and relevant comparator, it accounts for a small proportion of treatment regimens prescribed for chemotherapy-eligible patients with mCSPC in Canada. In the absence of direct evidence, the comparative efficacy between darolutamide plus docetaxel and ADT versus ARAT plus ADT, the most commonly prescribed treatment regimen for mCSPC, is unknown and represents a gap in evidence. The clinical experts considered the benefits of darolutamide plus docetaxel and ADT in survival and delaying disease progression to be clinically meaningful. The clinical relevance of the HRQoL outcome was uncertain considering the NCCN-FACT FPSI-17 instrument is not routinely administered in clinical practice, as noted by the clinical experts. Nevertheless, the instrument does capture common symptoms (e.g., pain, difficulty in urination) and treatment-related side effects (e.g., fatigue, weight gain, decreased sexual function), which are very relevant in the clinical assessment of these patients in practice, according to the clinical experts.

Indirect Comparisons

Description of Studies

As part of the development of a |||||| cost-effectiveness model for darolutamide in mCSPC, the sponsor conducted and submitted a network meta-analysis (NMA) that was used to inform these analyses. The sponsor-submitted indirect treatment comparison (ITC) first conducted a systematic literature review (SLR) to identify evidence for inclusion in a |||||| ITC. The relative efficacy of darolutamide plus ADT and docetaxel from the ARASENS trial was indirectly compared to alternative treatments for patients with mCSPC via |||||||||||||||||||||||||||||||||||| NMA. Comparators of interest for the sponsor-submitted NMA included abiraterone and prednisone, apalutamide, enzalutamide, and docetaxel, in combination with ADT. Outcomes of interest included OS, time to CRPC, and radiographic progression-free survival (rPFS).⁸

Two additional NMAs were identified in the CADTH literature search (Menges et al. [2022], Yanagisawa et al. [2022]).^9,10 The objective of the published SLR and NMA by Menges et al. was to assess the clinical effectiveness regarding survival and HRQoL, safety, and benefit-harm balance of metastatic hormone-sensitive prostate cancer (mHSPC) treatments, including docetaxel, abiraterone, enzalutamide, apalutamide, darolutamide, and radiotherapy (alone or in combination with ADT) via frequentist, random-effects NMA. Outcomes of interest evaluated in the study included OS, progression-free survival (PFS), HRQoL, and AEs; however, results for PFS and HRQoL were not available for comparisons involving darolutamide, and AEs were not available because comparisons were only made to ADT monotherapy,⁹ and therefore were not of interest to this review.

The objective of the published SLR and NMA by Yanagisawa et al. was to analyze the benefit of triplet combination therapies with androgen receptor signalling inhibitors (abiraterone acetate, apalutamide, darolutamide, and enzalutamide) in combination with docetaxel and ADT, compared to available treatment regimens in patients with mHSPC via frequentist, random-effects NMA. Outcomes of interest included OS, PFS, and AEs; however, results for PFS were not available for comparisons involving darolutamide, and comparisons for AEs were only made to ADT monotherapy¹⁰ and therefore were not of interest to this review.

Efficacy Results

The sponsor-submitted NMA included a total of |||||| trials. In the |||||||||||||||||| NMA for OS, darolutamide plus ADT and docetaxel was favoured over ADT and docetaxel (||||||||||||||||||||||||||||||||||||) and ADT alone (||||||||||||||||||||||||||||||||||||); however, comparisons between darolutamide plus ADT and docetaxel and enzalutamide (||||||||||||||||||||||||||||||||||||), apalutamide (||||||||||||||||||||||||||||||||||||), and abiraterone acetate (||||||||||||||||||||||||||||||||||||) were affected by imprecision, precluding conclusions about comparative efficacy. For time to CRPC, darolutamide plus ADT and docetaxel was favoured over apalutamide and ADT (||||||||||||||||||||||||||||||||||||), abiraterone and ADT (||||||||||||||||||||||||||||||||||||), docetaxel and ADT (||||||||||||||||||||||||||||||||||||), and ADT alone (||||||||||||||||||||||||||||||||||||), but over the comparison to enzalutamide and ADT (||||||||||||||||||||||||||||||||||||) was imprecise, precluding conclusions about comparative efficacy. Results for sensitivity analyses using |||||||||||| models were consistent with the base-case analyses, although 95% credible intervals (CrIs) were wider.

The NMA by Menges et al. included a total of 10 studies. In the frequentist, random-effects NMA for OS, darolutamide plus ADT and docetaxel was favoured over ADT and docetaxel (HR, 0.68 [95% CI, 0.57 to 0.81]). The comparisons of darolutamide plus ADT and docetaxel to abiraterone acetate plus ADT and prednisone, enzalutamide and ADT, apalutamide and ADT, and apalutamide plus ADT and docetaxel were affected by imprecision, precluding conclusions.

The NMA by Yanagisawa et al. included a total of 8 studies. In the frequentist, random-effects NMA for OS, darolutamide plus ADT and docetaxel was favoured over ADT and docetaxel (HR, 0.68 [95% 0.56 to 0.82]), and ADT and abiraterone acetate (HR, 0.74 [95% CI, 0.55 to 0.99]). No analyses were conducted for other comparators of interest.

Harms Results

Harms results were only reported for the published NMAs, although most comparisons were only conducted versus ADT monotherapy. No harms results were reported in the sponsor-submitted NMA. Results of the NMA focusing on AEs conducted by Yanagisawa et al. also showed wide CIs for the comparison of darolutamide plus ADT and docetaxel to abiraterone and ADT (OR, 26.62 [95% CI, 7.46 to 94.99]), precluding conclusions.

Critical Appraisal

Appraisal points across the 3 NMAs were similar. The sponsor-submitted NMA and both published NMAs were informed by SLRs; however, no information was provided in the sponsor-submitted report on the methods of study selection or data extraction (i.e., duplicate reviewers), or whether a risk-of-bias assessment was conducted. Both published NMAs followed appropriate methods for identification, inclusion, and assessment of studies. Both published NMAs also conducted a quality assessment using the Cochrane risk-of-bias 2.0 tool;¹¹ however, the results for the individual domains varied, despite authors reaching the same conclusion, that the studies were at a low risk of bias.

In general, treatments included in the NMAs were considered appropriate; however, there were some treatments, such as ADT monotherapy (included in all NMAs) and radiotherapy, which were not considered relevant comparators for this review. Additionally, the sponsor-submitted NMA did not consider the combination of abiraterone plus ADT and docetaxel, which was noted by the clinical experts consulted by CADTH as a relevant treatment option currently in Canada. Of note, the clinical experts reported that the combination has recently begun to be used by some clinicians in light of new clinical trial evidence⁴ and would not have been considered relevant at the time the NMA was conducted by the sponsor. The outcomes assessed across NMAs were also appropriate; however, important outcomes such as AEs and HRQoL were not considered in the sponsor’s NMA. While HRQoL and AEs were evaluated across the published NMAs, comparisons either did not include darolutamide plus ADT and docetaxel, or used ADT monotherapy as a reference, and were therefore not included in this report. Across NMAs, in most cases, comparisons were based on single trials and all evidence for comparisons to darolutamide plus ADT and docetaxel were indirect, increasing the uncertainty of the estimates of comparative efficacy. Additionally, results for OS were generally only in favour of darolutamide plus ADT and docetaxel over ADT and docetaxel alone and demonstrated wide 95% CIs, suggesting uncertainty and imprecision in the comparative efficacy estimates.

The main concern across the NMAs was the potential for heterogeneity across studies, which would result in violation of the underlying transitivity assumption, which introduces an unknown degree of bias into the results. In the sponsor-submitted NMA, potential effect modifiers were considered from results of the ARASENS subgroup analysis, but consultation with clinical experts or other empirical evidence was not reported. There were notable differences in study design (i.e., blinded versus open label) and baseline characteristics that could potentially result in changing relative treatment effects, and, although identified, these were not accounted for (i.e., PSA level and prostate cancer stage). Differences in prior treatment requirements, the time period during which the studies were undertaken, and follow-up duration were also evident across studies. Baseline characteristics were not available for all factors of interest across all studies. Although many baseline characteristics appeared similar across trials, other differences (e.g., study design, prior treatment, outcome definitions, length of follow-up, and time period during which the studies were undertaken) were not feasible to address. For Menges et al., the authors note that transitivity was assessed using epidemiological criteria and the presence of potential effect modifiers, along with considerations of clinical plausibility. No rationale or further discussion of the transitivity assessment is provided in the publication or supplementary material; therefore, it is difficult to conclude whether the transitivity assumption was met. Heterogeneity was assessed visually and by means of I² values and was reported to be low for OS in the overall NMA. Some potentially important effect modifiers were reported in the study but were not adjusted for or discussed (e.g., use of prior therapy, Gleason score), and follow-up duration varied across studies. In the Yanagisawa et al. NMA, potential sources of heterogeneity were evaluated in the initial meta-analysis via Cochran’s Q test; however, results did not suggest any important heterogeneity and therefore it was not explored for the NMA. As a result, it was unclear if the transitivity assumption was met. Additionally, no consideration was given to treatment effect modifiers, and thus the impact of any potential effect modifiers remains unknown. The authors noted that the publications included in the NMA included different patient populations with regards to the proportions of patients with de novo disease and disease burden. As such, the findings of all NMAs, although supportive of the ARASENS trial, were highly uncertain due to the methodological limitations as well as wide CrIs/CIs.

Other Relevant Evidence

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

Conclusions

In the ARASENS trial, darolutamide in combination with docetaxel and ADT demonstrated a clinically meaningful improvement of OS compared with docetaxel plus ADT in patients with mCSPC. Analyses of secondary outcomes supported the efficacy of darolutamide plus docetaxel and ADT in delaying progression to mCRPC, the need for subsequent antineoplastic therapy, worsening of pain, and SSEs, compared to docetaxel plus ADT. Due to limitations in the statistical analysis, no definitive conclusions can be drawn from the exploratory PSA and HRQoL analyses. The exclusion of patients with significant comorbidities and impaired performance status and the use of a comparator regimen that has low utilization in Canada may limit the generalizability of study findings. Across the included NMAs, the findings for the comparative survival benefit observed with darolutamide plus ADT and docetaxel compared to ADT and docetaxel supported the findings of the ARASENS trial; however, the findings of these NMAs are associated with uncertainty due to probable violation of the underlying transitivity assumption. Comparisons to other drugs of interest were uncertain due to the aforementioned methodological limitations as well as wide CIs. The safety profile of darolutamide plus docetaxel and ADT was overall similar to that of docetaxel and ADT in the ARASENS trial, with no additional serious safety concern.

Introduction

Disease Background

Prostate cancer is the most common cancer (excluding non-melanoma skin cancers) and the third leading cause of death from cancer in men living in Canada.¹ It is estimated that 24,600 men will be diagnosed with prostate cancer and that 4,600 deaths from prostate cancer will occur in Canada in 2022.¹

Approximately 9% of prostate cancers are metastatic at diagnosis.¹² mCSPC refers to metastatic prostate cancer that responds to ADT. The disease can occur de novo (newly diagnosed in the metastatic state) or present as metachronous disease (progressed from localized disease despite treatment). The most common sites of metastasis are bone (84%) and distant lymph nodes (11%).² The clinical presentations of mCSPC patients are highly variable. Some patients are asymptomatic after receiving local therapy of curative intent in the non-metastatic stage and were identified to have metachronous disease upon routine follow-up, while others present symptomatically, with bone pain and lower urinary tract symptoms being the most common symptoms, according to the clinical experts consulted by CADTH. In patients with bone metastasis, disease-related skeletal complications including pathological fracture and spinal cord compression may occur.³

mCSPC is an incurable disease. Progression to mCRPC, a disease state that is no longer responsive to ADT, is associated with increased disease burden and/or worsening symptoms. According to the clinical experts, the median survival for mCSPC is approximately 5 years, with a large variability in survival among patients.

The diagnosis of mCSPC involves physical examination, digital rectal exam, PSA test, biopsy, and imaging (e.g., CT, MRI, and bone scan).¹³

Standards of Therapy

ADT, achieved either by surgical castration (orchiectomy), or pharmacologically with GnRH agonists or antagonists, has been the cornerstone therapy for mCSPC.¹⁴ Treatment intensification with the addition of docetaxel or ARAT (i.e., enzalutamide, apalutamide, or abiraterone plus prednisone) to ADT is the current SoC in men with mCSPC and has been shown to improve survival and slow disease progression compared to ADT alone.^14,15 Recently, triplet therapy with ADT, docetaxel, and abiraterone plus prednisone is also observed in clinical practice in light of new clinical trial evidence,⁴ according to the clinical expert consulted by CADTH. This triplet therapy is currently under review at CADTH for mCSPC. Treatment intensification is the SoC; however, there is currently no consensus on the optimal treatment intensification strategy for mCSPC: the choice of therapy in general depends on functional assessment, comorbidities, patient preference, drug toxicity profiles, and drug accessibility, as per the clinical experts.

In select patients with a low-volume metastatic disease burden, ADT and external beam radiation therapy, directed toward the primary tumour(s) in the prostate, are considered.^14,15

The key treatment goals of mCSPC include improving survival, delaying progression to mCRPC and the need for subsequent antineoplastic therapy, and improving symptoms and HRQoL, according to the clinical expert consulted by CADTH.

Drug

Darolutamide is an androgen receptor pathway inhibitor that antagonizes the androgen receptor by binding directly to receptor ligand binding domain, leading to competitive inhibition of androgen binding, nuclear translocation of androgen receptor, and androgen receptor-mediated transcription. The drug has been shown to inhibit prostate cancer cell proliferation and tumour growth in animal models.

Darolutamide has been previously reviewed by CADTH for the treatment of non-mCRPC. Darolutamide was granted a Health Canada Notice of Compliance for the indication of the treatment of patients with mCSPC in combination with docetaxel on September 29, 2022. The sponsor is seeking reimbursement of darolutamide in combination with docetaxel for the treatment of mCSPC in patients who are chemotherapy-eligible.

As per the product monograph, the anticipated recommended dose of darolutamide is 600 mg (2 tablets of 300 mg) twice daily orally with food until disease progression or unacceptable toxicity.⁵ It is recommended that patients receive a GnRH analogue concurrently or have had a bilateral orchiectomy.⁵

The key characteristics of treatments for mCSPC are summarized in Table 3.

Table 3: Key Characteristics of Darolutamide, Enzalutamide, Apalutamide, and Abiraterone

ADT = androgen deprivation therapy; AR = androgen receptor; ARAT = androgen receptor-axis-targeted therapy; CV = cardiovascular; CYP17 = 17Alpha-hydroxylase/C17,20-lyase; GnRH = gonadotropin-releasing hormone; mCRPC = metastatic castration-resistant prostate cancer; mCSPC = metastatic castration-sensitive prostate cancer; nmCRPC = non-metastatic castration-resistant prostate cancer.

^aHealth Canada–approved indication.

^bThe drug has not been studied in patients with nmCRPC at low risk of developing metastases. The benefit and risk profile in these patients is unknown.

^cIn patients who are: chemotherapy-naive with asymptomatic or mildly symptomatic disease after failure of ADT; have received docetaxel therapy.

^dIn patients who are asymptomatic or mildly symptomatic after failure of ADT or who have received prior chemotherapy containing docetaxel after failure of ADT.

^eThe first of 6 cycles of docetaxel should be administered within 6 weeks after the start of darolutamide treatment.

Source: Nubeqa Product Monograph,⁵ Xtandi Product Monograph,¹⁶ Erleada Product Monograph,¹⁷ and Zytiga Product Monograph.¹⁸

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 has been included in the stakeholder section at the end of this report.

Two patient groups, the CCSN and the Canadian Cancer Society, submitted patient group input for this review. The CCSN gathered data through an online survey with responses from 24 patients with prostate cancer (6 of whom were diagnosed with metastatic disease) and from 2 caregivers. The Canadian Cancer Society conducted surveys and interviews with 39 patients with mCSPC and with 2 caregivers. Demographic information was provided for the Canadian Cancer Society respondents, with the majority identifying as male (98%), white (95%), and 70 to 79 years of age (49%). In both submissions, all patients were from Canada and a total of 8 patients had experience with darolutamide.

The Canadian Cancer Society respondents indicated that symptoms associated with mCSPC had a moderate-to-severe negative impact on their ability to engage in sexual activity (76%), work (49%), exercise (44%), and travel (32%); fulfill family obligations (32%); and maintain their mental health (32%). Common side effects following currently available treatments by patients in both submissions included changes in libido and sexual function, hot flashes, fatigue, loss of muscle mass, incontinence, and weight gain. Patients who had experience with darolutamide reported also currently taking or having taken ADT monotherapy, enzalutamide plus ADT, docetaxel plus ADT, and abiraterone acetate with prednisone plus ADT. Five patients from the CCSN submission rated how their experience with darolutamide compared to other treatments, with 4 of these patients indicating that darolutamide was easier to use and better addressed disease progression, 3 patients stating that they experienced a reduction in side effects compared to current treatments, and 1 indicating that it was better able to control their symptoms. AEs identified among respondents with experience with darolutamide were consistent with those of currently available treatments. Across both submissions, respondents reported that they would like to see future treatments that delay the onset of symptoms, delay the need for chemotherapy, have fewer side effects, improve survival, are easy to use, and allow them to maintain their HRQoL.

Clinician Input

Input From Clinical Experts Consulted by CADTH

All CADTH review teams include at least 1 clinical specialist with expertise in the diagnosis and management of the condition for which the drug is indicated. Clinical experts are a critical part of the review team and are involved in all phases of the review process (e.g., providing guidance on the development of the review protocol, assisting in the critical appraisal of clinical evidence, interpreting the clinical relevance of the results, and providing guidance on the potential place in therapy). The following input was provided by 2 clinical specialists with expertise in the diagnosis and management of mCSPC.

Unmet Needs

The clinical experts noted that the goals of therapy for mCSPC include improving survival, delaying disease progression, reducing symptoms, and improving HRQoL. The clinical experts noted that although there are currently 2 available treatment intensification strategies using chemotherapy or ARAT that can improve long-term outcomes beyond ADT alone, these are not curative. There remains a compelling need to extend survival longer while improving and maintaining quality of life overall, as per the clinical experts.

Place in Therapy

The clinical experts expected triplet therapy with darolutamide, docetaxel, and ADT to be considered as a first-line treatment option for men with mCSPC who are eligible for cytotoxic chemotherapy. The clinical experts noted that ARAT plus ADT regimens account for the majority of treatment regimens prescribed for mCSPC in Canada, and in the absence of head-to-head evidence comparing darolutamide plus docetaxel and ADT triplet therapy versus ARAT plus ADT, a major shift in the prescribing pattern is unlikely. The clinical experts noted that in patients for whom treatment with docetaxel plus ADT is considered, the addition of darolutamide is supported by the ARASENS trial and will likely be favoured over docetaxel-plus-ADT–only regimen by clinicians.

Patient Population

The clinical experts noted that, among clinicians in Canada, there is no consensus on who should be offered triplet therapy, and that it usually involves a case-by-case discussion between the patient and the treating physician. A clinical expert noted that in their clinical practice, triplet therapy is considered in patients with aggressive disease who are lacking in competing causes of mortality and eligible for chemotherapy. The clinical experts agreed that triplet therapy with darolutamide, docetaxel, and ADT should be available to all men with mCSPC who are candidates for cytotoxic chemotherapy.

Assessing Response to Treatment

The clinical experts noted that the approach to assessing treatment response is highly heterogeneous in clinical practice and that it usually involves assessment of clinical status, radiologic response, and PSA response. The clinical experts also indicated that the frequency of assessment is highly variable in clinical practice. One clinical expert noted that, in their practice, clinical and PSA assessments are conducted every 3 months in the first year and every 6 months thereafter.

Discontinuing Treatment

The clinical experts noted that treatment discontinuation is considered in patients who have unacceptably toxicity or disease progression, which is generally determined based on a composite of clinical, PSA, and radiologic factors at the physician’s discretion in clinical practice. Most physicians consider progression in at least 2 of 3 parameters to be indicative of disease progression, according to the clinical experts. In addition, the clinical experts noted that rapid progression and the absence of ongoing clinical benefit may warrant the use of a subsequent line of therapy.

Prescribing Conditions

The clinical experts noted that darolutamide plus docetaxel and ADT should be prescribed by a medical oncologist.

Clinician Group Input

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

Clinician group input was received from 7 groups: the BC Cancer Agency (12 clinicians); the Canadian Cancer Society (12 clinicians); genitourinary oncologists from the Maritime provinces (5 clinicians); the Allan Blair Cancer Centre (5 clinicians); the Ottawa Hospital Cancer Centre — Genitourinary Oncology Group (3 clinicians); the Ontario Health — Cancer Care Ontario Genitourinary Cancer Drug Advisory Committee (4 clinicians); and the Genitourinary Disease Site Group of the Cancer Centre of Southeastern Ontario (2 clinicians). The various submissions noted that current treatment goals are to reduce symptom burden, prolong survival, and delay disease progression. The clinician groups noted that current treatment for mCSPC includes either chemotherapy (docetaxel) or a second-generation androgen receptor inhibitor (i.e., abiraterone acetate plus prednisone, enzalutamide, or apalutamide) in combination with ADT. The clinician groups noted that mCSPC is incurable and that many patients fail systemic treatment shortly after treatment initiation. The groups emphasized a significant unmet need for treatments that further improve survival, increase HRQoL, and increase duration of treatment response while providing lower toxicity burden. In terms of place in therapy, the submissions stated that the darolutamide would be used as a first-line treatment for mCSPC in combination with ADT and docetaxel in patients who are fit for chemotherapy. The submissions stated that response to treatment would be assessed using PSA response, radiographic response, and clinical assessment (i.e., worsening pain or symptoms). Treatment would typically be discontinued upon disease progression (PSA, radiological, or symptomatic progression) or unacceptable toxicities.

Drug Program Input

The drug programs provide input on each drug being reviewed through CADTH’s reimbursement review processes by identifying issues that may impact their ability to implement a recommendation. The implementation questions and corresponding responses from the clinical experts consulted by CADTH are summarized in Table 4.

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

ADT = androgen deprivation therapy; ARAT = androgen receptor-axis-targeted therapy; ECOG = Eastern Cooperative Oncology Group; mCSPC = metastatic castration-sensitive prostate cancer; PSA = prostate-specific antigen.

Clinical Evidence

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

Systematic Review (Pivotal and Protocol-Selected Studies)

Objectives

To perform a systematic review of the beneficial and harmful effects of darolutamide 300 mg oral tablet in combination with docetaxel for the treatment of patients with mCSPC who are chemotherapy-eligible.

Methods

Studies selected for inclusion in the systematic review included pivotal studies provided in the sponsor’s submission to CADTH and Health Canada, as well as those meeting the selection criteria presented in Table 5. Of note, the systematic review protocol presented below was established before the granting of a Notice of Compliance from Health Canada. Table 5 shows the outcomes included in the CADTH review protocol, which reflect outcomes considered to be important to patients, clinicians, and drug plans.

Table 5: Inclusion Criteria for the Systematic Review

ADT = androgen deprivation therapy; AE = adverse event; ECOG = Eastern Cooperative Oncology Group; HRQoL = health-related quality of life; ORR = objective response rate; OS = overall survival; PSA = prostate-specific antigen; RCT = randomized controlled trial; SAE = serious adverse event; WDAE = withdrawal due to adverse event.

^aIn the input received by CADTH from patient groups, these outcomes were identified as being of particular importance to patients.

The literature search for clinical studies was performed by an information specialist using a peer-reviewed search strategy according to the PRESS Peer Review of Electronic Search Strategies checklist.¹⁹

Published literature was identified by searching the following bibliographic databases: MEDLINE All (1946–) via Ovid and Embase (1974–) via Ovid. All Ovid searches were run simultaneously as a multifile search. Duplicates were removed using Ovid deduplication for multifile searches, followed by manual deduplication in EndNote. The search strategy comprised both controlled vocabulary, such as the National Library of Medicine’s Medical Subject Headings (MeSH), and keywords. The main search concept was Nubeqa (darolutamide). Clinical trials registries were also searched: the US National Institutes of Health’s clinicaltrials.gov, WHO’s International Clinical Trials Registry Platform (ICTRP) search portal, Health Canada’s Clinical Trials Database, and the European Union Clinical Trials Register.

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

The initial search was completed on July 13, 2022. Regular alerts updated the search until the meeting of the CADTH pan-Canadian Oncology Drug Review Expert Committee (pERC) on November 9, 2022.

Grey literature (literature that is not commercially published) was identified by searching relevant websites from the Grey Matters: A Practical Tool For Searching Health-Related Grey Literature checklist.²⁰ Included in this search were the websites of regulatory agencies (US FDA and European Medicines Agency). Google was used to search for additional internet-based materials. Refer to Appendix 1 for more information on the grey literature search strategy.

These searches were supplemented through contacts with appropriate experts. In addition, the sponsor of the drug was contacted for information regarding unpublished studies.

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

Findings From the Literature

One study submitted by the sponsor (the ARASENS trial)⁶ was eligible for inclusion in the systematic review (Figure 1). The included study is summarized in Table 6.

Figure 1: Flow Diagram for Inclusion and Exclusion of Studies

Table 6: Details of Included Studies

ADT = androgen deprivation therapy; BPI-SF = Brief Pain Inventory (Short Form); CT = CT; ECOG = Eastern Cooperative Oncology Group; HRQoL = health-related quality of life; LHRH = luteinizing hormone-releasing hormone; NCCN-FACT FPSI-17 = National Comprehensive Cancer Network prostate cancer symptom index 17 item questionnaire/Functional assessment of cancer therapy; NYHA = New York Heart Association; PSA = prostate-specific antigen; RCT = randomized controlled trial.

^aMetastatic disease was defined as either malignant lesions in bone scan or measurable lymph nodes above the aortic bifurcation or soft tissue/visceral lesions according to Response Evaluation Criteria In Solid Tumors (RECIST) v1.1. Patients with regional lymph node metastases only (N1, below the aortic bifurcation) were not eligible for the study. Only patients with non-regional lymph node metastases (M1a) and/or bone metastases (M1b) and/or other sites of metastases with or without bone disease (M1c) were eligible.

^bSecond-generation androgen receptor inhibitors such as enzalutamide, apalutamide, darolutamide, and other investigational androgen receptor inhibitors.

^cCytochrome P17 enzyme inhibitor such as abiraterone acetate or oral ketoconazole.

^dUncontrolled hypertension was defined as a resting systolic blood pressure ≥ 160 mm Hg or diastolic BP ≥ 100 mm Hg despite medical management.

^eDocetaxel could be administered in combination with prednisone/prednisolone at the discretion of the Investigator.

^fAll patients received an ADT of the Investigator’s choice (LHRH agonist/antagonists or orchiectomy) because standard therapy started ≤ 12 weeks before randomization (if combined with a first-generation antiandrogen, such as bicalutamide, flutamide, nilutamide, or cyproterone acetate, it had to be stopped before randomization). For patients receiving LHRH agonists, treatment in combination with a first-generation antiandrogen for at least 4 weeks before randomization was recommended.

Source: ARASENS Clinical Study Report.⁷

Description of Studies

One study (ARASENS)^6,7 met the inclusion criteria for the CADTH systematic review. The ARASENS trial was a phase III, multicentre, double-blind RCT that aimed to compare the efficacy and safety of darolutamide in addition to ADT and docetaxel versus placebo in addition to ADT and docetaxel in adult males with mCSPC (n = 1,306). The study was initiated on November 30, 2016, and final analysis is complete. Patients were enrolled at 301 sites in 23 countries, including 5 Canadian sites.

The ARASENS trial consisted of 4 phases: screening, treatment, active follow-up, and long-term (survival) follow-up. Following a screening phase of up to 28 days to assess study eligibility, eligible patients entered the double-blind treatment phase, in which they were randomized in a 1:1 ratio to receive darolutamide or placebo, in addition to ADT and docetaxel. Randomization was stratified by the extent of disease (non-regional lymph nodes metastases only, bone metastases with or without lymph node metastases, visceral metastases with or without lymph node metastases or with or without bone metastases), and alkaline phosphatase (ALP) level (ALP less than the upper limit of normal [ULN]; ALP greater or equal to the ULN) using the Interactive Voice/Web Response System. Patients received study treatments until symptomatic disease progression, unacceptable toxicity, or change of antineoplastic therapy, and they were subsequently monitored in the active follow-up phase for safety and select efficacy outcomes for up to 1 year. Survival continued to be assessed in the long term follow-up period.

This report presents the final efficacy analysis in the ARASENS trial at the data cut-off on October 25, 2021.

Populations

Inclusion and Exclusion Criteria

The key inclusion and exclusion criteria of the ARASENS trial are summarized in Table 6. The study enrolled adult males aged 18 years or older with metastatic prostate adenocarcinoma who started ADT (luteinizing hormone-releasing hormone [LHRH] agonist or antagonist, or orchiectomy) no longer than 12 weeks before randomization and had an ECOG performance status of 0 or 1. Patients were excluded if they had regional lymph node metastases only or had prior treatment with 1 of the following: second-generation androgen receptor inhibitors (e.g., enzalutamide, apalutamide, darolutamide); cytochrome P17 enzyme inhibitors (e.g., abiraterone), chemotherapy or immunotherapy for prostate cancer; or radiotherapy within 2 weeks before the study.

Baseline Characteristics

The baseline patient demographic, disease characteristics, and prior cancer therapy were generally balanced between treatment arms, as shown in Table 7. Overall, the study population had a median age of 67.0 (range, 41 to 89) years. Fifty-two percent of patients were white, and 36.4% were Asian. Most patients had stage IV disease (87.6%) and a Gleason score of 8 or higher (78.2%) at initial diagnosis. At study baseline, the majority of patients had an ECOG performance status of 0 (71.1%) and bone metastases with or without lymph node metastases (82.8%). In 55.5% of patients, the ALP level was above the ULN, and the serum testosterone was at castrate level (< 0.5 ng/mL). Most patients had no prior local therapy and had prior systemic antineoplastic therapy (98.1%), all of which were ADTs. Almost all patients entered the study with ADT (99.8%), while others initiated ADT at randomization. The majority of patients were treated with an LHRH agonist or antagonist only (97.3%).

Table 7: Summary of Baseline Characteristics (FAS)

ADT = androgen deprivation therapy; ALP = alkaline phosphatase; ECOG = Eastern Cooperative Oncology Group; FAS = full analysis set; LHRH = luteinizing hormone-releasing hormone; PSA = prostate-specific antigen; SD = standard deviation; TNM = tumour, nodes, and metastases; ULN = upper limit of normal.

^aThe extent of metastatic disease was determined using the TNM classification system.

^bBased on data collected via Interactive Voice/Web Response System for randomization or stratification.

^cAccording to the American Joint Committee on Cancer Staging Manual, 7th edition, stage IV could be M1 or M0 disease. In this analysis, the stage IV M0 group was defined as the time interval of > 3 months between initial diagnosis and initial diagnosis of metastases. The stage IV M1 group was defined as the time interval of ≤ 3 months between initial diagnosis and initial diagnosis of metastases.

^dThis captures the proportion of patients who entered the study with ADT. Patients who initiated ADT at randomization were not captured here.

Data cut-off: October 25, 2021.

Source: ARASENS Clinical Study Report.⁷

Interventions

In the double-blind treatment period, patients were assigned to either darolutamide 600 mg (2 tablets of 300 mg) or a matching placebo, twice daily by mouth with food. All patients also received a background ADT (LHRH agonist or antagonists, or orchiectomy), which was initiated less than or equal to 12 weeks before randomization, and docetaxel 75 mg/m² via IV infusion every 21 days for 6 cycles starting within 6 weeks after the initiation of darolutamide or placebo. Docetaxel was administered in combination with prednisone or prednisolone at the discretion of the investigators. The study treatments were given until symptomatic progressive disease, change of antineoplastic therapy, or unacceptable toxicity. When clinically significant toxicities were present, the dosing of darolutamide was delayed for a maximum of 28 consecutive days or reduced to 300 mg twice daily, and dose reduction of docetaxel to 60 mg/m² was allowed. The darolutamide and placebo tablets were identical in appearance.

Initiation of the following medications was prohibited during the treatment period: radiopharmaceuticals, immunotherapy, cytotoxic chemotherapy (other than docetaxel), first- or second-generation antiandrogen (e.g., enzalutamide, apalutamide, bicalutamide, flutamide [other than darolutamide]), cytochrome P17 inhibitors (e.g., abiraterone, TAK-700), systemic ketoconazole as antineoplastic treatment for prostate cancer, and a switch from ADT to an LHRH agonist.

Concomitant use of analgesics (opioid and non-opioid), palliative radiation therapy or surgical intervention, bisphosphonate, and denosumab was allowed. Switching ADT to an LHRH antagonist, administering dexamethasone as a premedication for docetaxel infusion, and supportive care for docetaxel toxicity (e.g., granulocyte colony-stimulating factor) were permitted.

Outcomes

A list of efficacy end points identified in the CADTH review protocol that were assessed in the clinical trials included in this review is provided in Table 8. These end points are further summarized below. A detailed discussion and critical appraisal of these outcome measures is provided in Appendix 3.

Table 8: Summary of Outcomes of Interest Identified in the CADTH Review Protocol

AE = adverse event; BPI-SF = Brief Pain Inventory (Short Form); NCCN-FACT FPSI-17 = National Comprehensive Cancer Network prostate cancer symptom index 17 item questionnaire/Functional assessment of cancer therapy; PSA = prostate-specific antigen; SAE = serious adverse event; WDAE = withdrawal due to adverse event.

Efficacy Outcomes

The primary efficacy end point of the ARASENS trial was OS, which was defined as the time from the date of randomization until death from any cause.

The secondary efficacy end points that were of interest to this review include time to CRPC, time to initiation of subsequent systemic antineoplastic therapy, time to pain progression, time to first SSE, and SSE-FS. These end points were defined as follows:

• Time to CRPC, a composite end point of time to PSA progression, time to progression by soft-tissue lesions, or time to progression by bone lesions, was defined as the time from randomization to the time of the occurrence of 1 of the following events, whichever came first:

  • PSA progression, which was defined as a 25% or greater increase and an absolute increase of 2 ng/mL or more from the nadir in 2 consecutive assessments. This definition required serum testosterone at castrate levels of less than 0.50 ng/mL and a first assessment date at least 12 weeks from randomization.

  • Radiological progression by soft-tissue and visceral lesions, as per Response Evaluation Criteria In Solid Tumors (RECIST) v1.1.

  • Radiological progression by bone lesions, as per Prostate Cancer Clinical Trials Working Group 3.

• Time to initiation of subsequent systemic antineoplastic therapy was defined as the time from randomization to initiation of the first subsequent systemic antineoplastic therapy.

• Time to pain progression was defined as the time from randomization to the first date a patient experienced pain progression based on the Brief Pain Inventory (Short Form) (BPI-SF) questionnaire. Pain progression was defined as an increase of 2 or more points in the worst pain score from nadir (i.e., lowest at or after baseline) in 2 consecutive evaluations (and a worst pain score of 4 or above if patient was symptomatic), or initiation of short- or long-acting opioid use for pain.

  The BPI-SF is a patient-reported generic questionnaire for pain intensity and impact. Each item is scored on an 11-point scale from 0 to 10, where 0 is no pain/no interference and 10 is the worst pain/complete interference.²¹ The validity, reliability, responsiveness, and minimally important difference (MID) estimate of the BPI-SF instrument have not been evaluated in patients with mCSPC. A MID estimate of 2 or more points or 30% change in pain intensity items from baseline was previously used in 2 studies in mCRPC patients.^22,23 Detailed information on the scoring and validity of the BPI-SF instrument is provided in Appendix 3.

• Time to first SSE was defined as the time from randomization to the first occurrence of an SSE. An SSE was defined as the occurrence of 1 of the following: external beam radiation therapy to relieve skeletal symptoms, new symptomatic pathologic bone fracture, occurrence of spinal cord compression, or tumour-related orthopedic surgical intervention.

• SSE-FS was defined as the time from randomization to the first occurrence of an SSE or death from any cause.

The exploratory efficacy end points included HRQoL via NCCN-FACT FPSI-17, time to PSA progression, and absolute and relative PSA response rates. These end points were defined as follows:

• NCCN-FACT FPSI-17 is a 17-item, cancer-specific questionnaire developed to assess symptoms and well-being among patients with advanced prostate cancer who are undergoing treatment.²⁴ It consists of 4 main subscales that assess physical and emotional disease-related symptoms, treatment side effects, and function/well-being on a 5-point scale. Higher subscale scores and total scores indicate better outcomes on the aforementioned scales.²⁵ The validity, reliability, responsiveness, and MID estimate of the NCCN-FACT FPSI-17 have not been evaluated in patients with mCSPC. A study in mCRPC patients identified the MID estimates to be 4 to 6 points for the total score; 2 to 3.5 points for the disease-related symptoms, physical subscale; 0.5 points for the disease-related symptoms, emotional subscale; 1 to 1.5 points for the treatment side effects subscale; and 0.5 to 1 points for the function and well-being subscale.²⁵ Detailed information on the scoring and validity of the NCCN-FACT FPSI-17 instrument is provided in Appendix 3.

• Time to PSA progression was defined as the time from the date of randomization to the date of first PSA progression with testosterone at castrate level of less than 0.5 ng/mL.

• Absolute PSA response rate was defined as the number of patients with absolute PSA response divided by the total number of patients randomized, and was evaluated at 6 and 12 months. Absolute PSA response was defined as a postbaseline PSA level below 0.2 ng/mL in 2 consecutive evaluations.

• Relative PSA response rate was defined as the number of patients with relative PSA response divided by the total number of patients randomized, and was evaluated at 3, 6, and 12 months. Relative 30% PSA response was defined as a postbaseline 30% or greater reduction of the PSA level compared to the baseline value in 2 consecutive evaluations. Relative 50% and 90% PSA response were defined in the same way.

  The efficacy outcomes were assessed at baseline and then every 12 weeks until end of treatment in the treatment period. Outcomes related to SSE, pain, subsequent antineoplastic therapy, and survival continued to be assessed in the follow-up period.

Harms Outcomes

The harms outcomes included AEs, serious AEs, withdrawal due to AE, deaths, laboratory safety assessment, vital signs, and electrocardiogram. During the treatment period, these outcomes were assessed every 12 weeks until end of treatment. Study drug–related AEs and serious AEs were assessed every 12 weeks in the follow-up period.

Statistical Analysis

A summary of the statistical analyses of efficacy outcomes in the ARASENS trial is provided in Table 9.

Sample Size and Power Calculation

The sample size calculation was based on the primary end point of OS. The study required a sample size of 1,300 patients to achieve at least 509 death events to detect an HR of 0.75 with respect to OS at a 1-sided alpha of 0.025 with a power of 90% using a log-rank test, based on the following assumptions:

• Fifty patients were enrolled per month.

• There were exponential distributions of the OS event time.

• Median time of OS in the control group was 60 months.

• Dropout rate was 5%.

• There was a 6-month enrolment ramp-up period.

Interim and Final Analyses

One interim analysis for futility and 1 final (primary) analysis were planned to occur after 153 and 509 death events, respectively. A 1-sided overall beta of 0.1 was used for the interim and final analyses. For the interim analysis, the prespecified futility boundary of an HR for darolutamide over placebo greater than or equal to 1.166 was calculated using an O’Brien-Fleming beta-spending function. The independent data monitoring committee reviewed the interim results and recommended the study to proceed as planned.

Multiplicity Adjustment

Multiplicity adjustment using a hierarchical gatekeeping approach was implemented within the final efficacy analysis. The primary end point of OS was first tested with a 1-sided alpha of 0.025 using a stratified log-rank test. If the null hypothesis of OS was rejected, the secondary efficacy end points were tested hierarchically using the same significance level as the primary end point in the following order:

• Time to CRPC

• Time to pain progression

• SSE-FS

• Time to first SSE

• Time to initiation of subsequent systemic antineoplastic therapy

Handling of Missing Data

No imputation was performed to account for missing data in all efficacy end points (except for partial dates). Censoring rules were used to avoid excluding patients with missing or incomplete data from the analyses (Table 9).

Table 9: Statistical Analysis of Efficacy End Points

ALP = alkaline phosphatase; BPI-SF = Brief Pain Inventory (Short Form); CI = confidence interval; CMH = Cochran-Mantel-Haenszel; eCRF = electronic case report form; ePRO = electronic patient-reported outcome; IxRS = Interactive Voice/Web Response System; KM = Kaplan-Meier; NA = not applicable; NCCN-FACT FPSI-17 = National Comprehensive Cancer Network prostate cancer symptom index 17 item questionnaire/Functional assessment of cancer therapy; PSA = prostate-specific antigen; SSE = symptomatic skeletal event.

^aThis analysis was planned to occur only when there were more than 5% of patients with different values between IxRS and eCRF.

Source: ARASENS Statistical Analysis Plan.²⁶

Statistical Analysis for Primary Efficacy End Point

In the primary analysis of OS, the difference in OS between treatment arms in the full analysis set (FAS) population was tested using a 1-sided log-rank test stratified by extent of disease (non-regional lymph nodes metastases only, bone metastases with or without lymph node metastases, or visceral metastases with or without lymph node metastases or with or without bone metastases) and ALP level (ALP less than the ULN or ALP greater than or equal to the ULN). The HR and the 95% CI were based on a Cox proportional hazards model, stratified by extent of disease and ALP level. Median (95% CI) OS was estimated using the Kaplan-Meier (KM) method. The KM estimates for median OS, and the corresponding 95% CI, along with the estimated OS probability at every 12 months, were presented.

Three prespecified sensitivity analysis were conducted to assess the impact of different stratification strategies, including an unstratified analysis (using unstratified log-rank test and Cox model), an electronic case report form (eCRF)–variables stratified analysis (using stratification factors collected from eCRF, as opposed to from the Interactive Voice/Web Response System used in the primary analysis), and a stratified analysis using extent of disease data collected from the central imaging review (as opposed to from local imaging review in the primary analysis). A post hoc sensitivity analysis by docetaxel cycles was also performed.

Prespecified subgroup analyses with respect to OS were conducted using an unstratified Cox proportional hazards model. Descriptive statistics and HR estimates with 95% CI were presented. The subgroups that were assessed in the trial and of interest to this review included ECOG performance status at baseline (0 versus 1), Gleason score (less than 8, greater than or equal to 8) at initial diagnosis, and extent of disease (non-regional lymph node metastases only, bone metastases with or without lymph node metastases, or visceral metastases with or without lymph node metastases or with or without bone metastases). No adjustment for type I error was involved and no statistical testing was performed for treatment-by-subgroup interaction in the subgroup analyses.

Statistical Analyses for Secondary Efficacy End Points

The secondary efficacy end points were analyzed in the FAS population. Time-to-event end points were analyzed using the same statistical approach as the primary analysis of OS.

A sensitivity analysis was conducted for the secondary end point of time to CRPC using both central and local PSA data (as supposed to central PSA data only in the primary analysis). For time to pain progression end point, a sensitivity analysis using data from both electronic patient-reported outcome (ePRO) device and paper questionnaire were used (as opposed to from ePRO only in the primary analysis). Three additional sensitivity analyses were performed for time to pain progression, which aimed to evaluate the impact of using different definitions of time to pain progression (i.e., an increase of 2 or more points in the “worst pain in 24 hours” from either (1) baseline, (2) nadir after completion of docetaxel, or (3) baseline after completion of docetaxel).

Statistical Analyses for Exploratory Efficacy End Points

Time to PSA progression was analyzed using descriptive statistics based on KM estimates and their 95% CI. A sensitivity analysis based on both central and local PSA data (as supposed to central PSA data only in the primary analysis) was performed. Absolute and relative PSA response rates were evaluated descriptively and a stratified Cochran-Mantel-Haenszel model was used to estimate the risk difference between treatment arms, and the 95% CI. Descriptive statistics were presented for NCCN-FACT FPSI-17 score at each assessment time and for change from baseline by treatment arm.

Safety Analyses

The safety analyses were performed in the safety analysis set (SAF) population and were summarized using descriptive statistics.

Analysis Populations

• FAS: included all patients randomized in the study, and was analyzed by treatment assignment.

• SAF: included all randomized patients who received at least 1 dose of darolutamide or placebo, except for cases with critical Good Clinical Practice violations. Patients were analyzed by the actual treatment received.

Results

Patient Disposition

A summary of patient disposition is shown in Table 10. Of 1,686 screened patients, 380 (22.5%) were screen failures, the majority of which was due to failure to meet eligibility criteria (18.3%). A total of 1,306 patients were randomized to darolutamide plus docetaxel and ADT arm (N = 651), and placebo plus docetaxel and ADT arm (N = 654); 1 patient was excluded from the FAS due to a Good Clinical Practice violation (darolutamide plus docetaxel and ADT group). As of the data cut-off on October 25, 2021, the proportion of patients who discontinued from study treatment was higher in the placebo plus docetaxel and ADT arm (80.4%) than the darolutamide plus docetaxel and ADT arm (54.1%), with the most common reasons being clinical disease progression (41.6% versus 19.5.%) and radiological disease progression (20.2% versus 12.9%).

Table 10: Patient Disposition

ADT = androgen deprivation therapy; AE = adverse event; FAS = full analysis set; SAF = safety analysis set.

Note: In addition to the full analysis set and safety analysis set, the study also included a pharmacokinetics analysis set (PKS), which is not of interest to this review.

Data cut-off: October 25, 2021.

Source: ARASENS Clinical Study Report.⁷

Protocol Deviations

Important protocol deviations are summarized in Table 11. An important protocol deviation was reported in ||||||| of the study population, with the most common important deviation categories being procedure deviations (|||||||) and treatment deviations (|||||||). Procedure deviations were most frequently related to consent obtainment and study assessment (missed, incorrect, or incomplete procedures). Treatment deviations were most frequently related to incorrect dose, frequency, timing or method of drug delivery, and time frame between docetaxel administrations being less than 21 days. There was no imbalance in the important protocol deviations between treatment arms, with the exception that unblinding by error was notably more common in the placebo plus docetaxel and ADT arm (|||||||) than the darolutamide plus docetaxel and ADT arm (|||||||). Investigators were found to be unblinding patients to assist in determining the best subsequent treatment option. The sponsor indicated that the high number of protocol deviations was due to the use of a much broader and conservative approach in defining important protocol deviations than the definition by the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use and was not expected to compromise the integrity of study data.

Table 11: Important Protocol Deviations (FAS)

ADT = androgen deprivation therapy; FAS = full analysis set; PSA = prostate-specific antigen.

^aFrequency ≥ 10%.

Data cut-off: October 25, 2021.

Source: ARASENS Clinical Study Report.⁷

Exposure to Study Treatments

Exposure to study treatments is summarized in Table 12. In the darolutamide plus docetaxel and ADT arm, the median duration of darolutamide treatment was 41.0 (range = 0.1 to 56.5) months, while in the placebo plus docetaxel and ADT arm, the median duration of placebo treatment was 16.7 (range = 0.3 to 55.8) months. The median relative dose intensity was 100% in both arms (range = 32.1% to 100.1% in darolutamide plus docetaxel and ADT arm; range = 48.4% to 100.4% in placebo plus docetaxel and ADT arm).

The median duration of docetaxel treatment was 6 (range = 1 to 6) months in both treatment arms. The median relative dose intensity of docetaxel was 98.6% (range = 72.8% to 109.9%) in the darolutamide plus docetaxel and ADT arm, and 98.5% (range = 51.8% to 105.4%) in the placebo plus docetaxel and ADT arm.

All patients received ADT in the treatment period and most patients received a LHRH agonist or antagonist only, as summarized in Table 7.

Table 12: Study Treatment Exposure (SAF)

ADT = androgen deprivation therapy; SAF = safety analysis set; SD = standard deviation.

^aDuration of exposure = day of last dose of study drug – day of first dose of study drug + 1, divided by 30.44.

^bRelative dose intensity refers to the percent of planned dose, which incorporates treatment interruptions and dose reductions in the calculation.

Data cut-off: October 25, 2021.

Source: ARASENS Clinical Study Report.⁷

Concomitant Medications

All patients in the darolutamide plus docetaxel and ADT arm, and 99.1% in the placebo plus docetaxel and ADT arm, had concomitant medications. The most frequently used concomitant medication was ophthalmologicals (96.2%; similar in both groups). Bone-disease medications and analgesics were used concomitantly with the study treatments in 25.4% and 71.5% of patients, respectively. Of the patients who used concomitant corticosteroids (91.3%), prednisolone and prednisone were used in 20.3% and 20.8% of patients, respectively, while on docetaxel. Overall, there was no notable difference in the use of the concomitant medications between treatment arms.

Efficacy

Only those efficacy outcomes and analyses of subgroups identified in the review protocol are reported below. Refer to Appendix 2 for detailed efficacy data. The results presented are based on the data cut-off on October 25, 2021, and represent the final efficacy analysis of the ARASENS trial.

Overall Survival

The primary end point of OS is summarized in Table 13. A KM plot of OS is shown in Figure 2.

Table 13: Overall Survival Outcomes (FAS)

ADT = androgen deprivation therapy; CI = confidence interval; FAS = full analysis set; NR = not reached.

^aBased on Kaplan-Meier estimates.

^bThe hazard ratio and 95% CI were based on a Cox regression model, stratified by extent of disease (M1a vs. M1b vs. M1c) and alkaline phosphatase (< upper limit of normal vs. ≥ upper limit of normal).

^cP value was derived from the 1-sided log-rank test, stratified by extent of disease (M1a vs. M1b vs. M1c) and alkaline phosphatase (< upper limit of normal vs. ≥ upper limit of normal); the significance threshold was set at a 1-sided alpha of 0.025.

Data cut-off: October 25, 2021.

Source: ARASENS Clinical Study Report.⁷

The OS analysis was based on a median duration of follow-up of 43.7 (range not reported) months in the darolutamide plus docetaxel and ADT arm, and 42.4 months (range not reported) in the placebo plus docetaxel and ADT arm, when a total of 533 death events occurred (229 events [35.2%] and 304 events [46.5%], in respective arms). The KM estimate for the median OS was not reached in the darolutamide plus docetaxel and ADT arm, and was 48.9 (95% CI, 44.4 to NR) months in the placebo plus docetaxel and ADT arm, with an HR of 0.68 (95% CI, 0.57 to 0.80; P < 0.0001), in favour of darolutamide.

Sensitivity Analyses of OS

Results of the 3 prespecified sensitivity analyses of OS using different stratification approaches (unstratified analysis, using stratification data from eCRF, and using extent-of-disease stratification data from the central imaging review) were consistent with the results of the primary OS analysis.

Figure 2: Kaplan-Meier Plot of Overall Survival (FAS)

FAS = full analysis set.

Data cut-off: October 25, 2021.

Source: ARASENS Clinical Study Report.⁷

Table 14: Subgroup Analysis for OS (FAS)

ADT = androgen deprivation therapy; CI = confidence interval; ECOG = Eastern Cooperative Oncology Group; FAS = full analysis set.

^aThe hazard ratio and 95% CI were based on an unstratified Cox regression model.

^bBased on electronic case report form.^..

Data cut-off: October 25, 2021.

Source: ARASENS Clinical Study Report.⁷

A post hoc sensitivity analysis by number of cycles of docetaxel received postbaseline was conducted. The treatment effect was consistent across all subgroups (6 cycles versus 5 and fewer cycles; 6 and 5 cycles versus 4 and fewer cycles).

Subgroup Analyses of OS

A summary of the planned subgroup analyses of OS that are of interest to this review is presented in Table 14. Across the subgroups of interest, results were consistent across groups.

Time to CRPC

An analysis of time to CRPC, a secondary end point, is summarized in Table 15. A KM plot of CRPC events is presented in Figure 3.

The proportion of patients with a CPRC event was 34.6% in the darolutamide plus docetaxel and ADT arm, and 59.8% in the placebo plus docetaxel and ADT arm. The majority of CRPC events were attributed to PSA progression in both arms. The median time to CRPC was not reached in the darolutamide plus docetaxel and ADT arm, and was 19.1 (95% CI, 16.5 to 21.8) months in the placebo plus docetaxel and ADT arm, with an HR of 0.36 (95% CI, 0.30 to 0.42; P < 0.0001), in favour of darolutamide. Results of the sensitivity analysis performed based on both central and local PSA data were consistent with the primary analysis of time to CRPC.

Table 15: Time to CRPC (FAS)

ADT = androgen deprivation therapy; CI = confidence interval; CRPC = castration-resistant prostate cancer; FAS = full analysis set; NR = not reached; PSA = prostate-specific antigen.

^aPatients with multiple events were only counted for the category in which the first event occurred.

^bBased on the Prostate Cancer Clinical Trials Working Group 3 criteria.

^cBased on Response Evaluation Criteria In Solid Tumors (RECIST) version 1.1 criteria.

^dBased on Kaplan-Meier estimates.

^eThe hazard ratio and 95% CI were based on a Cox regression model, stratified by extent of disease (M1a vs. M1b vs. M1c) and alkaline phosphatase (< upper limit of normal vs. ≥ upper limit of normal).

^fP value was derived from the 1-sided log-rank test, stratified by extent of disease (M1a vs. M1b vs. M1c) and alkaline phosphatase (< upper limit of normal vs. ≥ upper limit of normal).

Data cut-off: October 25, 2021.

Source: ARASENS Clinical Study Report.⁷

Figure 3: Kaplan-Meier Plot of CRPC (FAS)

FAS = full analysis set.

Data cut-off: October 25, 2021.

Source: ARASENS Clinical Study Report.⁷

Time to Initiation of Subsequent Antineoplastic Therapy

An analysis of time to initiation of subsequent systemic antineoplastic therapy, a secondary end point, is summarized in Table 16. A KM plot of initiation of subsequent systemic antineoplastic therapy is presented in Figure 4.

Initiation of a new systemic antineoplastic therapy after discontinuation of study treatment occurred in 33.6% of patients in the darolutamide plus docetaxel and ADT arm, and 60.4% of patients in the placebo plus docetaxel and ADT arm. The median time to initiation of subsequent systemic antineoplastic therapy was not reached in the darolutamide plus docetaxel and ADT arm, and was 25.3 (95% CI, 23.1 to 28.8) months in the placebo plus docetaxel and ADT arm, with a HR of 0.39 (95% CI, 0.33 to 0.46; P < 0.0001), in favour of darolutamide. The most frequently used subsequent systemic antineoplastic therapies were abiraterone, enzalutamide, and cabazitaxel.

Table 16: Time to Initiation of Subsequent Systemic Antineoplastic Therapy (FAS)

ADT = androgen deprivation therapy; CI = confidence interval; FAS = full analysis set; NR = not reached.

^aBased on Kaplan-Meier estimates.

^bThe hazard ratio and 95% CI were based on a Cox regression model, stratified by extent of disease (M1a vs. M1b vs. M1c) and alkaline phosphatase (< upper limit of normal vs. ≥ upper limit of normal).

^cP value was derived from the 1-sided log-rank test, stratified by extent of disease (M1a vs. M1b vs. M1c) and alkaline phosphatase (< upper limit of normal vs. ≥ upper limit of normal).

^dSubsequent systemic antineoplastic therapy used in ≥ 2% of patients are summarized.

Data cut-off: October 25, 2021.

Source: ARASENS Clinical Study Report.⁷

Time to Pain Progression

An analysis of time to pain progression, a secondary end point, is summarized in Table 17. A KM plot of pain progression is presented in Figure 5. Pain progression occurred in 34.1% of patients in the darolutamide plus docetaxel and ADT arm, and in 37.9% of patients in the placebo plus docetaxel and ADT arm. The median time to pain progression was not reached (95% CI, 30.5 months to NR) in the darolutamide plus docetaxel and ADT arm, and was 27.5 (95% CI, 22.0 to 36.1) months in the placebo plus docetaxel and ADT arm, with an HR of 0.79 (95% CI, 0.66 to 0.95; P = 0.0058), in favour of darolutamide. Results of the sensitivity analyses (3 using different definitions of time to pain progression and 1 based on both ePRO device and paper questionnaires) were consistent with the primary analysis.

Figure 4: Kaplan-Meier Plot of Initiation of Subsequent Antineoplastic Therapy (FAS)

FAS = full analysis set.

Data cut-off: October 25, 2021.

Source: ARASENS Clinical Study Report.⁷

Table 17: Time to Pain Progression (FAS)

ADT = androgen deprivation therapy; CI = confidence interval; FAS = full analysis set; NR = not reached.

^aBased on Kaplan-Meier estimates.

^bThe hazard ratio and 95% CI were based on a Cox regression model, stratified by extent of disease (M1a vs. M1b vs. M1c) and alkaline phosphatase (< upper limit of normal vs. ≥ upper limit of normal).

^cP value was derived from the 1-sided log-rank test, stratified by extent of disease (M1a vs. M1b vs. M1c) and alkaline phosphatase (< upper limit of normal vs. ≥ upper limit of normal).

Data cut-off: October 25, 2021.

Source: ARASENS Clinical Study Report.⁷

Figure 5: Kaplan-Meier Plot of Pain Progression (FAS)

FAS = full analysis set.

Data cut-off: October 25, 2021.

Source: ARASENS Clinical Study Report.⁷

Health-Related Quality of Life

NCCN-FACT FPSI-17 questionnaire

HRQoL was an exploratory outcome; in the ARASENS trial, the NCCN-FACT FPSI-17 was used to approximate HRQoL. The mean total score and mean subscale scores (disease-related physical symptoms, disease-related emotional symptoms, treatment side effects, function, and well-being) over time are shown in Figure 6, Figure 11, Figure 12, Figure 13, and Figure 14, respectively.

The proportion of patients who completed the questionnaire in the FAS population was above 80% in both treatment arms at baseline. Other than week ||||||||||||||||, the proportion was above 50% up until week ||||||| in the darolutamide plus docetaxel and ADT arm, and up until week ||||||| in the placebo plus docetaxel and ADT arm.

The mean total score at baseline was ||||||||||||||||||||| in the darolutamide plus docetaxel and ADT arm, and ||||||||||||||||||||| in the placebo plus docetaxel and ADT arm. At the end of treatment, the mean change in score from baseline was ||||||||||||||||||||| in the darolutamide plus docetaxel and ADT arm, and ||||||||||||||||||||| in the placebo plus docetaxel and ADT arm.

Analyses of the subscales scores were consistent with the total score analysis, in which the mean baseline scores were similar between treatments arms. The mean score was stable at most time points but a decrease in score was observed near the end of treatment in both treatment arms. There was no notable difference in the mean change in score from baseline between treatment arms at most time points; however, this was not formally tested.

Objective Response Rate

This outcome was not measured in the study.

Time to Skeletal-Related Events

Analyses of time to first SSE and SSE-FS, which are secondary end points, are shown in Table 18.

Time to First SSE

The median time to first SSE was not reached in both treatment arms, with an HR of 0.71 (95% CI, 0.54 to 0.94; P = 0.0081), in favour of darolutamide. In both arms, most of the first SSE is attributed to the use of external beam radiation therapy to relieve skeletal symptoms. The KM plot of time to first SSE is shown in Figure 6.

Table 18: Time to First SSE and SSE-FS (FAS)

ADT = androgen deprivation therapy; CI = confidence interval; FAS = full analysis set; NR = not reached; SSE = symptomatic skeletal event; SSE-FS = symptomatic skeletal even-free survival.

^aBased on Kaplan-Meier estimates.

^bThe hazard ratio and 95% CI were based on a Cox regression model, stratified by extent of disease (M1a vs. M1b vs. M1c) and alkaline phosphatase (< upper limit of normal vs. ≥ upper limit of normal).

^cP value was derived from the 1-sided log-rank test, stratified by extent of disease (M1a vs. M1b vs. M1c) and alkaline phosphatase (< upper limit of normal vs. ≥ upper limit of normal).

^dPatients with multiple events were only counted for the category in which the first event occurred.

Data cut-off: October 25, 2021.

Source: ARASENS Clinical Study Report.⁷

Figure 6: Kaplan-Meier Plot of Time to First SSE (FAS)

FAS = full analysis set; SSE = symptomatic skeletal event.

Data cut-off: October 25, 2021.

Source: ARASENS Clinical Study Report⁷

Symptomatic Skeletal Event-Free Survival

The median SSE-FS was 51.2 (95% CI, 47.2 to NR) months in the darolutamide plus docetaxel and ADT arm, and 39.7 (95% CI, 36.0 to 42.3) months in the placebo plus docetaxel and ADT arm, with an HR of 0.61 (95% CI, 0.52 to 0.72; P < 0.0001), in favour of darolutamide. The most common event in both treatment groups was death.

PSA Outcomes

Time to PSA progression and PSA response rate are exploratory end points and are summarized in Table 19.

Table 19: Time to PSA Progression and PSA Response Rates (FAS)

ADT = androgen deprivation therapy; CI = confidence interval; FAS = full analysis set; NR = not reached; PSA = prostate-specific antigen.

^aBased on the Prostate Cancer Clinical Trials Working Group 3 criteria.

^bBased on Kaplan-Meier estimates.

^cThe hazard ratio and 95% CI were based on a Cox regression model, stratified by extent of disease (M1a vs. M1b vs. M1c) and alkaline phosphatase (< upper limit of normal vs. ≥ upper limit of normal).

^dThe P value has not been controlled for type I error and was derived from the 1-sided log-rank test, stratified by extent of disease (M1a vs. M1b vs. M1c) and alkaline phosphatase (< upper limit of normal vs. ≥ upper limit of normal).

^eAbsolute PSA response was defined as a postbaseline PSA level below 0.2 ng/mL in 2 consecutive evaluations.

^fThe risk difference and 95% CI were based on Cochran-Mantel-Haenszel test comparing between treatment arm, stratified by extent of disease (M1a vs. M1b vs. M1c) and alkaline phosphatase (< upper limit of normal vs. ≥ upper limit of normal).

^gRelative 90% PSA response was defined as a postbaseline 90% or greater reduction of the PSA level compared to the baseline in 2 consecutive evaluations.

Data cut-off: October 25, 2021.

Source: ARASENS Clinical Study Report.⁷

Time to PSA Progression

PSA progression occurred in 20.9% of patients in the darolutamide plus docetaxel and ADT arm, and 47.4% of patients in the placebo plus docetaxel and ADT arm. The median time to PSA progression was not reached in the darolutamide plus docetaxel and ADT arm, and was 22.4 (95% CI, 22.1 to 27.6) months in the placebo plus docetaxel and ADT arm, with an HR of 0.26 (95% CI, 0.21 to 0.31). Results of the sensitivity analysis based on central and local PSA data were consistent with the primary analysis.

PSA Response Rates

The absolute PSA response rate (the proportion of patients with PSA level below 0.2 ng/mL) was 48.7% at 6 months and 60.2% at 12 months in the darolutamide plus docetaxel and ADT arm, while in the placebo plus docetaxel and ADT arm, the rate was 23.9% at 6 months and 26.1% at 12 months. The risk difference between treatment arms was 25.0% ||||||||||||||||||||| at 6 months and 34.2% ||||||||||||||||||||| at 12 months.

The relative 90% PSA response rate (the proportion of patients with at least 90% PSA level reduction compared to baseline) was 67.6% at 3 months, 82.0% at 6 months, and 84.3% at 12 months in the darolutamide plus docetaxel and ADT arm, while in the placebo plus docetaxel and ADT arm, the rate was 42.8%, 54.4%, and 57.5% in the placebo plus docetaxel and ADT arm. The risk difference between treatment arms was 24.8% ||||||||||||||||||||| at 3 months, 27.6% ||||||||||||||||||||| at 6 months, and 26.8% ||||||||||||||||||||| at 12 months. Analyses of relative 30% and 50% response rates were consistent with the result of the relative 90% response rate analysis, in which there was a numerically higher proportion of PSA responders in the darolutamide plus docetaxel and ADT arm than the placebo plus docetaxel and ADT arm.

The relative 90% PSA response rate (the proportion of patients with at least 90% PSA level reduction compared to baseline) was 67.6% at 3 months, 82.0% at 6 months, and 84.3% at 12 months in the darolutamide plus docetaxel and ADT arm, while in the placebo plus docetaxel and ADT arm, the rate was 42.8%, 54.4%, and 57.5% in the placebo plus docetaxel and ADT arm. The risk difference between treatment arms was 24.8% ||||||||||||||||||||| at 3 months, 27.6% ||||||||||||||||||||| at 6 months, and 26.8% ||||||||||||||||||||| at 12 months. Analyses of relative 30% and 50% response rates were consistent with the result of the relative 90% response rate analysis, in which there was a numerically higher proportion of PSA responders in the darolutamide plus docetaxel and ADT arm than the placebo plus docetaxel and ADT arm.

Harms

Only those harms identified in the review protocol are reported here. Refer to Table 20 for detailed harms data.

Adverse Events

Almost all patients in the study reported at least 1 TEAE (99.5% in the darolutamide plus docetaxel and ADT arm; 98.9% in the placebo plus docetaxel and ADT arm). There is generally no notable difference in the incidence of all TEAEs between treatment arms, except that a numerically higher incidence of decreased appetite was observed in the darolutamide plus docetaxel and ADT arm (18.6%) compared to the placebo plus docetaxel and ADT arm (13.1%). The most common TEAEs in both treatment arms were alopecia (40.5% in the darolutamide plus docetaxel and ADT arm versus 40.6% in the placebo plus docetaxel and ADT arm), fatigue (33.1% versus 32.9%), and anemia (27.8% versus 25.1%).

Table 20: Summary of Harms (SAF)

ADT = androgen deprivation therapy; AE = adverse event; ALT = alanine aminotransaminase; AST = aspartate aminotransferase; SAF = safety analysis set; TEAE = treatment-emergent adverse event.

^aTEAEs were defined as any event(s) arising or worsening after the first dose of darolutamide or placebo until 30 days after the last dose of darolutamide or placebo administration.

^bFrequency ≥ 15%.

^cFrequency ≥ 3%.

^dFrequency ≥ 2%.

^eReported in ≥ 4 patients.

Data cut-off: October 25, 2021.

Source: ARASENS Clinical Study Report.⁷

At least 1 grade 3 or higher TEAE was reported in 70.2% of patients in the darolutamide plus docetaxel and ADT arm and in 67.5% in the placebo plus docetaxel and ADT arm. The most common TEAEs of grade 3 or higher (at least 10% of patients in either arm) were decreased white blood cell count (16.9% in the darolutamide plus docetaxel and ADT arm versus 14.9% in the placebo plus docetaxel and ADT arm), and neutropenia (8.6% versus 10.5%). The incidence of hypertension was numerically higher in the darolutamide plus docetaxel and ADT arm (6.4%) than the placebo plus docetaxel and ADT arm (3.2%).

Serious Adverse Events

At least 1 serious TEAE was reported in 44.8% of patients in the darolutamide plus docetaxel and ADT arm, and 42.3% of patients in the placebo plus docetaxel and ADT arm, with the most common serious TEAE in both arms being febrile neutropenia (6.1% in the darolutamide plus docetaxel and ADT arm, and 6.0% in the placebo plus docetaxel and ADT arm).

Withdrawals Due to AEs

Study treatment discontinuation due to TEAE was reported in 13.5% of patients in the darolutamide plus docetaxel and ADT arm, and 10.6% in the placebo plus docetaxel and ADT arm. The most common TEAEs that led to treatment discontinuation in the darolutamide plus docetaxel and ADT arm were increased aspartate aminotransferase level (0.9%) and increased alanine aminotransaminase level (0.6%).

Mortality

Death events were reported for 35.1% of patients in the darolutamide plus docetaxel and ADT arm, and 46.8% of patients in the placebo plus docetaxel and ADT arm. The majority of deaths in both arms were attributed to progressive disease (26.1% in the darolutamide plus docetaxel and ADT arm, and 36.0% in the placebo plus docetaxel and ADT arm). No specific TEAE was identified to account for the majority of deaths in either arm.

Critical Appraisal

Internal Validity

The ARASENS trial was a phase III, double-blind RCT. The methods of randomization, which involved stratification by extent of metastatic disease and ALP level, and Interactive Voice/Web Response System for concealment of the randomized assignment, were appropriate. The baseline characteristics of patients were generally balanced between treatment arms, suggesting randomization was adequate. A higher proportion of patients was unblinded by error during the study in the placebo plus docetaxel and ADT arm (|||||||) than the darolutamide plus docetaxel and ADT arm (|||||||), suggesting some inconsistencies in the implementation of blinding between treatment arms. Investigators were found to be unblinding patients to assist in determining the best subsequent treatment option. Bias in the measurement of the outcome in favour of darolutamide may result for subjective efficacy outcomes, including time to pain progression, time to first SSE, SSE-FS, and HRQoL. There is also a risk of bias in favour of the placebo group for harms outcomes (i.e., more frequent reporting of known harms in the darolutamide group). However, the extent of bias was likely to be small given the small number of unblinded patients. Reporting of objective outcomes—including OS, time to CRPC, time to initiation of subsequent systemic antineoplastic therapy, and PSA outcomes—would not be affected by unblinding.

A high number of important protocol deviations were reported in the study population (|||||||), which was mainly attributed to procedure deviations (|||||||) and treatment deviations (|||||||). Important protocol deviations were most frequently related to consent obtainment, study assessment (missed, incorrect, or incomplete procedures), incorrect dosing or method of drug delivery, and time frame between docetaxel administrations being too short. The sponsor indicated that the high number of protocol deviations was due to the use of a much broader and conservative approach in defining important protocol deviation than the definition by the ICH E3 classification and was not expected to compromise the integrity of study data. The sponsor noted that none of the important protocol deviations was considered to be major according to the old ICH E3 classification.

The statistical analyses were generally well designed. The efficacy analysis was conducted in the full analysis set, in which all randomized patients were included and analyzed by treatment assignment. The results correspond to the final efficacy analysis, which was conducted as planned a priori based on the number of death events occurred. A hierarchical gatekeeping approach was appropriately used to account for multiplicity in all secondary efficacy end points. No formal hypothesis testing was conducted for all exploratory efficacy end points (time to PSA progression, PSA response rates, HRQoL), and the statistical testing for PSA progression and response rates were associated with an increased risk of type I error due to the lack of control for multiplicity. For the prespecified OS subgroup analysis, there was a lack of sample size consideration, control for multiplicity, and treatment-by subgroup interaction analysis, which renders the findings exploratory. With respect to the Cox proportional hazards model for time-to-event analyses, visual inspection of the KM plots indicates that there do not appear to be any major violations of the proportional hazards assumption.

A higher proportion of patients discontinued treatment in the placebo plus docetaxel and ADT arm (80.4%) than the darolutamide plus docetaxel and ADT arm (54.1%). The most common reasons for treatment discontinuation were clinical disease progression (41.6% in the placebo plus docetaxel and ADT arm, and 19.5% in the darolutamide plus docetaxel and ADT arm) and radiologic disease progression (20.2% and 12.9%, respectively), while treatment discontinuation due to all other causes was balanced between treatment arms. There was no concern with the censoring rules.

A large amount of missing data was noted for the NCCN-FACT FPSI-17 questionnaire, which was used to approximate HRQoL by the sponsor. The completion rate declined to less than 50% after week ||||||| in the darolutamide plus docetaxel and ADT arm, and after week ||||||| in the placebo plus docetaxel and ADT arm. No data imputation was involved. There is a high risk of bias in measurement of the outcome from patients who remained in the trial, affecting the interpretability of HRQoL trends over time. The direction of the potential bias is difficult to predict. It should also be noted that the instrument has not been validated in patients with mCSPC, and most items in the questionnaire were measuring symptoms. As such, it is uncertain if the instrument can adequately capture HRQoL of the study population.

Time to pain progression was assessed based on the BPI-SF questionnaire and an MID estimate of 2 points was used to define pain progression. It should be noted that evidence for validity, reliability, and responsiveness of the BPI-SF instrument was not available for mCSPC and that most available evidence was for mCRPC. An MID for mCSPC has not yet been established. In consultation with the clinical experts, the MID estimate of 2 points used by the sponsor for defining time to pain progression was considered reasonable, although patients’ perspective on this is unknown.

External Validity

According to the clinical experts consulted by CADTH, the inclusion and exclusion criteria of the trial were reflective of the selection criteria for chemotherapy-eligible patients with mCSPC in clinical practice, except that the requirement for ECOG performance status is less stringent in clinical practice. The clinical experts noted that patients with significant comorbidities (e.g., cardiovascular diseases) and impaired performance status were excluded from the trial, limiting the generalizability of study results in this patient population commonly seen in clinical practice. With regard to baseline characteristics, the clinical experts noted that the study population had a younger median age and that there were more patients who identified as Asian, who had an ECOG performance status of 0, and who had de novo metastatic disease, compared to the general patient population in Canada. However, the clinical experts did not expect any of these differences to significantly affect the applicability of study findings in clinical practice.

The dosing of darolutamide, docetaxel, and ADT was consistent with the product monograph and clinical practice. Darolutamide was administered until disease progression (symptomatic disease progression or change in antineoplastic therapy) or unacceptable toxicity. The clinical experts noted that the treatment discontinuation criteria align with clinical practice where clinical presentation, radiographic findings, and PSA levels are factored into response evaluation and the decision to discontinue treatment. A limitation to note is that in the darolutamide plus docetaxel and ADT arm, ARATs were the most commonly prescribed subsequent antineoplastic therapy. This does not align with the clinical experts’ expectation, considering second- or later-line re-treatment with an alternate ARAT is currently not funded by most jurisdictions. The clinical experts noted that cabazitaxel, radium-223 chloride, and poly-(ADP-ribose)-polymerase inhibitors are used instead as subsequent therapies in most cases.

It should also be noted that ARAT (enzalutamide, apalutamide, or abiraterone plus prednisone) plus ADT accounts for the majority of systemic treatments currently prescribed for mCSPC in Canada according to the clinical experts. The clinical experts noted that docetaxel plus ADT was the SoC at the time the trial was designed; however, with new clinical evidence for ARAT plus ADT available in recent years, ARAT plus ADT became the most commonly prescribed regimen (and therefore the most relevant comparator). The generalizability of study results is therefore uncertain given that the use of docetaxel plus ADT is limited in the current practice. The absence of head-to-head evidence between darolutamide plus docetaxel and ADT versus ARAT plus ADT represents an evidence gap in mCSPC.

The trial included many outcomes that were indicated to be important in the clinician and patient input. The clinical experts noted that, from clinicians’ perspective, end points related to survival and disease progression (OS, time to mCRPC, and time to initiation of subsequent antineoplastic therapy) were the most important when evaluating treatment efficacy. Outcomes related to symptoms, SSE, PSA level, and HRQoL are important but generally serve as supporting evidence to inform treatment decisions, as per the clinical experts. The clinical experts agreed that the benefits of darolutamide plus docetaxel and ADT in survival and delaying disease progression were clinically meaningful. Of note, the clinical experts commented that the median time to mCRPC in the comparator arm (19.1 months) is shorter than anticipated (approximately 24 months) based on their clinical experience; as such, it is uncertain if the observed magnitude of benefit of darolutamide plus docetaxel and ADT will translate in clinical practice.

Objective response rate, an outcome of interest in the CADTH protocol, was not measured in the trial, although it was not a concern as per the clinical experts. The clinical experts commented that objective response rate is a RECIST criteria-based end point, and that it cannot be applied to the assessment of bone metastasis, which accounts for the majority of metastasis in mCSPC, due to the biology of the disease. The clinical experts noted that time to CRPC was a more appropriate end point for disease progression in the context of prostate cancer.

The clinical relevance of NCCN-FACT FPSI-17 instrument is uncertain, considering it is not routinely administered in clinical practice, according to the clinical experts. Nevertheless, the instrument does capture common symptoms (e.g., pain, difficulty in urination) and treatment-related side effects (e.g., fatigue, weight gain, decreased sexual function) indicated in the patient input, which are very relevant in the clinical assessment of patients with mCSPC in practice, as per the clinical experts.

Indirect Evidence

Objectives and Methods for the Summary of Indirect Evidence

As part of the development of a ||||||| cost-effectiveness model for darolutamide in mCSPC, the sponsor conducted and submitted an NMA that was used to inform these analyses, in the absence of direct evidence. The objective of this section is to summarize and critically appraise the methods and findings of the sponsor-submitted ITC on the relative efficacy of darolutamide and ADT in combination with docetaxel, which was indirectly compared to alternative treatments for patients with mCSPC.⁸

In addition to the sponsor-submitted ITC, a focused literature search for ITCs dealing with darolutamide and mCSPC was run in MEDLINE All (1946–) on July 13, 2022. No limits were applied. A total of 50 citations were identified and 8 full-text articles were retrieved for scrutiny. Two publications met the inclusion criteria for this review.^9,10

Description of Indirect Comparisons

The sponsor-submitted ITC first conducted an SLR to identify evidence for inclusion in a ||||||| ITC. The relative efficacy of darolutamide plus ADT and docetaxel from the ARASENS trial was indirectly compared to alternative treatments for patients with mCSPC via ||||||| NMA. Comparators of interest for the sponsor-submitted NMA included ADTs, abiraterone and prednisone, apalutamide, enzalutamide, and docetaxel. Outcomes of interest included OS, time to CRPC, and rPFS.⁸

Two additional NMAs were identified in the CADTH literature search (Menges et al. [2022] and Yanagisawa et al. [2022]).^9,10

The objective of the published SLR and NMA by Menges et al. was to assess the clinical effectiveness regarding survival and HRQoL, safety, and benefit-harm balance of mHSPC treatments, including docetaxel, abiraterone, enzalutamide, apalutamide, darolutamide, and radiotherapy (alone or in combination with ADT).⁹

The objective of the published SLR and NMA by Yanagisawa et al. was to analyze the benefit of triplet combination therapies with androgen receptor signalling inhibitors (abiraterone acetate, apalutamide, darolutamide, and enzalutamide) in combination with docetaxel and ADT, compared to available treatment regimens in patients with mHSPC.¹⁰

The populations, interventions, comparators, and designs of studies included in the sponsor-submitted NMA and the published NMAs are summarized in Table 21. Further details and results of these NMAs are provided subsequently.

Table 21: Study Selection Criteria and Methods for SLRs From ITCs

ADT = androgen deprivation therapy; AE = adverse event; ARSI = androgen receptor signalling inhibitor; CR = complete response; CRPC = castration-resistant prostate cancer; EBRT = external beam radiation therapy; HRQoL = health-related quality of life; ITC = indirect treatment comparison; LHRH = luteinizing hormone-releasing hormone; mHSPC = metastatic hormone-sensitive prostate cancer; mCSPC = metastatic castration-sensitive prostate cancer; NMA = network meta-analysis; ORR = objective response rate; OS = overall survival; nsAA = non-steroid antiandrogen; PFS = progression-free survival; PICO = Population, Intervention, Comparison, Outcomes; PR = partial response; PSA = prostate-specific antigen; RCT = randomized controlled trial; RoB = risk of bias; rPFS = radiographic progression-free survival; SLR = systematic literature review; SNA = nonsteroidal antiandrogen.

Source: Sponsor-submitted NMA,⁸ Menges et al. (2022),⁹ Yanagisawa et al. (2022).¹⁰

Methods of Sponsor-Submitted NMA

Objectives

The objective of the sponsor-submitted NMA was to inform the ||||||| cost-effectiveness model for darolutamide in mCSPC by indirectly comparing darolutamide and ADT in combination with docetaxel to alternative treatments ||||||||||||||, as well as to Canadian comparators of interest ||||||||||||||||||||.

Study Selection Methods

The sponsor-submitted NMA was informed by an SLR. Methods for identification of citations were limited to database searches of MEDLINE® In-Process (using Pubmed.com), Embase+ and MEDLINE (using Embase.com), and the Cochrane Library. The date of the search was not reported. No information on screening or data extraction methods, or methods for assessing risk of bias of the trials included in the SLR, were provided.

|||||||| Evidence Base

The SLR informing the sponsor-submitted NMA was originally conducted to inform the |||||||| NMA. There were |||||||| studies identified from the SLR as potentially relevant for inclusion in the |||||||| NMA. Minimal information on the selection of studies from the SLR for evaluation in the NMA were provided and appeared based solely on treatments evaluated in the trials identified by the SLR. Trials including docetaxel were grouped separately from those in which docetaxel was not included (other treatments for mCSPC including anti-androgens, radiotherapy, castration). In total, |||||||| trials were excluded from the |||||||| NMA based on treatment type. Reason for exclusion of these studies was provided and included irrelevancy of comparators ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||, or lack of comparator specification ||||||||||||||||||||||||.

Canadian Evidence Base

|||||||||||||||||||||||| some treatments that were not relevant to the Canadian market ||||||||||||||||||||||||||||||||||||||||:

• Enzalutamide and ADT with docetaxel: Studies evaluating enzalutamide with docetaxel and ADT (subgroups of the ENZAMET trial) were removed from consideration for the Canadian NMA as enzalutamide and ADT with docetaxel is not indicated, reimbursed, or commonly used in Canada, per the sponsor’s report.

• Abiraterone and ADT with docetaxel: The regimen of abiraterone and ADT with docetaxel (subanalysis of the PEACE-1 trial) is not indicated, reimbursed, or commonly used in Canadian clinical practice; and was therefore not considered an appropriate comparator. Additionally, the sponsor noted methodological limitations of the PEACE-1 trial that precluded its inclusion in the Canadian NMA.

• Standard nonsteroidal antiandrogens (SNAs): SNAs include bicalutamide, nilutamide, and flutamide. SNAs no longer represent the SoC in mCSPC since the availability of docetaxel and the androgen receptor-axis-targeted class of therapies. Current Canadian and international mCSPC guidelines do not recommend the use of SNAs in mCSPC and were not considered as appropriate comparators of interest for the NMA.

As such, the treatments of interest for the Canadian submission included the following:

• Darolutamide plus ADT and docetaxel

• Enzalutamide and ADT

• Apalutamide and ADT

• Abiraterone and prednisone with ADT

• Docetaxel and ADT

The sponsor also considered ADT monotherapy as a relevant comparator for the Canadian submission; however, results from this comparison were not reported in this review because it was not considered an appropriate comparator since it was not included in the CADTH systematic review protocol (Table 5).

In total, |||||||| additional studies were excluded ||||||||||||||||||||||||. A summary of studies identified for inclusion in the |||||||| NMA is depicted in Figure 7.

Although the ENZAMET study was a key trial of interest for enzalutamide, SNA was the main comparator in this study. The ENZAMET trial also included a subgroup of patients who received the concurrent combination of enzalutamide and ADT with docetaxel; however, this evaluation was not prespecified. The ENZAMET trial was excluded from this analysis because it does not provide evidence relative to other relevant treatments for Canada. To explore the implications of excluding this trial from the NMA, a comparison between the enzalutamide arms from the ENZAMET trial and the ARCHES trial was conducted. The ARCHES trial, the pivotal study of enzalutamide, included ADT as a comparator. ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. The resulting HR (95% CI) was |||||||||||||||||||||||| for the ARCHES trial versus the ENZAMET trial, favouring the ARCHES trial. As such, it was deemed conservative to exclude the ENZAMET trial, ||||||||||||||||||||||||||||||||||||||||||||||||||||||||.

Figure 7: Flowchart of Studies Identified for Inclusion in the NMA (Redacted)

Note that this figure has been redacted.

The end points of interest for the Canadian NMA were stricter than that of the SLR (Table 21) and included OS, time to CRPC, and rPFS. Time to CRPC was defined according to the ARASENS trial as the time from randomization to occurrence of the following events: PSA progression, radiological progression by bone lesions, or radiological progression by soft-tissue and visceral lesions, whichever occurred first. The sponsor considered time to CRPC as a more conservative definition of disease progression than rPFS because it has a closer representation of clinical practice, since clinical decisions in mCSPC are driven by disease progression through radiographic and biochemical means. Thus, time to CRPC was the preferred end point to model disease progression in the base-case NMA, and rPFS was included as a sensitivity analysis.

ITC Analysis Methods

In addition to the treatments evaluated and considered for the Canadian evidence base noted previously, an assessment of the heterogeneity of the eligible trials was conducted based on evaluation of trial design, the patient population, inclusion and exclusion criteria, and visual analysis of baseline characteristics, including age, ECOG performance status, Gleason score, PSA levels, and prostate cancer stage. No studies were excluded based on the heterogeneity assessment. An exploratory analysis of the individual patient data from the ARASENS trial was conducted to identify potential treatment effect modifiers.

The NMA was carried out using a ||||||||||||||||||||||||. All outcomes in the NMA were time-to-event outcomes (OS, time to CRPC, and rPFS). For these analyses, the reported HR and associated variance estimates (standard error or 95% CI) were used to derive the input data for the analysis. Additionally, the log HR and the standard error were also required from each study, and where these data were not reported, sufficient data to calculate these measures were required. ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. Both fixed and random-effects models were considered for the analysis of all outcomes, and model comparison methods were used to compare the goodness-of-fit. The preferred model was selected based on clinical plausibility of the estimated relative treatment effects and deviance information criterion, and/or the total residual deviance. For all outcomes, |||||||||||||||||||||||||||||||| were chosen as the base case based on the ||||||||||||||||.

Relative treatment effects were estimated using Markov chain Monte Carlo methods. A first series of iterations was discarded as “burn-in,” and the inferences were based on additional iterations using 3 chains with different starting values and checking convergence using the Brooks-Gelman-Rubin statistic and plots of posterior density.²⁸ The number of burn-in iterations discarded was not specified. If there was evidence that the Markov chain Monte Carlo had not converged, the number of burn-in iterations was increased. Once convergence had been achieved, further samples were generated from the posterior distribution to estimate treatment effects and 95% CrIs. Autocorrelation was assessed using autocorrelation plots to determine whether samples within each chain were highly correlated, and a suitable thinning interval was applied if needed to ensure that the chain mixed well and was representative of the posterior distribution.

In line with visual assessment of heterogeneity, statistical heterogeneity was evaluated based on the I² statistic derived from direct head-to-head meta-analysis of those treatment comparisons in each network that were reported by more than 1 study.

For all outcomes analyzed, inconsistency was assessed via “node-splitting” using the van Valkenhoef et al. method when closed loops were present.²⁹ Direct and indirect estimates along with CIs for each split treatment comparison were compared. Inconsistency was also presented as an inconsistency factor (IF) with an associated 95% CI. If the relative treatment effect was a ratio measure (e.g., OR HR), then the IF is the ratio of the direct and indirect estimates of the relative treatment effect such that if the 2 estimates are perfectly consistent then IF = 1.

Sensitivity Analysis

Sensitivity analyses were conducted on the choice of prior distribution by comparing 3 alternative prior distributions, which were non-informative and in line with those specified in the National Institute for Health and Clinical Excellence Decision Support Unit Technical Support Document 2 (NICE DSU TSD 2), to allow the posterior distribution to be primarily driven by the data.³⁰ The following prior distributions for the between-trials SD were assessed:

• Unif(0, 3): A uniform distribution indicating that any value of the between-trial SD between 0 and 3 is equally probable.

• Unif(0, 5): A uniform distribution indicating that any value of the between-trial SD between 0 and 5 is equally probable.

• Unif(0, 10): A uniform distribution indicating that any value of the between-trial SD between 0 and 10 is equally probable.

Radiographic PFS was not an outcome in the ARASENS trial. Additionally, time to CRPC was a composite of PSA progression and metastatic progression, while rPFS measures metastatic progression but not PSA. As such, rPFS was only included as a sensitivity analysis for modelling disease progression.

Results of Sponsor-Submitted NMA

Summary of Included Studies

A summary of potential sources of heterogeneity from the ||||| eligible studies is provided in Table 22. Each stage of the STAMPEDE trial was considered 1 study for a total of ||||| studies. No studies were excluded based on substantial heterogeneity. The studies were published between 2013 and 2020. Most studies were phase III (n = 6), 3 were open-label, and 4 were double-blind RCTs. Most studies had placebo/best SoC as comparator (n = 4), while 3 had active and placebo-controlled comparators. The median duration of follow-up varied across studies, ranging from 191 weeks to 364.5 weeks.

The definition of the population in each trial varied across the studies. The ARASENS, ARCHES, and CHAARTED trials defined the population as mHSPC, while the TITAN and LATITUDE trials describe the trial population as metastatic CSPC, although these definitions are generally used interchangeably. The GETUG-AFU 15 study included patients with non-castrate metastatic prostate cancer, and the STAMPEDE trial defined their trial population as patients with prostate cancer.

Baseline characteristics were visually assessed. Due to absence of data, it was not possible to assess the similarity of these baseline characteristics for all studies. Age, ECOG performance status, and Gleason score were generally similar across studies. All studies had a greater proportion of patients with ECOG performance status of 0. Per the sponsor’s analysis, of the studies that report ECOG, none was a clear outlier. There were notable differences in PSA levels in each study, ranging from just over 0 ng/mL to 100 ng/mL, and 1 study not reporting PSA level. Finally, prostate cancer stage was only reported in 4 studies, with the ARASENS trial differing from the 3 other studies that reported stage. Most patients in the ARASENS trial were T4, while most patients in the other studies were T3. The sponsor considered that T4 patients have a worse relative effect compared to T3 patients, and therefore the estimate using all patients for the ARASENS trial would be conservative. A notable difference across the trials was the presence and type of prior treatments; several trials did not report the proportion of patients having received specific prior treatments. Treatment with chemotherapy was also considered as an effect modifier that could have invalidated the transitivity assumption, as eligibility for chemotherapy is based on considerations including patient preference, age, performance status, and the presence of comorbidities. Outcomes of interest were not defined in the same way across all trials; therefore, in some cases, proxies needed to be used (e.g., failure-free survival in place of time to CRPC in some trials).

Table 22: Study Design and Other Characteristics Included in Heterogeneity Assessment

ADT = androgen deprivation therapy; BSC = best standard of care; DB = double-blind; ECOG = Eastern Cooperative Oncology Group; LHRH = luteinizing hormone-releasing hormone; NR = not reported; PBO = placebo; RCT = randomized controlled trial.

Source: Sponsor-Submitted NMA.⁸

Results of Sponsor-Submitted NMA

The network of evidence (Figure 8) for outcomes of interest including OS, time to CRPC, and rPFS were informed by the same set of studies. In total, ||||| studies comprising 6 treatment regimens were included in the analyses. Each stage of STAMPEDE was included as a separate study in the network.

The ||||||||||||| model was selected as the base case for OS, and time to CRPC based on the |||||||||||||||. Results of the |||||||||||||||||||| model comparing darolutamide plus ADT and docetaxel to relevant comparators for all end points is summarized in Table 23.

Figure 8: Sponsor-Submitted NMA Network Diagram

ADT = androgen deprivation therapy.

Each stage of the STAMPEDE trial was included as a separate study.

Source: Sponsor-submitted NMA.⁸

Table 23: ||||||||||||||||||||||||| Model for Relative Effect of Darolutamide Plus ADT and Docetaxel Compared to All Treatments for OS, Time to CRPC, and rPFS

ADT = androgen deprivation therapy; CrI = credible interval; CRPC = castration-resistant prostate cancer; HR = hazard ratio; OS = overall survival; rPFS = radiographic progression-free survival.

^aThe analysis of rPFS was considered a sensitivity analysis for modelling disease progression.

Source: Sponsor-Submitted NMA⁸

Overall Survival

For OS, darolutamide plus ADT and docetaxel was favoured over docetaxel and ADT (|||||||||||||||||||||||||||||||||||). There was insufficient evidence to show a difference between darolutamide plus ADT and docetaxel and enzalutamide, apalutamide, or abiraterone acetate, and the 95% CrIs were wide.

Inconsistency assessments between direct and indirect evidence for OS showed some differences (Figure 16). However, 95% CIs overlapped, and therefore no statistically significant differences were found. Inconsistency could not be tested specifically for comparisons to darolutamide plus ADT and docetaxel because these were informed only by indirect evidence. For all pairwise comparisons informed by at least 2 studies, the I² statistic was |||||.

Time to CRPC

For time to CRPC, darolutamide plus ADT and docetaxel was favoured over most treatments in the network, including apalutamide and ADT (|||||||||||||||||||||||||||||||||||), abiraterone and ADT (|||||||||||||||||||||||||||||||||||), and docetaxel and ADT (||||||||||||||||||||||||||||||||||), although the evidence was insufficient to show a difference compared to enzalutamide and ADT (||||||||||||||||||||||||||||||||||||) as the 95% CrIs were wide.

Inconsistency assessments for time to CRPC showed little differences between direct and indirect estimates (Figure 17). Inconsistency could not be tested specifically for comparisons with darolutamide plus ADT and docetaxel, because these were informed only by indirect evidence. The I² statistic was reported to be ||||||||||||||||||||||||||||||||||| for all pairwise comparisons informed by at least 2 studies except for |||||||||||||||||||||||||||||||||||.

Radiographic PFS

As a sensitivity analysis, for rPFS, darolutamide plus ADT and docetaxel was favoured over all other treatments in the network with HRs of ||||||||||||||||||||||||||||||||||| compared to docetaxel and ADT, ||||||||||||||||||||||||||||||||||| compared to apalutamide and ADT, ||||||||||||||||||||||||||||||||||| compared to abiraterone and ADT, and ||||||||||||||||||||||||||||||||||| compared to enzalutamide and ADT.

Inconsistency assessments for rPFS showed few differences between direct and indirect estimates (Figure 18). The I² statistic indicated that the variance due to heterogeneity across studies was reported to be ||||||||||||||||||||||||||||||||||| for all pairwise comparisons informed by at least 2 studies except for |||||||||||||||||||||||||||||||||||.

Additional sensitivity analyses for OS, time to CRPC, and rPFS using ||||||||||||||||||||||||| are summarized in Table 24. Results for sensitivity analyses were consistent with the base-case analyses, although 95% CrIs were wider.

Table 24: ||||||||||||||||| Model for Relative Effect of Darolutamide Plus ADT and Docetaxel Compared to All Treatments for OS, Time to CRPC, and rPFS

ADT = androgen deprivation therapy; CrI = credible interval; CRPC = castration-resistant prostate cancer; HR = hazard ratio; OS = overall survival; rPFS = radiographic progression-free survival.

^aThe analysis of rPFS was considered a sensitivity analysis for modelling disease progression.

Source: Sponsor-submitted NMA.⁸

Critical Appraisal of Sponsor-Submitted NMA

The sponsor-submitted NMA was informed by a ||||||||| SLR that included planned searches of multiple databases; other sources were not searched, but it is unclear whether this would have resulted in relevant studies being missed. Inclusion of studies was based on prespecified PICOS criteria; however, no information was provided on the methods of study selection or data extraction (e.g., duplicate reviewers), or conduct and results of a risk-of-bias assessment. As a result, there is a potential risk for bias and/or error in the SLR; however, the extent of this risk cannot be quantified. CADTH was unable to assess the quality of clinical studies included in the NMA, although, given that only phase III RCTs were included, the risk of bias of the trials was not expected to be high for objective outcomes.

There was no mention of an a priori protocol to inform the SLR and resulting NMA, and therefore it is unclear if the criteria for inclusion in the Canada-specific NMA were prespecified. Regardless, Canadian NMA criteria were more restrictive than for the ||||||||| SLR, and were based on appropriate Canadian comparators, including ADTs, abiraterone and prednisone, apalutamide, enzalutamide, and docetaxel. The combination of abiraterone plus ADT and docetaxel was considered a comparator of interest for this review by the clinical experts consulted by CADTH as well as the drug plans, but was not included in the sponsor’s analysis. Of note, the clinical experts reported that the combination has recently begun to be used by some clinicians in light of new clinical trial evidence⁴ and would not have been considered relevant at the time the NMA was conducted. As such, the comparative efficacy of darolutamide plus ADT and docetaxel and abiraterone plus ADT and docetaxel remains unknown.

The sponsor conducted a feasibility assessment to evaluate potential areas of heterogeneity based on study design and baseline characteristics, including age, ECOG performance status, Gleason score, PSA levels, and prostate cancer stage. No formal search for potential treatment effect modifiers was conducted; instead, the sponsor evaluated subgroups from the ARASENS trial, which were not powered to detect differences across subgroups. Overall, the sponsor did not exclude any studies based on heterogeneity, and although some heterogeneity was noted by the sponsor, it did not consider any baseline characteristics to be treatment-effect modifiers. The CADTH team and the clinical expert consulted by CADTH noted that important factors were considered; however, there remained notable differences in the baseline characteristics that could potentially result in changing relative treatment effects, including differences in PSA level, and prostate cancer stage. Additionally, there were differences in study design characteristics, including in blinding, in which 3 were open label and 4 were double blind. Open-label studies are more susceptible to bias, which could influence treatment effects. Differences in prior treatment requirements; a wide time span between trials (2013 to 2020), during which SoC may have changed; some differences in outcome definitions across the trials; and differences in follow-up duration were also evident. Although differences in study design are not feasible to address, there remains increased potential that the underlying transitivity assumption of the NMA has been violated.

In most cases, comparisons for almost all competing interventions were based on single trials, although 1 closed loop was formed between the connection of ADT, docetaxel and ADT, and abiraterone acetate (the LATITUDE, STAMPEDE-2, STAMPEDE-3, GETUG-AFU 15, CHAARTED, and STAMPEDE-4 studies) for which consistency between direct and indirect evidence was assessed. Direct and indirect evidence were not entirely consistent for OS. Statistical heterogeneity varied based on I² ||||||||| across outcome comparisons) and in general, there were few studies informing these comparisons; thus, they are subject to substantial bias and difficult to interpret. Comparisons to darolutamide plus ADT and docetaxel were informed entirely by indirect evidence, and therefore are associated with increased uncertainty.

Base-case analyses were conducted using |||||||||||||| models based on the ||||||||||||||, which was considered appropriate. However, |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. This assumption was derived from the ARASENS subgroup analysis, but the ARASENS trial was not powered to draw conclusions across subgroups. It does not appear that these assumptions were corroborated by other empirical evidence or expert opinion; thus, it is uncertain if |||||||||||||||||||||||||||| models were truly most appropriate. Regardless, sensitivity analyses using |||||||||||||||||||||| models with varying alternative prior distribution were conducted, for which results were consistent with the base case, although point estimates were associated with wide CrIs, highlighting the uncertainty in the results.

Outcomes included in the NMA were relevant to the treatment of mCSPC, focusing mainly on OS. In the sponsor-submitted NMA, time to CRPC was considered a proxy for modelling disease progression over rPFS because it has a closer representation of clinical practice, according to the sponsor. This assumption was also considered appropriate by the clinical experts consulted by CADTH. No outcomes related to HRQoL or safety were evaluated; thus, the impact of darolutamide plus ADT and docetaxel compared to other relevant treatments on HRQoL and safety remains unknown.

In the |||||||||||||| analysis comparing darolutamide plus ADT and docetaxel to other therapies available in Canada, the sponsor considered darolutamide plus ADT and docetaxel the most efficacious treatment versus comparators of interest for OS and time to CRPC. While point estimates trended in the direction favouring darolutamide plus ADT and docetaxel, the 95% CrIs were wide, often crossing the threshold of no effect, particularly for the main outcome of OS, thereby increasing the uncertainty in the comparative effect estimates. Results for sensitivity analyses using |||||||||||||| models were consistent with the ||||||||||||||, although 95% CrIs were even wider, suggesting increased uncertainty.

Methods of NMA by Menges et al. (2022)⁹

Objectives

This NMA aimed to comprehensively assess the clinical effectiveness regarding survival and HRQoL, as well as the safety and benefit-harm balance of combinations of docetaxel, ARATs, and radiotherapy with ADT compared to ADT alone in patients with mHSPC.

As previously noted, comparisons to ADT monotherapy are not of interest to this review because ADT monotherapy was not considered an appropriate comparator in the CADTH systematic review protocol (Table 5). As such, results for comparisons to ADT monotherapy were not reported in this review.

Study Selection Methods

A protocol for the systematic review for the NMA was registered on PROSPERO (CRD42020171129). Three databases (MEDLINE, EMBASE, and Cochrane Central Register of Controlled Trials [CENTRAL]) and ClinicalTrials.gov were searched for records of studies published up until February 25, 2020, with an update on March 1, 2022, for studies or updated results published between January 2020 and March 1, 2022. Relevant conference proceedings (American Society of Clinical Oncology [ASCO], American Urological Association [AUA], European Society for Medical Oncology [ESMO], European Association of Urology [EAU], AUA from 2016 to 2020, and APCCC from 2017 and 2019) were searched for publications between January 2016 and April 2020. Reference lists of recently published SLRs were also screened, although no date range for the year of publication was defined. Only RCTs, including multiarm, multistage RCTs, were considered for inclusion in the SLR. The full screening, eligibility assessment, data extraction, and risk-of-bias assessment was performed in duplicate by 3 reviewers and during the search update; this process was performed by 1 reviewer and the data and risk-of-bias assessment were verified by a second reviewer (studies including darolutamide were included at the stage of the update search). The risk of bias in the studies was assessed by reviewers using the Cochrane risk-of-bias 2.0 tool;¹¹ the information from the bias assessment was presented alongside the results. Appraisal using the Grading of Recommendations, Assessment, Development and Evaluations (GRADE) approach was undertaken to assess the certainty of evidence for direct, indirect, and NMA effect estimates, and was presented alongside the results.³¹

Eligible studies included men with newly diagnosed mHSPC who had not previously received systemic treatment for prostate cancer. Studies including patients pretreated with ADT alone in or combination with non-steroid antiandrogen (nsAA) for no longer than 24 months and at least 12 months before enrolment, or for less than 6 months before enrolment, were considered eligible. Patients with non-metastatic prostate cancer or with receipt of chemotherapy before enrolment were not considered eligible.

The following interventions were eligible for inclusion: ADT and docetaxel (75 mg/m² body surface area IV every 3 weeks for 6 cycles) in combination with prednisone 10 mg/day during 6 cycles, followed by ADT alone; ADT and abiraterone acetate (1,000 mg/day orally) in combination with prednisone 5 mg/day until disease progression; ADT and enzalutamide (160 mg/day orally) until disease progression; ADT and apalutamide (240 mg/day orally) until disease progression; ADT and darolutamide (1,200 mg/day orally) until disease progression; and ADT and radiotherapy (external beam radiation therapy to the prostate with unspecified dose and frequency), followed by ADT alone. Any concurrent or per-protocol immediate sequential combination of these treatments was also considered eligible. Comparators of interest included ADT alone or in combination with placebo (various administrations or orchiectomy); and ADT plus nsAA alone or in combination with placebo (various administrations including orchiectomy).

Outcomes included OS, PFS, HRQoL, and AEs. The PFS and HRQoL analysis did not include any studies that studied darolutamide and the network was not provided; therefore, the results for PFS and HRQoL are not discussed in this report. The AE analysis included darolutamide, but it was only compared to ADT in the analysis; thus, it was not included in this report.⁹ A quantitative benefit-harm assessment was also conducted; however, this was not considered an outcome of interest for this review and is therefore not discussed.

Analysis Methods

The list of eligible interventions and comparators included specific dosages and administration routes; therefore, individual studies with the same intervention and comparator arms were pooled and their effect estimates analyzed with an NMA to obtain pooled results at the intervention level. Effect estimates obtained from this analysis informed the direct estimates in the subsequent pairwise analysis and NMA.⁹

Of note, studies using ADT and nsAA therapy as a comparator arm (the ENZAMET and NCT02058706 trials) were not excluded from the main analysis but were combined with the ADT-only comparator in their group (which included the ARCHES trial using ADT alone as a comparator) and assumed to be equivalent. Authors noted that this was in line with existing evidence from systematic reviews and with approaches taken in other analysis groups. In addition, correlation matrices based on the overlap in patient accrual in different time periods assuming a constant variance were estimated for arms C and G in the STAMPEDE trial due to the multiarm, multistage design, and were used to estimate the overall NMA model; this assumed no relevant variation across different treatments or time periods.⁹

A frequentist, random-effects pairwise meta-analysis and NMA was conducted using study-level aggregate data. The frequentist framework was chosen for the meta-analysis based on the limited number of included studies and a recommendation by Langan et al.³² Random-effects models were selected because the authors assumed there would be important study-level variation.⁹

Transitivity was assessed according to epidemiological criteria based on the presence of potential effect modifiers and clinical plausibility. Heterogeneity was assessed visually and using the I² statistic. Direct estimates, indirect estimates, and the results from the NMA were assessed individually for certainty of evidence using the GRADE criteria and extension for network meta-analysis. The presence of inconsistency between direct and indirect estimates was evaluated using IFs (HR from direct evidence divided by HR from indirect evidence, with a variance equal to the sum of direct and indirect variances). Potential publication bias was considered by the authors as the likelihood that a trial would have been conducted but not published, and using a comparison-based funnel plot with ADT plus docetaxel and darolutamide as a reference.⁹

Preplanned subgroup analyses were conducted for 4 patient groups, which were selected based on published clinical guidelines and conversations with clinical exports: high and low volume (defined as visceral metastases or 4 or more bone lesions with 1 or more outside of the vertebral bodies and pelvis, per the CHAARTED trial); de novo mHSPC or mHSPC progressing after prior local therapy; low or high risk (defined as at least 2 of Gleason score ≥ 8 and/or ≥ 3 lesions on bone scan and/or presence of measurable visceral lesions, per the LATITUDE trial); and high or reduced ECOG score (0 and ≥ 1).⁹ All comparisons were made versus ADT alone and are therefore not reported.

In addition, several sensitivity analyses were performed to explore the influences of uncertainties relating to the evidence itself, including the exclusion of evidence from the LATITUDE trial due to high-risk de novo mHSPC population, the ENZAMET and NCT02058706 trials due to uncertainties regarding patient populations and ADT and nsAA as comparator, the GETUG-AFU 15 and CHAARTED trials due to likely incomplete recording of AEs suggested by substantially lower AE rates compared to other studies, and the ARASENS trial due to effect estimates being estimated only indirectly through NMA.⁹

Results of NMA by Menges et al.⁹

Summary of Included Studies

From the SLR, a total of 10 trials were eligible for analysis representing data from 11,557 patients, which informed the NMA. The STAMPEDE trial arms C, G, and H were reported as distinct trials based on their different interventions, resulting in a total of 12 studies.

Study and participant characteristics of included RCTs were summarized in a supplementary appendix. The authors noted several key differences between the studies in study design and baseline characteristics. The majority of included RCTs were open label (n = 8), with the remaining 4 trials double blinded. The 3 arms of the STAMPEDE trial were all multiarm, multistage design. In terms of interventions, 3 RCTs each investigated ADT plus docetaxel and ADT and enzalutamide, 2 RCTs each investigated ADT plus abiraterone acetate and ADT plus radiotherapy, and 1 RCT each investigated ADT plus apalutamide, and ADT plus docetaxel and darolutamide. All RCTs were found to be at low risk of bias, although the authors noted some uncertainty regarding overproportioned censoring in the ADT plus enzalutamide group and inconsistencies in reporting censoring and death events in 1 study. Follow-up times for OS ranged from 34 to 83 months.

Mean age was similar across all included RCTs, ranging from 63 to 70 years, with 6 RCTs reporting age ranges greater than 12 years. Race and other demographic characteristics were not reported. Across included studies, the proportion of patients with prior local therapy ranged from 0% to 42%, the proportion with de novo mHSPC ranged from 63% to 100% (not reported in 2 RCTs), 55% to 98% had a Gleason score between 8 and 10, and the proportion of patients with high-volume disease ranged from 41% to 82% (not reported in 1 RCT). The proportion of patients with an ECOG performance status greater than or equal to 1 ranged across RCT arms from 1% to 56%. The proportion of patients with visceral metastases was all under 20%.

Results

The network diagram for the NMA is presented in Figure 9. A total of 21 distinct intervention plus comparator combinations were evaluated for the primary outcome of OS. Performance characteristics for the models were not presented.

Overall Survival

Results of the NMA for OS for comparisons involving darolutamide are summarized in Table 25. For OS, darolutamide plus ADT and docetaxel was favoured over ADT and docetaxel (HR, 0.68 [95% CI, 0.57 to 0.81]). There was no difference in terms of OS between darolutamide plus ADT and docetaxel and other comparators consisting of abiraterone acetate plus ADT and prednisone, enzalutamide and ADT, apalutamide and ADT, and apalutamide plus ADT and docetaxel.

The GRADE assessments conducted on the direct and indirect estimates reported either “low” or “very low” ratings for all but 2 of the estimates of interest from the NMA; studies with a “low” or “very low” score were downgraded due to imprecision in the estimate.

Figure 9: Network Diagram for OS (Menges et al. [2022])

Abi = abiraterone; ADT = androgen deprivation therapy; Apa = apalutamide; Dar = darolutamide; Doc = docetaxel; Enz = enzalutamide; RT = radiotherapy.

Thickness represents the number of supporting studies (scale not defined in publication).

Source: Menges et al. (2022).⁹

Table 25: Findings and GRADE Assessment of Direct, Indirect, and NMA Evidence on OS (Menges et al.)

ADT = androgen deprivation therapy; CI = confidence interval; HR = hazard ratio; GRADE = Grading of Recommendations, Assessment, Development and Evaluations; NMA = network meta-analysis.

^aDirect estimates for individual treatments were informed by pairwise meta-analysis (results not shown).

Source: Menges et al. (2022).⁹

Critical Appraisal of NMA by Menges et al. (2022)

The NMA by Menges et al. was informed by an SLR in 2020 that included a comprehensive search comprising multiple databases and several cancer-focused conference proceedings, as well as searching for potential clinical trials in progress but not yet published. In addition, an update of the SLR was conducted on March 1, 2022, to include emerging therapies. Methods of screening, data extraction, and quality assessment were outlined and considered appropriate. A risk-of-bias assessment of the included studies was conducted per the Cochrane risk-of-bias 2.0 tool,¹¹ in which the authors considered all studies to have a low risk of bias for the main outcome of OS. The population, interventions, and outcomes of the NMA were also considered relevant to Canadian clinical practice. Funding for this study was obtained from the Swiss Medical Board and was not supported by the sponsor.

The authors note that transitivity was assessed using epidemiological criteria and the presence of potential effect modifiers, along with considerations of clinical plausibility. It is not clear how these effect modifiers were identified. Some potentially important effect modifiers were reported in the study but were not adjusted for nor discussed. Follow-up duration was noted but was not considered an effect modifier, or considered in the analyses. Given the outcome of OS and the use of HRs that rely on the assumption that a patient has survived to that point, this could have implications for the results, because studies with longer follow-up may have captured more robust results. Other potentially important factors would be the use of prior therapy, which may have an impact on treatment effects, because 2 studies permitted prior docetaxel and 1 permitted early use of docetaxel in the trial itself. Additionally, Gleason score, which may represent a treatment-effect modifier, ranged from 55% to 98% across studies for scores of 8 to 10 representing high-grade cancer; thus, differences by cancer grade may have varied but were not explored. As a result, it is likely that the underlying transitivity assumption of the NMA was violated.

Heterogeneity was assessed visually and by means of I² values. No information on the visual heterogeneity assessment was provided. Although minimal given the limited number of studies per comparison, there was evidence of moderate statistical heterogeneity in the comparison of the ENZAMET and NCT02058706 trials (I² = 33%). A sensitivity analysis was conducted excluding these 2 studies, which led to some changes in the overall results, although this did not impact the comparisons involving darolutamide. Reported heterogeneity in the overall NMA was low (I² = 0.0%) for OS. However, given the small number of studies, the I² value presented is prone to bias.

The authors conducted a frequentist, random-effects, pairwise meta-analysis and NMA. No information on model selection and fit was provided, and therefore it remains uncertain whether this choice was appropriate. However, given the evidence base, random-effects models were likely appropriate to allow for the possibility of heterogeneity between studies. No scenario analyses were conducted for different models. The authors used the GRADE criteria to assign certainty ratings to the direct and indirect estimates of the NMA results. In comparisons involving darolutamide, certainty of NMA estimates was often considered to be low or very low, suggesting that the comparative effects derived from the NMA are likely to be quite different than the true effects.

Only the end point of OS contained complete information for inclusion in this review. In general, comparisons were associated with wide 95% CIs, suggesting imprecision and uncertainty in the results. Sensitivity and subgroup analyses, while conducted and generally in line with the primary analysis, were only comparing treatments to ADT monotherapy. While PFS, HRQoL, and AEs were assessed as part of the objectives of this review, PFS and HRQoL did not include data for darolutamide plus ADT and docetaxel. Results for AEs were only reported in aggregate, by grade, and were only compared to a reference treatment of ADT; thus, they were not considered relevant to this review.

Methods of NMA by Yanagisawa et al. (2022)¹⁰

Objectives

The objective of the NMA by Yanagisawa et al. (2022) was to analyze the oncologic outcomes of combination therapy with androgen receptor signalling inhibitors plus docetaxel and ADT and to compare their efficacy to currently available treatments.

As previously noted, comparisons to ADT monotherapy are not of interest to this review because ADT monotherapy was not considered an appropriate comparator in the CADTH systematic review protocol (Table 5). As such, results for comparisons to ADT monotherapy were not reported in this review.

Study Selection Methods

The systematic review was conducted according to an a priori protocol that was registered on the PROspective Multicentre Study to Identify Diagnostic Key PERfOrmance Indicators in Upper GI Endoscopy (PROSPERO) (CRD42022298107). In April 2022, a literature search was performed in PubMed, Web of Science, and Scopus databases to identify publications investigating the oncologic outcomes of systemic therapy for mHSPC. Abstracts for relevant conference proceedings (ASCO and ESMO) were also searched. No restrictions to the search were placed on publication year or language. In addition, manual searches of the reference lists from relevant articles, as well as those from all papers ultimately included in the analysis, were performed to identify additional studies of interest. Only RCTs were considered for inclusion. Studies that did not include original patient data, review articles, letters, editorial comments, replies from authors, case reports, and articles not written in English were excluded. In cases where studies contained duplicate cohorts, the highest-quality or most recent publication was selected for inclusion.

Initial screening was performed by 2 authors based on titles and abstracts, and then potentially relevant studies were subject to a full-text review. Data were extracted independently by 2 authors. Study risk of bias was assessed independently by 2 authors using the Cochrane risk-of-bias 2.0 tool.¹¹ Results of the assessment were reported alongside the other results in the study. At all stages of screening and extraction, disagreements or discrepancies were resolved by consensus with coauthors.

Eligible studies included those that enrolled patients with mHSPC who were treated with triplet combination therapy using an androgen receptor signalling inhibitor plus docetaxel and ADT compared to patients treated with other currently available, systemic treatment strategies. No further information on eligible interventions or comparators was provided.

Outcomes included in the review consisted of OS, PFS, or AEs. Analysis of PFS in the NMA did not include any results from the ARASENS trial; thus, the results for PFS are not discussed in this report. Results for AEs from the NMA were presented by grade and by preferred term, although analyses were only conducted for comparisons to ADT, which was not a relevant comparator and therefore is not discussed in this review.

Analysis Methods

Following the SLR, an initial fixed-effects pairwise meta-analysis was conducted to analyze the relationships between combination therapy and PFS and OS using the HRs from included trials. Results for this meta-analysis are not summarized in this review.

Network diagrams were utilized to illustrate the connectivity of the treatment networks in terms of OS, PFS, and AEs. For OS and PFS outcomes, a frequentist NMA using random effects was performed for direct and indirect treatment comparisons. No justification was provided for either the choice of meta-analysis model or the choice of NMA model for outcomes. Contrast-level data from individual studies were used, including HRs and SD or 95% CIs. Results for OS were presented for each unique intervention against ADT alone, ADT and docetaxel, and ADT plus docetaxel and abiraterone. The relative effects were presented as HRs and 95% CIs. Subgroup analyses for PFS and OS included high- versus low-volume disease and de novo versus metachronous metastasis. Heterogeneity for the NMA was assessed using Cochrane’s Q statistic when more than 1 trial was available for comparisons. I² statistics were not reported for the NMAs, despite being prespecified.

Sensitivity analyses were conducted for patients with de novo/metachronous metastasis, and tumour burden due to differences in patient populations. It was unclear if this sensitivity analysis was prespecified.

Results of NMA by Yanagisawa et al. (2022)¹⁰

Summary of Included Studies

After title, abstract, and full-text screening, a total of 17 publications representing 11 RCTs and 7,679 patients were included in the meta-analysis and NMAs. Trials included in the analysis were the ARASENS, ARCHES, CHAARTED, ENZAMET, GETUG-AFU15, LATITUDE, PEACE-1, STAMPEDE (arms G, C, and G; and B, C, and E), and TITAN trials. Due to duplicate cohorts, only arms C versus G from the STAMPEDE trial were included in the primary analysis for OS, while arms B, C, E, and G were included from prior publications for the subgroup analyses. Patients treated with androgen receptor signalling inhibitor plus ADT and docetaxel in the PEACE-1, ARCHES, ENZAMET, and TITAN trials were extracted from their subgroup analyses. Only the abiraterone acetate and ADT arm from the LATITUDE trial was included, because none of the patients in the study received docetaxel.

Characteristics of all studies were reported descriptively; they consisted of dosages, ages, proportions with de novo disease, disease volume, the number of patients treated with docetaxel, HRs for all outcomes and for post-hoc subgroups, and the length of the follow-up period. Information on study designs of the included studies was not reported. Risk-of-bias assessments indicated that the majority of included RCTs had a high risk of bias in terms of outcome assessment and the blinding of participants and personnel. Follow-up time, representing another potential source of heterogeneity for outcome assessment, ranged from 34 to 83.9 months.

The interventions represented in these RCTs were abiraterone acetate and SoC (with or without radiotherapy), abiraterone acetate and ADT, apalutamide and ADT, docetaxel and ADT, darolutamide plus ADT and docetaxel, and enzalutamide and ADT. The comparator arms represented by these studies were SoC (with or without radiotherapy), placebo plus docetaxel and ADT, placebo and ADT, nsAA and ADT, docetaxel and ADT, and ADT alone.

Median age was similar across all arms of studies, ranging from 63 to 70 years. The proportion of patients with de novo disease ranged from 66% to 100%, and the proportion with high- and low-volume disease ranged from 47% to 82% and 18% to 53%, respectively. Out of 11 RCTs, 6 reported on de novo metastasis, 5 reported on prior local treatment, and 5 reported on docetaxel use during the trial.

Results

The network diagram for the NMA assessing OS is presented in Figure 10. A total of 7 individual interventions were analyzed in the NMA for OS. No performance measures were reported for the model fit and results were not adjusted for any effect modifiers.

Figure 10: Network Diagram for OS NMA (Yanagisawa et al. [2022])

ABI = abiraterone; ADT = androgen deprivation therapy; APA = apalutamide; DAR = darolutamide; DOC = docetaxel; ENZ = enzalutamide; NMA = network meta-analysis; OS = overall survival.

Source: Yanagisawa et al. (2022).¹⁰

Overall Survival

Table 26 demonstrates the full results of the NMA for OS for darolutamide from the primary analysis, including all patients, as well as from the subgroup analysis of de novo metastatic patients and metachronous patients, against the comparators of ADT and docetaxel, and ADT and abiraterone acetate.

For OS, darolutamide plus ADT and docetaxel was favoured over ADT and docetaxel (0.68 [0.56 to 0.82]) and ADT and abiraterone acetate (HR, 0.74 [95% CI, 0.55 to 0.99]). Among the subgroup of studies reporting on de novo patients, darolutamide plus ADT and docetaxel was favoured over ADT and docetaxel (HR, 0.71 [95% CI, 0.55 to 0.92]). There was no difference in OS between darolutamide plus ADT and docetaxel versus ADT and docetaxel.

Table 26: OS NMA Results (Yanagisawa et al. [2022])

ADT = androgen deprivation therapy; CI = confidence interval; HR = hazard ratio; NMA = network meta-analysis; OS = overall survival.

Source: Yanagisawa et al. (2022).¹⁰

Results for subgroup analysis in patients with high- or low-volume disease were not available for darolutamide.

Adverse Events

No network diagram was available for the NMA of AEs. For comparators of interest for the analysis of AEs, darolutamide plus ADT and docetaxel was only compared to abiraterone and ADT for the AE of febrile neutropenia, when darolutamide plus ADT and docetaxel was associated with greater odds of febrile neutropenia (OR, 26.62 [95% CI, 7.46 to 94.99]).

Critical Appraisal of NMA by Yanagisawa et al. (2022)¹⁰

The ITC conducted by Yanagisawa et al. was informed by an adequately conducted SLR that included planned searches of multiple databases, and conference proceedings. Screening was conducted based on standard methods, with studies selected independently in duplicate. A risk-of-bias assessment of the included studies was conducted, per the Cochrane risk-of-bias 2.0 tool,¹¹ in which, overall, the authors considered all included studies to be at a low risk of bias due to the prospective, randomized design. However, the authors also judged there to be a high risk of bias for outcome assessment blinding (detection bias) and participant or personnel blinding (performance bias). No adjustments or sensitivity analyses, excluding studies with these biases, were undertaken. Some authors mentioned competing conflicts of interest for this study; however, funding for this study was not related to the sponsor.

Networks were constructed based on the available studies; however, it was unclear which studies were informing comparisons based on the network structure, and whether direct or indirect evidence was assessed. One closed loop was formed in the network for OS; however, no assessment of consistency was conducted. For the NMA, potential sources of heterogeneity were only evaluated statistically using Cochrane’s Q test; however, results of the heterogeneity assessment were not reported. Heterogeneity of comparisons was also evaluated in the initial meta-analysis via Cochrane’s Q test, although results did not suggest any important heterogeneity (P = 0.5). Given that no significant heterogeneity was observed, it was not explored for the NMA. As a result, it was unclear if the transitivity assumption was met, because no assessment of clinical heterogeneity via study design or baseline characteristics was conducted. Additionally, no consideration was given to treatment-effect modifiers, and thus the impact of any potential effect modifiers remains unknown. The authors noted that the publications included in the NMA included different patient populations with regards to proportions of patients with de novo disease and disease burden. As a result, the authors conducted sensitivity analyses that excluded studies based on population; however, they noted that the results must be interpreted with caution due to the limited sample sizes, thus decreasing statistical power. Given the relative lack of attention to potential effect modifiers, it is likely that the transitivity assumption underlying the NMA was violated.

The NMA was conducted using a frequentist, random-effects model. No model-fit statistics or justification for model choice was provided, and therefore it remains unclear if random-effects models were the most appropriate choice. However, given the potential for increased, unexplored heterogeneity across studies, random-effects models were likely the most appropriate. No scenario analyses were conducted for different models. All OS estimates were informed by contrast-level data, which was considered appropriate given the time-to-event–related outcomes. This method also allows for preservation of randomization across comparisons. For AEs, arm-level data were used, which, given that the method of reporting of AEs in clinical trials was also considered appropriate, however, does not preserve randomization across comparisons. Additionally, for the STAMPEDE trial, only certain arms were included in the analysis, and therefore randomization throughout the NMA was not preserved for these comparisons.

Overall survival and AEs were analyzed in this NMA. The intervention of interest to this review, darolutamide plus ADT and docetaxel, was only compared to ADT and docetaxel, and to abiraterone plus ADT and docetaxel. The network diagram for this analysis also included comparators of enzalutamide and apalutamide, for which no comparative results were displayed; thus, comparative efficacy remains unknown for this analysis. In all cases, darolutamide plus ADT and docetaxel was favoured; however, this was generally associated with wide 95% CIs, suggesting imprecision and uncertainty in the estimates, although the reason for this was unexplored and remains unknown.

Summary

In total, 3 NMAs were eligible for inclusion in this review, 1 which was commissioned and submitted by the sponsor, and 2 of which were published NMAs. The sponsor-submitted NMA included a systematic review with unclear methodology and used a |||||||||||||| approach to evaluate the relative clinical efficacy of darolutamide plus ADT and docetaxel compared to enzalutamide and ADT, apalutamide and ADT, abiraterone plus prednisone and ADT, and docetaxel and ADT in adult patients with mHSPC. In the published NMAs by Menges et al. and Yanagisawa et al., SLRs were also conducted before evaluation using a frequentist approach in both cases. In the study by Menges et al., the authors aimed to assess the survival, HRQoL, and safety of mHSPC treatments including docetaxel, abiraterone, enzalutamide, apalutamide, darolutamide in combination with ADT. In the study by Yanagisawa et al. the authors aimed to analyze the survival and safety benefits of abiraterone acetate, apalutamide, darolutamide, and enzalutamide in combination with docetaxel and ADT, compared to other treatments in patients with mHSPC.

The evidence included across NMAs was generally similar and evaluated mostly the same treatments. There were ||||||||, 10, and 8 studies (combining STAMPEDE trials) included in the NMAs by the sponsor, Menges et al., and Yanagisawa et al., respectively. All NMAs included the ARASENS, ARCHES, CHAARTED, LATITUDE, TITAN, and STAMPEDE trials, although different arms of the STAMPEDE trial were included across NMAs. The GETUG-AFU15 study was included in both the sponsor-submitted NMA and Menges et al. The ENZAMET study was included in both published NMAs. The PEACE-1 trial was only included in the NMA by Yanagisawa et al. and the HORRAD and NCT02058706 trials were only included in Menges et al. The NCT02058706 study was included in the sponsor’s SLR but not included in the NMA based on the treatments included (enzalutamide and ADT versus bicalutamide and ADT). The HORRAD trial evaluated radiotherapy and ADT versus ADT alone, and therefore was not of interest to the other NMAs. Other reasons for differences in included studies across NMAs were unknown.

Both published NMAs conducted a risk-of-bias assessment using the same tool from Cochrane (Figure 19 and Figure 2). As previously noted, similar studies were included across NMAs. Although the authors of both papers concluded that the included studies in their NMAs were at low risk of bias, the results for the individual domains varied. Menges et al. considered there to be a low risk of bias in all domains for all studies. Conversely, Yanagisawa et al. noted that there was a high risk of bias in blinding of participants and outcome assessment (performance and detection biases). No risk of assessment was conducted for the sponsor-submitted NMA, which included similar studies; therefore, the risk of bias remains uncertain and open to interpretation.

The sponsor-submitted NMA reported on outcomes of OS, time to CRPC, and rPFS. Conversely, comparative efficacy results reported for darolutamide plus ADT and docetaxel were only available for OS from the published NMAs. The sponsor-submitted NMA used a |||||||||||||||||||||||||||| framework. Results of the sponsor-submitted NMA for the end point of OS reported that darolutamide plus ADT and docetaxel was favoured over ADT and docetaxel, while comparisons to enzalutamide, apalutamide, and abiraterone combinations were inconclusive due to wide CIs. For time to CRPC and rPFS, darolutamide plus ADT and docetaxel was generally favoured over all other treatment combinations. In contrast to the sponsor-submitted NMA, the 2 published NMAs used a frequentist, random-effects framework for their analyses. For the end point of OS, results were consistent for the NMA conducted by Menges et al. with darolutamide plus ADT and docetaxel favoured only over ADT and docetaxel; however, compared to enzalutamide, apalutamide, and abiraterone combinations, the evidence was inconclusive due to wide CIs. For Yanagisawa et al., comparative results for OS of darolutamide plus ADT and docetaxel were limited to ADT and docetaxel, and abiraterone and ADT, in which darolutamide plus ADT and docetaxel was favoured over both combinations.

A common critical appraisal among the NMAs was the choice of model fit, in which the sponsor-submitted NMA conducted |||||||||||||| analyses and the published NMAs conducted random-effects NMAs. Only the sponsor-submitted NMA reported model-fit statistics, while published NMAs did not; thus, it was unclear if the proper model was selected for these analyses, although given the potential for heterogeneity, the models chosen were likely appropriate. There were generally some differences noted across populations in each NMA regarding baseline disease characteristics that may have affected the results, including prostate cancer stage, PSA level, disease volume, and prior treatment; however, these were not explored. Overall, none of the NMAs considered there to be significant heterogeneity across studies; however, given the wide 95% CrIs and 95% CIs, it was believed that there was unaccounted-for heterogeneity within the analyses, in addition to the limited number of studies per comparison and the small sample sizes.

Other Relevant Evidence

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

Discussion

Summary of Available Evidence

This report summarizes the evidence for darolutamide in combination with docetaxel and ADT in the treatment of mCSPC based on 1 phase III RCT and 3 NMAs.

One trial was included in the CADTH systematic review. The ARASENS trial was a phase III, double-blind RCT that aimed to demonstrate superiority of darolutamide plus docetaxel and ADT compared to placebo plus docetaxel and ADT in patients with mCSPC (N = 1,306) in the first-line, metastatic setting. The primary end point was OS; secondary end points included time to CRPC, time to initiation of subsequent systemic antineoplastic therapy, time to pain progression, time to first SSE, and SSE-FS. In the final efficacy analysis (data cut-off on October 25, 2021), the median age was 67.0 (range = 41 to 89) years. The majority of patients were white (52.0%) or Asian (36.4%) and had stage IV disease at initial diagnosis (87.6%), ECOG performance status of 0 (71.1%), and bone metastases (82.8%) at baseline. Most patients did not receive prior local therapy and no patients had prior systemic antineoplastic therapy for prostate cancer other than ADTs.

In total, 3 NMAs were eligible for inclusion in this review. One was commissioned and submitted by the sponsor and evaluated the comparative efficacy of darolutamide plus ADT and docetaxel compared to enzalutamide and ADT, apalutamide and ADT, abiraterone plus prednisone and ADT, and docetaxel and ADT on OS, time to CRPC, and rPFS in adult patients with mHSPC. Two were published NMAs that evaluated comparative OS for darolutamide plus ADT and docetaxel compared to other available treatment regimens in patients with mHSPC.

Interpretation of Results

Efficacy

In the pivotal, phase III ARASENS trial, the primary end point of OS was met in the final efficacy analysis, supporting the superiority of darolutamide plus docetaxel and ADT compared to placebo plus docetaxel and ADT in patients with mCSPC. OS was an outcome noted by patients and clinicians to be highly important. The clinical experts consulted by CADTH considered the magnitude of OS benefit of darolutamide plus docetaxel and ADT combination to be clinically meaningful. It should be noted that the distribution of subsequent antineoplastic therapy used in the darolutamide plus docetaxel and ADT arm did not align with clinical practice as per the clinical experts, which may introduce uncertainties into the generalizability of OS results. It is unclear whether a different distribution of subsequent treatments that is more reflective of Canadian clinical practice would have resulted in different effect estimates. Subgroup analyses by baseline ECOG performance status, Gleason score, and extent of metastatic disease were performed; however, no conclusion can be drawn due to the exploratory nature, lack of control for multiplicity, and the risk of confounding in the absence of randomization in the comparisons.

In addition to improving OS, delaying disease progression to CRPC and the need for a subsequent line of therapy are important goals of treatment for mCSPC because mCRPC is associated with morbidity, poor HRQoL, and poor prognosis, according to the clinical experts. The magnitude of benefits observed for secondary end points of time to CRPC and time to initiation of subsequent systemic antineoplastic therapy was considered clinically meaningful by the clinical experts. It should be noted that the median time to CPRC in the placebo group was lower than anticipated by the clinical experts based on their clinical experience; as such, it is uncertain if the observed magnitude of benefit of darolutamide plus docetaxel and ADT will translate in clinical practice.

Results of the time to pain progression, time to first SSE, and SSE-FS analyses were in favour of the darolutamide plus docetaxel and ADT arm, supporting the efficacy of this combination in delaying pain progression and SSE, which are events noted to be burdensome by patients. Of note, unblinding occurred by error in some patients during the study (||||||| in the darolutamide plus docetaxel and ADT arm, and ||||||| in the control arm), which could potentially introduce reporting bias in favour of darolutamide (i.e., toward an inflated efficacy of darolutamide plus docetaxel and ADT) in these subjective efficacy outcomes, although the impact is likely to be small given that the number of unblinded patients was small.

HRQoL, measured using the NCCN-FACT FPSI-17 questionnaire, was an exploratory outcome of interest to patients and clinicians. The mean total and subscale scores were stable at most time points and trending toward deterioration by the end of treatment in both treatment arms. There was with no notable difference in the mean change in score from baseline between treatment arms, in general, based on the sponsor-identified MID estimate of 3 points; however, no statistical testing was conducted for the difference between treatment arms, which precludes conclusion on the effect of darolutamide on HRQoL outcomes when used in combination with docetaxel and ADT. Further, this outcome is subject to a high risk of bias resulting from unblinding by error and a large amount of missing data. The instrument is also not routinely administered in clinical practice and has not been validated in patients with mCSPC, and most items in the questionnaire are measuring symptoms, although the sponsor has used the tool to assess HRQoL. Overall, no definitive conclusion can be drawn about the effect of the darolutamide plus docetaxel and ADT combination on HRQoL based on available data.

Per the clinical experts, the biochemical factor PSA is assessed along with radiologic and clinical factors to inform treatment-response evaluations in patients with mCSPC. Results of the exploratory PSA analyses suggested that darolutamide plus docetaxel and ADT was associated with a delay in time to PSA progression and numerically higher PSA response rates compared to docetaxel and ADT; however, these results were not controlled for multiplicity and, therefore, were subject to an increased risk of false-positive findings.

Although the study results were generally in favour of darolutamide, it should be noted that there was a high number of important protocol deviations in the study, many of which pertain to missed study assessments, incorrect or incomplete assessment procedure, and incorrect dosing or method of drug delivery. These deviations could potentially compromise the completeness and reliability of study data, although none of the important protocol deviations was considered major, as per the sponsor. Further, the generalizability of the study results is limited by the exclusion of patients with significant comorbidities (e.g., cardiovascular diseases) and impaired performance status commonly seen in clinical practice and the fact that the comparator regimen, docetaxel and ADT, only accounts for a small proportion of systemic treatments prescribed for mCSPC. In the absence of direct evidence, the comparative efficacy between darolutamide plus docetaxel and ADT versus ARAT plus ADT, which is the most commonly prescribed regimen in Canada, according to the clinical experts, is unknown and represents a gap in evidence.

As part of the development of a ||||||| cost-effectiveness model for darolutamide in mCSPC, the sponsor conducted and submitted an NMA that was used to inform these analyses. The sponsor-submitted NMA aimed to indirectly compare darolutamide plus ADT and docetaxel from the ARASENS trial to alternative treatments for patients with mCSPC. Across NMAs, results for all end points were associated with wide 95% CrIs and 95% CIs, which failed to demonstrate differences across comparisons and resulted in uncertainty and loss of precision in estimates, precluding comparative efficacy conclusions for darolutamide plus ADT and docetaxel. Additionally, all comparisons of darolutamide plus ADT and docetaxel were informed by indirect evidence, resulting in greater uncertainty. The sponsor-submitted NMA reported that darolutamide plus ADT and docetaxel was favoured over ADT and docetaxel for OS benefit. Two published NMAs were also included, which also reported that darolutamide plus ADT and docetaxel was favoured over ADT and docetaxel despite differences in included studies and analysis methods. However, the sponsor-submitted NMA and the Menges et al. and Yanagisawa et al. NMAs demonstrated wide CIs that precluded conclusions for comparisons between darolutamide plus ADT and docetaxel and other androgen receptor signalling inhibitors, including abiraterone, apalutamide, or enzalutamide. Results from the ARASENS trial demonstrated improved survival over ADT and docetaxel; thus, the reviewed indirect evidence did not provide any additional comparative effectiveness insight regarding OS. One published NMA also demonstrated improved survival with darolutamide plus ADT and docetaxel over abiraterone and ADT. Time to CRPC and rPFS were also evaluated in the sponsor-submitted NMA, for which darolutamide plus ADT and docetaxel was favoured over all treatments except enzalutamide and ADT for the outcome of time to CRPC.

Harms

Overall, the safety profile of darolutamide plus docetaxel and ADT arm was similar to that of the placebo plus docetaxel and ADT arm in the ARASENS trial, with a similar incidence of TEAEs, treatment discontinuations due to TEAE, and deaths due to AE between treatment arms. Almost all patients experienced at least 1 TEAE in both treatment arms, with the most frequently reported TEAEs (in at least 25% of patients in either arm) being alopecia, fatigue, anemia, arthralgia, peripheral edema, decreased neutrophil count, and diarrhea, all of which are common AEs of docetaxel. Although decreased appetite (any grade) and hypertension of grade 3 or higher were more common AEs in the darolutamide plus docetaxel and ADT arm, they are known AEs of ARATs and are manageable in clinical practice, according to the clinical experts. Based on the safety findings in the ARASENS trial, it is unlikely that darolutamide plus docetaxel and ADT can fulfill the need for new treatments with reduced toxicity burden expressed by patient and clinician groups. In the NMA by Yanagisawa et al., darolutamide plus ADT and docetaxel was only compared to abiraterone and ADT for the AE of febrile neutropenia. Although darolutamide plus ADT and docetaxel was associated with greater odds of experiencing febrile neutropenia, the results were highly uncertain, given the wide 95% CIs.

Conclusions

In the ARASENS trial, darolutamide in combination with docetaxel and ADT demonstrated a clinically meaningful improvement of OS compared to docetaxel plus ADT in patients with mCSPC. Analyses of secondary outcomes supported the efficacy of darolutamide plus docetaxel and ADT in delaying progression to mCRPC, the need for subsequent antineoplastic therapy, worsening of pain, and SSEs compared to docetaxel plus ADT. Due to limitations in the statistical analysis, no definitive conclusions can be drawn from the exploratory PSA and HRQoL analyses. The exclusion of patients with significant comorbidities and impaired performance status and the use of a comparator regimen that has low utilization in Canada may limit the generalizability of study findings. Across the included NMAs, the findings for the comparative survival benefit observed with darolutamide plus ADT and docetaxel compared to ADT and docetaxel supported the findings of the ARASENS trial; however, the findings of these NMAs are associated with uncertainty due to probable violation of the underlying transitivity assumption. Comparisons to other drugs of interest were uncertain due to the aforementioned methodological limitations as well as wide CIs. The safety profile of darolutamide plus docetaxel and ADT was overall similar to that of docetaxel and ADT in the ARASENS trial, with no additional serious safety concern.

References

1.Prostate Cancer Statistics. Canadian Cancer Society; 2022: https://cancer.ca/en/cancer-information/cancer-types/prostate/statistics. Accessed 2022 Aug 16.

2.Gandaglia G, Abdollah F, Schiffmann J, et al. Distribution of metastatic sites in patients with prostate cancer: A population-based analysis. Prostate. 2014;74(2):210-216. PubMed

3.Overview of systemic treatment for advanced, recurrent and metastatic castration-sensitive prostate cancer and local treatment for patients with metastatic disease. UpToDate; 2022: https://www.uptodate.com/contents/overview-of-systemic-treatment-for-advanced-recurrent-and-metastatic-castration-sensitive-prostate-cancer-and-local-treatment-for-patients-with-metastatic-disease. Accessed 2022 Jul 04.

4.Fizazi K, Foulon S, Carles J, et al. Abiraterone plus prednisone added to androgen deprivation therapy and docetaxel in de novo metastatic castration-sensitive prostate cancer (PEACE-1): a multicentre, open-label, randomised, phase 3 study with a 2 x 2 factorial design. Lancet. 2022;399(10336):1695-1707. PubMed

5.Nubeqa (darolutamide): 300mg, oral tablets [product monograph]. Mississauga (ON): Bayer Inc; 2022 May 12.

6.Smith MR, Hussain M, Saad F, et al. Darolutamide and Survival in Metastatic, Hormone-Sensitive Prostate Cancer. N Engl J Med. 2022;386(12):1132-1142. PubMed

7.Clinical Study Report: ARASENS. A randomized, double–blind, placebo–controlled Phase III study of darolutamide (ODM–201) versus placebo in addition to standard androgen deprivation therapy and docetaxel in patients with metastatic hormone–sensitive prostate cance [internal sponsor's report]. Whippany (NJ): Bayer Inc; 2022 Feb 15.

8.Darolutamide for metastatic castration-sensitive prostate cancer. Network-meta analysis for public payer submissions in Canada. In: Drug Reimbursement Review sponsor submission: Nubeqa, 300 mg, Oral tablet. Mississauga (ON): Bayer Inc.; 2022 May 12.

9.Menges D, Yebyo HG, Sivec-Muniz S, et al. Treatments for Metastatic Hormone-sensitive Prostate Cancer: Systematic Review, Network Meta-analysis, and Benefit-harm assessment. Eur Urol Oncol. 2022;19:19. PubMed

10.Yanagisawa T, Rajwa P, Thibault C, et al. Androgen Receptor Signaling Inhibitors in Addition to Docetaxel with Androgen Deprivation Therapy for Metastatic Hormone-sensitive Prostate Cancer: A Systematic Review and Meta-analysis. Eur Urol. 2022;19:19. PubMed

11.Sterne JAC, Savović J, Page MJ, et al. RoB 2: a revised tool for assessing risk of bias in randomised trials. BMJ. 2019;366:l4898. PubMed

12.Canadian Cancer Statistics: A 2018 Special Report on Cancer Incidence by Stage. Canadian Cancer Society; 2018: https://cdn.cancer.ca/-/media/files/research/cancer-statistics/2018-statistics/canadian-cancer-statistics-2018-en.pdf. Accessed 2022 Aug 16.

13.Diagnosis of Prostate Cancer. Canadian Cancer Society; 2021: https://cancer.ca/en/cancer-information/cancer-types/prostate/diagnosis. Accessed 2022 Aug 16.

14.So AI, Chi KN, Danielson B, et al. Canadian Urological Association-Canadian Urologic Oncology Group guideline on metastatic castration-naive and castration-sensitive prostate cancer. Can Urol Assoc J. 2020;14(2):17-23. PubMed

15.NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines®) Prostate Cancer Version 4.0. National Comprehensive Cancer Network; 2022: https://www.nccn.org/professionals/physician_gls/pdf/prostate.pdf. Accessed 2022 Aug 16.

16.Xtandi (enzalutamide): 40mg, oral capsules [product monograph]. 2022: https://pdf.hres.ca/dpd_pm/00064474.PDF. Accessed 2022 Jul 30.

17.Erleada (apalutamide): 60mg, oral tablet [product monograph]. 2022: https://pdf.hres.ca/dpd_pm/00066114.PDF. Accessed 2022 Jul 30.

18.Zytiga (abiraterone acetate): 250mg uncoated oral tablet, 500mg film-coated oral tablet [product monograph]. 2018: https://pdf.hres.ca/dpd_pm/00044223.PDF. Accessed 2022 Jul 30.

19.McGowan J, Sampson M, Salzwedel DM, Cogo E, Foerster V, Lefebvre C. PRESS Peer Review of Electronic Search Strategies: 2015 Guideline Statement. J Clin Epidemiol. 2016;75:40-46. PubMed

20.Grey matters: a practical tool for searching health-related grey literature. Ottawa (ON): CADTH; 2019: https://www.cadth.ca/grey-matters. Accessed 2022 Jul 4.

21.Cleeland C, Ryan K. Pain assessment: global use of the Brief Pain Inventory. Annals, academy of medicine, Singapore. 1994. PubMed

22.Logothetis CJ, Basch E, Molina A, et al. Effect of abiraterone acetate and prednisone compared with placebo and prednisone on pain control and skeletal-related events in patients with metastatic castration-resistant prostate cancer: exploratory analysis of data from the COU-AA-301 randomised trial. Lancet Oncol. 2012;13(12):1210-1217. PubMed

23.Kuppen MCP, Westgeest HM, van den Eertwegh AJM, et al. Health-related Quality of Life and Pain in a Real-world Castration-resistant Prostate Cancer Population: Results From the PRO-CAPRI Study in the Netherlands. Clin Genitourin Cancer. 2020;18(3):e233-e253. PubMed

24.Victorson DE, Beaumont JL, Rosenbloom SK, Shevrin D, Cella D. Efficient assessment of the most important symptoms in advanced prostate cancer: the NCCN/FACT‐P Symptom Index. Psycho‐Oncology. 2011;20(9):977-983. PubMed

25.Beaumont JL, Butt Z, Li R, Cella D. Meaningful differences and validity for the NCCN/FACT‐P Symptom Index: An analysis of the ALSYMPCA data. Cancer. 2019;125(11):1877-1885. PubMed

26.Statistical Analysis Plan: ARASENS. A randomized, double–blind, placebo–controlled Phase III study of darolutamide (ODM–201) versus placebo in addition to standard androgen deprivation therapy and docetaxel in patients with metastatic hormone–sensitive prostate cance [internal sponsor's report]. 2021 Nov 11.

27.Guyot P, Ades AE, Ouwens MJNM, Welton NJ. Enhanced secondary analysis of survival data: reconstructing the data from published Kaplan-Meier survival curves. BMC Med Res Methodol. 2012;12(1):9. PubMed

28.Brooks SP, Gelman A. General Methods for Monitoring Convergence of Iterative Simulations. Journal of Computational and Graphical Statistics. 1998;7(4):434-455.

29.van Valkenhoef G, Dias S, Ades AE, Welton NJ. Automated generation of node-splitting models for assessment of inconsistency in network meta-analysis. Research Synthesis Methods. 2016;7(1):80-93. PubMed

30.Dias S WN, Sutton AJ, et al. NICE DSU Technical Support Document 2: A Generalised Linear Modelling Framework for Pairwise and Network Meta-Analysis of Randomised Controlled Trials [Internet]. London: National Institute for Health and Care Excellence (NICE). 2014: https://www.ncbi.nlm.nih.gov/books/NBK310366/.

31.Brignardello-Petersen R, Izcovich A, Rochwerg B, et al. GRADE approach to drawing conclusions from a network meta-analysis using a partially contextualised framework. BMJ. 2020;371:m3907. PubMed

32.Langan D, Higgins JPT, Jackson D, et al. A comparison of heterogeneity variance estimators in simulated random-effects meta-analyses. Res Synth Methods. 2019;10(1):83-98. PubMed

33.Robinson DW, Zhao N, Dawkins F, Qi M, Revicki D. Pain questionnaire performance in advanced prostate cancer: comparative results from two international clinical trials. Qual Life Res. 2013;22(10):2777-2786. PubMed

34.Gater A, Abetz-Webb L, Battersby C, et al. Pain in castration-resistant prostate cancer with bone metastases: a qualitative study. Health and quality of life outcomes. 2011;9(1):1-11. PubMed

35.Skaltsa K, Longworth L, Ivanescu C, Phung D, Holmstrom S. Mapping the FACT-P to the preference-based EQ-5D questionnaire in metastatic castration-resistant prostate cancer. Value Health. 2014;17(2):238-244. PubMed

36.Cleeland CS. The brief pain inventory user guide. Houston, TX: The University of Texas MD Anderson Cancer Center. 2009:1-11.

37.Fizazi K, De Bono J, Flechon A, et al. Randomised phase II study of siltuximab (CNTO 328), an anti-IL-6 monoclonal antibody, in combination with mitoxantrone/prednisone versus mitoxantrone/prednisone alone in metastatic castration-resistant prostate cancer. Eur J Cancer. 2012;48(1):85-93. PubMed

38.Chu FM, Picus J, Fracasso PM, Dreicer R, Lang Z, Foster B. A phase 1, multicenter, open-label study of the safety of two dose levels of a human monoclonal antibody to human αv integrins, intetumumab, in combination with docetaxel and prednisone in patients with castrate-resistant metastatic prostate cancer. Invest New Drugs. 2011;29(4):674-679. PubMed

39.Clark MJ, Harris N, Griebsch I, Kaschinski D, Copley-Merriman C. Patient-reported outcome labeling claims and measurement approach for metastatic castration-resistant prostate cancer treatments in the United States and European Union. Health Qual Life Outcomes. 2014;12:104. PubMed

40.FACIT. National Comprehensive Cancer Network/Functional Assessment of Cancer Therapy Prostate Cancer Symptom Index - 17 Item Version. 2021: https://www.facit.org/measures/NFPSI-17. Accessed July 27 2022.

41.Hoskin P, Sartor O, O'Sullivan JM, et al. Efficacy and safety of radium-223 dichloride in patients with castration-resistant prostate cancer and symptomatic bone metastases, with or without previous docetaxel use: a prespecified subgroup analysis from the randomised, double-blind, phase 3 ALSYMPCA trial. The Lancet Oncology. 2014;15(12):1397-1406. PubMed

Appendix 1: Literature Search Strategy

Note that this appendix has not been copy-edited.

Clinical Literature Search

Overview

Interface: Ovid

Databases:

• MEDLINE All (1946 to present)

• Embase (1974 to present)

• Note: Subject headings and search fields have been customized for each database. Duplicates between databases were removed in Ovid.

Date of search: July 13, 2022

Alerts: Weekly search updates until project completion

Search filters applied: No filters were applied to limit the retrieval by study type.

Limits:

• Publication date limit: none

• Language limit: none

• Conference abstracts: excluded

Table 27: Syntax Guide

Multi-Database Strategy

1. (darolutamide* or nubeqa* or darramamide* or bay-1841788 or bay1841788 or odm-201 or odm201 or orm-16497 or orm16497 or orm-16555 or orm16555 or X05U0N2RCO).ti,ab,kf,ot,hw,nm,rn.

2. 1 use medall

3. *darolutamide/ or (darolutamide* or nubeqa* or darramamide* or bay-1841788 or bay1841788 or odm-201 or odm201 or orm-16497 or orm16497 or orm-16555 or orm16555).ti,ab,kf,dq.

4. 3 use oemezd

5. 4 not (conference review or conference abstract).pt.

6. 2 or 5

7. remove duplicates from 6

Clinical Trials Registries

ClinicalTrials.gov

Produced by the US National Library of Medicine. Targeted search used to capture registered clinical trials.

Search: Nubeqa (darolutamide); metastatic castration sensitive prostate cancer (mCSPC)

WHO ICTRP

International Clinical Trials Registry Platform, produced by WHO. Targeted search used to capture registered clinical trials.

Search: Nubeqa (darolutamide); metastatic castration sensitive prostate cancer (mCSPC)

Health Canada’s Clinical Trials Database

Produced by Health Canada. Targeted search used to capture registered clinical trials.

Search: Nubeqa (darolutamide); metastatic castration sensitive prostate cancer (mCSPC)

EU Clinical Trials Register

European Union Clinical Trials Register, produced by the European Union. Targeted search used to capture registered clinical trials.

Search: Nubeqa (darolutamide); metastatic castration sensitive prostate cancer (mCSPC)

Grey Literature

Search dates: July 4, 2022 to July 13, 2022

Keywords: Nubeqa (darolutamide); metastatic castration sensitive prostate cancer (mCSPC)

Limits: Publication years: None

Updated: Search updated prior to the completion of stakeholder feedback period

Relevant websites from the following sections of the CADTH grey literature checklist Grey Matters: A Practical Tool for Searching Health-Related Grey Literature were searched:

• Health Technology Assessment Agencies

• Health Economics

• Clinical Practice Guidelines

• Drug and Device Regulatory Approvals

• Advisories and Warnings

• Drug Class Reviews

• Clinical Trials Registries

• Databases (free)

• Internet Search

• Open Access Journals

The complete search archive of sites consulted for this report is available on request.

Appendix 2: Detailed Outcome Data

Note that this appendix has not been copy-edited.

Figure 11: NCCN-FACT FPSI-17 Questionnaire — Total Score (Redacted)

Note that this figure has been redacted.

Figure 12: NCCN-FACT FPSI-17 Questionnaire — Disease-Related Symptom Subscale — Physical Score (Redacted)

Note that this figure has been redacted.

Figure 13: NCCN-FACT FPSI-17 Questionnaire — Disease-Related Symptoms Subscale — Emotional Score (Redacted)

Note that this figure has been redacted.

Figure 14: NCCN-FACT FPSI-17 Questionnaire — Treatment Side Effects Subscale Score (Redacted)

Note that this figure has been redacted.

Figure 15: NCCN-FACT FPSI-17 Questionnaire — Function and Well-Being Subscale Score (Redacted)

Note that this figure has been redacted.

Figure 16: Assessment of Inconsistency Based on Node-Splitting for OS (Redacted)

Note that this figure has been redacted.

Figure 17: Assessment of Inconsistency Based on Node-Splitting for Time to CRPC (Redacted)

Note that this figure has been redacted.

Figure 18: Assessment of Inconsistency Based on Node-Splitting for rPFS (Redacted)

Note that this figure has been redacted.

Figure 19: Risk of Bias in Studies Regarding Overall Survival (Menges et al. [2022])

Source: Menges et al. (2022).⁹

Figure 20: Risk of Bias Assessment of the Included RCTs (Yanagisawa et al. [2022])

Source: Yanagisawa et al. (2022).¹⁰

Appendix 3: Description and Appraisal of Outcome Measures

Note that this appendix has not been copy-edited.

Aim

To describe the following outcome measures and review their measurement properties (validity, reliability, responsiveness to change, and MID):

• Brief Pain Inventory - Short Form (BPI-SF)

• National Comprehensive Cancer Network/ Functional Assessment of Cancer Therapy-Prostate Symptom Index - 17 (NCCN-FACT FPSI-17)

Findings

Table 28: Summary of Outcome Measures and Their Measurement Properties

ALP = alkaline phosphatase; BPI-SF = Brief Pain Inventory - Short Form; DRS-P = disease-related symptoms–physical subscale; DRS-E = disease-related symptoms–emotional subscale; ECOG = Eastern Cooperative Oncology Group F/WB = Function/Well-Being subscale; HRQoL = health-related quality of life; mCRPC = metastatic castration-resistant prostate cancer; MID = minimal important difference; PSA = prostate-specific antigen; PPI = Present Pain Intensity (PPI) component of the McGill Pain Questionnaire; TSE: Treatment Side Effects subscale.

Brief Pain Inventory — Short Form (BPI-SF)

Description and Scoring

The BPI is a patient-reported pain questionnaire designed and validated to assess cancer pain,³⁶ and was used in the ARASENS trial to measure time to pain progression. It is used to assess the intensity of pain experienced, as well as the degree to which this pain interferes with function.²¹ The BPI is available as a long and short version, the latter of which has a 24-hour recall period for both worst pain and least pain items.³³ It consists of a diagram of a human body onto which the location of pain is recorded.³⁶ There is a section for reporting use of analgesics and the relief these provide. Pain measurement is divided into 2 categories: intensity and interference with function. The intensity category consists of 4 items: pain now, average pain, worst pain, and least pain. The interference with function category is divided into 7 items: general activity, mood, walking ability, normal work, relations with other persons, sleep, and enjoyment of life. Each item is scored on an 11-point scale from 0 to 10, where 0 is no pain or no interference and 10 is the worst pain or complete interference. With respect to the pain interference category, a mean can be reported if a minimum of 4 of the 7 items are assessed. There is no total score for the BPI-SF. A composite of the 4 pain items (a mean severity score) can be presented and pain interference is typically scored as the mean of the 7 interference items.³⁶

Validity and Reliability

Robinson et al. examined data from 2 trials evaluating siltuximab³⁷ (n = 93) or intetumumab in combination with docetaxel³⁸ (n = 131) among patients with mCRPC.³³ Results demonstrated good construct validity with strong Pearson correlations between the BPI-SF worst pain and average pain items (r = 0.79; P <0.006) as well as between the BPI-SF worst pain item and the Present Pain Intensity (PPI) component of the McGill Pain Questionnaire (r = 0.52; P < 0.006).³³ A moderate correlation was found between the BPI-SF worst pain item and the FACT-P total score (r = −0.42; P < 0.006).³³

Gater et al. conducted in-depth interviews with 17 patients (aged 53 to 86 years) with mCRPC and bone metastases to assess the content validity of the BPI-SF average pain and worst pain items.³⁴ Results indicated that patients understood the worst pain item but had some difficulty in determining their average pain. A total of 11 patients based their average pain on the value between their worst pain and least pain while 6 patients based their average pain value on the level of pain that they experienced most of the time. Recall periods for average pain varied between participants (e.g., since beginning treatment or in the last week). Therefore, evidence supports the content validity of the worst pain item as a measure of pain intensity for use in this patient population and notes limitations in patients’ interpretation of the average pain item.^34,39

Robinson et al. also examined the reliability of the BPI-SF pain intensity scale (4 items).³³ Internal consistency reliability was evaluated using Cronbach’s alpha which was reported as good to excellent, with values of 0.89 and 0.93 for each trial, respectively. Test-retest reliability was evaluated using intraclass correlation coefficient (ICC) among a cohort of patients in each trial who completed the questionnaire at 2 initial visits and were clinically stable between visits (with a mean (SD) of 12 (7) days between visits). Results found good internal consistency reliability for the BPI-SF pain intensity scale with ICC values of 0.73 and 0.91 for each trial, respectively.³³

Responsiveness

Responsiveness of the instrument was not assessed in the indicated patient population.

MID

An MID was not assessed in the indicated patient population. An MID estimate of 2 or more points or 30% change in pain intensity items from baseline was previously used in studies in mCRPC patients.^22,23

National Comprehensive Cancer Network/Functional Assessment of Cancer Therapy-Prostate Symptom Index 17 (NCCN-FACT FPSI-17).

Description and Scoring

The NCCN-FACT FPSI-17 is a cancer-specific questionnaire developed to examine the symptoms and function/well-being among people undergoing treatment for advanced prostate cancer,²⁴ and was used in the ARASENS trial to approximate HRQoL. Developed in consultation with medical experts (n = 66) and patients (n = 74) in 2010, the instrument includes a 17-item index that produces a total score and is also divided into 4 main subscales that assess disease-related symptoms: Disease-Related Symptoms–Physical (DRS-P), Disease-Related Symptoms–Emotional (DRS-E), Treatment Side Effects (TSE), and Function/Well-Being (F/WB). Each item is ranked on a 5-point scale ranging from 0 (not at all) to 4 (very much).⁴⁰ Higher subscale and total scores indicate lesser symptoms and better function/well-being.²⁵

Validity and Reliability

Beaumont et al., used data from the ALSYMPCA trial⁴¹ among mCRPC patients to assess the construct validity of the instrument using the known groups approach.²⁵ All baseline sources significantly differed among groups based on ECOG performance status, except for the DRS-E subscale. Also, all baseline scores, except for the NFPSI–DRS-E, significantly differed between total ALP scores (< 220 U/L versus ≥ 220 U/L), providing support for construct validity of the instrument in this patient population.⁴¹

Cronbach’s alpha was used to estimate the internal consistency reliability of the instrument, with good reliability for the total score (alpha = 0.80) as well as the F/WB (alpha = 0.78) and DRS-P (alpha = 0.67) subscales.²⁵ The TSE subscale had poor internal consistency reliability (alpha = 0.47) likely due to the diverse set of common AEs that comprise the subscale.²⁵

Responsiveness

The instrument’s total score and all subscale scores were shown to be responsive to change in ECOG performance status in the aforementioned study.²⁵ Effect sizes for those who worsened were ≥ 0.50 for the instrument’s total score and all subscale scores except for the DRS-E subscale which had a small effects size of 0.12. The instrument was also responsive to the presence of a skeletal-related event during treatment the with effect sizes ≥ 0.54 for all scales except for the DRS-E subscale. The instrument was responsive to patients who did not experience a PSA response before the end of treatment with effect sizes ranging from 0.36 to 0.47 for all scales except for the DRS-E. Effect sizes for PSA responders improved by a similar magnitude for all scales except for TSE subscale.²⁵

MID

Both distribution- and anchor-based methods were used to determine an MID for the instrument based on trial data among mCRPC patients.²⁵ Results indicated an MID of 4 to 6 points for the total score, 2 to 3.5 points for the DRS-P, 0.5 points for the DRS-E, 1 to 1.5 points for TSE, and 0.5 to 1 point for the F/WB subscales.²⁵

Pharmacoeconomic Review

Executive Summary

The executive summary comprises 2 tables (Table 1 and Table 2) and a conclusion.

Table 1: Submitted for Review

mCSPC = metastatic castration-sensitive prostate cancer; NOC = Notice of Compliance.

Table 2: Summary of Economic Evaluation

ADT = androgen deprivation therapy; AE = adverse event; ARAT = androgen receptor-axis-targeted therapy; ICER = incremental cost-effectiveness ratio; LY = life-year; mCSPC = metastatic castration-sensitive prostate cancer; NMA = network meta-analysis; OS = overall survival; QALY = quality-adjusted life-year; RCT = randomized controlled trial.

Conclusions

The CADTH Clinical Review concluded that darolutamide plus docetaxel and androgen deprivation therapy (ADT) demonstrated clinically meaningful improvement of overall survival (OS) compared to docetaxel and ADT in patients with metastatic castration-sensitive prostate cancer (mCSPC) in the ARASENS trial. Analyses of secondary outcomes supported the efficacy of darolutamide plus docetaxel and ADT in delaying progression to metastatic castration-resistant prostate cancer (mCRPC), the need for subsequent antineoplastic therapy, worsening of pain, and symptomatic skeletal events compared to docetaxel and ADT. Regarding the sponsor-submitted indirect treatment comparison (ITC), the CADTH Clinical Review noted that, for OS, there was insufficient evidence to show a difference between darolutamide plus docetaxel and ADT and either abiraterone and ADT, apalutamide and ADT, and enzalutamide and ADT. For time to mCRPC, darolutamide plus docetaxel and ADT was favoured over all treatments except enzalutamide and ADT. The reviewers noted the potential for biases related to heterogeneity in study design, prior treatment requirements, and follow-up duration across included studies. The clinical review noted that there remains increased potential that the underlying transitivity assumption of the network meta-analysis (NMA) has been violated.

CADTH undertook reanalyses to the sponsor’s economic evaluation; these included using a gamma extrapolation for the OS results and updating costs for docetaxel.

The results of the CADTH reanalysis were in line with those submitted by the sponsor: darolutamide plus docetaxel and ADT is not a cost-effective treatment for mCSPC at a willingness-to-pay threshold of $50,000 per quality-adjusted life-year (QALY). Based on the CADTH sequential analysis, darolutamide plus docetaxel and ADT is more costly and more effective than abiraterone and ADT (incremental costs: $121,237; incremental QALYs: 0.77), resulting in an ICER of $156,172 per QALY. A price reduction of 58% would be required for darolutamide in order for the treatment regimen (i.e., darolutamide plus docetaxel and ADT) to be cost-effective compared to abiraterone and ADT at a $50,000-per-QALY threshold. The sponsor’s OS extrapolations were uncertain and resulted in overestimates of survival according to clinical experts, who indicated that a Weibull extrapolation might also be appropriate. CADTH conducted a scenario analysis in which a Weibull extrapolation was used for OS, which resulted in an incremental cost-effectiveness ratio (ICER) of $180,113 per QALY compared to abiraterone and ADT.

Results of the CADTH base case and scenario analyses are uncertain given the results of the NMA, which found no statistical difference in OS for darolutamide plus docetaxel and ADT compared to abiraterone, apalutamide, or enzalutamide and ADT. The results were imprecise, and the 95% credible intervals were wide. This imprecision was evident in the sponsor’s base-case model, which predicted incremental QALYs ranging from –1.4 to 3.3 for the comparison of darolutamide plus docetaxel and ADT to abiraterone and ADT across different probabilistic iterations. The CADTH base case predicts an incremental LY gain of 0.80 for darolutamide plus docetaxel and ADT compared to abiraterone and ADT, despite the lack of strong conclusions that can be drawn from the results of the NMA. Furthermore, in the CADTH base case, 72% of the incremental QALYs gained for darolutamide plus docetaxel and ADT compared to abiraterone and ADT were obtained after the maximum follow-up in the ARASENS trial. The results of the model are dependent on a survival benefit for darolutamide plus docetaxel and ADT being realized relative to comparators, which is uncertain given the aforementioned limitations. The scenario analysis in which no differences in OS were assumed between darolutamide plus docetaxel and ADT and comparators resulted in an ICER of $520,548 per QALY compared to abiraterone and ADT and a price reduction of 75% for darolutamide. This scenario highlights the impact of OS assumptions on the ICER and resulting price reduction required to achieve cost-effectiveness at a $50,000-per-QALY threshold.

Stakeholder Input Relevant to the Economic Review

This section is a summary of the feedback received from the patient groups, registered clinicians, and drug plans that participated in the CADTH review process.

Patient input was received from 2 groups: the Canadian Cancer Society and the Canadian Cancer Survivor Network. The Canadian Cancer Society conducted surveys and interviews of 39 patients and 2 caregivers. The Canadian Cancer Survivor Network conducted online surveys with 24 patients and 2 caregivers. All survey respondents lived in Canada. Patients reported previous experience with hormone therapy, radiation, antiandrogens, surgery, and chemotherapy. The most common side effects of currently available therapy that had a significant impact on daily life were sexual dysfunctions, hot flashes, and fatigue. Patient expectations for a new therapy were for it to result in improved quality of life (reduced side effects) and delayed symptom onset. Six of the patients surveyed had current experience with darolutamide, and 1 was interviewed in detail. Side effects with darolutamide included sexual dysfunction, fatigue, weight gain, and loss of muscle mass, among others. Most patients with experience on darolutamide reported improved outcomes in terms of prostate-specific antigen and quality of life, and stated that the benefits of the treatment outweigh the side effects.

CADTH received 7 registered clinician input submissions for this review, which consisted of clinician groups from Southeastern Ontario; BC Cancer; the Canadian Cancer Society; Ontario Health; the Maritime provinces; Regina, Saskatchewan; and The Ottawa Hospital. The baseline therapy for all patients with mCSPC is ADT, with clinical trial evidence showing a benefit for the addition of docetaxel, or an androgen receptor-axis-targeted therapy (ARAT), namely, abiraterone, enzalutamide, and apalutamide. Although the standard care a decade ago was ADT alone, clinicians currently treat patients with ADT in addition to docetaxel, or in addition to an ARAT, with some clinicians advocating for the use of “triplet therapy” (ARAT plus docetaxel and ADT). Darolutamide would fall into this category of triplet therapy in that it would be given, first-line, in combination with docetaxel and ADT. The goals of treatment are to improve and extend disease control for as long as possible, improve and extend quality of life, and extend OS.

Drug plan input for this review was interested in whether patients unable to tolerate 6 cycles of docetaxel should be eligible to continue with darolutamide and ADT, or whether patients unable to tolerate darolutamide would be eligible to switch another ARAT plus docetaxel and ADT. Plans wondered whether patients who recently initiated docetaxel and ADT for mCSPC would be eligible to add on darolutamide, or whether patients currently receiving an ARAT and ADT would switch to darolutamide plus docetaxel and ADT at the time of funding. Plans noted the presence of confidential, negotiated prices for comparators.

Several of these concerns were addressed in the sponsor’s model:

• The sponsor included various comparators in their model, such as abiraterone, apalutamide, docetaxel, and enzalutamide in combination with ADT, along with ADT alone.

• CADTH was unable to address the following concerns raised from stakeholder input:

• Sexual dysfunction, noted to be 1 of the more common side effects of therapy, was not included as an adverse event in the sponsor’s model.

• CADTH reanalyses are based on publicly available prices and do not incorporate the presence of confidential, negotiated prices.

Economic Review

The current review is for darolutamide (Nubeqa) in combination with docetaxel for patients with mCSPC.

Economic Evaluation

Summary of Sponsor’s Economic Evaluation

Overview

The sponsor submitted a cost-utility analysis of darolutamide in combination with docetaxel compared to docetaxel, abiraterone and prednisone, apalutamide, enzalutamide, or ADT alone. All other treatments were also given alongside ADT. The modelled population consisted of patients with mCSPC eligible for chemotherapy, which is aligned with the eligibility criteria for the ARASENS clinical trial.¹ The Health Canada indication states that darolutamide is indicated for “the treatment of patients with mCSPC in combination with docetaxel” without the explicit stipulation that patients must be eligible for chemotherapy;² the reimbursement request represents patients with mCSPC who are chemotherapy-eligible, which is aligned with the clinical trial.^1,3

Darolutamide is available in 300 mg oral tablets. The recommended dose is 600 mg twice daily, equivalent to a total daily dose of 1,200 mg.³ The dose may be reduced to 300 mg twice daily in the case of serious adverse events (AEs).³ Darolutamide is given in combination with docetaxel and ADT. Docetaxel has a recommended dose of 75 mg/m² every 3 weeks for a maximum of 6 cycles and is given in combination with dexamethasone. ADT is included as a basket treatment received by all patients and consists of the following drugs, weighted by market share: degarelix, 12.6%; leuprorelin, 54.0%; goserelin, 31.9%; triptorelin, 1.5%. Market shares were obtained from IQVIA public claims data.^3,4 The cost for darolutamide is $28.34 per tablet, leading to a 28-day cost of $3,175.³

The comparators for this analysis have been previously stated. The 28-day drug costs for the comparators used in the sponsor’s model are as follows: docetaxel, $97; abiraterone and prednisone, $872; apalutamide, $3,175; and enzalutamide, $3,270.¹ All patients receive ADT; the 28-day costs for ADT are $417 in the first cycle and $356 in the second cycle due to degarelix requiring a higher loading dose.³

The clinical outcomes of interest were QALYs and life-years over a lifetime horizon of 25 years. Discounting (1.5% per annum) was applied to both costs and outcomes and a cycle length of 4 weeks was used. The base-case perspective was that of the Canadian publicly funded health care payer.

Model Structure

The sponsor submitted a partitioned survival model consisting of 3 mutually exclusive health states, mCSPC (i.e., progression-free), mCRPC (i.e., postprogression), and death. All patients entered the model in the progression-free state and received darolutamide plus docetaxel and ADT or a comparator. The proportion of patients in the progression-free state was defined by the progression-free survival (PFS) curves for the darolutamide plus docetaxel and ADT, and docetaxel and ADT treatments received in the ARASENS trial.¹ Since the ARASENS trial did not capture radiographic PFS directly, progression was modelled using the secondary end point, time to castration resistance or death (TTCROD). TTCROD is a composite end point defined by any prostate-specific antigen progression, radiographic progression by soft-tissue or visceral lesions or radiographic progression by bone lesions, or death. The proportion of patients in the postprogression health state was equal to the difference between the OS and the PFS (defined by TTCROD) curves from the ARASENS trial. For comparators, the effect of treatment on OS and PFS was expressed in the model in terms of hazard ratios (HRs) based on an ITC. Patients in the postprogression state can receive subsequent line therapy. Patients transitioning into the death state remained there until the end of the model time horizon. A figure of the sponsor’s model structure is available in Appendix 3 (Figure 1).

Model Inputs

The target population of the economic evaluation was based on the ARASENS clinical trial, a phase III, double-blind, randomized controlled trial comparing darolutamide in combination with docetaxel and ADT to docetaxel and ADT.¹ The mean age of the population was 66.8 years, mean weight was 77.5 kg, mean body surface area was 1.8 m², and 100% of patients were male.¹

For the comparison of darolutamide plus docetaxel and ADT and docetaxel and ADT, PFS, OS, and time on treatment (ToT) Kaplan-Meier data were derived directly from the ARASENS trial. Parametric extrapolations were performed to extrapolate the Kaplan-Meier data out to the model time horizon of 25 years. Curves were fitted independently for PFS (based on TTCROD), OS, and ToT, though the same extrapolations were chosen for both treatments. The sponsor’s base case extrapolations for PFS, OS, and ToT were the log-normal, log-logistic, and Gompertz extrapolations, respectively.

In the absence of direct comparative evidence to other comparators not included in the ARASENS trial, comparative effectiveness data were derived from a sponsor-commissioned ITC.⁵ |||||||||| studies, including the ARASENS trial, were identified as being relevant to the decision problem, and were included in an NMA.⁵ The NMA was used to generate HRs, which were applied to the parametric extrapolations for darolutamide plus docetaxel and ADT to generate comparator-specific survival curves. The OS HRs for the comparison of darolutamide plus docetaxel and ADT to other comparators ranged from |||||||||| (95% CI, |||||||||||||||||||| for ADT alone to |||||||||| 95% CI, |||||||||||||||||||| for both apalutamide and ADT and abiraterone and ADT (95% CI, |||||||||||||||||||| , favouring darolutamide plus docetaxel and ADT. The results were not significant (i.e., confidence interval includes the null) for abiraterone and ADT, apalutamide and ADT, and enzalutamide and ADT. The sponsor presented HRs for PFS according to TTCROD (base case) and radiographic PFS (scenario analysis); the HRs for the base case ranged from |||||||||| (95% CI, |||||||||||||||||||| for ADT alone to |||||||||| (95% CI, |||||||||||||||||||| for enzalutamide and ADT. ToT was assumed not to exceed PFS.

Regarding safety outcomes, the model considered grade 3+ AEs that occurred for at least 5% of patients for any of the treatments; these consisted of increased alanine aminotransferase, decreased neutrophil count, decreased white blood cell count, febrile neutropenia, hypertension, hypokalemia, and neutropenia.

The ARASENS trial collected health-related quality of life data from patients through various questionnaires but not via a generic, preference-based HRQoL measure such as the EQ-5D. Thus, EQ-5D utility data were collected from various sources in the literature. For the preprogression health state, a utility value of 0.847 was used based on the enzalutamide clinical trial ARCHES.⁶ The utility values for the first line of postprogression treatment in the mCRPC health state (0.782), and end-of-life (0.543) health states were obtained from the ARCHES trial using EQ-5D-5L measurements that applied Canadian tariffs.⁷ Finally, a value of 0.663 was used for second- and third-line mCRPC health states based on the mean of these 2 values. Disutilities, derived from various published sources, were included for all AEs except increased alanine aminotransferase and hypokalemia.³

The economic model included costs related to drugs (acquisition, administration, background ADT, and subsequent therapy), monitoring, AE management, and terminal care. The drug acquisition costs of darolutamide and comparators have been previously stated, as well as those for ADT. Costs for subsequent treatments were considered (e.g., abiraterone, enzalutamide, docetaxel, radium-223, cabazitaxel), with subsequent treatment breakdown obtained from patients in the ARASENS trial.⁸ However, these data were adjusted to account for the fact that some jurisdictions do not allow ARAT use in subsequent lines post-ARAT therapy for mCSPC. As such, enzalutamide was assumed not to be used after darolutamide and abiraterone use was reduced by 43% as its use post-ARAT therapy is only available in Ontario, Prince Edward Island, and New Brunswick. The subsequent treatment distribution for darolutamide plus docetaxel and ADT was assumed to apply to the other ARAT treatments, while the distribution for the docetaxel and ADT arm was assumed to apply to ADT alone.³ In addition to acquisition costs, an administration cost of $300 was included per administration of infusion therapies (e.g., docetaxel).⁹ Vial sharing was assumed to occur in the base case for docetaxel. Costs of AEs were applied as a one-off cost based on a weighted average of all AEs experienced in a given treatment. Regarding medical resource use, the following direct medical costs have been included in the model: cost of outpatient treatment (e.g., visits to urologist and/or oncologist, laboratory examinations, and emergency treatment); cost of drug therapies and concomitant medications, if applicable; administration costs; monitoring costs; hospitalization costs; and all follow-up treatment costs and costs for nursing care, with costs obtained from various published sources.^3,10,11 Lastly, a one-off terminal care cost of $6,626 was applied when a patient dies, to reflect the costs of terminal care in the last 3 months of life.¹²

Summary of Sponsor’s Economic Evaluation Results

All analyses were run probabilistically (5,000 iterations for the base-case and scenario analyses). The deterministic and probabilistic results were similar. The probabilistic findings are presented below.

Base-Case Results

In the sponsor’s base case, darolutamide plus docetaxel and ADT was associated with an estimated cost of $219,925 and 5.78 QALYs over a lifetime horizon. In a sequential analysis, darolutamide plus docetaxel and ADT was associated with an ICER of $113,782 compared to abiraterone and ADT (incremental cost: $118,972, incremental QALYs: 1.05). Both apalutamide and enzalutamide were extendedly dominated in the sponsor’s base case, resulting in higher ICERs than the next most costly treatment. In the sponsor’s sequential analysis, darolutamide plus docetaxel and ADT had a 1% probability of being cost-effective at a willingness-to-pay threshold of $50,000 per QALY. In the sponsor’s base case, 79% of the incremental QALYs gained for darolutamide plus docetaxel and ADT compared to abiraterone and ADT were obtained after the maximum follow-up in the ARASENS trial (approximately 5 years). Additional results from the sponsor’s base case are available in Appendix 3.

Table 3: Summary of the Sponsor’s Economic Evaluation Results

ADT = androgen deprivation therapy; ICER = incremental cost-effectiveness ratio; LY = life-year; QALY = quality-adjusted life-year.

Note: The submitted analyses are based on the publicly available prices of comparators and may not reflect confidential, negotiated prices. Only treatments on the cost-effectiveness frontier are reported in this table; apalutamide + ADT and enzalutamide + ADT were extendedly dominated by abiraterone + ADT and darolutamide + docetaxel + ADT.

Source: Sponsor’s pharmacoeconomic submission.³

Sensitivity and Scenario Analysis Results

The sponsor conducted several scenario analyses involving alternative time horizons, discount rates, subsequent treatment distribution based directly on the ARASENS trial, half-cycle correction, optimistic/pessimistic survival extrapolations, utilities, drug wastage, and PFS HRs based on radiographic PFS. The scenarios involving a discount rate of 5%, and shorter survival extrapolation resulted in the largest changes in the ICERs for darolutamide plus docetaxel and ADT compared to abiraterone, about $144,000 per QALY in both. Results from other scenarios had minimal impact on the overall results.

CADTH Appraisal of the Sponsor’s Economic Evaluation

CADTH identified several key limitations to the sponsor’s analysis that have notable implications on the economic analysis:

• OS extrapolations are overestimated compared to general Canadian estimates. The sponsor’s base case used a log-logistic extrapolation for OS, which resulted in unrealistic OS estimates for patients receiving darolutamide plus docetaxel and ADT. The log-logistic extrapolation predicted that after 30 years, 5% of patients will still be alive, and after 40 years, 3% of patients will still be alive, with an average age of 107 years. For patients receiving docetaxel, the log-logistic extrapolation predicted that after 40 years, 2% of patients would still be alive. This is much higher than estimates of life expectancy in the general population in Canada, in which less than 1% of males are expected to still be alive at age 102.¹³ To address this issue, the sponsor adjusted the modelled OS to match the maximum age and sex-matched Canadian general population-based mortality hazard. This adjustment implies that the sponsor’s chosen extrapolation is inaccurate after a certain point — in the sponsor’s model, this occurs after 18.3 years. The adjustment increases the uncertainty in the analysis because it arbitrarily assumes a higher mortality for some patients that is not reflective of the chosen extrapolation, suggesting that the chosen extrapolation is flawed and unrepresentative of the data it is intended to model. Moreover, the adjustment also suggests that after 18.3 years, darolutamide plus docetaxel and ADT has “cured” the disease, such that patients with mCSPC experience the same mortality hazard as the general population. Finally, 5% of patients on darolutamide plus docetaxel and ADT and 3% of patients on docetaxel and ADT were still alive after 25 years (with an average age of 92), which was deemed unreasonable, according to clinical experts.

  CADTH solicited clinician feedback to help determine the most appropriate OS curves. The feedback indicated that the sponsor’s chosen extrapolations were overestimated and emphasized that after 20 years, few patients would remain alive. Clinicians indicated that the gamma and Weibull extrapolations were more reflective of clinical expectations for this disease area. Clinicians agreed that the gamma extrapolation was most representative of clinical expectations.

  • As part of the base case, in consultation with clinical experts, CADTH selected a gamma extrapolation for both darolutamide plus docetaxel and ADT and docetaxel and ADT. These extrapolations resulted in less than 1% of patients remaining alive after 25 years. The Weibull extrapolation was tested in a scenario analysis.

• There is uncertainty regarding clinical effects from the ITC. To inform the comparisons between darolutamide plus docetaxel and ADT and the other comparators abiraterone and ADT, apalutamide and ADT, and enzalutamide and ADT, the sponsor performed an ITC and NMA.¹ As noted in the CADTH Clinical Review, results of this analysis indicated that there was insufficient evidence to show a difference between darolutamide plus docetaxel and ADT and these comparators for OS. The results were imprecise, as seen in the 95% credible intervals being wide. This imprecision was evident in the sponsor’s base-case model, which predicted incremental QALYs ranging from –1.4 to 3.3 for the comparison of darolutamide plus docetaxel and ADT to abiraterone and ADT across different probabilistic iterations. The clinical review also highlighted the lack of information regarding data extraction and risk of bias assessment, which results in a potential risk of bias and/or error in the systematic literature review that could not be quantified. Regarding heterogeneity, the clinical review stated that, while the heterogeneity assessment was reasonable and important factors were considered, there remained notable differences in the baseline characteristics that could potentially result in changing relative treatment effects that were not accounted for, including prostate-specific antigen level and prostate cancer stage. There were also differences in blinding, prior treatment requirements, and follow-up duration across included studies, and there remains increased potential that the underlying transitivity assumption of the NMA has been violated.

  • CADTH was unable to address this limitation due to a lack of evidence. As part of a scenario analysis, CADTH assumed no difference in OS between darolutamide plus docetaxel and ADT and abiraterone and ADT, apalutamide and ADT, and enzalutamide and ADT. This scenario analysis may be considered conservative.

• Docetaxel costs are underestimated in the sponsor’s analysis. In their economic analysis, the sponsor used a cost per mg of docetaxel of $0.54 based on a published economic evaluation.¹⁴ However, the original pricing source could not be located and the cost of $0.54 per mg could not be verified. Because this value is substantially different than pricing data obtained from the IQVIA DeltaPA database,¹⁵ CADTH updated the cost of docetaxel based on the IQVIA database.

  • As part of the base case, CADTH updated the cost of docetaxel according to the unit costs summarized in the CADTH cost table (Table 8).

• Subsequent therapy in the ARASENS trial does not align with expected clinical practice. In the sponsor’s model, costs of treatments for additional lines of therapy were added as a lump sum upon patients leaving the initial mCSPC health state. To estimate the costs of subsequent therapy, the proportions of subsequent treatment use postprogression in the ARASENS trial were used.⁸ In this trial, patients received various therapies following darolutamide plus docetaxel and ADT, including abiraterone, enzalutamide, docetaxel, radium-223, and cabazitaxel. The sponsor performed an adjustment to the subsequent therapy proportions whereby enzalutamide use was not assumed and abiraterone use was reduced by 43% based on the assumption that only patients in Ontario, Prince Edward Island, and New Brunswick could receive this therapy subsequent to an ARAT (darolutamide). These assumptions align partially with clinician feedback, which indicated that subsequent ARAT use in patients already having progressed on an ARAT is uncommon in Canada and unlikely to occur. However, in the ARASENS trial, patients did receive the ARATs enzalutamide (15.2%) and abiraterone (35.6%) post-darolutamide, which may have resulted in additional benefits to PFS and/or OS that were not considered or adjusted for. By excluding the costs of enzalutamide and 43% of the costs of abiraterone, the sponsor has artificially reduced the costs associated with the darolutamide plus docetaxel and ADT treatment while still maintaining any residual effectiveness. This potentially biases the results in favour of darolutamide plus docetaxel and ADT compared to docetaxel and ADT because patients are able to receive an ARAT post-docetaxel therapy.

  • The sponsor’s model provided an option to use the proportions of subsequent therapy observed directly in the ARASENS trial without an adjustment. This option was explored in a scenario analysis.

• All relevant AEs were not included. The sponsor’s model included grade 3 or 4 AEs observed in the ARASENS trial, consisting of increased alanine aminotransferase, decreased neutrophil count, decreased white blood cell count, febrile neutropenia, hypertension, hypokalemia, and neutropenia.⁸ However, the patient input noted several other side effects of currently available therapies that have significant impact on daily life, including sexual dysfunctions, hot flashes, and fatigue. In addition, the clinical expert noted that enzalutamide, in particular, may be associated with neurocognitive toxicities such as dysgeusia, decreased concentration focus, and exacerbation of pre-existing depression or anxiety, all of which can affect quality of life in this population. These AEs were not included in the sponsor’s model, and their costs and quality-of-life impacts were not considered.

  • CADTH was unable to address this limitation due to a lack of data on the prevalence of AEs.

• Treatment waning was not considered. The sponsor used HRs derived from the NMA and applied them to OS and PFS data from the ARASENS trial for the entirety of the model. This implicitly assumes that treatment effects observed in the NMA will be maintained throughout the model, up to 25 years, which is unreasonable given that the median follow-up of included studies ranged from 3.7 to 7.0 years. Without long-term data to substantiate this assumption, the survival benefits of darolutamide plus docetaxel and ADT are likely overestimated relative to comparators.

  • CADTH was unable to address this limitation due to a lack of long-term data. In scenario analyses, CADTH tested the impact of applying a full treatment waning effect (i.e., no further additional OS or PFS benefits) for darolutamide after 5 years.

Additional limitations were identified but were not considered to be key limitations. There were some small discrepancies between the costs for apalutamide and enzalutamide used by the sponsor and those listed in the Ontario Exceptional Access Program Formulary. However, the discrepancies were slight and did not affect the conclusions of the economic evaluation or the comparators on the efficiency frontier. Furthermore, dexamethasone is required to be given alongside docetaxel; the cost of dexamethasone was not included by the sponsor.

Additionally, the following key assumptions were made by the sponsor and have been appraised by CADTH (refer to Table 4).

CADTH Reanalyses of the Economic Evaluation

Base-Case Results

CADTH undertook reanalyses that addressed limitations within the model, as summarized in Table 5. The CADTH base case was derived by making changes in model parameters and assumptions, in consultation with clinical experts. All CADTH probabilistic reanalyses were based on 5,000 iterations.

CADTH undertook a stepped analysis, incorporating each changed proposed in Table 5 to the sponsor’s base case to highlight the impact of each change. A summary is presented in Table 6, while the full stepwise analysis is presented in Appendix 4 (Table 11). Disaggregated results of the CADTH base case are also presented in Appendix 4. In the CADTH base case, 72% of the incremental QALYs gained for darolutamide + docetaxel and ADT compared to abiraterone and ADT were obtained after the maximum follow-up in the ARASENS trial (approximately 5 years).

Scenario Analysis Results

CADTH undertook price-reduction analyses based on the sponsor’s and CADTH’s base case. The CADTH base case suggested that a price reduction of 58% would be required to achieve cost-effectiveness of darolutamide + docetaxel and ADT relative to abiraterone and ADT at a $50,000-per-QALY threshold (Table 7).

Table 4: Key Assumptions of the Submitted Economic Evaluation (Not Noted as Limitations to the Submission)

AE = adverse event; mCRPC = metastatic castration-resistant prostate cancer; PFS = progression-free survival; PSA = prostate-specific antigen; QALY = quality-adjusted life-year; rPFS = radiographic progression-free survival; TTCROD = time to castration resistance or death.

Table 5: CADTH Revisions to the Submitted Economic Evaluation

OS = overall survival.

Table 6: Summary of the CADTH Reanalysis Results

ADT = androgen deprivation therapy; ICER = incremental cost-effectiveness ratio; QALY = quality-adjusted life-year.

Note: Only products on the efficiency frontier are presented in this table. In the sponsor’s and CADTH’s base case, enzalutamide + ADT and apalutamide + ADT were dominated. In the CADTH base case only, docetaxel + ADT was also dominated. Full results are available in Appendix 4.

Table 7: CADTH Price Reduction Analyses

ADT = androgen deprivation therapy; ICER = incremental cost-effectiveness ratio.

CADTH undertook several scenario analyses to determine the impact of alternative assumptions on the cost-effectiveness of darolutamide plus docetaxel and ADT in the base case, which are outlined as follows:

• Weibull extrapolations of the OS data from the ARASENS trial were used for darolutamide plus docetaxel and ADT, and docetaxel and ADT.

• Proportions of subsequent therapy observed in the ARASENS trial were used directly without an adjustment.

• In the sponsor-submitted NMA, PFS was defined by radiographic progression-free survival rather than time to CRPC.

• In alignment with the results of the NMA, CADTH assumed no difference in OS between darolutamide plus docetaxel and ADT and abiraterone and ADT, apalutamide and ADT, and enzalutamide and ADT.

• Full treatment waning was considered after 5 years.

The scenario analysis involving a Weibull extrapolation resulted in an ICER of $180,113 per QALY, indicating that results are sensitive to the choice of OS extrapolation. The scenario involving treatment waning at 5 years resulted in an ICER of $269,855 for darolutamide plus docetaxel and ADT compared to abiraterone and ADT, with an associated price reduction of 70% to ensure cost-effectiveness at a $50,000-per-QALY threshold. The scenario in which no OS benefit was assumed resulted in an ICER of $520,548 compared to abiraterone and ADT and a price reduction of 75% for darolutamide. Other scenarios tested did not meaningfully affect the results.

Issues for Consideration

Abiraterone plus docetaxel and ADT, while not a Health Canada–approved comparator, may become a relevant triplet therapy comparator soon, according to clinical experts. This comparator was not included in the sponsor-submitted NMA and, as such, the cost-effectiveness of darolutamide plus docetaxel and ADT compared to abiraterone plus docetaxel and ADT is unknown.

Overall Conclusions

The CADTH Clinical Review concluded that darolutamide plus docetaxel and ADT demonstrated clinically meaningful improvement of OS compared to docetaxel and ADT in patients with mCSPC in the ARASENS trial. Analyses of secondary outcomes supported the efficacy of darolutamide plus docetaxel and ADT in delaying progression to mCRPC, the need for subsequent antineoplastic therapy, worsening of pain, and symptomatic skeletal events compared to docetaxel and ADT. Regarding the sponsor-submitted ITC, the CADTH Clinical Review noted that, for OS, there was insufficient evidence to show a difference between darolutamide plus docetaxel and ADT and either abiraterone and ADT, apalutamide and ADT, or enzalutamide and ADT. For time to CRPC, darolutamide plus docetaxel and ADT was favoured over all treatments except enzalutamide and ADT. The reviewers noted the potential for biases related to heterogeneity in study design, prior treatment requirements, and follow-up duration across included studies. The CADTH Clinical Review noted that there remains increased potential that the underlying transitivity assumption of the NMA has been violated.

CADTH undertook reanalyses to the sponsor’s economic evaluation, which included using a gamma extrapolation for the OS results and updating costs for docetaxel.

The results of the CADTH reanalysis were in line with those submitted by the sponsor: darolutamide plus docetaxel and ADT is not a cost-effective treatment for mCSPC at a willingness-to-pay threshold of $50,000 per QALY. Based on the CADTH sequential analysis, darolutamide plus docetaxel and ADT is more costly and more effective than abiraterone and ADT (incremental costs $121,237; incremental QALYs: 0.77), resulting in an ICER of $156,172 per QALY. A price reduction of 58% would be required for darolutamide for the treatment regimen (i.e., darolutamide plus docetaxel and ADT) to be cost-effective compared to abiraterone and ADT at a $50,000-per-QALY threshold. The sponsor’s OS extrapolations were uncertain and resulted in overestimates of survival, according to clinical experts, who indicated that a Weibull extrapolation might also be appropriate. CADTH conducted a scenario analysis in which a Weibull extrapolation was used for OS; this resulted in an ICER of $180,113 per QALY compared to abiraterone and ADT.

Results of the CADTH base case and scenario analyses are uncertain given the results of the NMA, which found no statistical difference in OS for darolutamide plus docetaxel and ADT compared to abiraterone, apalutamide, or enzalutamide and ADT. The results were imprecise, and the 95% credible intervals were wide. This imprecision was evident in the sponsor’s base-case model, which predicted incremental QALYs ranging from –1.4 to 3.3 for the comparison of darolutamide plus docetaxel and ADT to abiraterone and ADT across different probabilistic iterations. The CADTH base case predicts an incremental life-years gain of 0.80 for darolutamide plus docetaxel and ADT compared to abiraterone and ADT, despite the lack of strong conclusions that can be drawn from the results of the NMA. Furthermore, in the CADTH base case, 72% of the incremental QALYs gained for darolutamide plus docetaxel and ADT compared to abiraterone and ADT were obtained after the maximum follow-up in the ARASENS trial. The results of the model are dependent on a survival benefit for darolutamide plus docetaxel and ADT being realized relative to comparators, which is uncertain given the aforementioned limitations. The scenario analysis in which no differences in OS were assumed between darolutamide plus docetaxel and ADT and comparators resulted in an ICER of $520,548 per QALY compared to abiraterone and ADT and a price reduction of 75% for darolutamide. This scenario highlights the impact of OS assumptions on the ICER and the resulting price reduction required to achieve cost-effectiveness at a $50,000-per-QALY threshold.

References

1.Smith M, Hussain M, Saad F, et al. Darolutamide and survival in metastatic, hormone-sensitive prostate cancer. N Engl J Med. 2022;386(12):1132-1142. PubMed

2.Nubeqa (darolutamide): 300mg, oral tablets [product monograph]. Mississauga (ON): Bayer Inc; 2022 May 12.

3.Pharmacoeconomic evaluation [internal sponsor's report]. In: Drug Reimbursement Review sponsor submission: Nubeqa (darolutamide), 300 mg tablets for the treatment of patients with metastatic castration-sensitive prostate cancer (mCSPC). Mississauga (ON): Bayer Inc; 2022 Jun 15.

4.PharmaStat. Ottawa (ON): IQVIA; 2022: https://www.iqvia.com/. Accessed 2022 Sep 25.

5.Darolutamide for metastatic castration-sensitive prostate cancer. Network-meta analysis for public payer submissions in Canada [internal sponsor's report]. In: Drug Reimbursement Review sponsor submission: Nubeqa (darolutamide), 300 mg tablets for the treatment of patients with metastatic castration-sensitive prostate cancer (mCSPC). Mississauga (ON): Bayer Inc; 2022 Jun 15.

6.Armstrong A, Szmulewitz R, Petrylak D, al. e. Arches: a randomized, phase III study of androgen deprivation therapy with enzalutamide or placebo in men with metastatic hormone-sensitive prostate cancer. J Clin Oncol. 2019;37(32):2974-2986. PubMed

7.Saad F, Chilelli A, Hui B, al. e. Cost-effectiveness of enzalutamide versus apalutamide versus androgen deprivation therapy alone for the treatment of metastatic castration-sensitive prostate cancer in Canada. J Med Econ. 2022;25(1):583-590. PubMed

8.Clinical Study Report: ARASENS. A randomized, double–blind, placebo–controlled Phase III study of darolutamide (ODM–201) versus placebo in addition to standard androgen deprivation therapy and docetaxel in patients with metastatic hormone–sensitive prostate cance [internal sponsor's report]. Whippany (NJ): Bayer Inc; 2022 Feb 15.

9.Ng R, Hasan B, Mittmann N, al. e. Economic analysis of NCIC CTG JBR.10: a randomized trial of adjuvant vinorelbine plus cisplatin compared with observation in early stage non-small-cell lung cancer--a report of the Working Group on Economic Analysis, and the Lung Disease Site Group, National Cancer Institute of Canada Clinical Trials Group. J Clin Oncol. 2007;25(16):2256-2261. PubMed

10.Schedule of benefits for physician services under the Health Insurance Act: effective October 1, 2021. Toronto (ON): Ontario Ministry of Health; 2022: https://www.health.gov.on.ca/en/pro/programs/ohip/sob/physserv/sob_master.pdf. Accessed 2022 Aug 26.

11.Schedule of benefits for laboratory services: effective July 1, 2020. Toronto (ON): Ontario Ministry of Health; 2020: https://www.health.gov.on.ca/en/pro/programs/ohip/sob/lab/lab_mn2020.pdf. Accessed 2022 Aug 26.

12.de Oliveira C, Pataky R, Bremner K, al. e. Phase-specific and lifetime costs of cancer care in Ontario, Canada. BMC Cancer. 2016;16(1):809. PubMed

13.Table: 13-10-0114-01: Life expectancy and other elements of the life table, Canada, all provinces except Prince Edward Island. Ottawa (ON): Statistics Canada; 2022: https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=1310011401. Accessed 2022 Aug 26.

14.Beca J, Majeed H, Chan K, Hotte S, Loblaw A, Hoch J. Cost-effectiveness of docetaxel in high-volume hormone-sensitive metastatic prostate cancer. Can Urol Assoc J. 2019;13(12):396-403. PubMed

15.DeltaPA. Ottawa (ON): IQVIA; 2022: https://www.iqvia.com/. Accessed 2022 Sep 6.

16.Docetaxel: 10 mg/ mL as 20 mg/2 mL, 80 mg/8mL, 160 mg/16mL for injection [product monograph]. Boucherville (QC): Sandoz Canada Inc; 2020 Nov 17: https://pdf.hres.ca/dpd_pm/00058887.PDF. Accessed 2022 Sep 6.

17.Apo-abiraterone (abiraterone acetate): 250 mg and 500 mg, film-coated tablets [product monograph]. Toronto (ON): Apotex Inc; 2021 Apr 1: https://pdf.hres.ca/dpd_pm/00060444.PDF. Accessed 2022 Sep 6.

18.Xtandi (enzalutamide): 40 mg capsules [product monograph]. Markham (ON): Astellas Pharma Canada, Inc; 2022 Jan 24: https://pdf.hres.ca/dpd_pm/00064474.PDF. Accessed 2022 Sep 9.

19.Erleada (apalutamide): 60 mg oral tablets [product monograph]. Toronto (ON): Janssen Inc; 2022 May 31: https://pdf.hres.ca/dpd_pm/00066114.PDF. Accessed 2022 Sep 9.

20.Firmagon (degarelix): 120 mg, 80 mg per vial for injection [product monograph]. North York (ON): Ferring Phamaceuticals; 2016 Mar 18: https://pdf.hres.ca/dpd_pm/00034229.PDF. Accessed 2022 Sep 6.

21.Zoladex (goserelin depot): 3.6 mg [product monograph]. Mississauga (ON): AstraZeneca Canada Inc; 2017 Dec 21: https://pdf.hres.ca/dpd_pm/00042728.PDF.

22.Lupron (leuprolide acetate): 5 mg/mL sterile solution subcutaneous injection; Lupron Depot (leuprolide acetate for depot suspension): 3.75 mg/syringe (1-Month slow release), 7.5 mg/syringe (1-Month slow release), 11.25 mg/syringe (3-Month slow release), 22.5 mg/syringe (3-Month slow release), 30 mg/syringe (4-Month slow release), pre-filled dual-chamber syringe containing sterile lyophilized microspheres intramuscular injection [product monograph]. St-Laurent (QC): AbbVie Corporation; 2021 Nov 18: https://pdf.hres.ca/dpd_pm/00063626.PDF. Accessed 2022 Sep 6.

23.Trelstar (triptorelin): 3.75 mg per vial (1 month sustained-release formulation), 11.25 mg per vial (3 month sustained-release formulation), 22.5 mg per vial (6 month sustained-release formulation), intramuscular injection [product monograph]. Montreal (QC): Knight Therapeutics Inc; 2022 Sep 8: https://pdf.hres.ca/dpd_pm/00067340.PDF. Accessed 2022 Sep 6.

24.Nova Scotia Drug Formulary. Halifax (NS): Nova Scotia Department of Health; 2022. Accessed 2022 Sep 25.

25.Budget Impact Analysis [internal sponsor's report]. In: Drug Reimbursement Review sponsor submission: Nubeqa (darolutamide), 300 mg tablets for the treatment of patients with metastatic castration-sensitive prostate cancer (mCSPC). Mississauga (ON): Bayer Inc; 2022 Jun 15.

26.Statistics Canada. Population projections for Canada, provinces and territories. 2015; https://www150.statcan.gc.ca/n1/pub/91-520-x/2010001/t147-eng.pdf. Accessed 2022 Sep 25.

27.Robles C, A M. Female population. Ottawa (ON): Statistics Canada; 2011: https://www150.statcan.gc.ca/n1/pub/89-503-x/2010001/article/11475-eng.htm. Accessed 2022 Sep 25.

28.Canadian Cancer Society. Prostate cancer statistics. 2022; https://cancer.ca/en/cancer-information/cancer-types/prostate/statistics. Accessed 2022 Sep 25.

29.Canadian Cancer Statistics Advisory Committee. Canadian cancer statistics. Toronto (ON): Canadian Cancer Society; 2021: https://cdn.cancer.ca/-/media/files/research/cancer-statistics/2021-statistics/2021-pdf-en-final.pdf. Accessed 2022 Sep 25.

30.Francini E, Gray KP, Xie W, et al. Time of metastatic disease presentation and volume of disease are prognostic for metastatic hormone sensitive prostate cancer (mHSPC). Prostate. 2018;78(12):889-895. PubMed

Appendix 1: Cost Comparison Table

Note that this appendix has not been copy-edited.

The comparators presented in the following table have been deemed to be appropriate based on feedback from clinical experts. Comparators may be recommended (appropriate) practice or actual practice. Existing Product Listing Agreements are not reflected in the table and, as such, the table may not represent the actual costs to public drug plans.

Table 8: CADTH Cost Comparison Table for Darolutamide Plus Docetaxel and ADT for the Treatment of Metastatic Castration-Sensitive Prostate Cancer (mCSPC)