Which Patients Are Eligible for Coverage?:

Qulipta should only be reimbursed for adults with episodic migraine, according to the criteria used by public drug plans for other calcitonin gene-related peptide (CGRP) inhibitors for the prevention of episodic migraine.

What Are the Conditions for Reimbursement?:

In addition to following the pre-existing criteria for other CGRP inhibitors, Qulipta should not be used in combination with other CGRP inhibitors. Qulipta should only be reimbursed if the cost is reduced such that the total treatment cost of Qulipta does not exceed the total treatment cost of the least costly CGRP inhibitor reimbursed for the preventive treatment of episodic migraine in adults.

Why Did CADTH Make This Recommendation?

• Evidence from 3 clinical trials demonstrated that Qulipta reduced the mean number of days patients experienced migraines and headaches per month compared with placebo. The trials showed that Qulipta reduced migraine symptoms.
• Qulipta may meet some needs that are important to patients, as it is another treatment option that results in fewer migraine days, has manageable side effects, and is taken by mouth rather than injection.
• Based on CADTH’s assessment of the health economic evidence, Qulipta does not represent good value to the health care system at the public list price, and there is not enough robust evidence to justify a greater cost for Qulipta compared with relevant anti-CGRP comparators reimbursed for the prevention of episodic migraine.
• Based on public list prices, Qulipta is estimated to cost the public drug plans approximately $1.1 million over the next 3 years, if reimbursed according to the criteria used by public drug plans for other CGRP inhibitors for episodic migraine.

What Is Migraine?:

Migraine is a neurologic disease characterized by recurrent episodes of pulsating headache pain that can be accompanied by sensitivity to light or sound, nausea, vomiting, numbness, and auras. It affects 1 in 10 people in Canada, and women are affected more than others.

Unmet Needs in Migraine:

Many patients have trouble finding effective treatments that reduce migraine frequency and need to try several medications before realizing benefit. Furthermore, conventional migraine prevention treatments are associated with unwanted side effects.

How Much Does Qulipta Cost?:

Treatment with Qulipta is expected to cost approximately $6,735 per patient per year.

Input From the Clinical Expert Consulted by CADTH

The clinical expert consulted by CADTH emphasized that currently available treatments for patients with migraine have several issues, notably that not all patients respond to current treatments, and that — in the case of monoclonal antibodies (mAbs) for migraine prevention — patients become refractory to treatment, requiring a change of treatment. Migraine attacks are treated with abortive agents as well as preventive strategies. mAbs (fremanezumab or galcanezumab) are used following the failure of prior prophylactic migraine prevention therapy such as antidepressants, antihypertensives, or anticonvulsants. The clinical expert noted that atogepant is not a cure, but rather reduces symptomatic events of EM.

The expert noted that patients with higher migraine headache frequency and major functional disability are more likely to receive atogepant, though it likely would not be considered the first choice for most patients and would likely be considered in patients who have not responded to or are intolerant of anti-CGRP mAbs. The clinical expert noted that identifying patients who would have better response to atogepant was unlikely, though patients least likely to benefit from atogepant are those with a history of poor compliance. Aside from intolerability due to side effects or poor compliance, the clinical expert stated that failure to reach a 50% reduction in MMDs without any improvements in HRQoL would be the main reason to discontinue treatment with atogepant.

The clinical expert noted that clinicians concentrate on what can be quickly quantified and understood from similar metrics used in studies, with 50% reduction in MMDs, coupled with change in consumption of abortive medications. The expert did note that change in MMDs is not a perfect metric, as some patients have no change in daily frequency but may have significant reductions in severity or duration of migraine.

The clinical expert highlighted that no specialized settings are required, and that neurologists or other experts (pain clinic specialists, family medicine physicians with expertise) with headache expertise should prescribe atogepant.

Clinician Group Input

One clinician group — the Canadian Headache Society (CHS), consisting of 5 headache specialists — provided input to CADTH for the review of atogepant. The CHS is a scientific society of health care professionals dedicated to research, education of residents and physicians, and promotion of better care for patients with headache disorders.

The clinician group emphasized that migraine is often underdiagnosed and undertreated, with limited access to specialized care for migraine in Canada. Along with unmet needs similar to those identified by the clinical expert consulted by CADTH — notably that current treatments are not effective for all patients (response rate of 40% to 50% for oral medications) and may lose effectiveness over time — the clinician group also highlighted difficulties in access due to limited coverage and regional variation in funding by province, particularly for triptans, onabotulinumtoxinA, and CGRP mAbs.

The clinician group indicated that atogepant could be used as a first-line treatment option for the prevention of migraines, but noted its place in therapy will be determined in part by its cost, and it thus could be considered before other CGRP antibodies. Moreover, the clinician groups emphasized that atogepant could be provided in primary care, increasing access to patients in need. In contrast, the clinical expert consulted by CADTH indicated atogepant would be considered as last-line therapy or used in specific circumstances, such as nonresponse, intolerance, or contraindication to first-line treatment options, and where risks outweigh the benefits with other first-line treatment options (i.e., monoclonals) in individuals of child-bearing potential. The clinician group and clinical expert consulted by CADTH considered the potential for concurrent use of atogepant with mAbs or onabotulinumtoxinA.

Description of Studies

A total of 3 studies were included in this review: ADVANCE, CGP-MD-01, and ELEVATE. Two of the studies, ADVANCE and CGP-MD-01, were provided to CADTH when the submission was initially provided by the sponsor, while the ELEVATE study was provided to CADTH during the later stages of the review. The ADVANCE trial was a phase III, double-blind RCT evaluating the safety and tolerability of atogepant for the preventive treatment of migraine in patients with EM. Patients in the ADVANCE trial were required to have a 1-year history of migraine consisting of 4 to 14 migraine days per month, with or without aura, and migraine onset before 50 years of age. A total of 910 patients were randomized 1:1:1:1 to atogepant 10 mg once daily (n = 222), atogepant 30 mg once daily (n = 230), atogepant 60 mg once daily (n = 235), or placebo (n = 223). The primary outcome of the ADVANCE trial was change from baseline in mean MMDs, with key secondary end points of change from baseline in MHDs, change from baseline in acute MUDs, greater than or equal to 50% reduction in 3-month average of MMDs, change from baseline in Migraine-Specific Quality of Life Questionnaire (MSQ) v2.1 Role Function-Restrictive domain score, and change from baseline in the Performance of Daily Activities domain score and mean monthly Physical Impairment domain score of the Activity Impairment in Migraine – Diary (AIM-D). The ADVANCE trial was conducted at 136 sites in the US. There were no Canadian investigative sites included. No interim analyses were conducted.

The CGP-MD-01 study was a phase II/III, double-blind RCT evaluating the safety and tolerability of 10 mg once daily, 30 mg once daily, 30 mg twice daily, 60 mg once daily, and 60 mg twice daily dose regimens of atogepant for the prevention of EM. Included patients for the CGP-MD-01 trial were similar to the ADVANCE trial, although diagnosis of migraine was based on International Classification of Headache Disorders (ICHD) 2013. A total of 834 patients were randomized to 1 of 6 different arms in a 2:1:2:1:2:1 randomization sequence of placebo (n = 186), atogepant 10 mg once daily (n = 94), atogepant 30 mg once daily (n = 185), atogepant 30 mg twice daily (n = 89), atogepant 60 mg once daily (n = 187), or atogepant 60 mg twice daily (n = 93). Only Health Canada–approved dosages are summarized in this report; thus, any results for the atogepant 30 mg twice daily and atogepant 60 mg twice daily doses are not discussed. The primary outcome of the CGP-MD-01 trial was the same as the ADVANCE trial: change from baseline in mean MMDs, with 3 secondary end points of change from baseline in mean MHDs, proportion of patients with at least a 50% reduction in mean MMDs, and change from baseline in mean monthly acute MUDs. The CGP-MD-01 trial was conducted at 78 sites in the US. There were no Canadian investigative sites included. No interim analyses were conducted.

The ELEVATE trial was a phase III, randomized, double-blind, placebo-controlled study. The objective of the ELEVATE study was to evaluate the efficacy and safety of atogepant 60 mg once daily for the prevention of migraine in adult patients with EM who have previously had inadequate response to 2 to 4 classes of oral medications for the prophylaxis of migraine. A total of 315 patients were randomized 1:1 to atogepant 60 mg once daily (n = 157) or placebo (n = 158). The primary and key secondary outcomes of the ELEVATE study were identical to the ADVANCE study. A total of 73 sites in North America and Europe screened patients for eligibility, and 6 patients were included from Canada. No interim analyses were conducted.

Demographic and baseline characteristics in all studies were well balanced. Most patients were female (ADVANCE: 86.1% to 90.5%; CGP-MD-01: 82.8% to 90.7%; ELEVATE: ||||| ||| |||||) and white (ADVANCE: 81.1% to 89.2%; CGP-MD-01: 71.5% to 79.2%; ELEVATE: ||||| ||| |||||). The median age ranged from 38.5 years to 42.0 years in the ADVANCE study, 38.0 years to 40.5 years in the CGP-MD-01 study, and |||| || |||| ||||| in the ELEVATE study. The included studies differed in the proportion of patients who had received prior migraine prevention medicine, with ||||| || ||||| of patients in the ADVANCE study receiving prior migraine therapy, and only 25.1% to 31.2% of patients receiving prior migraine therapy in the CGP-MD-01 study, while all patients in the ELEVATE study received prior migraine therapy.

Efficacy Results

The primary efficacy end point of the included studies was change from baseline in MMDs to week 12. In all trials, atogepant resulted in statistically significant differences compared with placebo in the reduction of mean MMDs across the 12-week treatment period. In the ADVANCE trial, the least squares mean difference (LSMD) in mean change from baseline in MMDs at 12 weeks compared to placebo was –1.21 days (95% CI, –1.78 to –0.64) for the atogepant 10 mg group; –1.38 days (95% CI, –1.94 to –0.82) for the atogepant 30 mg group; and –1.72 days (95% CI, –2.28 to –1.15) for the atogepant 60 mg group (all P < 0.0001). In the CGP-MD-01 trial, the LSMD for mean change from baseline in MMDs at 12 weeks compared to placebo was –1.15 days (95% CI, –1.93 to –0.37; P = 0.0039) for the atogepant 10 mg group; –0.91 days (95% CI, –1.55 to –0.27; P = 0.0056) for the atogepant 30 mg group; and –0.70 days (95% CI, –1.35 to –0.06; P = 0.0325) for the atogepant 60 mg group. In the ELEVATE trial, the LSMD in mean change from baseline in MMDs between atogepant 60 mg once daily and placebo at 12 weeks was ||||| |||| |||| ||| ||||| || |||||| | | |||||||. Results for the subgroup analyses of the ADVANCE study in patients with or without prior exposure to migraine prevention therapy were consistent with the primary analysis. A post hoc subgroup analysis of the ADVANCE trial by number of prior preventive treatment failures exhibited similar results to the primary analysis, although the mean difference from placebo was higher in the subgroup of patients who had inadequate response to more than 2 prior treatments, and results were consistent with the ELEVATE study.

Results for key secondary outcomes were in line with the primary end point, with atogepant demonstrating statistically significantly greater efficacy compared to placebo. In the ADVANCE trial, a greater proportion of patients achieved a greater than or equal to 50% reduction in mean MMDs with atogepant (55.6%, 58.7%, and 60.8% for the atogepant 10 mg, 30 mg, and 60 mg groups, respectively) compared to placebo (29.0%). In the CGP-MD-01 trial, a greater proportion of patients achieved a greater than or equal to 50% reduction in mean MMDs with atogepant (57.6%, 53.3%, and 52.0% in the atogepant 10 mg, 30 mg, and 60 mg groups, respectively) compared to placebo (40.4%). In the ELEVATE trial, a greater proportion of patients achieved a greater than or equal to 50% reduction in mean MMDs with atogepant 60 mg once daily (|||||) compared to placebo (|||||). Post hoc subgroup analysis from the ADVANCE trial for patients who had inadequate response to at least 2 prior treatments were |||||||||| |||| ||| ||||||| |||||||| ||| |||| ||| ||||||| |||||| ||||||||||||| ||||||| ||||||| for the proportion of patients achieving at least a 50% reduction in mean MMDs with atogepant (ranging from ||||| || ||||| across atogepant treatment groups) compared to a lower placebo group rate (|||||).

Results for secondary outcomes of MHDs and acute MUDs were consistent with the primary analysis for all studies, demonstrating statistically significant efficacy compared to placebo. In the ADVANCE trial, the LSMD in change from baseline in MHDs and acute MUDs compared to placebo was –1.42 days (95% CI, –2.03 to –0.81) and –1.31 days (95% CI, –1.81 to –0.82) for atogepant 10 mg; –1.53 days (95% CI, –2.13 to –0.92) and –1.33 days (95% CI, –1.82 to –0.83) for atogepant 30 mg; and –1.71 days (95% CI, –2.32 to –1.10) and –1.50 days (95% CI, –2.00 to –1.01) for atogepant 60 mg (all P < 0.0001). In the CGP-MD-01 trial, the LSMD in change from baseline in MHDs and acute MUDs compared to placebo was –1.38 days (95% CI, –2.23 to –0.54; P = 0.0014) and –1.30 days (95% CI, –1.99 to –0.60; P = 0.0002) for atogepant 10 mg; –1.24 days (95% CI, –1.94 to –0.55; P = 0.0005) and –1.44 days (95% CI, –2.01 to –0.87; P < 0.0001) for atogepant 30 mg; and –0.94 days (95% CI, –1.64 to –0.24; P = 0.0087) and –1.11 days (95% CI, –1.68 to –0.54; P = 0.0001) for atogepant 60 mg. In the ELEVATE trial, the LSMD in change from baseline in MHDs and acute MUDs compared to placebo was ||||| |||| |||| ||| ||||| || |||||| || ||||||| ||| ||||| |||| |||| ||| ||||| || |||||| || ||||||||, respectively.

Change from baseline at week 12 in MSQ v2.1 Role Function-Restrictive domain score was a key secondary end point of the ADVANCE and ELEVATE studies. In the ADVANCE trial, the LSMD in change from baseline versus placebo was statistically significant in favour of atogepant with a mean difference of 9.90 points (95% CI, 5.45 to 14.36) for atogepant 10 mg; 10.08 points (95% CI, 5.71 to 14.46) for atogepant 30 mg; and 10.80 points (95% CI, 6.42 to 15.18) for atogepant 60 mg (all P < 0.0001). In the ELEVATE trial, the LSMD in change from baseline versus placebo was ||||| |||||| |||| ||| ||||| || |||||| || ||||||| || |||||| || ||||||||| || || |||

Change from baseline in the HIT-6 total score was an additional efficacy outcome end point of the ADVANCE, CGP-MD-01, and ELEVATE studies. In the ADVANCE study, the LSMD change from baseline in HIT-6 total score compared to placebo at week 12 was ||||| |||||| |||| ||| ||||| || |||||| for atogepant 10 mg; ||||| |||||| |||| ||| ||||| || |||||| for atogepant 30 mg; and ||||| |||||| |||| ||| ||||| || |||||| for atogepant 60 mg. Higher proportions of HIT-6 responders (defined as patients who had at least a 5-point improvement [decrease] from baseline in the HIT-6 total score) were observed for the atogepant 10 mg (|||||), 30 mg (|||||), and 60 mg (|||||) groups, compared to placebo (|||||). In the CGP-MD-01 study, the LSMD change from baseline in HIT-6 scores was greater for all atogepant doses compared to placebo at all time points. Over 12 weeks, the LSMD versus placebo was |||| |||||| |||| ||| |||| || |||| for atogepant 10 mg, |||| |||||| |||| ||| |||| || ||||| for atogepant 30 mg, and |||| |||||| |||| ||| |||| || |||| for atogepant 60 mg. In the ELEVATE study, the LSMD change from baseline in HIT-6 scores was ||||| |||||| |||| ||| ||||| || |||||| in favour of atogepant 60 mg once daily over 12 weeks compared to placebo.

Harms Results

The incidence of treatment emergent adverse events (TEAEs) was generally consistent between patients treated with atogepant versus placebo, as well as across trials. At least 1 TEAE was experienced by 52.9%, 52.2%, 53.7%, and 56.8% with atogepant 10 mg, 30 mg, 60 mg, and placebo, respectively, in the ADVANCE study; by 65.6%, 62.8%, 57.5%, and 49.5% with atogepant 10 mg, 30 mg, 60 mg, and placebo, respectively, in the CGP-MD-01 study; and ||||| ||| ||||| of patients in the atogepant 60 mg and placebo groups, respectively, in the ELEVATE study. The most frequently reported TEAEs in the ADVANCE trial were constipation (7.7%, 7.0%, 6.9%, and 0.5%), nausea (5.0%, 4.4%, 6.1%, and 1.8%), and upper respiratory tract infections (4.1%, 5.7%, 3.9%, and 4.5%) in the atogepant 10 mg, 30 mg, 60 mg, and placebo groups, respectively. The most frequently reported TEAEs in the CGP-MD-01 trial were nausea (5.4%, 7.1%, 11.8%, and 4.8%), upper respiratory tract infection (6.5%, 7.7%, 5.4%, and 8.1%), nasopharyngitis (3.2%, 6.0%, 7.5%, and 2.2%), and constipation (2.2%, 5.5%, 4.8%, and 2.2%) for the atogepant 10 mg, 30 mg, and 60 mg, and placebo groups, respectively. The most frequently reported TEAEs in the ELEVATE study were |||||||||||| |||||| ||| |||||| |||||| ||||| ||| |||||| |||||||| ||||| ||| |||||| ||| ||||||||||||||| ||||| ||| |||||. In all studies, most TEAEs were mild to moderate in severity.

Serious adverse events (SAEs) in the ADVANCE, CGP-MD-01, and ELEVATE trials were infrequent, occurring in only 2 patients (0.9%) in both the atogepant 10 mg and placebo groups in the ADVANCE trial, and ||||| SAEs occurring in 7 patients in the CGP-MD-01 trial (1 [1.1%] with atogepant 10 mg, 2 [1.1%] with atogepant 30 mg, 2 [1.1%] with atogepant 60 mg once daily , and 2 [1.1%] with placebo), and in | |||||| | || ||||| patients in the atogepant and placebo groups in the ELEVATE study, respectively.

In the ADVANCE study, the incidence of withdrawals due to adverse events (WDAEs) was similar across treatment groups, occurring in 4 to 9 patients (1.8% to 4.1%) treated with atogepant and 6 patients (2.7%) in the placebo group. In the CGP-MD-01 trial, WDAEs were more common in the atogepant groups (4.3%, 6.0%, and 3.2%) than in the placebo group (2.7%). In the ELEVATE trial, | |||||| ||| | |||||| patients in the atogepant and placebo groups had WDAEs. There were no deaths reported during any of the included studies.

In the ADVANCE study, 1 patient in the placebo group reported suicidal behaviour during the double-blind treatment period. No patients reported suicidal ideation with intent to act via their Columbia-Suicide Severity Rating Scale (C-SSRS) assessments. In the CGP-MD-01 trial, no patients reported suicidal behaviour during the study; however, 1 patient in the placebo group reported suicidal ideation limited to “wish to be dead” during the double-blind treatment period. In the ELEVATE study, | |||||| and | |||||| patients in the atogepant and placebo groups, respectively, reported suicidal behaviours during the study.

Critical Appraisal

The ADVANCE, CGP-MD-01, and ELEVATE studies were all double-blind RCTs. Appropriate methods for randomization (via interactive web response system [IWRS]), treatment allocation, and maintenance of blinding to treatment assignment were used in all studies, reducing the possibility for selection, performance, and detection biases. There was a high proportion of screen failures in the ADVANCE, CGP-MD-01, and ELEVATE studies (60%, 53%, and |||), mostly due to patients not meeting eligibility criteria. In the CGP-MD-01 study, more patients discontinued based on withdrawal of consent or withdrawal by patient in the placebo arm; however, it is unclear how such discontinuations would have affected blinding or the study results. The rate of constipation was more frequent in the atogepant groups across trials, which may have led to unblinding. Given that the overall rates were generally low, it is unclear what effect this would have on the results. Sensitivity analyses to account for missing data were conducted on the primary end point in all studies, and were in line with the primary results, suggesting that missing data had little impact. Acceptable methods to account for multiplicity were used in all trials. In the ADVANCE and ELEVATE studies, the primary end point and 6 key secondary end points were controlled for multiplicity using the overall familywise error rate at the 0.05 level. One prespecified subgroup analysis of the ADVANCE study was conducted, which included patients with or without prior migraine prevention medication with proven efficacy. An additional post hoc subgroup analysis of ADVANCE was submitted to CADTH by request for patients in the ADVANCE study, with at least 1 and at least 2 prior migraine prevention treatment failures, which represents the population for the reimbursement request. Given that this subgroup was conducted post hoc and was not part of the randomization scheme or statistically powered to detect within-group or between-group differences, the results from the subgroup analysis may confound the observed results and should only be interpreted as supportive evidence for the overall effect of atogepant. Moreover, missing data were unaccounted for, and the analyses did not adjust for multiplicity. The population for this post hoc subgroup analysis was the target population for the ELEVATE study, which also included 3 prespecified subgroups, including 2 of interest to this review (inadequate response to prior oral prophylactic treatment, and migraine days at baseline).

The inclusion and exclusion criteria for the ADVANCE, CGP-MD-01, and ELEVATE studies were appropriate and generalizable to the population in Canada, according to the clinical expert consulted by CADTH. As part of the inclusion and exclusion criteria for the ADVANCE and CGP-MD-01 studies, patients were required to have an inadequate response to no more than 3 medications prescribed for the prevention of migraine, and patients were excluded who had previous exposure to CGRP mAbs. Conversely, the ELEVATE study enrolled patients who had inadequate response to 2 to 4 oral prophylactic migraine medications and was the only trial that was reflective of the population included in the reimbursement request. One of the major differences between the ADVANCE, CGP-MD-01, and ELEVATE studies was the proportion of patients who had received prior migraine prevention medications, where ||||| || ||||| of patients received prior migraine therapy in the ADVANCE trial compared to 25.1% to 31.2% of patients in the CGP-MD-01 trial and |||| of patients in the ELEVATE trial. As noted in the post hoc subgroup analysis for the ADVANCE trial, only 119 patients had inadequate response to 2 or more prior preventive migraine treatments; however, given that baseline characteristics for this subgroup were not presented, it was unclear if any of these patients had received prior anti-CGRP mAbs. Thus, the full population of the ADVANCE study does not entirely represent the population for the reimbursement request and may not be generalizable to this population in Canada. All included trials were placebo-controlled and did not include an active comparator, which allows for adequate evaluation of the treatment effect of atogepant; however, this may overestimate the treatment effects. In all studies, there was a high placebo response, impacting the ability to interpret the efficacy of atogepant.

Baseline demographic and clinical characteristics, including the average number of MMDs and MHDs days at baseline, were noted to provide a true reflection of what would be seen in Canadian clinical practice, as noted by the clinical expert; however, it is worth noting that patients enrolled in the studies had to have history of 4 to 14 migraine days per month on average in the 3 months before the first visit. Therefore, all studies excluded patients with 1 to 3 migraine days per month, and it is uncertain if results from the ADVANCE, CGP-MD-01, and ELEVATE trials are generalizable to patients with fewer than 4 migraine days per month. Outcomes of the ADVANCE and CGP-MD-01 trials were similar to those reported in other clinical trials for migraine and are reflective and important in guiding treatment decisions in Canadian clinical practice.

Description of Studies

For the purposes of the Canadian submission, the sponsor-submitted NMA included 2 analysis scenarios from the original NMA that were updated to reflect the relevant comparators and reimbursement request:

• Scenario 2: CGRP inhibitors and key oral preventives approved in the US as a treatment for EM.
• Scenario 4: Global patients who have experienced 2 or more prior preventive treatment failures, versus CGRP preventives.

The objective of the sponsor-submitted report was to evaluate the relative efficacy, safety, and tolerability of atogepant compared with injectable CGRP inhibitors and key oral preventives approved for the treatment of EM. The sponsor-submitted NMA was informed by a systematic literature review (SLR) (updated to August 9, 2021) to identify all existing RCTs assessing the efficacy, safety, and tolerability of preventive treatments for adults with EM compared to other preventive treatments, placebo, or standard care. The analyses were conducted using a Bayesian NMA. Selection of both fixed and random effects was conducted. In analysis scenario 2, random-effects models for the analyses excluding Japanese studies were selected as the base-case analysis given the larger evidence base. In analysis scenario 4, fixed-effects models were selected as the base case due to the lower number of trials and the lower deviance information criterion. In the updated NMAs, where available, efficacy analyses included 50% response in MMDs, between-treatment change from baseline in MMDs, and between-treatment change from baseline in monthly migraine MUDs. Safety outcomes included all-cause discontinuation and TEAEs.

Efficacy Results

In analysis scenario 2, ||| ||| |||||||| |||||||||| ||||| ||| |||||||| ||||| |||| ||| |||||||||| |||| |||||||| |||| |||||||| |||||||| ||||| ||| || |||||||||| ||||||| ||||||||| ||| ||| ||||| |||||| ||||||||||| ||||||| ||| ||||||||||| |||||||| ||||||||| |||||||| |||||||| |||| |||||||||| |||| ||| |||| |||| |||||||||

In analysis scenario 4, ||| |||| |||||||||| |||| |||||||| |||| ||||||| ||| ||| |||||||| |||||||||| |||| |||||||||||| ||| || ||| ||| ||| || |||||||||||| ||||||| || ||||||||||||| ||| |||||||| ||||||| |||| |||||||| |||||||| || |||||||| ||| ||| |||||||||| ||||| ||| || |||||||||| ||||||| |||||| |||||||||||

In analysis scenario 2| ||||| ||| || |||||||||| ||||||| ||| ||||||||| |||||||||| ||| ||||||||| |||||||||||||||| ||| ||||| ||| ||||||||| || |||||||||| ||||||| |||||||||| ||| |||||| |||||| ||||||||||||| |||||| || ||| |||||||||| ||| || |||| |||||||| |||| ||| ||||| || ||||||||||

||||| |||| ||| ||||||||| || ||| |||||||| || |||||||| |||| |||||||| |||||||| |||| || ||||||| |||| |||||||||||| || |||||||| |||| ||||||| |||| || ||||| || |||||||||| |||||||||

Critical Appraisal

There were several limitations associated with the sponsor-submitted NMA, particularly the clinical and methodological heterogeneity, which resulted in limited interpretability and generalizability of the results. The SLR and feasibility assessment were generally well conducted; however, the list of treatments for the NMA was narrower than that of the SLR. The NMA did not include valproic acid or candesartan, which, according to the clinical expert consulted by CADTH, could be considered relevant comparators for the treatment of EM. Important outcomes such as HRQoL were not considered based on a low availability of data. Following the submission of the ELEVATE study to CADTH, the SLR and NMA were not updated to include this relevant study in this patient population.

Analysis scenario 2 evaluated CGRP inhibitors and key oral preventives, while analysis scenario 4 evaluated patients who have experienced 2 or more prior preventive treatment failures in only CGRP inhibitors. In analysis scenario 2, it is unclear how the number of prior treatment failures as a factor of heterogeneity may have impacted the results, and the direction of bias remains uncertain. In analysis scenario 4, trial populations often included small sample sizes, ranging from 19 patients to 137 patients per treatment arm, with the ADVANCE trial including only 122 patients total in the subgroup who had inadequate response to 2 or more treatments, which limits the precision and generalizability of the treatment effect. Follow-up duration of the included trials generally varied and was also a significant source of heterogeneity across trials, with treatment periods ranging from 12 weeks to 56 weeks. For the primary efficacy end point, the time of assessment of 1 week to12 weeks was chosen, as this was the time frame of the primary efficacy end point in the ADVANCE trial. However, other included studies varied on when change from baseline was assessed.

Clinical heterogeneity was assessed visually for baseline characteristics, including age, sex, race, body mass index (BMI), baseline MMDs, and baseline MHDs, as well as for time points and end point availability. The sponsors reported that, in general, the studies were similar, including mostly patients of the same age group, sex, and gender. The sponsor considered the main difference between studies to be with regard to race, whereby Japanese studies were excluded from the base case of the primary analysis in the original NMA, with 2 other Japanese studies excluded in the NMA update due to a lower or negligible placebo response compared to other studies, potentially due to unaccounted-for baseline or study centre characteristics that varied. Consideration was given to many baseline characteristics as treatment effect modifiers or prognostic factors; however, it was unclear how this was managed in any statistical analyses. While not reported, there may have been several differences in study and baseline characteristics across the trials that remain unaccounted for, including study design — which included RCTs, open-label studies, and crossover studies — as well as varying definitions of MMD and MHD, with some trials not reporting any MMD or MHD inclusion criteria; and, as noted by the sponsor, none of the trials published before 2001 reported MMD or MHD inclusion criteria.

All studies included in the NMA were believed to be statistically heterogeneous based on the considerable range of I² values, although it is unclear what the source of heterogeneity was, as it was not explored. Though the authors relied on visual inspection of clinical heterogeneity, the observed heterogeneity is likely due to the observed and unobserved differences in patient populations across the included studies, data imputation analysis methods, and the specific prior or background treatments allowed or received.

In the analyses comparing atogepant to all other treatments, there was generally no difference between atogepant and any of the other treatments in analysis scenario 2, or the other CGRP inhibitors in analysis scenario 4. Moreover, there were wide credible intervals (CrIs) that crossed the null threshold, further challenging the precision of the results. The general results in analysis scenario 2 displayed a reverse dose-response relationship for atogepant, whereby the atogepant 10 mg dose demonstrated the greatest response, while the atogepant 60 mg dose demonstrated less of a response, which was the opposite of what was seen in the ADVANCE trial. No rationale for this observation was provided, and the reason for this remains uncertain; however, it may be due to the pooling of estimates from the ADVANCE and CGP-MD-01 trials. This effect was not observed in analysis scenario 4.

Description of Studies

Two studies, Study 309 and Study 302, were included as other relevant evidence for the review of atogepant. Study 309 was a phase III, open-label extension study that examined the long-term safety and tolerability of oral atogepant 60 mg once daily in adult patients with EM, for up to 40 weeks of treatment. Patients were eligible to enrol into Study 309 if they completed the lead-in ADVANCE study. A total of 685 patients received at least 1 dose of atogepant 60 mg once daily and 511 (74.6%) patients completed the study. The mean age of patients in the study was 41.8 years (standard deviation [SD] = 12.3). Most patients (43.9%) were diagnosed with migraine without aura and the mean duration of the migraine disorder was 21.6 years (SD = 12.8). The mean number of migraine and headache days per month in the last 3 months were ||| |||| |||| ||| ||| |||| |||), respectively.

Study 302 was a phase III, randomized, open-label study that examined the long-term safety and tolerability of oral atogepant 60 mg once daily in adult patients with EM, for up to 52 weeks of treatment. Patients were eligible to enrol into Study 302 if they had completed the lead-in CGP-MD-01 study, in addition to new patients who met the eligibility criteria. Patients were randomized at visit 2 to receive atogepant 60 mg once daily or standard of care (SOC) (oral migraine preventive medication) in a 5:2 ratio. The SOC treatment arm only served to provide context for interpreting the safety results of atogepant. A total of 543 and 196 patients received at least 1 dose of atogepant 60 mg once daily and SOC, respectively. The mean age of patients was 42.5 years (SD = 12.0) in the atogepant arm and 41.1 years (SD = 12.1) in the SOC arm. Most (||||| ||| ||||| in the atogepant and SOC arms, respectively) patients were diagnosed with migraine without aura. The mean duration of the migraine disorder was |||| years (SD: ||||) and |||| years (SD = ||||) in the atogepant and SOC arms, respectively. The mean number of migraine and headache days per month in the last 3 months were 7.3 (SD = 2.6) ||| ||| |||| ||||, respectively, in the atogepant arm. The mean number of migraine and headache days per month in the last 3 months were ||| |||| |||| and ||| |||| ||||, respectively, in the SOC arm. A total of ||| ||||||| and ||| ||||||| patients completed the open-label treatment period in the atogepant and SOC arms, respectively.

Efficacy Results

Study 309 did not evaluate the efficacy of atogepant 60 mg once daily.

Efficacy outcomes in Study 302 were collected daily at home via an electronic diary and at clinic visits via an electronic tablet from patients in the atogepant arm only. The mean number of MMDs decreased at weeks 49 to 52 from baseline; mean MMDs at baseline were 7.28 (SD = 2.70) and least squares (LS) mean change was –5.19 (standard error [SE] = 0.16; 95% CI, –5.50 to –4.87). The proportion of patients who achieved a greater than or equal to 50%, 75%, and 100% reduction in MMDs at weeks 49 to 52 was |||||| |||||| ||| |||||, respectively. The mean number of MDHs decreased at weeks 49 to 52 from baseline; mean MHDs at baseline were 8.33 (SD = 2.97) and LS mean change was ||||| |||| ||||| ||| ||| ||||| || ||||||. At weeks 49 to 52, the LS mean change from baseline in the number of monthly moderate to severe and severe headache days was ||||| ||| |||||, respectively. The LS mean change from baseline in the number of monthly cumulative headache hours was |||||| hours at weeks 49 to 52. The mean number of MUDs decreased at weeks 49 to 52 from baseline; mean MUD at baseline was |||| |||| ||||| and LS mean change was ||||| |||| ||||| ||| ||| ||||| || ||||||. The LS mean change from baseline in the number of monthly triptan use days was ||||| days at weeks 49 to 52. The LS mean change from baseline in the MSQ v2.1 Role Function-Restrictive domain score was ||||| (SE = |||||| ||| ||| ||||| || |||||) at week 52. The LS mean change from baseline in the AIM-D Performance of Daily Activities domain score was –||||| |||| ||||| ||| ||| |||||| || |||||| at weeks 49 to 52. The LS mean change from baseline in the AIM-D Physical Impairment domain score was ||||| |||| ||||| ||| ||| ||||| || |||||) at weeks 49 to 52.

Harms Results

In Study 309, TEAEs were reported in 428 patients (62.5%) during open-label treatment, and included upper respiratory tract infection (5.5%) and urinary tract infection (5.3%). SAEs were reported in 23 patients (3.4%) and no deaths were reported during the open-label treatment. Premature discontinuation due to at least 1 TEAE was reported in 22 patients (3.2%) during the open-label treatment. For notable harms, 23 patients (3.4%) reported constipation and 4 patients (0.6%) reported alanine or aspartate aminotransferase greater than or equal to 3 times the upper limit of normal value. No Hy’s law cases or suicidal ideation were reported.

In Study 302, TEAEs were reported in 364 patients (67.0%) during the open-label treatment, including upper respiratory tract infection (10.3%), constipation (7.2%), nausea (6.3%), and urinary tract infection (5.2%). For context, TEAEs were reported in 154 patients (78.6%) in the SOC arm. SAEs were reported in 24 patients (4.4%) and 7 patients (3.6%) during the open-label treatment with atogepant and SOC, respectively. Two deaths were reported in the safety population in the atogepant arm; no deaths were reported in SOC arm. Premature discontinuation due to at least 1 TEAE was reported in 31 patients (5.7%) and 5 patients (2.6%) during the open-label treatment with atogepant and SOC, respectively. Notable harms identified in the atogepant arm included constipation in 39 patients (7.2%), suicidal ideation in 3 patients (0.6%), and elevations in alanine or aspartate aminotransferase that were greater than or equal to 3 times the upper limit of normal value in 13 patients (2.4%). No Hy’s law cases were reported.

Critical Appraisal

The open-label study design of the long-term extension study, Study 309, can bias the reporting of end points, particularly any subjective measures included in the safety (and efficacy in Study 302) parameters, due to the unblinding of the study drug during the treatment period. Because patients were required to have completed the lead-in study without any significant deviations from the protocol (i.e., noncompliance with procedures) and did not experience any adverse events (AEs) that may indicate an unacceptable safety risk per investigator judgment, the resultant population may be more tolerant of atogepant — leading to an underreporting of AEs — and are also more likely to experience benefits with atogepant, overestimating the efficacy of treatment, as those without benefit are unlikely to continue. In the absence of an active comparator or placebo group, the interpretation of the results is limited. This is compounded using descriptive statistics only.

The limitations can also be applied to Study 302. The enrolment of new patients without prior experience with atogepant and patients who have completed a lead-in study further limits the interpretation of the results. It should be noted that the SOC treatment arm only served to provide context for interpreting the safety results of atogepant. The oral migraine preventives were prescribed in a manner that reflected routine clinical practice. A flexible treatment paradigm was used that permitted the discontinuation of or switching from 1 drug to an alternative for migraine prevention as needed and per investigator judgment. Regardless of the type of change made, patients in the SOC arm were permitted to continue with the study. Thus, AE reporting in the SOC arm could be influenced by investigator choice, as the AEs may differ based on the oral migraine preventive selected.