CADTH Reimbursement Review

Galcanezumab (Emgality)

Sponsor: Eli Lilly Canada Inc.

Therapeutic area: Prevention of migraine

This multi-part report includes:

Clinical Review

Pharmacoeconomic Review

Clinical Review

Executive Summary

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

Introduction

Migraine is a neurologic disease characterized by recurrent episodes of pulsating headache pain of at least moderate severity.¹ The type of migraine can be refined based on the frequency of monthly migraine headache days (MHDs) and monthly headache days.¹ The International Classification of Headache Disorders, third edition (ICHD-3), describes chronic migraine (CM) as a headache (tension-type-like or migraine-like) occurring on 15 or more days per month for more than 3 months with the features of migraine headaches on at least 8 days per month.² In episodic migraine (EM), individuals experience headaches on 14 or fewer days per month for more than 3 months with the features of migraine headaches on at least 4 days per month.² In Canada (2010 to 2011), 9.6% of the population over 18 years of age experienced migraine attacks, with more females (13.8%) than males (5.3%) having had migraine.³

Comprehensive migraine therapy includes management of lifestyle factors and triggers, acute and preventive (or prophylactic) medications, and migraine self-management strategies.^4,5 The goals of migraine treatments are to relieve pain, restore function, improve health-related quality of life (HRQoL), reduce headache frequency, and prevent the progression of EM to CM.⁶ Preventive medications include a variant of the botulinum toxin (onabotulinumtoxinA; for CM only), inhibitors of the calcitonin gene-related peptide receptor (CGRP), beta-blockers, calcium-channel blockers, tricyclic antidepressants, serotonin-norepinephrine reuptake inhibitors, angiotensin-converting enzyme inhibitors and angiotensin receptor blockers, anticonvulsants, and pizotifen. Only topiramate, onabotulinumtoxinA, and the CGRP inhibitors have been approved by Health Canada for the prevention of migraine, and of these, onabotulinumtoxinA is indicated only for the prevention of CM. Migraine prophylaxis is an important part of the overall approach for a proportion of individuals with migraine.⁴ Of patients with migraine who have received preventive medications, 87% have an inadequate response to 2 or more preventive therapies.⁷

Galcanezumab is a humanized immunoglobin G4 monoclonal antibody that binds to CGRP and prevents its biologic activity.⁸ Galcanezumab is administered by subcutaneous injection. Galcanezumab received a Health Canada Notice of Compliance on July 30, 2019, with an indication for the prevention of migraine in adults who have at least 4 migraine days per month. The recommended dose is a loading dose of 240 mg (administered as 2 consecutive subcutaneous injections of 120 mg) followed by once-monthly doses of 120 mg (1 injection). Galcanezumab has not been previously reviewed by CADTH.

The objective of this report is to perform a systematic review of the beneficial and harmful effects of galcanezumab (120 mg/mL solution in a 1 mL single-use, pre-filled syringe or pen) for the prevention of migraine in adults who have at least 4 MHDs per month and have experienced an inadequate response, intolerance, or contraindication to at least 2 prophylactic migraine medications.

Stakeholder Perspectives

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

Patient Input

CADTH received a joint submission from Migraine Canada and Migraine Québec.

Migraine Canada and Migraine Québec identified the following as key impacts on the lives of those living with migraine and their families: inability to work resulting in financial stress and reliance on a spouse or family members to compensate, childcare and needing additional help, restricted social activities and difficulties with relationships, and lack of understanding from others. Patients described how their spouse and/or family often must bear the burden of household and financial responsibilities, how families miss out on spending time together, difficulties with starting new relationships and stress on existing relationships, and the lack of support available for caregivers.

Patients indicated that it is important to control the frequency and severity of migraine as well as reduce or eliminate the need for acute medications (i.e., triptans and opioids). Patients indicated that they wanted a preventive medication that allowed them to be more productive at work and home. They also felt it was important that a new medication would allow them to fully participate in daily life, work, improve family and social relationships, and reduce exhaustion and side effects. In general, survey respondents felt that nearly any degree of relief would be a successful outcome for a preventive therapy. When asked about methods of administration, 73% of participants stated they would prefer a monthly injection to a daily pill.

Clinician Input

Input From Clinical Experts Consulted by CADTH

According to the clinical expert consulted by CADTH, the most important goal of treatment is a reduction in the frequency of headache. In the clinical expert’s experience, a trial of 2 to 3 oral preventive medications is often required before a patient experiences a benefit. The older preventive medications have important side effects that often limit their use and affect adherence to treatment.

If the cost of the patient’s prescription drugs is reimbursed, the patient will typically be offered 2 or 3 oral preventive medications before receiving either onabotulinumtoxinA (if they have CM) or an anti-CGRP monoclonal antibody.

Patients with a diagnosis of migraine with or without aura, more than 4 headache days per month, failure on 2 or more daily preventers used at an appropriate dose for an appropriate period of time, and no contraindication of the use of an anti-CGRP monoclonal antibody would be suitable for galcanezumab. In the clinical expert’s opinion, patients with frequent EM without medication overuse could be most likely to respond to an anti-CGRP monoclonal antibody and patients with CM may realize the greatest benefit. The clinical expert added that patients least suitable for treatment with galcanezumab would include those contemplating pregnancy, and some physicians may be reluctant to prescribe it (or any anti-CGRP monoclonal antibody) to patients with known active peripheral vascular, cardiovascular, or cerebrovascular disease. In addition, use of this family of medications in patients with Raynaud phenomenon who were taking triptans may be associated with digital ischemia leading to digit amputation, and caution is therefore needed in that area.

Reduction in frequency and/or severity of headaches, reduced use of abortive medications, improved function, and quality of life are important measures of treatment response. The clinical expert reported that the goal is to reduce the frequency of headache, ideally to fewer than 4 headache days per month. A 50% or greater reduction in headache frequency is also considered successful. A reduction in the severity of the headaches, as measured by the Migraine Disability Assessment (MIDAS) or 6-item Headache Impact Test (HIT-6), even if the frequency of headaches remains unaltered, would also be considered successful, according to the clinical expert. The clinical expert reported that patients should be assessed for response to galcanezumab treatment at 2 to 3 weeks after their third injection, and those who have not shown improvement at that time should be assessed 2 to 3 weeks after their sixth injection. If there is no improvement after 6 injections, treatment would be discontinued. The clinical expert indicated that development of intolerable side effects, comorbidities that preclude the patient being on treatment (e.g., stroke or heart attack), or pregnancy would also lead to discontinuation of treatment.

The clinical expert noted that galcanezumab could be prescribed by headache specialists, general neurologists, family doctors on the advice of a neurologist, and family doctors experienced with the use of the medication in other patients. The clinical expert cautioned against requiring patients to be supervised by a headache specialist, as few such specialists practise in Canada.

Clinician Group Input

No input from clinician groups was received for this review.

Drug Program Input

The drug programs noted that 2 other CGRP inhibitors (erenumab and fremanezumab) received recommendations to reimburse with conditions under the CADTH review process and were undergoing pan-Canadian Pharmaceutical Alliance negotiations at the time of this review. The drug programs pointed out that galcanezumab has the same Health Canada indication as both fremanezumab and erenumab, although the reimbursement requests from the sponsors for the 3 medications differ.

The drug programs noted that the reimbursement request for galcanezumab specifies patients who have tried at least 2 prophylactic migraine medications, and indicated that it would be helpful to jurisdictions for this to be defined further (e.g., at least 2 prophylactic medications of different classes and clarification of the optimal dose and duration of the trials). The clinical expert consulted by CADTH indicated that, for an adequate trial, patients are ideally within the therapeutic range for at least 8 weeks before deciding whether treatment has failed. Furthermore, although the reimbursement request specifies failure on at least 2 prophylactic drugs, the drug plans indicated it would be helpful to clarify if there is a maximum number of prophylactic drugs that would be accepted before consideration of coverage. In the clinical expert’s experience, only a small proportion of patients have tried 4 or more preventive medications, and they did not think that a maximum number of prophylactic agents would be necessary.

The drug programs added that it would be helpful to outline whether failure on or intolerance to another CGRP inhibitor would exclude patients from coverage of galcanezumab. The clinical expert reported that no data currently indicate whether failure on or intolerance to 1 or more CGRP inhibitor would exclude patients from trying a treatment with another CGRP inhibitor such as galcanezumab. In addition, the clinical expert noted that the CGRP inhibitors have different properties and can work through different mechanisms (i.e., bind to the CGRP ligand versus the receptor).

The drug programs noted that there is potential in some jurisdictions for galcanezumab to be used in combination with onabotulinumtoxinA. The clinical expert noted that anti-CGRP monoclonal antibodies work to inhibit CGRP activity on nerves different from those affected by onabotulinumtoxinA. The clinical expert therefore suggested there may be reason to consider using both medications at the same time in some patients.

Clinical Evidence

Pivotal Studies and Protocol-Selected Studies

Description of Studies

Four phase III, multi-national, double-blind, randomized, placebo-controlled trials were identified and included in the systematic review: EVOLVE-1,⁹ EVOLVE-2,¹⁰ REGAIN,¹¹ and CONQUER.¹² In all trials, galcanezumab and matching-administration placebo were supplied as an injectable solution in 1 mL pre-filled manual syringes designed to deliver 120 mg of galcanezumab each. The injections were administered by study-site personnel once monthly at dosing visits. The primary outcome in all trials was the overall mean change from baseline in the number of monthly MHDs during double-blind treatment.

The EVOLVE-1 (N = 862) and EVOLVE-2 (N = 922) trials were identically designed studies of patients with EM. In both studies, patients were randomized in a 2:1:1 ratio to placebo, galcanezumab 120 mg (loading dose of 240 mg), or galcanezumab 240 mg. The double-blind treatment phase of the studies was 6 months in duration. The EVOLVE-1 and EVOLVE-2 trials excluded patients who had previously failed at least 3 classes of migraine-preventive treatments. Key secondary outcomes controlled for multiplicity included the Migraine-Specific Quality of Life Questionnaire version 2.1 (MSQ v2.1) role function – restrictive (RF-R) domain and monthly MHDs with acute headache pain medication intake.

The REGAIN trial (N = 1,117) was conducted in patients with CM. Patients were randomized in a 2:1:1 ratio to placebo, galcanezumab 120 mg (loading dose of 240 mg), or galcanezumab 240 mg. The study had a 3-month-long double-blind treatment phase. The REGAIN trial excluded patients who had previously failed at least 3 classes of migraine-preventive treatments. Key secondary outcomes that controlled for multiplicity included the MSQ v2.1 RF-R domain and monthly MHDs with acute headache pain medication intake.

The CONQUER trial (N = 463) was conducted in patients with EM or CM who had a documented history of 2 to 4 migraine-preventive medication category failures due to inadequate efficacy or tolerability in the past 10 years. Patients were randomized in a 1:1 ratio to placebo or galcanezumab 120 mg (loading dose of 240 mg). The study had a 3-month double-blind treatment phase. The key secondary outcome of interest to the systematic review was the MSQ v2.1 RF-R domain.

This review only reports the results for the galcanezumab 120 mg treatment arms because it is the Health Canada–approved dose. Results for the galcanezumab 240 mg treatment arms are not reported.

In all trials, most patients were female and White, and the mean age of patients was between 39 and 46 years. Most patients (> 60%) in the EVOLVE and REGAIN studies and all patients in the CONQUER study had received prior preventive treatment. Mean baseline MIDAS total scores of 33.2 and 33.0, which reflect severe disability, were reported in EVOLVE-1 and EVOLVE-2, respectively.¹² In the REGAIN trial, 29.5% of patients had failed 2 or more such treatments due to lack of efficacy in the past 5 years, and the mean baseline MIDAS total score was 67.2, which reflects very severe disability.¹² Overall, 15.5% of patients in the REGAIN trial had concurrent prophylaxis use with topiramate or propranolol. In the CONQUER trial, most patients had 2 (58.2%) or 3 (30.1%) prior medication category failures and the mean baseline total MIDAS score was 50.93, which reflects very severe disability.¹² During the baseline period, the mean number of monthly MHDs was 9.1 in both the EVOLVE-1 and EVOLVE-2 studies. In the REGAIN trial’s baseline period, patients had a monthly average of 21.4 headache days, of which an average of 19.4 were MHDs. During the baseline period in the CONQUER trial, patients had a monthly average of 15.0 headache days, of which an average of 13.2 were MHDs.

Efficacy Results

Key efficacy results from the double-blind treatment periods of the EVOLVE-1, EVOLVE-2, REGAIN, and CONQUER trials are summarized in Table 2 and Table 3.

Migraine-Specific Quality of Life Questionnaire Version 2.1

The mean change from baseline in the MSQ RF-R domain score was a key secondary outcome in the trials and controlled for multiplicity. In the EVOLVE-1 trial, the mean change from baseline in the MSQ RF-R domain score during months 4 to 6 of double-blind treatment was 7.74 points greater in the galcanezumab 120 mg group compared with placebo (95% confidence interval [CI], 5.20 to 10.28; P < 0.001). Similarly, in the EVOLVE-2 trial, the mean change from baseline in the MSQ RF-R domain score during months 4 to 6 was 8.82 points greater in the galcanezumab 120 mg group compared with placebo (95% CI, 6.33 to 11.31; P < 0.001). In the REGAIN trial, the least squares (LS) mean change from baseline was 5.06 points greater in the galcanezumab 120 mg arm compared with the placebo arm (95% CI, 2.12 to 7.99); however, the difference could not be tested for statistical significance based on the predefined multiple-testing procedure. In the CONQUER trial, the mean change from baseline in the MSQ v2.1 RF-R domain score at month 3 was 12.53 points greater in the galcanezumab group compared with placebo (95% CI, 9.19 to 15.87; P < 0.0001).

Migraine Headache Days With Symptoms

The overall change from baseline in number of monthly MHDs with symptoms (nausea and/or vomiting, photophobia and phonophobia, aura, and prodromal symptoms other than aura) during the double-blind treatment phase was an exploratory outcome in all pivotal trials. These outcomes were not controlled for multiplicity within the trials’ multiple-testing procedures. Results are summarized in Table 2. The LS mean change differences in the galcanezumab 120 mg arm versus the placebo arm indicated that galcanezumab treatment was associated with a numerically greater reduction in the number of monthly MHDs with all types of symptoms in all trials.

Migraine Disability Assessment

Change in MIDAS total score was a secondary outcome in all pivotal trials. This outcome was not controlled for multiplicity within the trials’ multiple-testing procedures. In the EVOLVE-1 trial, the mean change from baseline to the end of the double-blind treatment phase (month 6) in the galcanezumab 120 mg arm was −6.29 points (95% CI, −9.45 to −3.13), indicating a numerically greater reduction compared with placebo. In the EVOLVE-2 trial, the mean change from baseline to the end of the double-blind treatment phase (month 6) in the galcanezumab 120 mg arm was −9.15 points (95% CI, −12.61 to −5.69), indicating a numerically greater reduction compared with placebo. In the REGAIN trial, the mean change from baseline to the last observation carried forward (LOCF) end point for the galcanezumab 120 mg arm was −8.74 points (95% CI, −16.39 to −1.08), indicating a numerically greater reduction compared with placebo. In the CONQUER trial, the mean change in the galcanezumab 120 mg arm from baseline to the LOCF end point was −17.8 points (95% CI, −25.6 to −10.0), indicating a numerically greater reduction compared with placebo.

Number of Monthly Migraine Headache Days

The overall change from baseline in the number of monthly MHDs during the double-blind treatment phase was the primary outcome in each of the pivotal trials. The reduction in the overall LS mean change from baseline in the number of monthly MHDs during the double-blind treatment phase for galcanezumab 120 mg was statistically significantly greater compared to placebo in all studies.

In the EVOLVE-1 trial, the overall mean change from baseline in the number of monthly MHDs in the galcanezumab 120 mg arm during the double-blind treatment phase was −1.92 days (95% CI, −2.48 to −1.37; P < 0.001), indicating a greater reduction compared to placebo. In the EVOLVE-2 trial, the overall mean change in the galcanezumab 120 mg arm was −2.02 days (95% CI, −2.55 to −1.48; P <  0.001), indicating a greater reduction compared to placebo. In the REGAIN trial, the overall mean change in the galcanezumab 120 mg arm was −2.09 days (95% CI, −2.92 to −1.26; P < 0.001), indicating a greater reduction compared with placebo. In the CONQUER trial, the overall mean change in the galcanezumab arm was −3.12 days (95% CI, −3.92 to −2.32; P < 0.0001), indicating a greater reduction compared to placebo.

Number of Monthly Headache Days

The overall change from baseline in number of monthly headache days was a secondary outcome in the 4 pivotal trials. This outcome was not included in the trials’ multiple-testing procedures and therefore was not adjusted for multiplicity. In the EVOLVE-1 trial, the mean change difference in the galcanezumab 120 mg arm was −1.66 days (95% CI, −2.25 to −1.07), a numerically greater reduction compared to placebo. In the EVOLVE 2 trial, the mean change difference for the galcanezumab 120 mg arm was −2.00 days (95% CI, −2.58 to −1.42), a numerically greater reduction compared with placebo. In the REGAIN trial, the LS mean change difference in the galcanezumab 120 mg arm was −1.84 days (95% CI, −2.65 to −1.02) compared to placebo. In the CONQUER trial, the mean change difference in the galcanezumab 120 mg arm compared to placebo was a numerically greater −3.13 days (95% CI, −3.96 to −2.29).

Acute Headache Pain Medication Intake

In the EVOLVE-1, EVOLVE-2, and REGAIN trials, the overall change from baseline in monthly MHDs with acute medication use during the double-blind treatment phase was a key secondary outcome and was included in the trials’ multiple-testing procedures. In the EVOLVE-1 trial, the mean change difference from placebo in the galcanezumab 120 mg arm was a numerically greater −1.81 days (95% CI, −2.28 to −1.33; P < 0.001). In the EVOLVE-2 trial, the mean change difference from placebo in the galcanezumab 120 mg arm was a numerically greater −1.82 days (95% CI, −2.29 to −1.36; P < 0.001). In the REGAIN trial, the mean change difference in the galcanezumab 120 mg arm was −2.51 days (95% CI, −3.27 to −1.76), a numerically greater reduction compared to placebo. The difference between the galcanezumab 120 mg and placebo arms could not be tested for statistical significance based on the REGAIN trial’s predefined multiple-testing procedure. In the CONQUER trial, the overall change from baseline in MHDs with acute medication use during the double-blind treatment phase was a secondary outcome. This outcome was not included in the trial’s multiple-testing procedure, and therefore is not adjusted for multiplicity. The mean change difference in the galcanezumab 120 mg arm was −3.40 days (95% CI, −4.14 to −2.65), a numerically greater reduction compared to placebo.

Time to First Loss of Response in the Post-Treatment Phase

Time to the first loss of 50% response in patients who were 50% responders (defined as a reduction of 50% or more from baseline in monthly MHDs) in the last month of double-blind treatment and who entered the post-treatment phase was assessed in the EVOLVE-1, EVOLVE-2, and REGAIN trials. In EVOLVE-1 and EVOLVE-2, approximately half of the patients in all treatment groups had first loss of 50% response by 4 months after the end of the double-blind treatment phase.¹³ In the REGAIN study, the percentage of patients with first loss of 50% response at month 1 of the post-treatment phase was 24.3%.¹³ By month 4 of the post-treatment phase, 48.2% patients had first loss of 50% response.

Time to Initiation of Migraine-Prevention Treatment in the Post-Treatment Phase

Time to initiation of migraine-preventive medication during the post-treatment phase was assessed in EVOLVE-1, EVOLVE-2, and REGAIN. In the EVOLVE-1 study, in which 12 patients (< 2%) initiated treatment with a migraine-prevention medication, no significant differences between placebo and galcanezumab-treated patients were observed in the time of initiation.¹³ In the EVOLVE-2 study, 3.2% of patients in the placebo arm initiated treatment with a migraine-preventive medication, compared with 1.4% patients in the galcanezumab 120 mg arm.¹³ There were no significant differences between placebo and galcanezumab-treated patients with regard to the time of initiation. In the REGAIN study, |||| of patients who entered the post-treatment follow-up phase started a migraine-preventive medication during the post-treatment period.¹³

Health Care Resource Utilization

Health care resource utilization (HCRU) was a secondary outcome in the REGAIN and CONQUER trials. These outcomes were not adjusted for multiplicity. The number of HCRU events was recorded in the baseline period (the 6 months before randomization) and the 3-month double-blind treatment periods. Because low rates of HCRU events were observed and because of the different time periods assessed at baseline and post-baseline, rates for the migraine-related events were standardized per 100 patient-years. Migraine-related HCRU per 100 patient-years is summarized in Table 2.

Work Productivity and Activity Impairment Questionnaire

The Work Productivity and Activity Impairment (WPAI) questionnaire was assessed in the CONQUER trial. The LS mean changes for activity impairment were −20.71% (standard error [SE] = 1.95%) and −8.64% (SE = 1.92%) in the galcanezumab and placebo arms, respectively. For presenteeism, the LS mean changes were −12.50% (SE = 2.37%) and −2.56% (SE = 2.32%) in the galcanezumab and placebo arms, respectively. For overall work impairment, the LS mean changes were −14.31% (SE = 2.51%) and −3.46% (2.41%) in the galcanezumab and placebo arms, respectively. For absenteeism, the LS mean changes were −4.22% (1.29%) and −2.90% (SE = 1.24%) in the galcanezumab and placebo arms, respectively.

Harms Results

Harms results from the double-blind treatment periods of the EVOLVE-1, EVOLVE-2, REGAIN, and CONQUER trials are summarized in Table 2 and Table 3.

Adverse Events

During the double-blind treatment period, most patients in the EVOLVE-1 and EVOLVE-2 trials experienced at least 1 treatment-emergent adverse event (AE), with a numerically smaller proportion of patients experiencing at least 1 AE in the placebo arm compared to the galcanezumab 120 mg arm (60% versus 66%, respectively, in EVOLVE-1; 62% versus 65%, respectively, in EVOLVE-2). The most frequently reported AEs in the EVOLVE-1 trial were injection site pain and nasopharyngitis. The most frequently reported AEs in the EVOLVE-2 trial were injection site pain and upper respiratory tract infections.

During the double-blind treatment period of the REGAIN trial, 50% of patients in the placebo arm and 58% of patients in the galcanezumab 120 mg arm experienced at least 1 treatment-emergent AE. The most frequently reported AEs were injection site pain and nasopharyngitis.

During the double-blind treatment period of the CONQUER trial, 53% of patients in the placebo arm and 51% of patients in galcanezumab 120 mg arm experienced at least 1 treatment-emergent AE. The most frequently reported AEs were nasopharyngitis and influenza.

Serious Adverse Events

Fewer than 3% of patients experienced a serious adverse event (SAE) in the studies.

Withdrawals Due to Adverse Events

As with SAEs, a small percentage of patients (< 4%) discontinued double-blind treatment due to AEs.

Mortality

No patients died during the trials.

Notable Harms

Anaphylaxis and hypersensitivity reactions: No patients in the trials experienced an anaphylactic reaction. In the EVOLVE trials, 2% to 4% of patients in the placebo arms and 4% to 6% of patients in the galcanezumab 120 mg arms experienced a hypersensitivity event. One patient in the placebo arm experienced angioedema in each of the EVOLVE trials. In the REGAIN trial, 2% and 4% of patients experienced a hypersensitivity event in the placebo and galcanezumab 120 mg arms, respectively, during the double-blind treatment phase. Three patients in the placebo arm and 2 patients in the galcanezumab 120 mg arm experienced angioedema during double-blind treatment. In the CONQUER trial, 3% of patients in the placebo arm and 3% in the galcanezumab 120 mg arm experienced a hypersensitivity event during the double-blind treatment phase. One patient in the placebo arm and 0 patients in the galcanezumab 120 mg arm experienced angioedema during double-blind treatment.

Injection site reactions: Injection site reactions during double-blind treatment were reported in a numerically greater proportion of patients in the EVOLVE-1 study compared to the other trials. In the EVOLVE-1 trial, injection site reactions were reported by 20% of patients in the placebo arm and 28% of patients in the galcanezumab 120 mg arm. In the other trials, injection site reactions were reported by 9% to 10% and 7% to 18% of patients in the placebo and galcanezumab 120 mg arms, respectively.

Antibody formation: During the double-blind treatment periods of the EVOLVE-1, EVOLVE-2, and REGAIN studies, up to 9.4% of patients treated with galcanezumab 120 mg and up to 1.7% of patients treated with placebo were positive for treatment-emergent antidrug antibodies (ADAs). The formation of ADAs was not assessed proactively in the CONQUER study and is therefore not reported.

Vascular events: In the EVOLVE trials, approximately 2% and 3% of patients in the placebo and galcanezumab 120 mg arms, respectively, experienced a vascular disorder. In the REGAIN trial, 1.79% and 1.10% of patients in the placebo and galcanezumab 120 mg arms, respectively, experienced a vascular disorder during the double-blind treatment phase. The most frequently reported vascular disorders in EVOLVE-1, EVOLVE-2, and REGAIN were hypertension and hot flushes. In the CONQUER trial, 2.61% and 0.43% of patients in the placebo and galcanezumab 120 mg arms, respectively, experienced a vascular disorder during the double-blind treatment phase. Hypertension was the only vascular disorder experienced by more than 1 patient (1.30% in the placebo arm, 0.43% in the galcanezumab arm).

Critical Appraisal

The pivotal trials were designed to assess the superiority of galcanezumab over placebo. The trials were double-blind and the blinding methods used were appropriate. Although the efficacy analyses used what appeared to be a modified intention-to-treat (ITT) population, which included randomized patients who received at least 1 dose of the study drug, the numerical differences between the randomized (ITT) population and the analyzed population were generally small and unlikely to be a major source of bias for most outcomes. The analyzable sample sizes for the MSQ v2.1 and MIDAS were notably smaller than the randomized sample sizes in each study except CONQUER. There was no discernable difference between treatment groups in any study regarding disproportionate missing patients, but characteristics of the patients who were not included in the analyzable set were not reported and it could not be determined what impact this reduced sample size had on the results for these outcome measures. The safety population included data from all randomized patients who received at least 1 dose of the investigational drug, with analyses conducted based on modal treatment. A numerically greater proportion of patients discontinued from double-blind treatment in the EVOLVE-1 and EVOLVE-2 trials compared to the other trials. Furthermore, the clinical expert consulted by CADTH indicated that the discontinuation rates in the EVOLVE trials were higher than is typically seen in their clinical practice, and higher than expected with anti-CGRP monoclonal antibodies. The most frequent reason for discontinuing double-blind treatment was withdrawal by the patient. Other reasons for discontinuation included lost to follow-up, lack of efficacy, and AEs. Patient compliance with the electronic patient-reported outcomes (ePRO) diary, which was used to collect data for migraine and headache-related end points, was high during the double-blind treatment period across trials. Multiple testing procedures were used in all the trials to control type I error for the primary and key secondary outcomes. Each of the trials met their primary objective for the galcanezumab 120 mg arm, which statistically significantly reduced the overall mean number of monthly MHDs during double-blind treatment. The EVOLVE-1, EVOLVE-2, and CONQUER trials also met all of their key secondary objectives. In the galcanezumab 120 mg arm in the REGAIN trial, only 1 key secondary objective (50% response rate), which was not included in the CADTH systematic review protocol, was statistically significant after multiplicity adjustment. in the REGAIN trial, monthly MHDs with acute medication use and MSQ v2.1 RF-R could not be tested for statistical significance within the pre-specified multiple-testing procedure; thus, these results cannot be interpreted to draw conclusions for this end point.

All trials were placebo-controlled. No direct comparative effect between galcanezumab and other available migraine-preventive treatments (e.g., other CGRP medications) was studied. Patients randomized to the galcanezumab 120 mg arm in the studies received a loading dose of 240 mg, which aligns with the Health Canada–approved dose. The clinical expert consulted by CADTH indicated that the numbers of monthly MHDs and headache days at baseline in the REGAIN and CONQUER trials were lower than would be expected for patients with CM typically treated in regular clinical practice, although it was acknowledged that the patients enrolled in the CM populations of the trials met the criteria for CM (i.e., headaches occurring on 15 or more days per month and MHDs on 8 or more days per month). The CONQUER trial was conducted in a treatment-resistant migraine-patient population (i.e., history of failures to 2 to 4 prior migraine-preventive medication categories due to inadequate efficacy or safety and tolerability), including patients with either EM or CM. This most closely aligns with the patient population specified in the reimbursement request. In contrast, the EVOLVE-1, EVOLVE-2, and REGAIN trials excluded patients who failed to have an efficacy response to at least 3 classes of migraine-preventive treatment, and most patients enrolled in these trials had not previously failed 2 or more prior migraine-preventive treatments. All the pivotal trials excluded patients who had been previously treated with a CGRP inhibitor. In the REGAIN trial, patients could be on 1 stable dose of prophylactic therapy (topiramate or propranolol only). This aligns with regular practice in Canada, according to the clinical expert consulted by CADTH, who indicated that galcanezumab could be used concurrently with other migraine-preventive medications.

Indirect Comparisons

Description of Studies

Two indirect treatment comparison (ITCs) were included. One is the sponsor-submitted ITC,¹⁴ which compared galcanezumab (120 mg and 240 mg) with erenumab (70 mg and 140 mg) in the prophylaxis of EM and CM in patients with a history of prior preventive treatment failures. The other identified by CADTH literature search was a network meta-analysis (NMA) by the Institute of Clinical and Economic Reviews (ICER),¹⁵ which compared galcanezumab with other migraine therapies in the treatment of patients with EM.

Efficacy Results

The sponsor-submitted ITC¹⁴ reported there was insufficient evidence to show a difference between galcanezumab and erenumab for all reported efficacy outcomes in patients with EM or CM who failed at least 2 or at least 3 preventive medications.

The ICER meta-analysis found that patients with EM in the galcanezumab 120 mg group experienced fewer days of acute medication use per month compared with the erenumab 70 mg group. No treatments were favoured when comparing galcanezumab with fremanezumab in all assessed outcomes. Patients in the galcanezumab 120 mg group experienced fewer monthly migraine days when compared with topiramate (200 mg/day and 50 mg/day) and placebo. In terms of 50% response (i.e., 50% reduction in MHDs), the results favoured galcanezumab 120 mg over placebo, with a higher proportion of patients on galcanezumab 120 mg achieving a 50% response.

Harms Results

No data on harms (AEs, SAEs, notable AEs) were reported in either of the ITCs.

Critical Appraisal

For the sponsor-submitted ITC, the main limitations were poor reporting of methods (i.e., details of the literature search, the process of study selection and data extraction, and the risk of bias assessment of individual studies were not clearly described), as well as clinical heterogeneity in the included studies (e.g., the definition of treatment response varied across the included studies). For the ICER meta-analysis, the key limitations were the population was not specifically aligned with the population indicated in the reimbursement request to CADTH, a relatively small number of trials were included for each outcome, clinical heterogeneity (e.g., the definition of the response varied), and the NMA was limited to patients with EM (i.e., no NMA evidence for patients with CM). In addition, outcomes important to patients, such as headache-related disability (as measured by HIT-6, migraine-related disability scores as measured by the MIDAS, work productivity, loss of workdays WPAI), adherence, and HCRU (hospitalizations), were not assessed in either ITC. Due to methodological limitations and clinical heterogeneities across the included studies in both ITCs, the findings from both ITCs do not clearly indicate whether galcanezumab is inferior, similar, or superior to the comparators of interest for this review.

The clinical expert CADTH consulted for this review indicated that, while there is a lack of direct clinical trial evidence, reliable and robust indirect comparison evidence, and clinical experience to compare galcanezumab with other CGRPs (e.g., erenumab or fremanezumab), the mechanism of the different CGRPs suggests it is reasonable to expect similar efficacy and safety among CGRPs approved by Health Canada. It should be noted that, clinically, not all patients respond to any individual CGRP drug, and different CGRP drugs may work for different patients. Galcanezumab may therefore provide an option for patients who do not respond to other CGRP inhibitors.

Other Relevant Evidence

Description of Studies

One long-term study (CGAJ¹⁶) was summarized to provide additional evidence regarding the safety of galcanezumab 120 mg (with a 240 mg loading dose) for patients with EM or CM. The CGAJ study was a multi-centre, phase III, randomized, open-label study. The trial consisted of a screening period, 12 months of open-label treatment with galcanezumab, and 4 months of follow-up. In total, 270 patients with migraine were enrolled and randomized in a 1:1 ratio to either galcanezumab 120 mg (with a 240 mg loading dose) or 240 mg in CGAJ. Because galcanezumab 240 mg is outside of the Health Canada–recommended dose, this summary of the CGAJ study focuses on the results of the 120 mg dose. Patients who received galcanezumab 120 mg had a single initial loading dose of galcanezumab 240 mg (2 injections of galcanezumab 120 mg) then 120 mg for the remainder of the 12-month treatment period. The study drug was administered subcutaneously using a pre-filled syringe (80.9%) or an autoinjector (19.1%), when the latter was available. Study-site personnel administered the first dose while patients and/or caregivers were trained to administer all subsequent doses. The primary outcome was long-term safety and tolerability for 12 months of treatment, which included assessments of the number of AEs, SAEs, and discontinuation rates. Secondary outcomes included long-term efficacy and HRQoL measured by change in MIDAS score, MSQ v2.1 score, and HCRU and employment status.

For patients who received galcanezumab 120 mg, the mean age was 40.2 years (standard deviation [SD] = 11.68), and most patients were female (81.5%) and White (76.3%). This group had a mean baseline MIDAS score of 45.8 (SD = 42.06), which reflects severe disability. In terms of medication history, 60% of patients had tried at least 1 prior preventive treatment, with 43% and 20% having failed at least 1 and 2 prior treatments, respectively.

Efficacy Results

In the galcanezumab 120 mg arm, the overall LS mean change from baseline was −5.61 (95% CI, −6.27 to −4.95) for MHDs and −2.17 (95% CI, −2.76 to −1.58) for headache days. The overall mean change from baseline in use of acute migraine or headache treatment was −5.09 days per month (95% CI, −5.83 to −4.35). The overall LS mean change from baseline was −33.58 (95% CI, −37.73 to −29.42) for the MIDAS total score and 28.27 (95% CI, 25.98 to 30.56) for the MSQ v2.1 total score. In months 7 to 12, the numbers of patients who had at least 1 health care visit or emergency room visit related to migraine were 12 and 3, respectively, compared with 52 and 11 patients at baseline, respectively.

Harms Results

Most patients (82.2%) experienced at least 1 treatment-emergent AE during the treatment phase of the CGAJ trial. The most frequently reported AEs were nasopharyngitis (17.8%), injection site pain (17.1%), injection site reaction (11.6%), sinusitis (10.9%), and back pain (9.3%). Three patients reported having a single SAE in the form of each of the following: lumbar radiculopathy, migraine, and osteoarthritis. In total, 6 patients (4.7%) discontinued the study due to an AE and no deaths were reported.

Notable harms included anaphylactic reaction (6.2%), hypersensitivity reaction (14.7%), injection site reaction (11.6%), and vascular disorder (4.7%). The presence of ADAs was detected in 8 (6.3%) patients at baseline and 16 (12.5%) patients during the treatment phase. A patient was considered ADA-positive when a post-baseline titre was 4 times greater than the baseline value or a post-baseline titre was greater than 1:20 if the baseline ADA tests were negative. Neutralizing antibodies were present for all ADA-positive patients at both baseline and during treatment (n = 8 and n = 16, respectively).

Critical Appraisal

The CGAJ trial did not have a control group. Additionally, the open-label design may have influenced the perception of improvement by patients and clinicians, which could affect the reporting of harms and efficacy measures. All information on migraine frequency and acute medication use was collected retrospectively by direct questioning from study personnel during visits. This form of data collection could affect the accuracy of harms and efficacy reporting, although it is unknown if this would be in favour of or against the study treatment. Limitations in the study design make it challenging to interpret the results and form conclusions with certainty. The CGAJ trial sample size was small. Patients were predominantly female and White, which the clinical expert consulted by CADTH confirmed means they were similar to patients treated in Canadian clinics. The clinical expert consulted by CADTH suggested that, ideally, patients are tried on a medication within the therapeutic range for at least 8 weeks before deciding whether the treatment had failed. The time on treatment (maximum 12 months) was therefore acceptable.

Conclusions

The EVOLVE-1 and EVOLVE-2 trials provide direct evidence regarding the efficacy and safety of galcanezumab relative to placebo for adult patients with EM. The REGAIN trial provides direct evidence regarding the efficacy and safety of galcanezumab relative to placebo for adult patients with CM. The CONQUER trial provides direct evidence regarding the efficacy and safety of galcanezumab relative to placebo in adult patients with EM or CM who have previously failed 2 to 4 classes of migraine-preventive treatments. Compared to placebo, patients who were treated with galcanezumab 120 mg showed benefits in the form of a reduction in monthly MHDs during double-blind treatment periods in all trials (6 months in the EVOLVE-1 and EVOLVE-2 trials; 3 months in the REGAIN and CONQUER trials). In addition, reductions in monthly MHDs with acute medication use and improvement in the MSQ v2.1 RF-R domain were observed in the EVOLVE-1, EVOLVE-2, and CONQUER trials. The effect of galcanezumab on MHDs with symptoms, headache days, other patient-reported outcomes, and HCRU and employment status remains uncertain due to a lack of control for multiplicity. Many study patients reported treatment-emergent AEs. The most frequently reported AEs across the trials were injection site pain and nasopharyngitis. Galcanezumab was related to a numerically higher frequency of becoming ADA-positive in the 3 trials that assessed immunogenicity. Few patients discontinued double-blind treatment due to AEs and few patients experienced SAEs. No patients died. The pivotal trials did not provide direct evidence regarding the relative efficacy and safety of galcanezumab versus other migraine-preventive medications in adults with EM or CM.

The results of the CGAJ trial, a randomized, open-label, long-term safety study of galcanezumab 120 mg and 240 mg, supported the safety of galcanezumab 120 mg. The results also supported the beneficial effect of galcanezumab in terms of monthly MHDs, headache days, use of acute medication, and HRQoL. However, limitations of this study, such as a small sample size and lack of a control group, contribute uncertainty to the results.

Neither the sponsor-submitted ITC nor the ICER meta-analysis provided sufficient evidence to conclude that galcanezumab differed in efficacy compared to other CGRPs in terms of response rate, change from baseline of monthly MHDs, change from baseline of monthly MHDs with acute medication use, discontinuation from all causes, and discontinuation from AEs in patients with CM and/or EM. Due to several limitations of both ITCs (i.e., methodological issues and clinical heterogeneity), no definitive conclusion can be drawn on the clinical efficacy and safety galcanezumab 120 mg compared with erenumab or fremanezumab in the treatment of patients with migraine (EM or CM) who failed at least 2 preventive treatments.

Introduction

Disease Background

Migraine is a neurologic disease characterized by recurrent episodes of pulsating headache pain of at least moderate severity.¹ Migraine episodes may last from 4 to 74 hours and can be accompanied by symptoms such as photophobia, phonophobia, nausea, and vomiting.⁵ The type of migraine can be refined by the frequency of monthly MHDs and monthly headache days.¹ The ICHD-3 describes CM as a headache (tension-type-like or migraine-like) occurring on 15 or more days per month for more than 3 months with the features of migraine headaches on at least 8 days per month.² In EM, individuals experience headaches on 14 or fewer days per month for more than 3 months with the features of migraine headaches on at least 4 days per month.² A diagnosis of migraine is made using a history, physical examination, and neurologic examination.⁵

In Canada (2010 to 2011), 9.6% of the population over 18 years of age experienced migraine attacks, with more females (13.8%) than males (5.3%) having had migraine.³ In a longitudinal web-based panel study of migraine in the US (N = 16,789), 91.2% of patients had EM and 8.8% had CM.¹⁷ An estimated 2.5% of patients with EM transition to having CM.¹⁸

Among those who had migraine in Canada (aged ≥ 15 years, 2011), 38.2% reported that migraine at least moderately affected their life and 25.5% reported that the pain prevented them from activities.¹⁹ In a cross-sectional, web-based observational survey of patients with migraine (N = 8,726), nearly half of all respondents reported moderate or severe disability, with more headache days per month associated with more severe disability.²⁰ Among the respondents, 5.7% had CM and 94.3% had EM.²⁰ Patients with CM reported longer, more painful headaches, and more comorbidities than those with EM.²⁰ Additionally, patients with CM reported worse headache-related disability compared with those with EM, as measured by the MIDAS, which is a validated tool that measures disability in patients with migraine.²⁰ Migraine attacks are often disabling. Headache disorders are among the 3 highest causes of years lived with disability worldwide (1990 to 2017), with migraine accounting for 47,245.4 years lost to disability (thousands) in 2017.²¹

Migraine attacks are associated with missed activities at work, school, and/or at home.¹⁹ Additionally, prevalence is highest during peak productive years (i.e., around 30 to 64 years of age),⁵ which maximizes the impact on the sufferer, family, and society.^19,22-24 Migraine reduces productivity, leading to missed work days and substantial economic costs. Loss of productivity accounts for up to 70% of total migraine-related annual costs.²⁵ In Canada, 34% of individuals with migraine reported limitations in job opportunities due to migraine in 2011; 36% of those currently employed reported missing at least 1 day of work in the past 3 months due to migraine; and 18% who had previously been employed reported that, due to migraine, they had changed their work activities (hours, type of work, or stopped work) for 3 months or longer.¹⁹

Standards of Therapy

Comprehensive migraine therapy includes management of lifestyle factors and triggers, acute and preventive (or prophylactic) medications, and migraine self-management strategies.^4,5 The goals of migraine treatments are to relieve pain, restore function, improve HRQoL, reduce headache frequency, and prevent the progression of EM to CM.⁶ The Canadian Headache Society has guidelines for the acute treatment of migraine and for preventing attacks.⁵

Preventive medications include a variant of the botulinum toxin (onabotulinumtoxinA; for CM only), inhibitors of CGRP (e.g., galcanezumab, fremanezumab, erenumab), beta-blockers (e.g., propranolol, metoprolol), calcium-channel blockers (e.g., flunarizine, verapamil), tricyclic antidepressants (e.g., amitriptyline, nortriptyline), serotonin-norepinephrine reuptake inhibitors (e.g., venlafaxine), anticonvulsants (e.g., topiramate, gabapentin, or divalproex), angiotensin-converting enzyme inhibitors and angiotensin receptor blockers (e.g., candesartan), and a serotonin antagonist (pizotifen). Only topiramate, onabotulinumtoxinA, and CGRP inhibitors have been approved by Health Canada for the prevention of migraine, and onabotulinumtoxinA is only indicated for the prevention of CM. Migraine prophylaxis is an important part of the overall approach for a proportion of individuals with migraine.⁴ Of patients with migraine who have received preventive medications, 87% with migraine have an inadequate response to 2 or more preventive therapies.⁷

Drug

Galcanezumab is a humanized immunoglobin G4 monoclonal antibody that binds to CGRP and prevents its biologic activity.⁸ Galcanezumab does not bind to the CGRP receptor. Galcanezumab is administered by subcutaneous injection. The recommended protocol is a loading dose of 240 mg (administered as 2 consecutive subcutaneous injections of 120 mg) followed by once-monthly, single-injection doses of 120 mg. Galcanezumab received a Health Canada Notice of Compliance on July 30, 2019, with an indication for the prevention of migraine in adults who have at least 4 migraine days per month. Galcanezumab also is indicated for the reduction in the frequency of attacks throughout a cluster period in adults with episodic cluster headache with prior cluster headache periods lasting at least 6 weeks and who have had an inadequate response to, tolerated poorly, or had contraindications to conventional preventive therapies established by Canadian practice guidelines. Galcanezumab has not been previously reviewed by CADTH.

The sponsor requested reimbursement of galcanezumab for the prevention of migraine in adults who have at least 4 MHDs per month and have experienced an inadequate response, intolerance, or contraindication to at least 2 prophylactic migraine medications, which differs from the Health Canada indication.²⁶

Stakeholder Perspectives

Patient Group Input

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

About the Patient Group(s) and Information Gathered

CADTH received a joint submission from Migraine Canada and Migraine Québec for the review of galcanezumab (Emgality). Migraine Canada and Migraine Québec are not-for-profit patient organizations that provide support and education to patients, caregivers, and health care professionals, and raise awareness about the impact of migraine. The groups are advocates for optimal care for those living with migraine and support improved quality of life and research toward a cure.

The information for the patient-group submission was collected through 2 online surveys conducted in 2018 and 2021. The 2018 survey was promoted through Migraine Canada’s Facebook community, Twitter platform, and in migraine clinics across Canada. In total, 597 respondents participated, the majority of whom were between 26 and 54 years old. The survey results indicated that 26% of participants had low-frequency (1 to 6 days per month) EM, 32% had high-frequency (7 to 14 days per month) EM, and 42% had CM (15 or greater days per month). The 2021 survey captured input from 115 participants and was specific to CGRP medications and galcanezumab. In this group, 97% were female, nearly 90% had been diagnosed with migraine, and 73% had experienced at least 15 migraine attacks each month. Both surveys were designed and analyzed by Migraine Canada. Direct input from 2 Canadian patients, a 37-year-old male and a 25-year-old female, was also included in the submission.

Disease Experience

Based on the survey responses, Migraine Canada and Migraine Québec identified the following key impacts on the lives of those living with migraine and their families: inability to work resulting in financial stress and relying on a spouse or family members to compensate, childcare and needing additional help, restricted social activities and difficulties with relationships, and lack of understanding from others.

The patient groups emphasized that many patients take time away from work due to active attacks (ictal state) but then try to compensate by doing more to catch up during the time between attacks (interictal state), which does not allow for rest and recovery. They added that lost productivity should include both missed working days and “presenteeism,” when patients attend work despite having migraine symptoms. When asked to rate how migraine affected their daily lives in the past 3 months, 1% of respondents indicated having no limitations, 24% avoided triggers and missed personal activities but not work, 45% missed at least 1 day of full-time work or household work, 5% were working part-time, and 25% reported being disabled or not working.

Migraine also affected patients’ family members and relationships, with 48% indicating it had a minor impact, 40% noting a major impact, and 9% attributing their lack family or intimate relationships primarily to migraine. The remaining 3% responded that migraine had no impact on their relationships. Patients described how their spouse and/or family often must bear the burden of household and financial responsibilities, how families miss out on spending time together, difficulties with starting new relationships and stress on existing relationships, and the lack of support available for caregivers.

Quality of life is severely affected during interictal states as patients often live in fear of their next attack, avoiding potential triggers, which limits their activities and social interactions, and may be continually managing medication side effects. This last impact is unacceptable to patients aiming to be active and contributing members at work, home, and in their communities. Mental health is also greatly affected by migraine and 80% of respondents stated that they have anxiety or depression.

According to the patient submission, 27% of respondents had been to the emergency department at least 4 times since the start of their condition. Patients noted that living with migraine stressful, and they also felt stigmatized and blamed for wasting health care resources and health care providers’ time when making trips for urgent care.

The following quotes are representative of patient and caregiver experiences of living with migraine:

“It is invisible. It is stigmatised. It isolates and diminishes you. Constant pain symptoms from migraine wear on the body and soul. You are trapped unable to escape or explain. Unable to live your life or enjoy simple moments.”

“Physically and emotionally exhausting! The stress of going to work (nurse) and suddenly having to leave my patients and coworkers is upsetting. The unpredictability of migraine is stressful. The loss of freedom and independence is stressful. This mostly invisible chronic pain is stressful, almost overwhelmingly so.”

“My migraines are the main reason my husband and I do not have children. My husband currently has to work full-time outside the house to support us since I am not working.”

“Chronic migraine has turned my wife's life upside down. She suffers with bouts of depression, anxiety, and even suicidal thoughts. I can't even begin to be imagine the pain she suffers on an almost daily basis.”

Experiences With Currently Available Treatments

Patients reported having experience with triptans, amitriptyline, venlafaxine, topiramate, Emtec, ibuprofen, almotriptan, ketorolac tromethamine, onabotulinumtoxinA, lidocaine, and galcanezumab, along with broader categories of medications, such as tricyclic antidepressants, beta-blockers, calcium-channel blockers, anti-epileptics, CGRP inhibitors, and nonspecific compounds such as cannabidiol oil, magnesium, and riboflavin. Many still struggle to find an effective treatment or access to 1. One patient shared, “I have daily migraine and can barely work. Amitriptyline is not working anymore which is why I am currently looking into injection types of treatment recommended by my neurologist but a lot so [sic] them are not covered by my insurance or pharmacare.”

When specifically discussing CGRP medications, patients noted the advantages of the medications’ convenience and effectiveness with minimal side effects. Unfortunately, costs and access were identified as major limitations: “They are a godsend to those who have had migraine relief with them. The costs makes [sic] them unavailable to many unless they have a private insurance plan. Side effects for some migraineurs make it impossible to take them even if they provide relief” and “Advantages are getting migraines under control with only one dose a month. Feel better in-between migraines. Only disadvantage is access to them and cost.”

The survey responses indicated that 22% had experience with 1 or 2 preventive medications, 22% had tried 3 or 4 drugs, 45% had used at least 5 medications, and the remaining 11% had not tried any. Of those who had tried oral preventive medications, 68% of respondents had experienced side effects leading to discontinuation, 25% reported having tolerable side effects, and 7% stated having no side effects. The most frequently reported side effects included somnolence, dizziness, weight gain, cognitive problems, gastrointestinal issues, and mood problems. One patient reported that, although triptans and opioids were the most effective at controlling pain, a number of side effects followed. With triptans, they experienced burning or tingling of the extremities, muscle weakness, medication overuse that made it less effective or led to medication overuse headaches, frequent urination and urine leakage, extreme irritability, and not wanting to be around other people. The patient also stated that opioids caused drowsiness that prevented them from being able to drive or work.

Not only do patients have to deal with treatment options that inadequately control migraine, but they are also faced with limited access to care. For instance, 27% of patients reported being on wait lists for neurologists or specialists that are more than a year long, and 54% of patients were dissatisfied with the care they have received.

Because there is no cure for migraine, patients are told that a 50% improvement in frequency and intensity is acceptable. Despite this, 74% of respondents did not indicate that they had experienced an improvement of at least a 50%, whereas only 6% indicated that a drug they have tried provided a 75% improvement.

Improved Outcomes

In general, survey respondents reported that nearly any degree of relief would be a successful outcome for a preventive therapy. More specifically, patients indicated that they would prefer a preventive medication that allows them to be more productive at work and home. They also raised concerns over the frequency of administration, potential side effects, affordability, accessibility, and impact on quality of life.

According to 1 patient, the most important aspects of migraine requiring control are the frequency and severity of migraine attacks, as well as reducing or eliminating the need for triptans and opioids. The patient added that it was important that a new medication would allow them to fully participate in daily life, work, improve family and social relationships, and reduce exhaustion and side effects. Another patient had similar interests and added that having more energy on migraine-free days, reducing nausea, and being able to maintain a healthy weight would be important to them.

When asked about methods of administration, 73% of participants preferred a monthly injection to a daily pill. One respondent replied that, “If I had access to a preventive that I could take only once monthly, I would be thrilled.”

Experience With the Drug Under Review

Among the participants in the 2021 survey, 36 had experience with galcanezumab. In general, patients appeared to be pleased with the drug and reported that it effectively reduced migraine severity and frequency, that it was convenient, and that acute medications offered a better response when taken. Despite the benefits, 1 patient stated that “the possibility of having to stop treatment due to cost is very stressful.” Although most responses indicated that patients had minimal to no side effects, dizziness, vertigo, vomiting, constipation, and temporary injection site pain were noted in the patient submission.

The following quotes have been selected to illustrate how galcanezumab has impacted patients’ lives:

“Emgality has been my miracle drug after trialing every other treatment available. While I still have daily headaches, it has helped reduce the severity.”

“This past year on Emgality has been life changing. I am functioning well at work, my sick time has drastically improved. My mood is upbeat, I look forward to life so much more without constantly feeling like I can’t do the things I enjoy. I feel more reliable and pleasant to be around.”

“Benefits were awesome, my migraines severity went down. I went from daily migraine/headache to 9-10 Free Days!! Severity I rank from 0-5 I went from average of 3-5s to 1-3 with only a few 4 or 5. The BIG disadvantage was vertigo. The last month I took Emgality I lost 15 pounds due to vertigo...vomiting and unable to eat.”

“Some increased dizziness in the first week and a lit bit of constipation but very well tolerated.”

Patients also reported that the benefits of treatment with galcanezumab extended to those around them as it allowed them to spend more time with their family and friends and participate in activities together:

“Greatly improved by quality of life which in turn has made my families lives happier.” [sic]

“My husband and my son are elated at the difference because I am able to be more present in our family and participate in activities that were impossible before my first Emgality shot. The symptoms get better with each subsequent injection as well.”

The 2 patients who provided direct input to the patient groups for this submission had experience with galcanezumab. One, a 37-year-old male, gained access through a prescription from a neurologist and with help from Eli Lilly patient support coverage. Treatment over the past year reportedly reduced migraine frequency from daily to only a couple of attacks per month without any reduction in effectiveness and allowed him to “reduce his effective dose of triptan to one half or one quarter” of his previous dose “to get rid of the migraine which has resulted in less side effects from the triptan.” Similar to the other experiences described, administration was convenient and side effects were minimal. It was noted that the possibility of not being able to access or afford treatment in the future has caused him to have severe anxiety.

The other patient, a 25-year-old female, accessed galcanezumab through the patient support program. Although she reported that she does not like needles, she found the drug to be convenient and easy to self-administer. For the first 6 to 7 months, there was minor irritation at the injection site but she has not experienced it since. After the first administration of galcanezumab, there was an improvement within a few days and, “Since the first month, after injection I would … have 3 weeks with only a couple migraines and by 4th week I’d start to experience an increase in migraine frequency and a few days after injection, migraine’s free up and don’t see same frequency. This pattern has held for 8 months.” She reported that “being able to increase dosage would maybe be helpful in reducing frequency.” Overall, the effects have been positive, and she shared that:

Emgality has given me back my life. I can take care of myself. I can cook. I can go out and socialize with friends. I feel like I actually have a life versus being bedridden and stuck at home.” and “My mental health is better now because she is not in pain all the time, I have more energy, less fatigue. I have a better perspective on life, and am more hopeful about the future.

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 1 clinical specialist with expertise in the diagnosis and management of migraine.

Unmet Needs

According to the clinical expert consulted by CADTH, the most important goal of treatment is to reduce the frequency of headaches. In the clinical expert’s experience, patients have an approximately 20% chance of responding well to the first preventive medication tried. As a result, a trial of 2 to 3 preventive medications is often required before a patient experiences benefit. The clinical expert noted that often the medications have not been given in the correct dose or for a sufficient period of time to make a valid judgment as to whether they were effective or not. Ideally, patients are within the therapeutic range for at least 8 weeks before deciding whether a treatment has failed. In the clinical expert’s experience, approximately 10% to 20% of patients have not tried any preventive medications when referred to their headache clinic, 40% have tried 1 to 2 medications, and most others have tried 2 to 3 medications. Only a small proportion of Canadian patients with migraine have tried 4 or more medications, in the clinical expert’s experience.

There are limitations with the older oral medications used to prevent migraine. The clinical expert reported that many patients cannot take beta-blockers or calcium-channel blockers because these medications cause hypotension. Weight gain is a problem with gabapentin, amitriptyline, divalproex, sodium valproate, and nortriptyline. Somnolence and dry mouth limit the use of amitriptyline for many patients. Cognitive slowing and word-finding difficulty often limit topiramate use. In addition, the clinical expert reported that adherence is an issue with oral preventive medications that are taken daily, even when the patient is receiving a benefit. Furthermore, cost is a barrier for patients who might benefit from anti-CGRP monoclonal antibodies and onabotulinumtoxinA.

Place in Therapy

The clinical expert consulted by CADTH indicated that 1 of the first lines of treatment for EM and CM is to identify and eliminate or reduce factors that trigger headaches. In addition, cognitive behavioural therapy and mindfulness can be help patients manage their migraine headaches.

For headaches that occur more than 4 times per month, the clinical expert reported that a variety of preventive medications are used. Prescription medications to prevent headaches include beta-blockers (e.g., propranolol, nadolol, metoprolol, atenolol), calcium-channel blockers (e.g., verapamil, flunarizine), angiotensin-converting enzyme inhibitors (e.g., lisinopril), angiotensin receptor blockers (e.g., candesartan), tricyclic antidepressants (e.g., amitriptyline and nortriptyline), serotonin-norepinephrine reuptake inhibitors (e.g., venlafaxine), anticonvulsants (e.g., topiramate, gabapentin, and divaloproex or sodium valproate), and anti-CGRP monoclonal antibodies (e.g., erenumab, fremanezumab, galcanezumab, and eptinezumab). For CM (headaches occurring 15 days per month, with 8 meeting the criteria for migraine), onabotulinumtoxinA injections may be used.

Before the advent of anti-CGRP monoclonal antibodies, patients often tried 4 to 6 daily preventers. If they had coverage, they would try 2 before receiving either onabotulinumtoxinA (if they have CM) or an anti-CGRP monoclonal antibody. The clinical expert noted that anti-CGRP monoclonal antibodies such as galcanezumab would be a first-line treatment in practice, but, for reasons of cost-effectiveness and funding status, these are typically used in patients who first fail on 2 or more oral preventive medications.

The clinical expert consulted by CADTH indicated that galcanezumab could be used concurrently with other migraine-preventive medications if the patient had a partial response from a daily preventer. In such cases, the clinician would keep the patient on the medication if they were not having side effects and add an anti-CGRP monoclonal antibody. The clinical expert noted that anti-CGRP monoclonal antibodies inhibit CGRP activity on nerves that are different from those affected by onabotulinumtoxinA. The clinical expert indicated that there may be mechanistic reasons to consider using both medications at the same time in some patients.

Patient Population

According to the clinical expert, patients with frequent EM without medication overuse would be the most likely to respond to an anti-CGRP monoclonal antibody. Patients with severe and/or CM may realize the greatest benefit from galcanezumab. However, the clinical expert noted that many patients with severe migraine are disabled or do not have good drug coverage, and therefore do not currently have access to anti-CGRP monoclonal antibodies. The clinical expert reported that there are no disease characteristics that signal who is more likely to respond to treatment with galcanezumab.

The clinical expert reported that suitability for galcanezumab treatment is determined based on patient history. Those with a diagnosis of migraine with or without aura, more than 4 headache days per month, and failure on 2 or more daily preventers used at an appropriate dose for an appropriate period of time, and no contraindication to the use of an anti-CGRP monoclonal antibody would be suitable for galcanezumab. No tests are necessary to determine suitability for treatment. The clinical expert noted that misdiagnosis of migraine in clinical practice would be unlikely.

Patients least suitable for treatment include women contemplating pregnancy. In addition, it would be unlikely for patients with a needle phobia to be able to inject themselves, although the clinical expert noted that autoinjectors can make self-injection easier. The clinical expert indicated there would be hesitancy to prescribe an anti-CGRP monoclonal antibody to patients with known active peripheral vascular, cardiovascular, or cerebrovascular disease. The clinical expert reported that colleagues in the US prescribe anti-CGRP monoclonal antibodies to patients after a year of stability following a cardiovascular event or onset of morbidity. In addition, use of this family of medications in patients with Raynaud phenomenon who were taking triptans has been associated with digital ischemia leading to digit amputation, and the clinical expert indicated that caution is needed in these patients.

Assessing Response to Treatment

When assessing response to treatment, the clinical expert indicated that it is common practice to ask the patient for their impression (i.e., Do they like the medication and do they want to stay on it?).

Reduction in frequency and/or severity of headaches is important, which can be assessed using headache diaries. The clinical expert reported that the goal of treatment is to reduce the frequency of headache, ideally to fewer than 4 headache days per month. A 50% or greater reduction in headache frequency is also considered a successful response. The clinical expert noted that medications may only reduce the severity of headaches as measured by MIDAS or HIT-6 scores, but not the frequency, which is also considered a successful response. Reduced use of abortive medications (e.g., triptans) may also be indicative of improvement. The clinical expert reported that a change within 2 days of acute medication use is clinically significant.

In addition, the clinical expert reported that improvement in function is important. Improvement in the MIDAS and HIT-6 scores are reflective of improvement in the patient’s life. A reduction of 6 points on the HIT-6 for EM and 5 points for CM were considered significant improvements by the clinical expert. A reduction in the MIDAS score by 50% or by 5 points was considered significant.

The clinical expert reported that patients should be assessed for response to galcanezumab treatment 2 to 3 weeks after their third injection, and those who have not shown improvement at that time should be assessed 2 to 3 weeks after their sixth injection. If there is no improvement after 6 injections, then treatment would be discontinued, according to the clinical expert.

Discontinuing Treatment

The clinical expert consulted by CADTH indicated that the following would lead to discontinuation of treatment: intolerable side effects, development of comorbidities that preclude the patient being on treatment (e.g., stroke or heart attack), pregnancy, and failure to respond after 6 injections.

Prescribing Conditions

The clinical expert indicated that all physicians and nurse practitioners can make a diagnosis of EM and CM. Galcanezumab would be prescribed by headache specialists, general neurologists, family doctors, or nurse practitioners on the advice of a neurologist, and family doctors or nurse practitioners who have experience with the use of the medication in other patients. The clinical expert cautioned against requiring patients to be supervised by a headache specialist because there are too few such specialists in Canada, and this could prevent patients from accessing treatment. The clinical expert added that it is within the scope of practice for a general neurologist, nurse practitioner, or family physician to supervise management after a recommendation for use is made. Patients would self-administer the medication at home.

Clinician Group Input

No input from clinician groups was received for this review.

Drug Program Input

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

The drug programs noted that 2 other CGRP inhibitors (erenumab and fremanezumab) recently received positive reimbursement recommendations under the CADTH review process and were the subjects of pan-Canadian Pharmaceutical Alliance negotiations at the time of this review. The drug programs noted that galcanezumab has the same Health Canada indication as both fremanezumab and erenumab, although the reimbursement requests for the 3 medications differ. The drug programs noted that the reimbursement request for galcanezumab is for patients who have tried at least 2 prophylactic migraine medications, and that it would be helpful to jurisdictions for this to be defined further. In addition, the drug programs indicated that it would be helpful to outline whether failure on or intolerance to another CGRP inhibitor would exclude patients from coverage of galcanezumab.

Clinical Evidence

The clinical evidence included in the review of galcanezumab 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 a sponsor-submitted safety study.

Systematic Review (Pivotal and Protocol-Selected Studies)

Objectives

To perform a systematic review of the beneficial and harmful effects of galcanezumab (120 mg/mL solution in a 1 mL single-use pre-filled syringe or pen) for the prevention of migraine in adults who have at least 4 migraine days per month and have experienced an inadequate response, intolerance, or contraindication to at least 2 prophylactic migraine medications.

Methods

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

The literature search for clinical studies was performed by an information specialist using a peer-reviewed search strategy according to the Peer Review of Electronic Search Strategies checklist.²⁸

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 multi-file search. Duplicates were removed using Ovid deduplication for multi-file searches, followed by manual deduplication in Endnote. The search strategy comprised both controlled vocabulary, such as the National Library of Medicine’s MeSH (Medical Subject Headings), and keywords. The main search concepts were Emgality (galcanezumab). Clinical trials registries included the US National Institutes of Health’s clinicaltrials.gov, 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. Appendix 1 provides detailed search strategies.

The initial search was completed on June 29, 2021. Regular alerts updated the search until the meeting of the CADTH Canadian Drug Expert Committee on October 27, 2021.

Grey literature (literature that is not commercially published) was identified by searching relevant websites from the Grey Matters: A Practical Tool For Searching Health-Related Grey Literature checklist.²⁹ 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. Appendix 1 provides more information on the grey literature search strategy.

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.

A focused literature search for NMAs dealing with migraine was run in MEDLINE All (1946–) on June 29, 2021. No limits were applied to the search.

Findings From the Literature

A total of 24 reports presenting data from 4 unique studies from the literature were identified for inclusion in the systematic review (Figure 1). The included studies are summarized in Table 7. A list of excluded studies is presented in Appendix 2.

Figure 1: Flow Diagram for Inclusion and Exclusion of Studies

Description of Studies

Four phase III, multi-national, double-blind, randomized, placebo-controlled trials were included in the systematic review: EVOLVE-1,⁹ EVOLVE-2,¹⁰ REGAIN,¹¹ and CONQUER.¹² In all studies, patients were randomized using an interactive web response system.

EVOLVE-1 and EVOLVE-2

The EVOLVE-1 and EVOLVE-2 trials were identically designed studies conducted in patients with EM.^9,10 The primary objective of the EVOLVE-1 and EVOLVE-2 studies was to test the hypothesis that at least 1 dose of galcanezumab (120 or 240 mg) is superior to placebo in the prevention of migraine headaches in patients with EM. The EVOLVE-1 trial had 90 sites in the US and Canada. The EVOLVE-2 trial had 109 study sites in 11 countries in Europe, South America, Asia, and the US. In both studies, patients were randomized in a 2:1:1 ratio to placebo, galcanezumab 120 mg (loading dose of 240 mg), or galcanezumab 240 mg. In the EVOLVE-1 trial, randomization was stratified by geographic region (eastern half of the US, western half of the US, Puerto Rico, and Canada) and migraine frequency at baseline (< 8 versus ≥ 8 MHDs per month). In the EVOLVE-2 trial, randomization was stratified by country and migraine frequency at baseline (< 8 versus ≥ 8 MHDs per month). The EVOLVE-1 trial randomized 862 patients: 433 to placebo, 213 to galcanezumab 120 mg, and 212 to galcanezumab 240 mg. The EVOLVE-2 trial randomized 922 patients: 461 to placebo, 231 to galcanezumab 120 mg, and 223 to galcanezumab 240 mg. Efficacy and harms data will not be presented for the galcanezumab 240 mg arm because this dose is not aligned with the Health Canada–approved dose.

The study design of the EVOLVE-1 and EVOLVE-2 trials is summarized in Figure 2. The studies comprised 4 study periods: screening; a prospective 30- to 40-day baseline phase to determine patient eligibility; a 6-month double-blind treatment phase; and a 4-month post-treatment follow-up phase. During the prospective baseline period, patients logged in daily to the ePRO system to answer questions about the occurrence of headaches, headache duration, headache features, severity of headaches, and use of headache medication. The purpose of the prospective baseline period was to confirm that the patient had between 4 and 14 MHDs and at least 2 migraine attacks during the 30- to 40-day period, and to establish baseline data for the comparison of end points during the treatment phase.

The data cut-off date for the main reports on the EVOLVE-1 trial was March 22, 2017, which was the last patient-visit date of the double-blind phase. The database was locked for analysis on April 28, 2017. The data cut-off date for the reports on the EVOLVE-2 trial was March 29, 2017, which was the last patient-visit date of the double-blind phase. The database was locked for analysis on May 5, 2017. According to the Clinical Study Reports, a data lock date after the data cut-off date was chosen because additional time was needed to resolve and close all data queries, process laboratory samples, and complete sign-off by investigators. These analyses are considered the final analyses of the primary and key efficacy end points during the double-blind treatment phase. In addition, Clinical Study Report addenda reporting data from the post-treatment phases were provided by the sponsor following CADTH’s request for additional information. The EVOLVE-1 Clinical Study Report Addendum was dated December 18, 2017, and the last patient completed the study on July 19, 2017. The EVOLVE-2 Clinical Study Report Addendum was dated December 18, 2017, and the last patient completed the study on August 4, 2017.

Figure 2: EVOLVE-1 and EVOLVE-2 Study Design

LY2951742 = galcanezumab; SP = study period.

* Eligibility period was a minimum of 30 days and a maximum of 40 days in length. Investigators may have had up to 5 additional days (beyond the 40 days) if needed to schedule a patient’s visit 3 appointment.

^a Patients randomized to the 120 mg dose received a loading dose of 240 mg at the first injection only (visit 3).

Source: EVOLVE-1 Clinical Study Report⁹ and EVOLVE-2 Clinical Study Report.¹⁰

REGAIN

The REGAIN trial was conducted in patients with CM.¹¹ The primary objective was to test the hypothesis that at least 1 dose of galcanezumab (120 or 240 mg) is superior to placebo in the prevention of migraine headaches in patients with CM. Patients were enrolled from 116 study sites in 12 countries in North America (including Canada), Europe, South America, and Asia. A total of 1,117 patients were randomized in a 2:1:1 ratio to placebo (N = 558), galcanezumab 120 mg (loading dose of 240 mg; N = 278), or galcanezumab 240 mg (N = 277). Randomization was stratified by country, acute headache medication overuse (yes versus no), and use of concurrent migraine prophylactic medication (yes versus no). Efficacy and harms data will not be presented for the galcanezumab 240 mg arm because this dose is not aligned with the Health Canada–approved dose.

The study design of the REGAIN trial is summarized in Figure 3. The study comprised 5 study periods: screening; a prospective 30- to 40-day baseline phase to determine patient eligibility; a 3-month double-blind treatment phase; an optional 9-month open-label extension phase; and a 4-month post-treatment follow-up phase. During the prospective baseline period, patients logged in daily to an ePRO system to answer questions about the occurrence of headaches, headache duration, headache features, severity of headaches, and use of headache medication. This prospective baseline period was to confirm that the patient had at least 15 headache days during the 30- to 40-day period, 8 of which must have had the features of a migraine headache, as well as to establish baseline data for comparison of end points during the double-blind treatment phase.

The data cut-off date for the reports on the REGAIN trial was March 16, 2017, which was the last patient-visit date of the double-blind phase. The database was locked for analysis on May 5, 2017. According to the Clinical Study Report, the data lock date was after the data cut-off date to allow for additional time needed to resolve and close all data queries, process laboratory samples, and complete sign-off by investigators. This analysis is considered the final analysis of the primary and key efficacy end points during the double-blind treatment phase. In addition, a Clinical Study Report Addendum that contained the final results of the open-label treatment and post-treatment phases was provided by the sponsor following CADTH’s request for additional information. This addendum was dated February 5, 2019, and the last patient completed the study on May 1, 2018.

Figure 3: REGAIN Study Design

LY2951742 = galcanezumab; OLE = open-label extension; SP = study period.

* Eligibility period determined between a minimum of 30 days and maximum of 40 days. Investigators may have had up to 5 additional days (beyond the 40 days) if needed to schedule a patient’s visit 3 appointment.

^a Patients randomized to the 120 mg dose received a loading dose of 240 mg at the first injection only (visit 3).

^b At visit 7, all patients who entered the open-label extension received galcanezumab at a dose of 240 mg.

^c At visit 8, all patients received galcanezumab at a dose of 120 mg.

^d Starting at visit 9, dosing was flexible (galcanezumab 120 or 240 mg) at the discretion of the investigator.

Source: REGAIN Clinical Study Report.¹¹

CONQUER

The CONQUER trial was conducted in patients with EM or CM who had 2 to 4 migraine-preventive medication category failures due to inadequate efficacy or tolerability in the previous 10 years.¹² The primary objective of the CONQUER trial was to test the hypothesis that galcanezumab is superior to placebo in the prevention of migraine in patients with treatment-resistant migraine. Patients were enrolled from 64 sites in 12 countries in North America (including Canada), Europe, and Asia. A total of 463 patients were randomized in a 1:1 ratio to placebo (N = 230) or galcanezumab 120 mg (loading dose of 240 mg; N = 232). One patient who was a screening failure was inadvertently randomized and immediately discontinued. Randomization was stratified by country and migraine frequency from the prospective baseline period (low-frequency episodic versus high-frequency episodic versus chronic). The sponsor stopped enrolment of patients with CM when the number of patients exceeded approximately 40% of the planned sample size, which was planned per-protocol.

The study design of the CONQUER trial is summarized in Figure 4. The study comprised 4 study periods: screening, a 30- to 40-day prospective baseline period, a 3-month double-blind treatment phase, and an optional 3-month open-label treatment phase. During the prospective baseline period, patients logged in daily to an ePRO system to answer questions about the occurrence of headaches, headache duration, headache features, severity of headache, and use of headache medication. This prospective baseline period was to confirm that the patient had at least 4 MHDs and at least 1 headache-free day per 30-day period, and to establish baseline data for comparison of end points during the double-blind treatment phase.

The data cut-off date for the reports on the CONQUER trial was June 19, 2019. The reporting database was validated and locked for analysis on July 22, 2019. The reason the data lock date was after the data cut-off date was not reported in the Clinical Study Report. This analysis is considered the final analysis of the primary and key efficacy end points, as well as all other efficacy and safety analyses of the double-blind treatment phase of the CONQUER trial. In addition, a Clinical Study Report Addendum containing the final results of the open-label treatment phase was provided by the sponsor following CADTH’s request for additional information. This addendum was dated December 18, 2019, and the last patient completed the study on September 19, 2019.

Figure 4: CONQUER Study Design

SP = study period.

^a Eligibility period was a minimum of 30 days and maximum of 40 days in length, with up to 5 additional days to schedule randomization visit, if necessary.

^b Patients randomized to galcanezumab 120 mg received a loading dose of 240 mg at the first injection only (visit 3).

^c Patients randomized to placebo who entered the open-label treatment phase received a loading dose of galcanezumab 240 mg at the first injection only of study period IV.

^d First injection of the open-label treatment phase occurred at visit 6 once all study procedures for the double-blind phase were completed.

Source: CONQUER Clinical Study Report.¹²

Populations

Inclusion and Exclusion Criteria

The EVOLVE-1 and EVOLVE-2 trials included adult patients (18 to 65 years of age) with a diagnosis of EM as defined by IHS ICHD-3 beta guidelines (1.1 or 1.2), with a history of migraine headaches for at least 1 year before screening, migraine onset before age 50, and a monthly frequency of 4 to 14 MHDs. The EVOLVE-1 and EVOLVE-2 trials excluded patients who had previously failed to have an efficacy response to 3 or more classes of migraine-preventive treatments, and those who were currently receiving medication or other treatments for the prevention of migraine headaches.

The REGAIN trial included adult patients (18 to 65 years of age) with a diagnosis of CM as defined by IHS ICHD-3 beta guidelines (1.3), with a history of migraine headaches for at least 1 year before screening, and migraine onset before age 50. Patients who had been on a stable dose of either topiramate or propranolol for at least 2 months before beginning the prospective baseline period could continue to take that preventive medication throughout the trial. Patients who had previously failed to have an efficacy response to 3 or more classes of migraine-preventive treatments were excluded.

The CONQUER trial included adult patients (18 to 75 years of age) with a diagnosis of EM or CM; a history of migraine headaches for at least 1 year before screening with onset before age 50; a history of at least 4 MHDs with at least 1 headache-free day per month on average within the past 3 months; and documentation of 2 to 4 migraine-preventive medication category failures in the past 10 years. Specifically, patients were required to have documentation of previous failure of 2 to 4 migraine-preventive medication categories from the following list due to inadequate efficacy (maximum tolerated dose for at least 2 months) and/or safety or tolerability reasons: propranolol or metoprolol, topiramate, valproate or divalproex, amitriptyline, flunarizine, candesartan, botulinum toxin A or B (if it was documented that botulinum toxin was taken for CM), and medication locally approved for prevention of migraine. The CONQUER trial excluded patients who had previously failed more than 4 migraine-preventive medication categories in the past 10 years from the list in the inclusion criteria due to inadequate efficacy. Patients currently receiving medication or other treatments for the prevention of migraine headaches were also excluded.

All 4 pivotal trials excluded patients with prior exposure to a CGRP antibody.

Baseline Characteristics

Demographic and baseline characteristics of patients enrolled in the EVOLVE studies, REGAIN study, and CONQUER study are summarized in Table 8, Table 9, and Table 10, respectively. In all trials, most patients were female and White, and the mean number of comorbidities other than migraine was 3 to 4.

In the EVOLVE studies, the mean age of patients was between 39 and 42 years and most patients (> 60%) had received prior preventive treatment. The mean baseline MIDAS total scores were 33.2 and 33.0, which reflect severe disability, in the EVOLVE-1 and EVOLVE-2 studies, respectively. In the EVOLVE studies, approximately 60% to 68% of patients had used at least 1 prior preventive therapy. The most frequently reported prior medications were topiramate, ibuprofen, Thomapyrin N, sumatriptan, paracetamol, propranolol, and amitriptyline. In the REGAIN trial, the mean age of patients was 41.0 years. Most patients (77.8%) in the REGAIN study reported using a prior migraine-preventive treatment, with 29.5% having failed 2 or more such treatments in the past 5 years due to lack of efficacy. In the REGAIN trial, the most frequently used medications were topiramate, amitriptyline, propranolol, ibuprofen, botulinum toxin type A, sumatriptan, paracetamol, Thomapyrin N, valproate sodium, amitriptyline hydrochloride, and nortriptyline. The mean baseline MIDAS total score in the REGAIN trial was 67.2, which reflects very severe disability. Overall, 15.5% of patients in the REGAIN trial had concurrent prophylaxis use with topiramate or propranolol. In the CONQUER trial, the mean age of patients was 45.8 years. Most patients had 2 (58.2%) or 3 (30.1%) prior medication category failures and the mean baseline total MIDAS score was 50.93, which reflects very severe disability. In the CONQUER trial, the most common medication categories failed in past 10 years were topiramate, amitriptyline, propranolol or metoprolol, valproate or divalproex, botulinum toxin A or B, candesartan, and flunarizine.

For all studies, the baseline for efficacy outcomes was derived from headache information collected using the ePRO system during the prospective baseline period. During the baseline period, the mean number of monthly MHDs was 9.1 in both the EVOLVE-1 and EVOLVE-2 studies. In the REGAIN trial, patients had an average of 21.4 headache days per month in the baseline period, of which an average of 19.4 were MHDs. During the baseline period in the CONQUER trial, patients had an average of 15.0 headache days per month, of which an average of 13.2 were MHDs.

Aside from some small demographic differences, baseline characteristics were generally similar between groups within studies.

Interventions

Administration of Study Drug

Double-Blind Treatment phase

Each of the EVOLVE-1, EVOLVE-2, and REGAIN trials had 3 study arms: placebo, galcanezumab 120 mg (with a loading dose of 240 mg), and galcanezumab 240 mg. The CONQUER trial had 2 study arms: placebo and galcanezumab 120 mg (with a loading dose of 240 mg).

In all trials, the galcanezumab or matching-administration placebo injections were administered by study-site personnel once monthly at dosing visits.^12,50 Galcanezumab and matching-administration placebo were supplied as an injectable solution in 1 mL, single-dose, pre-filled, disposable manual syringes. Each syringe of galcanezumab was designed to deliver galcanezumab 120 mg. The syringes (and their contents) containing either galcanezumab or placebo were visibly indistinguishable from each other. Subcutaneous injection sites included the abdomen, thigh, upper arm, or buttocks.

In the EVOLVE-1, EVOLVE-2, and REGAIN trials, all treatment groups received 2 injections of the study drug at each dosing visit to maintain the blind (2 placebo injections, 2 galcanezumab 120 mg injections, or 1 placebo injection and 1 galcanezumab 120 mg injection). In the CONQUER study, patients in both treatment groups received 2 injections (2 placebo injections or 2 galcanezumab 120 mg injections) at the loading-dose visits during the double-blind and open-label phases to maintain the blind and 1 injection at subsequent visits.

Open-Label phase

The EVOLVE-1 and EVOLVE-2 studies did not include an open-label treatment phase.

In the REGAIN trial, doses administered were galcanezumab 240 mg at the first visit in the open-label phase (visit 7), galcanezumab 120 mg at the next visit (visit 8), and either 120 or 240 mg, thereafter, at the discretion of the investigator. Dosing and dose changes could only occur at the regular once-monthly visits.

At the first visit in the open-label treatment phase (visit 6) in the CONQUER trial, patients in the galcanezumab 120 mg arm received 120 mg galcanezumab (1 injection of 120 mg plus 1 injection of placebo to maintain the blind) and patients in the placebo arm received a loading dose of 240 mg (2 injections of 120 mg). At the subsequent dosing visits in the open-label phase, all patients received 120 mg galcanezumab (1 injection).

Treatment Duration

In the EVOLVE studies, the double-blind treatment period was 6 months (6 administrations). In the REGAIN study, the double-blind treatment period was 3 months (3 administrations) and the optional open-label treatment phase was 9 months (9 administrations). In the CONQUER trial, the double-blind treatment phase was 3 months (3 administrations) and the optional open-label treatment phase was 3 months (3 administrations).

Concomitant Medications

In all trials, acute (abortive) medications were allowed and recorded in the patient ePRO diaries.^9-12 These included acetaminophen (paracetamol), nonsteroidal anti-inflammatories, triptans, ergotamine and derivatives, isometheptene mucate, dichloralphenazone and acetaminophen combination (Midrin), or combinations thereof.

In the EVOLVE-1, EVOLVE-2, and REGAIN studies, the use of concomitant medication for acute treatment of migraine was subject to some limitations, such as a restriction on the use of opioid- and barbiturate-containing medications for no more than 3 days per month, and a single dose of injectable steroids was allowed only once during the study, in an emergency setting. For patients who completed the double-blind treatment phase of the study, treatments used for the prevention of migraine were allowed 1 month after the patient entered the post-treatment phase, if clinically warranted due to a worsening of symptoms.

In the REGAIN trial, treatments used for the prevention of migraine were generally not allowed during the double-blind and open-label treatment phases. However, the study allowed approximately 1-third of enrolled patients to continue migraine prophylactic treatment with either topiramate or propranolol if the patient had been on a stable dose for at least 2 months before the prospective baseline period and if dosing was expected to remain stable throughout the double-blind treatment phase.

In the CONQUER trial, the concomitant use of acute medications to treat migraine was allowed, with some limitations, such as restricting the use of opioid- and barbiturate-containing medications to no more than 4 days per month, and a single dose of injectable steroids was allowed only once during the study, in an emergency setting. Treatments used for the prevention of migraine, including nutraceuticals and nonpharmacological interventions, were not allowed at any time during the prospective baseline, double-blind treatment, and open-label treatment phases of the study. Patients were to wash out all migraine-preventive treatments at least 5 days before beginning the prospective baseline period (visit 2). Botulinum toxin A or B in the head or neck area for therapeutic use was not allowed within 3 months before visit 2. Nerve blocks or use of therapeutic devices, such as transcranial magnetic stimulation, in the head or neck area or for migraine prevention, were not allowed within 30 days before visit 2.

In all studies, concomitant medications used for the acute treatment of migraine were recorded by the patient in the ePRO diary and all other concomitant medications taken during the study were recorded via electronic case report forms.

Discontinuation Criteria

In all studies, patients were to be discontinued from the study if the investigator or patient decided the patient should be withdrawn.

In EVOLVE-1, EVOLVE-2, and REGAIN, discontinuation of the study drug for abnormal liver tests was required when a patient met 1 of the following conditions and the event was at least possibly related to the study drug: alanine aminotransferase (ALT) or aspartate aminotransferase (AST) levels were 8 times the upper limit of normal (ULN); ALT or AST levels were greater than 5 times the ULN for more than 2 weeks; ALT or AST levels were greater than 3 times ULN and total bilirubin levels were greater than 2 times the ULN or prothrombin time was greater than 1.5 times the ULN; ALT or AST was greater than 3 times the ULN, with the appearance of fatigue, nausea, vomiting, right upper-quadrant pain or tenderness, fever, rash, and/or eosinophilia (> 5%); alkaline phosphatase levels were greater than 3 times the ULN; and alkaline phosphatase levels were greater than 2.5 times the ULN and total bilirubin was greater than twice the ULN; alkaline phosphatase was greater than 2.5 times the ULN, with the appearance of fatigue, nausea, vomiting, right-quadrant pain or tenderness, fever, rash, and/or eosinophilia (> 5%). Discontinuation of study treatment was also required in the following cases, if the event was at least possibly related to the study drug: serious allergic reaction to the study drug, serious injection site reaction, serious event of suicidality or depression, and serious cerebrovascular event.

In the CONQUER trial, discontinuation of the study drug for abnormal liver tests was to be considered by the investigator, in consultation with the sponsor. Patients must have been discontinued from the study drug if the patient requested to discontinue, in the event of pregnancy, or if the investigator decided the patient should be withdrawn.

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 11. These end points are summarized in the following section. A detailed discussion and critical appraisal of the outcome measures is provided in Appendix 4.

Efficacy Outcomes

Assessments of migraine and headache-related end points (i.e., MHDs with symptoms, number of monthly MHDs, number of monthly headache days, and acute headache pain medication intake) were based on headache information captured in ePRO diaries. Patients were asked to use a daily ePRO diary to record headache information (e.g., characteristics, duration, and severity) and whether any acute headache medication was taken. The system also was used to collect information about migraine-associated symptoms (e.g., photophobia, phonophobia, nausea, and/or vomiting). Based on patient responses in the daily ePRO diaries, days were defined as no headache, headache, non-migraine headache, ICHD-3 migraine headache, and MHDs (primary measure) using an automated algorithm according to the definitions in Table 12. Each month was defined as a 30-day period with migraine or headache measures normalized to a 30-day period from the actual visit intervals.

In the EVOLVE-1, EVOLVE-2, and REGAIN trials, time to first loss of response among patients who met the 50% response criteria at the end of their treatment interval and time to initiation of treatment with a migraine-prevention medication were assessed in the post-treatment follow-up periods. A 50% response was defined as a reduction of least 50% from baseline in MHDs during the month. Time to first loss of 50% response was analyzed among patients who met the 50% response criterion during the last month of treatment and also entered the post-treatment phase. Patients who discontinued or completed the study without loss of 50% response were censored in the time-to-event analysis of first loss of response.

A detailed discussion of the MSQ v2.1, MIDAS, HRCU employment status, EQ-5D 5-Levels questionnaire (EQ-5D-5L), WPAI questionnaire, Patient Health Questionnaire-9 (PHQ-9), and Generalized Anxiety Disorder 7-item (GAD-7) scale is provided in Appendix 4.

The MSQ v2.1 is a self-administered questionnaire that addresses the physical and emotional impact of migraine on functioning with a 4-week recall period.^9-12 It is a 14-item instrument covering 3 domains: RF-R (impact on performance of normal activities), role function – preventive (RF-P; complete functional impairment), and emotional function (EF; feelings related to disabling migraine). Responses are captured using a 6-point Likert scale, ranging from “none of the time” to “all of the time.” The total raw score is the sum of responses for each domain, which is converted to a 0-to-100 scale. Higher scores indicate a better health status, and an increase in score reflects improvement in HRQoL. A literature by CADTH identified the following minimal important differences (MIDs): 3.2 to 5.0 for RF-R, 4.6 to 7.9 for RF-P, and 7.5 to 10.6 for EF in patients with EM, and 10.9 for RF-R, 8.3 for RF-P, and 12.2 for EF in patients with CM.^52,53

The MIDAS is a 5-item instrument that measures headache-related disability over a 3-month period.⁵⁴ The instrument evaluates the number of days missed or with reduced productivity at work or school, at home, and in social settings. Total scores are translated to a 4-point grading system: grade 1 (scores from 0 to 5) is for minimal or infrequent disability; grade 2 (scores from 6 to 10) is for mild or infrequent disability; grade 3 (scores from 11 to 20) is for moderate disability; and grade 4 (scores of 21 or greater) is for severe disability.¹² Grade 4 can be further divided into 2 subcategories: grade 4-A (scores 21 to 40) is for severe disability and grade 4-B (scores 41 to 270) is for very severe disability. A higher value is indicative of more disability; a decrease in grade indicates a decrease in disability. CADTH identified an MID of 3.7 points in the literature.⁵⁵ The clinical expert consulted by CADTH indicated a 5-point or 50% change in total MIDAS score would be considered clinically significant.

The EQ-5D-5L is a patient-reported, generic, HRQoL instrument that that assesses current health status.^56-58 The recall period is “today” and the instrument consists of 2 parts: a health utility index and the EQ Visual Analogue Scale (EQ VAS). The health utility component consists of 5 dimensions (mobility, self-care, usual activities, pain/discomfort, and anxiety/depression) and patients respond to each dimension using 5 levels (1 indicates no problems, 5 indicates extreme problems or unable to perform). Results from the EQ-5D-5L descriptive system can be converted into a single index score using a scoring algorithm taking the local patient and population preferences into account. The index score therefore provides a country-specific value, which is a major feature of the instrument. The health state index score provides a single value on a scale from less than 0 to 1; negative values are for states worse than dead, 0 is a health state equivalent to death, and 1 is a health state equivalent to perfect health. For the EQ VAS, patients rate their perceived health state from 0 (the worst health imaginable) to 100 (the best health imaginable). A Canadian-specific estimate of a MID in the general population is 0.056.⁵⁹ No MID was identified in populations with migraine.

The 4-item Migraine Interictal Burden Scale (MIBS-4) is a self-reported questionnaire that assesses the burden of illness related to migraine during the time in between attacks (interictal state) with a 4-week recall period.¹² The 4 items address disruption at work and school, diminished family and social life, difficulty planning, and emotional difficulty. Responses range from 0 (don’t know/not applicable or never) to 3 (much or most/all of the time). The individual item scores are summed, producing a total score ranging from 0 to 12. A higher score indicates a greater interictal burden. No MID was identified in populations with migraine.

The PHQ-9 is a self-reported questionnaire used for screening patients for depression and measuring the severity of depressive symptoms over the past 2 weeks.^12,60 The instrument consists of 9 items corresponding to criteria for diagnosing major depressive disorder from the Diagnostic and Statistical Manual for Mental Disorders, fourth edition. Patients score each item on a 4-point scale for how frequent symptoms occurred during a 2-week recall period (0 = not at all, 1 = several days, 2 = more than half the days, 3 = nearly every day). The total score ranges from 0 to 27, with a higher score indicating greater severity of depressive symptoms (0 to 4 = none to minimal depression, 5 to 9 = mild depression, 10 to 14 = moderate depression, 15 to 19 = moderately severe depression, and 20 to 27 = severe depression).⁶¹ No MID was identified in patients with migraine.

The GAD-7 is a patient-administered questionnaire used to screen for and measure the severity of anxiety symptoms within a 2-week recall period.¹² The questionnaire consists of 7 items, including feelings of nervousness, uncontrollable worrying, excessive, worrying, trouble relaxing, restlessness, irritability, and fearfulness. Patients answer based on a 4-point rating scale (0 = not at all, 1 = several days, 2 = more than half the days, and 3 = nearly every day). Total scores range from 0 to 21, with a higher score indicating greater anxiety (0 to 4 = minimal anxiety, 5 to 9 = mild, 10 to 14 = moderate, and 15 to 21 = severe). No specific MID was identified in patients with migraine.

Data on HCRU and employment status were collected by study-site personnel in the REGAIN and CONQUER trials. Patients were asked 3 questions about the number of hospital emergency visits, overnight stays at the hospital, and other visits with a health care professional aside from study-related visits that occurred (all-cause and migraine-specific) since the patient’s last study visit. At baseline, the same questions were asked with a recall period of 6 months. A higher number (i.e., events or visits) indicates greater utilization of health care resources. Employment status information was also collected while documenting patient responses about the number of health care events (all-cause and migraine-specific) since the patient’s last study visit. No literature was found that assessed HCRU or employment status for validity, reliability, or responsiveness in patients with migraine. No MID was identified in populations with migraine.

The WPAI is a patient-reported instrument that measures the impact on work productivity and regular activities attributable to a specific health problem.⁶² The instrument consists of 6 items and measures impairments on both paid and unpaid work within a 7-day recall period. Four scores are calculated for absenteeism, presenteeism, work productivity loss, and activity impairment. No MID was identified in patients with migraine. The scores are reported as impairment percentages, with a higher number indicating greater impairment or lower productivity. No literature was found that assessed the WPAI for validity, reliability, or responsiveness in patients with migraine.

Harms Outcomes

In all studies, AEs related to injection sites were defined using terms from the Medical Dictionary for Regulatory Activities (MedDRA) high-level term of “injection site reactions.”

Potential treatment-emergent hypersensitivity events (immediate and non-immediate) were identified from a review of preferred terms generated from the standardized MedDRA query for anaphylactic reaction, angioedema, and hypersensitivity. Query results were narrowed by using preferred terms that are highly likely to represent the condition of interest and are reported in the CADTH systematic review.

Adverse events related to injection sites were defined using the MedDRA high-level term of “injection site reactions.”

In the EVOLVE-1, EVOLVE-2, and REGAIN studies, data on the incidence of treatment-emergent ADA were prospectively collected and reported. Treatment-emergent ADA-positive was defined as a negative baseline and a positive post-baseline ADA result with a titre of 20 or greater, or a positive baseline and a positive post-baseline result with a titre increase of at least 4-fold (e.g., baseline titre of 10 increasing to 40 or greater following baseline). A list of patients who were treatment-emergent ADA-positive at any time during the study was created and included neutralizing ADA status. Immunogenicity was not proactively evaluated in the CONQUER trial.

Vascular events included treatment-emergent AEs categorized in the MedDRA system order class of vascular disorders.

Statistical Analysis

The statistical analysis of efficacy end points conducted in the 4 trials included in the systematic review is summarized in Table 13.

All Pivotal Trials

In all trials, analyses were conducted primarily using 2 analysis populations: an ITT population and a safety population. Efficacy analyses were conducted on the ITT population. Safety analyses were conducted on the safety population.

Treatment effects were evaluated based on an overall 1-sided significance level of 0.025 (equivalent to a 2-sided 0.05 significance level) for all efficacy and safety analyses. The 95% CIs for the difference in LS means between treatment groups were presented.

Change from baseline of continuous variables with repeated measures was analyzed using a mixed model for repeated measures (MMRM) analysis.

For other continuous variables, the change from baseline to the LOCF end point was analyzed using an analysis of variance (ANOVA) or analysis of covariance (ANCOVA) model. Unless otherwise specified, when an ANOVA or ANCOVA model was used to analyze a continuous efficacy variable, type III sum of squares for the LS means was used for statistical comparisons.

For categorical efficacy variables without repeated measures, comparisons between treatment groups were performed using logistic regressions.

Comparisons between galcanezumab and placebo refer to comparisons of each galcanezumab treatment group with placebo.

Approaches to Handling Missing Data

To handle missing data, 2 statistical approaches were applied. Repeated measures analyses were used where the model parameters were simultaneously estimated using restricted-likelihood estimation incorporating all observed data.; estimates have been shown to be unbiased when the missing data are missing at random. For ANOVA and ANCOVA analyses, the LOCF was used.

In the EVOLVE-1, EVOLVE-2, and REGAIN trials, sensitivity analyses were conducted to assess the robustness of the primary analysis conclusions to deviations from missing at random assumption. The approach for these analyses was to vary the assumptions of missing data for the primary analysis in a systematic way. The method was to predict the missing outcomes, then add values (Δ₁₂₀, Δ₂₄₀, and Δ_P) to the predictions in the galcanezumab 120 mg, galcanezumab 240 mg, and placebo treatment groups, respectively, regardless of the reason the data were missing.

The primary outcome measure (i.e., the number of monthly MHDs) was summarized from the daily ePRO data for each patient (including daily data from the baseline period before randomization and 6 months of daily data during the treatment phase). The daily data were aggregated, and the number of MHDs was provided for each of the monthly visit or dose intervals. In calculating the number of MHDs for each monthly interval, the number of MHDs was normalized to a 30-day period by multiplying the number of MHDs by 30 and dividing by the total number of non-missing diary days in the monthly interval. This approach to missing ePRO diary data assumed that the rate of migraine headache per day was the same for days with missing and non-missing ePRO diary days. The same approach was applied to secondary and exploratory efficacy measures that were derived from ePRO data. If the compliance rate for a monthly interval was no more than 50%, then all end points to be derived from the ePRO diary data for that 1-month period were considered missing.

Primary Outcome

The primary outcome (i.e., the number of monthly MHDs) was summarized from the daily ePRO data for each patient. The primary analysis evaluated the efficacy of galcanezumab compared with placebo on the overall mean change from baseline in the number of monthly MHDs during the double-blind treatment phase. The primary analysis was performed using a MMRM analysis. The overall mean change in monthly MHDs was estimated as the main effect of treatment (each galcanezumab dose group and placebo) from the MMRM analysis during the double-blind treatment phase. This provided the average treatment effect across the duration of the double-blind treatment phase. The repeated measures analysis included data from all treatment groups. The type I error rate for the study was controlled at a 2-sided 0.05 level (equivalent to a 1-sided 0.025 level).

EVOLVE-1 and EVOLVE-2

Statistical analysis was performed as previously described.

The MMRM for efficacy analyses included the fixed, categorical effects of treatment, geographical region, month, and treatment-by-month interaction, as well as the continuous, fixed covariates of baseline value and baseline-by-month interaction. The baseline value and baseline-by-month interaction were included to account for the differential influence over time that the baseline value has on the post-baseline values. The MMRM for the safety analyses included the fixed, categorical effects of treatment, month, and treatment-by-month interaction, as well as the continuous, fixed covariates of baseline value and baseline-by-month interaction.

When an ANOVA model was used to analyze a continuous efficacy variable at an LOCF end point, the model contained the main effects of treatment and region. The ANCOVA model used to analyze a continuous efficacy variable contained the same main effects as the ANOVA model, but included an appropriate baseline value as a covariate. When an ANOVA or ANCOVA model was used to analyze a continuous safety variable, the term for region was excluded from the model.

For categorical variables without repeated measures, comparisons between treatment groups were performed using Fisher’s exact test.

With the exception of efficacy analyses on MHDs or categorical analysis of response rates derived from MHDs where the continuous value of baseline MHDs was used as covariate, all other efficacy analyses included a baseline monthly MHD category (< 8 versus ≥ 8) as a covariate in the model.

Multiple Testing Procedure

Control of the familywise type I error rate for the primary and key secondary end points was achieved through an overall superchain procedure (Kordzakhia and Dmitrienko [2013]), a multiple-testing method that employs alpha recycling techniques. As part of this procedure in the EVOLVE-1 and EVOLVE-2 studies, a Dunnett test (Dunnett [1955]) was used for the primary hypotheses (galcanezumab 120 mg versus placebo, galcanezumab 240 mg versus placebo). Tests for key secondary hypotheses were conducted following rejection of at least 1 of the 2 primary null hypotheses. Specifically, the Hochberg procedure (Hochberg [1988]) was used for a 50% and 75% response rate in MHDs at the corresponding dose(s). Following rejection of the previous secondary null hypotheses, the Bonferroni-Holm procedure (Holm [1979]) was used for change in the use of acute (abortive) migraine treatment and change in functioning for the MSQ v2.1. Following rejection of these secondary null hypotheses, the remaining hypotheses of 100% response rate and Patient Global Impression of Severity (PGI-S) were tested sequentially. Alpha could be recycled between doses following testing of all secondary null hypotheses for either dose.

The multiple-testing procedure is depicted in Figure 5. Propagation weights are denoted along the edges between boxes, which represent hypothesis tests for families of hypotheses. According to this testing procedure, the parallel branches (dose sequences) are initially tested simultaneously, and the available alpha is then recycled between the branches to retest end-point families containing non-rejected null hypotheses.

Multiple testing procedure results for the EVOLVE-1 and EVOLVE-2 studies are presented in Appendix 3 (Figure 19 and Figure 20, respectively).

There were no adjustments for multiplicity for analyses of other end points.

Figure 5: Multiple Testing Procedure in the EVOLVE-1 and EVOLVE-2 Trials

acute meds = migraine headache days with use of acute (abortive) treatment; MHD = number of monthly migraine headache days (mean change from baseline); MSQ = Migraine-Specific Quality of Life Questionnaire version 2.1 role function – restrictive domain; PGI-S = Patient Global Impression of Severity; RR = response rate in monthly migraine headache days (mean change from baseline).

Source: EVOLVE-1 Clinical Study Report⁹ and EVOLVE-2 Clinical Study Report.¹⁰

Primary Outcome Analysis

The primary outcome analysis was performed as previously described for all pivotal trials. The following subgroup analyses were planned for the primary outcome measure: sex, race, ethnicity, baseline monthly MHD category (< 8 or ≥ 8), baseline treatment-resistant status (previously failed 2 or more prophylactic treatments: yes or no), and having aura or not (during the baseline period). The subgroup analyses were conducted using the same MMRM applied to the primary efficacy analysis, with terms of subgroup, subgroup-by-treatment, subgroup-by-month, and subgroup-by-treatment-by-month interactions added as additional covariates. A post hoc subgroup analysis also performed for patients who failed 3 or more prophylactic treatments.

Sensitivity Analyses for the Primary Outcome: Sensitivity analyses for missing-data assumptions were planned to be performed in the EVOLVE-1 and EVOLVE-2 studies. Delta adjustments as suggested in Permutt (2016) were conducted for change from baseline in MHDs. These analyses were performed to assess the potential impact of missing-data assumptions by examining the primary conclusions under a set of plausible, worst-case scenarios around the distribution of missing outcome data. Missing outcome values were imputed to be worse than expected based on the observed values (at the extreme, even markedly worse than the baseline values) with no treatment benefit at all seen in missing outcomes data. In total, 9 delta sets were used.

In both the EVOLVE-1 and EVOLVE-2 studies, a sensitivity analysis for the raw number of MHDs (i.e., the total number of MHDs for each interval without normalization to a 30-day period) was conducted with a repeated measures negative binomial regression analysis to assess the validity of the primary MMRM results with respect to deviations from normality assumption.

In addition, a sensitivity analysis was conducted in both EVOLVE studies to assess the validity of the primary results if a minor change was made to the definition of migraine headache. This minor change was to add an additional condition that, for any diary day, even if the patient did not meet the migraine headache definition as defined by criteria A and B (Table 12) but was taking a triptan or ergot derivative (“criterion C”), this diary day for this patient was still considered an MHD. The rationale was to replicate the use of a similar criterion that is part of the ICHD-3 criteria for a diagnosis of CM. This analysis was implemented by repeating the primary analysis using the number of monthly MHDs derived based on the modified definition of migraine headache.

A post hoc sensitivity analysis was conducted to assess the impact of inadvertently enrolled patients (i.e., identified as having an important protocol deviation of inclusion or exclusion criteria not being met) on the primary efficacy analysis in the EVOLVE-1 and EVOLVE-2 studies.

In the EVOLVE-1 study, an additional post hoc sensitivity analysis was conducted to assess the impact of excluding the patients from a site that was closed due to Good Clinical Practice compliance issues.

Secondary and Exploratory End point Analyses

For the continuous efficacy measures, the change from baseline to each post-baseline period was estimated for each treatment from repeated measures analyses as described for analysis of the primary outcome, with the addition of baseline monthly MHD category (< 8 versus ≥ 8) as a covariate. For the continuous secondary efficacy measures for which the objective was to assess overall mean change during the 6-month double-blind treatment phase, the end point for comparing galcanezumab with placebo was estimated as the main effect of treatment from the MMRM analysis across months 1 to 6 (inclusive of month 6). In addition to the repeated measures analyses, the mean change from baseline to the average monthly measures or LOCF end point during the 6-month treatment phase was estimated for the continuous efficacy measures using ANCOVA models with covariate adjustments as previously described.

The mean change from baseline to each post-baseline visit for MSQ v.2.1 total and domain scores averaged across months 4 to 6 (inclusive of month 6) was evaluated using an MMRM. The mean change from baseline to each post-baseline visit for MIDAS total and item scores was also evaluated using an MMRM.

For time-to-event analyses, a stratified log-rank test was used.

Determination of Sample Size

Each study planned to enrol approximately 825 patients. Eligible patients were to be randomized in a 2:1:1 ratio to placebo (approximately 413 patients), galcanezumab 120 mg (target of 206 patients), or galcanezumab 240 mg (target of 206 patients). With the assumption of a 26% discontinuation rate and an effect size of 0.33 in the last month of the 6-month treatment phase, it was estimated that this sample would provide approximately 95% power that at least 1 dose of galcanezumab would separate from placebo at a 1-sided 0.025 significance level based on simulations using a Dunnett test. Assumptions were based on data from 2 double-blind, placebo-controlled, phase II studies, adjusted to reflect the longer treatment duration and greater variability expected in a larger phase III study.

REGAIN

Statistical analysis was performed as previously described. In addition to the primary efficacy analyses conducted on the ITT population, some analyses were conducted using the open-label treatment population, which included all patients who entered the open-label treatment phase as indicated by receiving any injections starting from visit 7.

The MMRM for efficacy analyses included the fixed, categorical effects of treatment, country, month, baseline medication overuse (yes versus no), concurrent prophylaxis (yes versus no), and treatment-by-month interaction, as well as the continuous, fixed covariates of baseline value and baseline-by-month interaction. The baseline value and baseline-by-month interaction were included to account for the differential influence over time that the baseline value has on the post-baseline values. The MMRM for the safety analyses included the fixed, categorical effects of treatment, month, baseline medication overuse, concurrent prophylaxis, and treatment-by-month interaction, as well as the continuous, fixed covariates of baseline value and baseline-by-month interaction.

When an ANOVA model was used to analyze a continuous efficacy variable, the model contained the main effects of treatment, baseline medication overuse, concurrent prophylaxis use, and country. The ANCOVA model used to analyze a continuous efficacy variable contained the same main effects as the ANOVA model, but included an appropriate baseline value as a covariate. When an ANOVA or ANCOVA model was used to analyze a continuous safety variable, the term for country was excluded from the model.

For categorical variables without repeated measures, comparisons between treatment groups were performed using a Cochran-Mantel-Haenszel or Fisher exact test. Unless otherwise stated, the former test was controlled for baseline medication overuse (yes versus no) and concurrent prophylaxis use (yes versus no). Unless specified otherwise, the Fisher exact test was used for comparisons of baseline measures.

When a logistic regression was used to analyze a binary variable, the model included the main effect of treatment, baseline medication overuse, concurrent prophylaxis use, and country, and an appropriate baseline value as a covariate. The country could have been excluded from the model in the event of non-convergence.

Multiple Testing Procedure

Control of the familywise type I error rate for the primary and key secondary end points was achieved through an overall superchain procedure (Kordzakhia and Dmitrienko [2013]), a multiple-testing method that employs alpha recycling techniques. As part of this procedure in the present study, a Dunnett test was used for the primary hypotheses (galcanezumab 120 mg versus placebo, galcanezumab 240 mg versus placebo). Tests for key secondary hypotheses were to be conducted following rejection of at least 1 of the 2 primary null hypotheses. Specifically, following rejection of at least 1 primary null hypothesis, the Hochberg procedure (Hochberg [1988]) was to be used for the 50% and 75% response rate in MHDs at the corresponding dose(s). Following rejection of the previous secondary null hypotheses, the Bonferroni-Holm procedure (Holm [1979]) was to be used for change in the use of acute (abortive) migraine treatment and change in functioning for the MSQ v2.1. Following rejection of these secondary null hypotheses, the remaining hypotheses of PGI-S and 100% response rate were to be tested sequentially. Alpha could be recycled between doses following testing of all secondary null hypotheses for either dose. This multiple-testing procedure is depicted in Figure 6. Propagation weights are denoted along the edges between boxes, which represent tests for families of hypotheses. According to this testing procedure, the parallel branches (dose sequences) are initially tested simultaneously, and then the available alpha is recycled between the branches to retest end-point families containing non-rejected null hypotheses.

Multiple testing procedure results for the REGAIN trial are presented in Appendix 3 (Figure 21 and Figure 22).

There were no adjustments for multiplicity for analyses of other end points.

Figure 6: Multiple Testing Procedure in the REGAIN Trial

acute meds = migraine headache day with the use of acute (abortive) treatment; MHD = the number of monthly migraine headache days (mean change from baseline); MSQ = Migraine-Specific Quality of Life Questionnaire version 2.1 role function – restrictive domain; PGI-S = Patient Global Impression of Severity; RR = response rate in monthly migraine headache days (mean change from baseline).

Source: REGAIN Clinical Study Report.¹¹

Primary Outcome Analysis

The primary outcome analysis was performed as previously described for all pivotal trials. The following subgroup analyses were planned for the primary outcome measure: sex, race, ethnicity, region, treatment-resistant status (i.e., failed at least 2 prophylactic treatments [yes versus no]), having an aura or not (during the baseline period), baseline medication overuse, and concurrent prophylaxis use. The subgroup-by-treatment interaction was tested at a 2-sided 0.10 significance level. Treatment-group differences were evaluated within each category of the subgroup variable. For all the subgroup variables, the subgroup analysis for change from baseline to each monthly interval in the number of MHDs was conducted with the same MMRM in terms of subgroup, subgroup-by-treatment, subgroup-by-month, and subgroup-by-treatment-by-month interactions added as additional covariates. In this analysis, the P values for the subgroup-by-treatment, subgroup-by-month, and subgroup-by-treatment-by-month interactions at the last monthly interval of the 3-month treatment (month 3) were reported. For subgroup analysis with both the MMRM and ANCOVA method, the LS mean and LS mean change estimate as well as the treatment comparisons within each subgroup were analyzed using the data within that specific subgroup only. The MMRM and ANCOVA models were the same as the primary analysis. A post hoc exploratory subgroup analysis was also performed for patients who failed at least 3 prophylactic treatments (yes versus no).

Sensitivity Analyses for the Primary Outcome: Two sensitivity analyses were planned in the REGAIN trial. Sensitivity analyses with delta adjustments as suggested in Permutt (2016) were conducted for change from baseline in MHDs to assess the potential impact of missing data assumptions by examining the primary conclusions under a set of plausible, worst-case scenarios around the distribution of missing outcome data. In addition, a sensitivity analysis for the raw number of MHDs (i.e., the total number of MHDs for each interval without normalization to a 30-day period) was conducted with a repeated measures negative binomial regression analysis to assess the validity of the primary MMRM results with respect to deviations from normality assumption.

Two post hoc sensitivity analyses were also conducted. One sensitivity analysis assessed the impact of inadvertently enrolled patients (i.e., identified as having an important protocol deviation of inclusion or exclusion criteria not being met) on the primary efficacy analysis. Another sensitivity analysis was conducted to assess the impact of the interactive web response system stratification error on the primary efficacy analysis.

Secondary and Exploratory End point Analyses

For the continuous efficacy measures, the change from baseline to each post-baseline period was estimated for each treatment from repeated measures analyses as described for analysis of the primary outcome. For the continuous secondary efficacy measures in which the objective was to assess overall mean change during the 3-month double-blind treatment phase, the end point for comparing galcanezumab with placebo was estimated as the main effect of treatment from the MMRM analysis assessing the average treatment effect across months 1, 2, and 3. In addition to the repeated measures analyses, the mean change from baseline to the average monthly measures or LOCF end point during the 3-month double-blind treatment phase was estimated for the continuous efficacy measures using ANCOVA models with covariate adjustments.

The mean change from baseline to each post-baseline visit for MSQ v2.1 total and domain scores was evaluated using an MMRM. The change from baseline to month 3 on the MIDAS total and item scores was evaluated using the ANCOVA model.

For time-to-event analysis (including time to first loss of 50% response and time to start of preventive treatment), a stratified log-rank test was used with the baseline monthly MHD category (< 8 versus ≥ 8) and region as covariates.

For the HCRU, numbers of the following health-related visits were enumerated and analyzed: hospital emergency room visits, overnight hospital stays, other visits with health care professional, hospital emergency room visits related to migraine headache, overnight hospital stays related to migraine headache, and other visits with health care professionals related to migraine headache. The HCRU rate during the 3-month double-blind phase was compared with the baseline 3-month rate. The baseline 3-month rate was derived from the baseline HCRU measurement, which assessed the previous 6 months. The 6-month baseline rate was then divided by 2 to establish a time frame comparable to the 3-month double-blind treatment phase. These measures were analyzed with a repeated measures negative binomial regression analysis. The model included country, month, baseline medication overuse, concurrent prophylaxis use, and treatment-by-month interaction, as well as the continuous fixed covariates of baseline and baseline-by-month interaction, and log (number of actual number of days within every 3 months divided by 90) as the offset in the model. In case of non-convergence, country and/or baseline-by-month interaction may not have been included in the model. Due to the low rates of HCRU, summary statistics were also provided. Employment status at each visit was summarized separately for each treatment group.

For the outcome of overall change in MHDs with acute migraine medication use from baseline, a post hoc exploratory subgroup analysis was performed for the number of failed prophylactic treatments (≥ 2 versus ≥ 3). For the outcome of overall change in the MSQ v2.1 RF-R domain score from baseline to month 3, a post hoc exploratory subgroup analysis was performed for the number of failed prophylactic treatments (≥ 2 versus ≥ 3).

Determination of Sample Size

A sample size re-estimation procedure was originally planned for the study due to uncertainty in the effect size of galcanezumab in CM. The final sample size could have been between the minimum (825 patients) and maximum (1,140 patients) sample sizes, as informed by the results of the planned interim sample size re-estimation computation:

• Based on the assumption of an effect size of 0.33 and a dropout rate of approximately 15%, an initial minimum sample size of 825 was expected to provide more than 90% power that at least 1 dose of galcanezumab would separate from placebo at a 2-sided significance level of 0.05 based on simulations using a Dunnett test (Dunnett 1955).

• The maximum sample size of 1,140 was based on the assumption of a 15% discontinuation rate and an effect size of 0.30 in the last month of the 3-month treatment phase; it was estimated that this sample size would provide approximately 95% power that at least 1 dose of galcanezumab would separate from placebo at a 1-sided 0.025 significance level.

To preserve blinding, full details of the sample size and power calculations were omitted from the protocol and were provided to the ethics review board in a separate document. Sites therefore remained blinded to the potential maximum as well as final sample size throughout the trial.

However, the planned sample size re-estimation was not conducted because the rapid rate of enrolment made this adaptive feature of the trial impossible. At the time that interim sample size re-estimation results would have been available, the study would have already surpassed the number of patients needed for screening to enrol the predefined maximum sample size. Therefore, the sample size target was set to the originally planned maximum of approximately 1,140 patients, and the sample size adaptive feature of the trial was removed. The decision to enrol approximately 1,140 patients made the sample size for this study comparable to those of the sponsor’s other global CM studies, according to publicly available records. The final sample size fell slightly short of the 1,140 target due to the variability of screening fail rates that can occur when using a prospective baseline period.

CONQUER

Statistical analysis was performed as described in the preceding section.

The MMRM included the fixed, categorical effects of treatment, baseline migraine frequency category (low-frequency EM, high-frequency EM, and CM), pooled country, month, and treatment-by-month interaction, as well as the continuous, fixed covariates of baseline value and baseline value-by-month interaction. Baseline value and baseline-by-month interaction were included to account for the differential influence over time that the baseline value has on the post-baseline values.

The ANOVA model for continuous efficacy and health outcome variables contained the main effects of treatment, baseline migraine frequency category, and pooled country. The ANCOVA model for continuous efficacy and health outcome variables contained the main effects of treatment, baseline migraine frequency category, pooled country, and appropriate baseline value.

For categorical efficacy variables without repeated measures, the logistic model included the main effect of treatment, baseline migraine frequency category, and appropriate baseline value as a covariate.

For some health outcome variables without repeated measures during the double-blind treatment phase, the change from baseline score was analyzed using nonparametric tests as the scores of the categories or number of events in the HCRU do not usually satisfy normal assumptions. The Kruskal-Wallis test (Kruskal and Wallis [1952]) was used for treatment comparison and Wilcoxon signed rank test (Wilcoxon [1945]) was used to compare the change from baseline within each treatment group.

Control for Multiplicity

To provide control of the type I error, the key secondary analyses were to be tested using a gated approach at a 2-sided alpha level of 0.05. If the null hypothesis was rejected for the primary end point, key secondary end points were to be sequentially tested following the gatekeeping hierarchy shown in Figure 7. Following the primary objective, the sequential procedure for key secondary objectives began with the comparison between treatment groups in the number of MHDs based on the ITT episodic subpopulation. If the null hypothesis was rejected for that comparison, then the comparison of the 50% response rate between treatment groups was to be tested in the ITT population. If that null hypothesis was rejected, then the next comparison in the sequence was to be tested (50% response rate in the ITT episodic subpopulation), following this same pattern until all hypotheses were tested or until the null hypothesis was accepted for an end point, at which point any further testing stopped for the key secondary objectives.

Multiple testing procedure results for the CONQUER trial are presented in Appendix 3 (Figure 23).

No adjustments were made for multiplicity for analyses of the other secondary or exploratory end points.

Figure 7: Multiple Testing Procedure in the CONQUER Trial

CM = chronic migraine; EM = episodic migraine; MHD = the number of monthly migraine headache days; MSQ RR = Migraine-Specific Quality of Life Questionnaire version 2.1 role function – restrictive domain; response = response rate.

Note: All testing was conducted at a 2-sided alpha of 0.05.

Source: CONQUER Clinical Study Report.¹²

Primary Outcome Analysis

Subgroup analyses were planned for the primary outcome measure for the ITT patients in the double-blind treatment phase. Subgroup variables for the primary efficacy measure included sex, race, age, region, baseline migraine frequency category, and number of failed preventive migraine medication categories in the past 10 years. The subgroup-by-treatment interaction was tested at a 2-sided 0.10 significance level. Treatment-group differences were evaluated within each category of the subgroup variable. The subgroup analyses were conducted with the same MMRM.

Sensitivity Analyses for the Primary Outcome: Multiple sensitivity analyses were planned in the CONQUER trial. Sensitivity analyses with delta adjustment as suggested in Permutt (2016) were conducted for change from baseline in the number of monthly MHDs. These analyses were performed to assess the potential impact of missing-data assumptions by examining the primary conclusions under a set of plausible, worst-case scenarios around the distribution of missing outcome data. A sensitivity analysis for the raw number of MHDs (i.e., the total number of MHDs for each interval without normalization to a 30-day period) was conducted with a repeated measures negative binomial regression analysis to assess the validity of the primary MMRM results with respect to deviations from normality assumption. A third sensitivity analysis was conducted after removing outliers. In this analysis, the Studentized residuals from the primary analysis model were examined and patients with outlier residuals were identified as those who had an absolute value of a Studentized residual greater than or equal to 2 at any month of the double-blind treatment phase.

Last, a per-protocol analysis of the primary efficacy end point was conducted post hoc. The per-protocol population included all ITT patients who did not have any important protocol deviations during the baseline and double-blind treatment phases.

Secondary and Exploratory End-Point Analyses

The MSQ v2.1, MIDAS, MIBS-4, WPAI, EQ-5D-5L, PHQ-9, and GAD-7 were evaluated using MMRM when there were repeated measures. When there was a single post-baseline measure, they were evaluated using an ANCOVA model.

As HCRU data are count data with excess zeros for migraine patients, they were summarized as the number of events per 100 patient-years. A Wilcoxon signed rank test was performed for comparisons within treatment group and a Kruskal-Wallis test was used for comparisons between treatment groups.

Subgroup analyses were planned for patients with 3 or more prior preventive medication category failures. These subgroup analyses were performed in the ITT population for the following outcome measures: MSQ v2.1 RF-R and MHDs with acute medication use.

Determination of Sample Size

The study planned to enrol approximately 420 patients. With the assumption of a 10% discontinuation rate and an effect size of 0.39, it was estimated that this sample size would provide approximately 96% power that galcanezumab will separate from placebo at a 2-sided significance level of 0.05 for the ITT population.¹³ The study was also powered for the subpopulation of patients with EM. The study aimed to enrol approximately 250 patients with EM as determined during the prospective baseline period. With the assumption of a 10% discontinuation rate and an effect size of 0.46, it was estimated that this would provide approximately 93% power that galcanezumab will separate from placebo at a 2-sided significance level of 0.05 for the EM subpopulation.

Analysis Populations

In all trials, efficacy analyses were performed on a modified ITT population. The modified ITT population included all randomized patients who received at least 1 dose of the study drug. For efficacy analyses, patients were analyzed according to the treatment to which they were randomized. In the CONQUER trial, efficacy analyses were also performed on the modified ITT episodic subpopulation (randomized patients diagnosed with EM) and modified ITT chronic subpopulation (randomized patients diagnosed with CM).

In all trials, the safety population included data from all randomized patients who received at least 1 dose of the study drug. Analyses in the safety population were conducted based on modal treatment the patient received during the double-blind treatment phase. Modal treatment was used to account for situations in which the patient accidentally received a treatment other than the treatment assigned, or cases in which a patient was randomized to the galcanezumab 120 mg group but discontinued after administration of the loading dose such that their exposure was predominantly to the 240 mg loading dose.

The EVOLVE-1, EVOLVE-2, and REGAIN trials also had a post-treatment population, which was used for analyses of the post-treatment phase only (excluding the double-blind treatment phase). This population included all patients who entered the post-treatment phase, as indicated by entering any post-treatment visit. Efficacy analyses using this population were conducted according to the ITT principle, whereas safety analyses were conducted based on modal treatment the patient received during double-blind treatment.

Results

Patient Disposition

Patient disposition in the EVOLVE-1 and EVOLVE-2 trials is summarized in Table 14. In the EVOLVE-1 trial, 1,671 patients were screened, and 862 patients were randomized: 433 to placebo, 213 to galcanezumab 120 mg, and 212 to galcanezumab 240 mg, with 4 excluded from the study’s ITT population because they did not receive at least 1 dose of study drug. The most common reason for screen failure was patients not meeting criteria for study enrolment based on migraine headache information collected in the ePRO diary during the prospective baseline phase. A total of 155 patients (18.1%) discontinued from the double-blind treatment phase. The most frequent reason for discontinuing from double-blind treatment was withdrawal by patient, which occurred in a similar percentage of patients across treatment groups. Overall, 67.1% of patients entered the post-treatment phase of the study as of the data cut-off date of March 22, 2017, for the final efficacy analyses for the double-blind treatment phase.

In the EVOLVE-2 trial, 1,696 patients were screened, and 922 patients were randomized: 461 to placebo, 231 to galcanezumab 120 mg, and 223 to galcanezumab 240 mg, with 7 excluded from the study’s ITT population because they did not receive at least 1 dose of study drug. The most common reason for screen failure was patients not meeting criteria for study enrolment based on migraine headache information collected in the ePRO diary during the prospective baseline phase. A total of 129 patients (14.1%) discontinued from the double-blind treatment phase. The most frequent reason for discontinuing from double-blind treatment was withdrawal by patient, which occurred in 8.5% of the placebo arm, 4.8% of the galcanezumab 120 mg arm, and 6.3% of the galcanezumab 240 mg arm. Overall, 68.5% of patients entered the post-treatment phase of the study as of the data cut-off date of March 29, 2017, for the final efficacy analyses for the double-blind treatment phase.

Patient disposition in the REGAIN trial is summarized in Table 15. A total of 1,903 patients were screened, and 1,117 patients were randomized: 558 to placebo, 278 to galcanezumab 120 mg, and 277 to galcanezumab 240 mg, with 4 excluded from the study’s ITT population because they did not receive at least 1 dose of study drug. Overall, 93.2% of patients completed the double-blind treatment phase, with 91.0% completing the placebo arm, 94.6% the galcanezumab 120 mg arm, and 96.0% the galcanezumab 240 mg arm. The most frequent reason for discontinuing from the double-blind treatment phase was withdrawal by patient (3.4% placebo, 1.4% galcanezumab 120 mg, and 2.5% galcanezumab 240 mg). Of the patients who completed the double-blind treatment phase, 98.5% entered the optional open-label treatment phase as of the data cut-off date of March 16, 2017, for the final efficacy analyses for the double-blind treatment phase.

Patient disposition in the CONQUER trial is summarized in Table 16. A total of 610 patients were screened, and 463 patients were randomized: 230 to placebo and 232 to galcanezumab 120 mg. One patient who was a screen failure was inadvertently randomized and immediately discontinued. A total of 462 randomized patients received at least 1 dose of study drug and were included in the ITT population. Overall, 97.6% of patients completed the double-blind treatment phase. The most frequent reason for discontinuing from the double-blind treatment phase was protocol deviation (1.7% galcanezumab and 0.4% placebo). Of the patients who completed the double-blind treatment phase, 99.6% entered the optional open-label treatment phase as of the data cut-off date of June 19, 2019, for the final efficacy analyses for the double-blind treatment phase.

Exposure to Study Treatments

Exposure to study treatments during the double-blind treatment phases of the 4 pivotal trials is summarized in Table 17.

Most patients in the EVOLVE-1 and EVOLVE-2 trials, which included a 6-month double-blind treatment phase, received 6 doses of the study drug (> 82% in all treatment arms). Most patients in the REGAIN and CONQUER trials, which had a 3-month double-blind treatment phase, received 3 doses of the study drug (> 92% in both arms of the REGAIN trial, > 97% in both arms of the CONQUER trial).

As of the data cut-off for the REGAIN trial’s Clinical Study Report Addendum,¹³ the mean duration of exposure to the study drug in the open-label treatment phase (which does not include exposure in the double-blind treatment phase) was 239.5 days. Approximately 80.9% of patients (N = 827) received all 9 doses of open-label galcanezumab. A total of 1,013 patients (99.3%) received the loading dose of 240 mg galcanezumab at visit 7 (first injection of open-label treatment phase) as per protocol. Seven patients received only 1 injection at visit 7, and 2 patients did not receive any injection at visit 7. All patients (100%) who continued to visit 8 received 1 injection (120 mg galcanezumab) at visit 8, as per protocol. Most patients received 2 injections (240 mg) at the flexibly dosed visits, ranging from 64.3% of patients (at visit 9) to 75.0% (at visit 14).

As of the data cut-off for the CONQUER trial’s Clinical Study Report Addendum,¹³ the mean duration of exposure to galcanezumab in the galcanezumab-treated population was 130.1 days. The most common number of galcanezumab doses received was either 3 or 6 doses (47.5% and 47.7%, respectively).

Adherence With the ePRO Diary

In the EVOLVE-1 and EVOLVE-2 trials, a majority of patients (> 80%) completed at least 80% of their daily diary entries at each month during the double-blind treatment phase. In the EVOLVE-1 trial, mean compliance with the ePRO diary averaged over the 6-month treatment phase was similar across treatment groups (89.8% for placebo, 91.9% for galcanezumab 120 mg, and 90.8% for galcanezumab 240 mg). In the EVOLVE-2 trial, mean compliance with the ePRO diary averaged over the 6-month treatment phase was similar across treatment groups (91.3% for placebo, 90.9% for galcanezumab 120 mg, and 92.7% for galcanezumab 240 mg).

In the REGAIN trial, most patients (> 85%) completed at least 80% of their daily diary entries at each month during the double-blind treatment phase. Mean compliance with the ePRO diary averaged over the 3-month double-blind treatment phase was 91.0% for the placebo group and 91.8% for the galcanezumab 120 mg group.

In the CONQUER trial, > 93% of patients completed at least 80% of their daily ePRO diary entries during each month of the double-blind treatment phase. Mean compliance with the ePRO diary averaged over the 3-month double-blind treatment phase was approximately 96% in both treatment groups.

Efficacy

Only those efficacy outcomes and analyses of subgroups identified in the review protocol are reported. Efficacy data are not presented for the galcanezumab 240 mg arm of the EVOLVE-1, EVOLVE-2, and REGAIN trials because this dose is not aligned with the Health Canada–approved dose.

Results for the key efficacy outcomes for the EVOLVE-1 and EVOLVE-2 trial are summarized in Table 18. Results for the key efficacy outcomes for the REGAIN trial are summarized in Table 20. Results for the key efficacy outcomes for the CONQUER trial are summarized in Table 23.

Health-Related Quality of Life

Migraine-Specific Quality of Life Questionnaire Version 2.1

In the EVOLVE-1 and EVOLVE-2 trials, the mean change from baseline in the MSQ v2.1 RF-R domain score during months 4 to 6 (the final 3 months of double-blind treatment) was a key secondary outcome and controlled for multiplicity. In the EVOLVE-1 trial, the mean change from baseline in the MSQ RF-R during months 4 to 6 was 7.74 points greater in the galcanezumab 120 mg group compared with placebo (95% CI, 5.20 to 10.28; P < 0.001). Similarly, in the EVOLVE-2 trial, the mean change from baseline in the MSQ RF-R during months 4 to 6 was 8.82 points greater in the galcanezumab 120 mg group compared with placebo (95% CI, 6.33 to 11.31; P < 0.001).

In the REGAIN trial, the mean change from baseline in the MSQ RF-R at month 3 was a key secondary outcome and controlled for multiplicity. The mean numerical change from baseline in the MSQ RF-R at month 3 was 5.06 points greater in the galcanezumab 120 mg arm compared to placebo (95% CI, 2.12 to 7.99). The galcanezumab 120 mg arm could not be tested for statistical significance based on the predefined multiple-testing procedure because of a previously failed end point in the testing sequence (75% response rate; Figure 21 and Figure 22 in Appendix 3). Results of a post hoc exploratory subgroup analysis for the number of failed prophylactic treatments (≥ 2 versus ≥ 3) are summarized in Table 22. The results of this analysis were consistent with the analysis of overall change in the MSQ RF-R from baseline to month 3 in the ITT population.

In the CONQUER trial, the mean change from baseline in the MSQ RF-R at month 3 in the ITT population was a key secondary outcome and controlled for multiplicity. The mean change from baseline in MSQ RF-R at month 3 was 12.53 points greater in the galcanezumab group compared with placebo (95% CI, 9.19 to 15.87; P < 0.0001). The results of a planned subgroup analysis in patients with 3 or more prior preventive-medication category failures are summarized in Table 25, and are consistent with the ITT population.

EQ-5D 5-Levels Questionnaire

The EQ-5D-5L was not assessed in the EVOLVE-1, EVOLVE-2, and REGAIN trials. In the CONQUER trial, the LS mean change from baseline to the LOCF end point in EQ VAS scores was 3.38 (95% CI, not reported; SE = 1.31). The outcome was a secondary end point, but was not controlled for multiplicity.

Headache Symptoms

Migraine Headache Days with Symptoms

The overall change from baseline in number of monthly MHDs with symptoms during the double-blind treatment phase was an exploratory outcome in all pivotal trials. These outcomes were not controlled for multiplicity within the trials’ multiple-testing procedures.

In the EVOLVE-1 trial, the mean change difference from baseline in the galcanezumab 120 mg arm compared to placebo was −0.74 days (95% CI, −1.10 to −0.39) for nausea and/or vomiting, −1.39 days (95% CI, −1.90 to −0.89) for photophobia and phonophobia, −0.43 days (95% CI, −0.70 to −0.16) for aura, and −0.61 days (95% CI, −0.93 to −0.28) for prodromal symptoms other than aura.

In the EVOLVE-2 trial, the mean change difference from baseline in the galcanezumab 120 mg arm compared to placebo was −1.14 days (95% CI, −1.50 to −0.79) for nausea and/or vomiting, −1.76 days (95% CI, −2.25 to −1.27) for photophobia and phonophobia, −0.48 days (95% CI, −0.81 to −0.16) for aura, and −0.83 days (95% CI, −1.18 to −0.47) for prodromal symptoms other than aura.

In the REGAIN trial, the mean change difference from baseline in the galcanezumab 120 mg arm compared placebo was −1.21 days (95% CI, −1.82 to −0.59) for nausea and/or vomiting, −1.56 days (95% CI, −2.37 to −0.75) for photophobia and phonophobia, 0.03 days (95% CI, −0.53 to 0.58) for aura, and −0.66 (95% CI, −1.29 to −0.02) days for prodromal symptoms other than aura.

In the CONQUER trial, the mean change difference from baseline in the galcanezumab 120 mg arm compared to placebo was |||||||||||||||||||||||||||||||||||| for nausea and/or vomiting, |||||||||||||||||||||||||||||||||||||||| for photophobia and phonophobia, ||||||||||||||||||||||||||||||| for aura, and ||||||||||||||||||||||||||| for prodromal symptoms other than aura.

Other Patient-Reported Outcomes

Migraine Disability Assessment

Change in the MIDAS total score from baseline was a secondary outcome in all pivotal trials. This outcome was not controlled for multiplicity within the trials’ multiple-testing procedures.

In the EVOLVE-1 trial, the mean change from baseline to the end of double-blind treatment (month 6) was −6.29 points (95% CI, −9.45 to −3.13), indicating a numerically greater reduction in the galcanezumab 120 mg arm compared to the placebo arm.

In the EVOLVE-2 trial, the mean change from baseline to the end of double-blind treatment (month 6) was −9.15 points (95% CI, −12.61 to −5.69), indicating a numerically greater reduction in the galcanezumab 120 mg arm compared to the placebo arm.

In the REGAIN trial, the mean change from baseline to LOCF end point was −8.74 points (95% CI, −16.39 to −1.08), indicating a numerically greater reduction in the galcanezumab 120 mg arm compared to the placebo arm. The mean change from baseline to the end of double-blind treatment (month 3) was −8.74 (95% CI, −16.39 to −1.08), indicating a numerically greater reduction in the galcanezumab 120 mg arm compared to the placebo arm.

In the CONQUER trial, the mean change from baseline to LOCF end point was −17.8 points (95% CI, −25.6 to −10.0), indicating a numerically greater reduction in the galcanezumab 120 mg arm compared to the placebo arm.

4-item Migraine Interictal Burden Scale

The MIBS-4 was not assessed in the EVOLVE-1, EVOLVE-2, and REGAIN trials.

In the CONQUER trial, change in the MIBS-4 total score was assessed as a secondary outcome. The outcome was not included in the CONQUER trial’s multiple-testing procedure and therefore not controlled for multiplicity. The mean change in MIBS-4 total score from baseline during the double-blind treatment period was −1.06 points (95% CI, −1.58 to −0.54), indicating a numerically greater reduction in the galcanezumab arm compared to the placebo arm.

Patient Health Questionnaire-9

The PHQ-9 was not assessed in the EVOLVE-1, EVOLVE-2, and REGAIN trials.

In the CONQUER trial, the PHQ-9 was an exploratory outcome. The outcome was not included in the CONQUER trial’s multiple-testing procedure. The mean change in PHQ-9 total score from baseline to LOCF end point in the double-blind treatment period was −2.12 (SE = 0.32) in the galcanezumab arm compared to −1.13 (SE = 0.31) in the placebo arm.

Generalized Anxiety Disorder 7-Item

The GAD-7 was not assessed in the EVOLVE-1, EVOLVE-2, and REGAIN trials.

In the CONQUER trial, the GAD-7 was an exploratory outcome. The outcome was not included in the CONQUER trial’s multiple-testing procedure. The mean change from baseline to LOCF end point for the GAD-7 total score was |||||||||||||||||| in the galcanezumab arm and |||||||||||||||||| in the placebo arm.

Headache or Migraine Frequency

Number of Monthly Migraine Headache Days

The overall change from baseline in the number of monthly MHDs during the double-blind treatment phase was the primary outcome and included in the multiple-testing procedure in each of the pivotal trials. Galcanezumab 120 mg was associated with a statistically significant greater reduction in the overall LS mean change from baseline in the number of monthly MHDs during the double-blind treatment phase compared to placebo in all 4 trials.

EVOLVE-1 and EVOLVE-2: In the EVOLVE-1 trial, the overall mean change from baseline in the number of monthly MHDs during the double-blind treatment phase was −1.92 days (95% CI, −2.48 to −1.37; P < 0.001), indicating a greater reduction in the galcanezumab 120 mg arm compared to placebo. Results of all sensitivity analyses were consistent with the primary efficacy analysis.

In the EVOLVE-2 trial, the overall mean change from baseline in the number of monthly MHDs during the double-blind treatment phase was −2.02 days (95% CI, −2.55 to −1.48; P < 0.001), indicating a greater reduction in the galcanezumab 120 mg arm compared to placebo. Results of all sensitivity analyses were consistent with the primary efficacy analysis.

Results of the subgroup analyses conducted for overall change in monthly MHDs in the EVOLVE-1 and EVOLVE-2 are summarized in Table 19. The results of these subgroup analyses (baseline number of monthly MHDs [< 8 versus ≥ 8], failed 2 or more prophylactic treatments [yes versus no], baseline headache medication overuse [yes versus no]) were consistent with the results of the primary analysis. The subgroup analyses were not included in the multiple-testing procedure. The analysis of the subgroup of patients who failed 3 or more prophylactic treatments was a post hoc exploratory analysis.

REGAIN: In the REGAIN trial, the overall mean change from baseline in the number of monthly MHDs during the double-blind treatment phase was −2.09 days (95% CI, −2.92 to −1.26; P < 0.001), indicating a greater reduction in the galcanezumab 120 mg arm compared to placebo. Results of all sensitivity analyses were consistent with the primary efficacy analysis.

Results of the subgroup analyses conducted for overall change in monthly MHDs in the REGAIN trial are summarized in Table 21. The results of these subgroup analyses (failed 2 or more prophylactic treatments [yes versus no], failed 3 or more prophylactic treatments [yes versus no], baseline headache medication overuse [yes versus no]) were consistent with the results of the primary analysis. The subgroup analyses were not included in the multiple-testing procedure. The subgroup analyses of those who failed 2 or more prophylactic treatments (yes versus no) and baseline headache medication overuse (yes versus no) were specified a priori. The analysis of the subgroups that failed 3 or more prophylactic treatments (yes versus no) was a post hoc exploratory analysis.

CONQUER: In the CONQUER trial, the overall mean change from baseline in the number of monthly MHDs in the ITT population during the double-blind treatment phase was −3.12 days (95% CI, −3.92 to −2.32; P < 0.0001), indicating a greater reduction in the galcanezumab arm compared to placebo. Results of all sensitivity analyses were consistent with the primary efficacy analysis.

Planned subgroup analyses were conducted for baseline migraine frequency category and number of prior preventive medication category failures. Results, which are summarized in Table 24 and Table 25, were consistent with the primary analysis.

Number of Monthly Headache Days

The overall change from baseline in number of monthly headache days was a secondary outcome in the 4 pivotal trials. This outcome was not included in the trials’ multiple-testing procedures and was not adjusted for multiplicity.

In the EVOLVE-1 trial, the overall mean change from baseline in the number of monthly headache days during the double-blind treatment phase was −1.66 days (95% CI, −2.25 to −1.07), indicating a numerically greater reduction in the galcanezumab 120 mg arm compared to placebo.

In the EVOLVE 2 trial, the overall mean change from baseline in the number of monthly headache days during the double-blind treatment phase was −2.00 days (95% CI, −2.58 to −1.42), indicating a numerically greater reduction in the galcanezumab 120 mg arm compared to placebo.

In the REGAIN trial, the overall mean change from baseline in the number of monthly headache days during the double-blind treatment phase was −1.84 days (95% CI, −2.65 to −1.02), indicating a numerically greater reduction in the galcanezumab 120 mg arm compared to placebo.

In the CONQUER trial, the overall mean change from baseline in the number of monthly headache days in the ITT population during the double-blind treatment phase was −3.13 days (95% CI, −3.96 to −2.29), indicating a numerically greater reduction in the galcanezumab 120 mg arm compared to placebo.

Acute Headache Pain Medication Intake

In the EVOLVE-1, EVOLVE-2, and REGAIN trials, the overall change from baseline in monthly MHDs with acute medication use during the double-blind treatment phase was a key secondary outcome and included in the trials’ multiple-testing procedures.

In the EVOLVE-1 trial, the overall mean change was −1.81 days (95% CI, −2.28 to −1.33; P < 0.001), indicating a greater reduction in the galcanezumab 120 mg arm versus the placebo arm.

In the EVOLVE-2 trial, the overall mean change was −1.82 days (95% CI, −2.29 to −1.36; P < 0.001), indicating a greater reduction in the galcanezumab 120 mg arm compared to the placebo arm.

In the REGAIN trial, the overall mean change was −2.51 days (95% CI, −3.27 to −1.76), indicating a numerically greater reduction in the galcanezumab 120 mg arm compared to the placebo arm. The difference between the galcanezumab 120 mg and placebo arms could not be tested for statistical significance based on the predefined multiple-testing procedure because of a previously failed end point in the testing sequence (75% response rate; Figure 21 and Figure 22 in Appendix 3). Results of a post hoc exploratory subgroup analysis for the number of failed prophylactic treatments (≥ 2 versus ≥ 3) are summarized in Table 22. The results of this analysis were consistent with the analysis of overall change in MHDs with acute migraine medication use from baseline in the ITT population.

In the CONQUER trial, the overall change from baseline in MHDs with acute medication use during the double-blind treatment phase was a secondary outcome. The outcome was not included in the trial’s multiple-testing procedure, and therefore is not adjusted for multiplicity. The overall mean change from baseline in the number of monthly days with acute headache medication intake during the double-blind treatment phase was −3.40 days (95% CI, −4.14 to −2.65), indicating a numerically greater reduction in the galcanezumab arm compared to the placebo arm. The results of a planned subgroup analysis in patients with 3 or more prior preventive-medication category failures are summarized in Table 25, and are consistent with the results observed in the ITT population.

Duration of Effect and Re-Treatment Intervals

Time to First Loss of Response in the Post-Treatment phase

In the EVOLVE-1 and EVOLVE-2 trials, time to first loss of response among patients who met the 50% response rate criteria during their last month of double-blind treatment was a secondary outcome. This outcome was not controlled for multiplicity. Time to the first loss of 50% response in patients who were 50% responders at the last month of double-blind treatment and who entered the post-treatment phase in the EVOLVE-1 trial is shown in Figure 8. In the EVOLVE-1 study, approximately half of the patients in all treatment groups had first loss of 50% response by 4 months after the end of the double-blind treatment phase.¹³

Time to the first loss of 50% response in patients who were 50% responders at the last month of double-blind treatment and who entered the post-treatment phase in the EVOLVE-2 trial is shown in Figure 9. In the EVOLVE-2 study, approximately half of the patients in all treatment groups had first loss of 50% response by 4 months after the end of the double-blind treatment phase.¹³

In the REGAIN trial, time to first loss of response among patients who met the 50% response rate criteria during their last month of treatment (double-blind or open-label phase) was a secondary outcome. This outcome was not controlled for multiplicity. Time to the first loss of 50% response in patients who were 50% responders at the last month of double-blind treatment and who entered the post-treatment phase in the REGAIN trial is shown in Figure 10. In the REGAIN study, the percentage of patients with first loss of 50% response at month 1 of the post-treatment phase was 24.3%.¹³ By month 4 of the post-treatment phase, 48.2% patients had first loss of 50% response.

The CONQUER study design did not include a post-treatment follow-up phase; thus, time to first loss of response after stopping treatment was not assessed.

Figure 8: Time to First Loss of 50% Response After Double-Blind Treatment — EVOLVE-1 Trial, Patients Who Were 50% Responders at the Last Month of Double-Blind Treatment and Who Entered the Post-Treatment phase

Evts = events; LY = galcanezumab; Pts = patients; SE = standard error; t = times.

Note: This analysis only includes patients who were 50% responders at the last month of double-blind treatment and who entered the post-treatment phase. Patients who were sustained 50% responders during the entire post-treatment follow-up phase are censored.

Source: EVOLVE-1 Clinical Study Report Addendum.¹³

Figure 9: Time to First Loss of 50% Response After Double-Blind Treatment — EVOLVE-2 Trial, Patients Who Were 50% Responders at the Last Month of Double-Blind Treatment and Who Entered the Post-Treatment phase

Evts = events; LY = galcanezumab; Pts = patients; SE = standard error; t = times.

Note: This analysis only includes patients who were 50% responders at the last month of double-blind treatment and who entered the post-treatment phase. Patients who are sustained 50% responders during the entire post-treatment follow-up phase are censored.

Source: EVOLVE-2 Clinical Study Report Addendum.¹³

Figure 10: Time to First Loss of 50% Response After Galcanezumab Treatment — REGAIN, Patients Who Were 50% Responders in the Last Month of Treatment Before Entering the Post-Treatment phase

LY = galcanezumab.

Note: This analysis only includes patients who are 50% responders in the last month of treatment before entering the post-treatment phase. Patients who maintained 50% response during the entire post-treatment phase were censored at the end of the post-treatment phase.

Source: REGAIN Clinical Study Report Addendum.¹³

Time to Initiation of a Migraine-Prevention Treatment in Post-Treatment phase

Time to initiation of treatment with a migraine-prevention medication in the post-treatment follow-up phase was a secondary outcome in the EVOLVE-1, EVOLVE-2, and REGAIN trials. The outcome was not controlled for multiplicity.

Time to initiation of migraine-preventive medication during the post-treatment phase of the EVOLVE-1 trial is shown in Figure 11. In the EVOLVE-1 study, 12 patients (< 2%) initiated treatment with a migraine-prevention medication.¹³

Time to initiation of migraine-preventive medication during the post-treatment phase of the EVOLVE-2 trial is shown in Figure 12. In the EVOLVE-2 study, 3.2% of patients in the placebo arm initiated treatment with a migraine-prevention medication compared with 1.4% patients in the galcanezumab 120 mg arm.¹³

Time to initiation of migraine-preventive medication during the post-treatment phase of the REGAIN trial is shown in Figure 13. In the REGAIN study, 5.7% of patients who entered the post-treatment follow-up phase started a migraine-preventive medication during the post-treatment period.¹³

The CONQUER study design did not include a post-treatment follow-up phase; thus, time to initiation of a migraine-prevention treatment in after stopping galcanezumab treatment was not assessed.

Figure 11: Time to Initiation of Preventive Treatment After Double-Blind Treatment — EVOLVE-1, Patients Who Entered Post-Treatment Follow-Up

Evts = events; LY = galcanezumab; Pts = patients; SE = standard error; t = times.

Note: This analysis only includes patients who entered the post-treatment phase. Patients who never started preventive treatment are censored.

Source: EVOLVE-1 Clinical Study Report Addendum.¹³

Figure 12: Time to Initiation of Preventive Treatment After Double-Blind Treatment — EVOLVE-2, Patients Who Entered Post-Treatment Follow-Up

Evts = events; LY = galcanezumab; Pts = patients; SE = standard error; t = times.

Note: This analysis only includes patients who entered the post-treatment phase. Patients who never started preventive treatment are censored.

Source: EVOLVE-2 Clinical Study Report Addendum.¹³

Figure 13: Time to Initiation of Preventive Treatment After Galcanezumab Treatment — REGAIN, Patients Who Entered Post-Treatment Follow-up

Evts = events; LY = galcanezumab; Pts = patients; SE = standard error; t = times.

Note: This analysis only includes patients who entered the post-treatment phase. Patients who never started preventive treatment are censored.

Source: REGAIN Clinical Study Report Addendum.¹³

Health Care Resource Utilization and Employment Status

The EVOLVE-1 and EVOLVE-2 trials did not assess HCRU and employment status; both were secondary outcomes in the REGAIN and CONQUER trials. The outcomes were not controlled for multiplicity.

In the REGAIN trial, the percentages of patients with health care professional visits related to migraine in the baseline period (6-month period before randomization) were 20.4% in the galcanezumab 120 mg arm and 18.3% in the placebo arm. Emergency room visits related to migraine were reported for 6.6% of patients in the galcanezumab 120 mg arm and 4.9% of patients in the placebo arm. Two patients (0.73%) in the galcanezumab arm and 4 patients (0.72%) in the placebo arm had an overnight hospital admission related to migraine during the baseline period. In the 3-month double-blind treatment period, the percentages of patients with health care professional visits related to migraine were 6.0% in the placebo group and 3.7% in the galcanezumab 120 mg group. Emergency room visits related to migraine were reported for 2.4% of placebo patients and 1.8% of galcanezumab 120 mg patients during the double-blind treatment phase. No patients had an overnight hospital admission related to migraine during the double-blind treatment period.

In the CONQUER trial, the percentages of patients with health care professional visits related to migraine during the 3-month double-blind treatment phase were |||| in the galcanezumab group and |||| in the placebo group. In the baseline period (i.e., 　|　 months before randomization), the percentages of patients with health care professional visits related to migraine were |||| in the placebo group and |||| in the galcanezumab group. Few patients had emergency room visits related to migraine (|||||||||||||||||||||||| |||| ||||||) or hospitalizations related to migraine (|||||||||||||||||||||||||||||||) during the double-blind treatment phase, and 　|　|||||| had an overnight hospital admission related to migraine.

Because of the low rates of HCRU events and because of the different time periods assessed at baseline (a period of 6 months before randomization) and post-baseline in both the REGAIN and ||||||||||||||||||, rates for the migraine-related events were standardized per 100 patient-years. Migraine-related HCRU per 100 patient-years is summarized in Table 20 for the REGAIN trial and Table 23 for the CONQUER trial.

Patients’ employment status at month 0 and month 3 of the double-blind treatment phase is summarized in Table 20 for the REGAIN trial and Table 23 for the CONQUER trial.

Work Productivity and Activity Impairment Questionnaire

The WPAI questionnaire was not assessed in the EVOLVE-1, EVOLVE-2, and REGAIN trials.

In the CONQUER trial, change from baseline in the WPAI was a secondary outcome. The outcome was not controlled for multiplicity. The LS mean change from baseline to the LOCF end point and SE was reported for percent of activity impairment due to health, percent of impairment while working due to health (presenteeism), percent of overall work impairment due to health, and percent of work time missed due to health (absenteeism). The 95% CIs were not reported. The LS mean change for activity impairment was −20.71% (SE = 1.95%) and −8.64% (SE = 1.92%) in the galcanezumab and placebo arms, respectively. For presenteeism, the LS mean change (SE) was −12.50% (SE = 2.37%) and −2.56% (SE = 2.32%) in the galcanezumab and placebo arms, respectively. For overall work impairment, the LS mean change was −14.31% (SE = 2.51%) and −3.46% (SE = 2.41%) in the galcanezumab and placebo arms, respectively. For absenteeism, the LS mean change was −4.22% (SE = 1.29%) and −2.90% (SE = 1.24%) in the galcanezumab and placebo arms, respectively.