Background and objective As new migraine prevention treatments are developed, the onset of a preventive effect, how long it is maintained and whether patients initially non-responsive develop clinically meaningful responses with continued treatment can be assessed.
Methods Analyses were conducted post-hoc of a double-blind, placebo-controlled, phase II-a study in patients with episodic migraine receiving galcanezumab 150 mg or placebo biweekly for 12 weeks (Lancet Neurol 13:885, 2014). The number of migraine headache days per week, and onset of efficacy measured as the first week galacanezumab separated from placebo were determined. Patients with ≥50%, ≥75% and 100% reduction in migraine headache days from baseline at months 1, 2 and 3 were calculated and defined as sustained responses. Non-responders (<50% response) at month 1 or 2 who then showed ≥50%, ≥75% and 100% response at later time-points were calculated.
Results Patients were randomised to galcanezumab (n=107) or placebo (n=110). A significant (p=0.018) change of −0.89±0.11 (galcanezumab) vs −0.53±0.11 (placebo) migraine headache days indicated onset at week 1. Forty-seven per cent of galcanezumab and 25% of placebo patients responding at month 1 maintained response through months 2 and 3. Of non-responders at month 1, 27% on galcanezumab and 20% on placebo responded on months 2 and 3, and 50% of galcanezumab non-responders in months 1 and 2 responded on month 3, vs 24% on placebo.
Conclusions The onset of efficacy of galcanezumab is within 1 week in a majority of patients, and patients receiving galcanezumab are twice more likely to maintain responses than placebo patients. Early non-responders may respond by month 2 or month 3.
Trial registration number NCT01625988.
- calcitonin gene-related peptide
- clinical trial
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
Novel preventive treatments that are effective and well tolerated are highly desirable given the number of individuals who suffer from frequent migraine1 and the burden migraine causes.2 Few drugs are approved by regulatory authorities for migraine prevention; with the exception of the recently approved humanised monoclonal antibodies, none are migraine-specific,3 only a small fraction of patients receive preventive treatment,1 and they are associated with undesirable side effects and low adherence rates.4 Although the postmarketing side effects and adherence are not yet known for the monoclonal antibodies, they were well tolerated in clinical trials.
Advances in our understanding of the pathogenesis of migraine5 have unveiled several potential drug targets for both acute and preventive treatment.6 The neuropeptide calcitonin gene-related peptide (CGRP) is found throughout the trigeminovascular complex7 and in central brain regions.8–11 This neuropeptide is regarded as important in the pathophysiology of migraine.11 CGRP’s role in migraine has been supported by several experimental and clinical findings: during spontaneous migraine attacks, the jugular venous blood concentration of CGRP increases.12 Interictal blood concentrations are significantly elevated in patients with episodic and chronic migraine, suggesting that elevated CGRP levels are not simply symptomatic, but may even serve as a biomarker for disease activity.13 Intravenous infusion of recombinant human CGRP can trigger a migraine attack that is indistinguishable from a spontaneous attack,14 and raised CGRP serum concentrations can be reversed with triptan administration—an effect that coincides with migraine symptom relief.15 16 In addition, small-molecule CGRP receptor antagonists have been shown to be effective in the acute17–22 and preventive23 treatment of migraine headache in double-blind, randomised, placebo-controlled trials.
Challenges in small molecule development and understanding the important role of CGRP in migraine have led to the development of monoclonal antibodies against the CGRP pathway. One of these monoclonal antibodies against the CGRP pathway is galcanezumab, a humanised monoclonal antibody that potently and selectively binds to CGRP. In a phase II-a study, the primary analyses evaluated the safety and efficacy of galcanezumab for the preventive treatment of episodic migraine24; the results provided preliminary evidence that galcanezumab was effective and generally well tolerated for the prevention of episodic migraine. These data have been substantiated by subsequent studies that have shown efficacy of galcanezumab in migraine prevention.25–27
As evidence for the efficacy of galcanezumab and other CGRP pathway monoclonal antibodies for the prevention of migraine has become available, with galcanezumab, fremanezumab and erenumab having received approval from the Food and Drug Administration for this indication,28–32 questions that are important to patients and clinicians remain, such as how quickly a patient can expect relief, whether relief will be sustained over the course of treatment, and what proportion of the patient population can expect to see relief over the course of treatment, even if they fail to respond initially by having at least a 50% reduction in migraine headache days (MHDs) within the first month or first 2 months of treatment.
Here, we present post-hoc analyses from the phase II-a study data24 to explore how early efficacy of galcanezumab can be observed using weekly measures, in contrast to efficacy assessments made using monthly measures previously presented, to evaluate sustained response over the 3-month study period, and to determine whether a cumulative response occurs over time. These post-hoc data have been presented in abstract form previously.33 34
Post-hoc analyses were conducted using data from a randomised, double-blind, placebo-controlled, phase II-a study (beginning June 2012 and concluding September 2013) with patients at 35 centres in the USA randomly assigned (1:1) to galcanezumab (150 mg subcutaneous injection every 2 weeks) or placebo for 12 weeks. The study consisted of four phases: (1) a screening period to assess inclusion/exclusion criteria and discontinuation of any excluded medications; (2) a baseline period to record the frequency and symptoms of migraine headaches; (3) double-blind, placebo-controlled treatment period of 12 weeks (month 1, 2 or 3) with injection of study drug every 2 weeks; and (4) a post-treatment follow-up period of an additional 12 weeks for continued safety assessment (figure 1).24
Key inclusion criteria included men and women aged 18–65 years with at least 1-year history of migraine as defined by the International Classification of Headache Disorders (ICHD-II)35 and experiencing 4–14 MHDs per month. In addition, the patient must have discontinued any drug or other treatment to prevent migraine headaches for at least 30 days before visit 2 (baseline). Specific details of the study design, patient population, procedure and primary outcomes of this phase II-a study have been previously reported.24
An MHD was defined as a calendar day on which a headache lasting >30 min and that met the remaining criteria for migraine as defined by ICHD-II24 35 occurred. Probable migraine was also defined according to the standard ICHD-II definition as a headache of >30 min duration, with or without aura, fulfilling all but one of the criteria for migraine headache and not attributed to another disorder.35 Probable MHD was defined as a calendar day when a probable MHD occurred. Responders were defined as patients who had a ≥50% reduction in the number of MHD in a 28-day period during treatment.
Statistical analysis for onset of efficacy
The aim of this analysis was to determine the onset of efficacy of galcanezumab. To assess the onset of efficacy, daily diary data indicating whether the patient had a migraine headache (‘yes’ vs ‘no’) during the 12-week treatment period were aggregated into the number of MHD for each weekly interval. Onset of efficacy was defined as the first week in which galcanezumab was statistically superior to placebo in reducing MHD per week (p<0.05).
For derivation of weekly MHD, weekly baseline value was calculated as the number of MHD per 28-day period at baseline divided by 4. For postbaseline data, first biweekly injection schedules at weeks 0, 2, 4, 6, 8 and 10 were used to cut daily data into biweekly intervals. Then biweekly intervals are divided into two equal weekly intervals by identifying the midpoint. Change from baseline in weekly MHD was analysed with mixed model repeated measures analyses with fixed covariate of treatment, weeks, treatment-by-week interaction and baseline weekly number of MHD. Least squares means of weekly treatment differences were calculated based on the mixed model repeated measures analysis. An unstructured covariance structure was used to model the within-patient errors.
Statistical analysis for response outcomes
Here, analyses assessed the proportion of patients with ≥50%, ≥75% and 100% reduction in MHD from baseline at month 1 who sustained those response levels for months 2 and 3 (defined as ‘sustained response’). The difference for the proportion of patients meeting 50% sustained response between galcanezumab and placebo was analysed using a categorical, pseudo-likelihood-based, repeated measures analysis of binary outcome indicating whether patients met the ≥50% (or the ≥75% or the 100%) sustained response criteria (SAS GLIMMIX).
A patient was defined as a non-responder if the percentage improvement from baseline in MHD is <50% or the patient discontinued early from the treatment phase. Furthermore, to characterise subsequent response outcomes for non-responders at month 1, the proportions of patients with ≥50%, ≥75% and 100% response at months 2 and 3 were calculated and compared between treatment groups; for non-responders at months 1 and 2, the proportions of patients with ≥50%, ≥75% and 100% response at month 3 were calculated and compared between treatment groups with logistic regression. All statistical analyses—for both onset and response outcomes—were conducted using SAS V.9.2. All p values presented are nominal without multiplicity adjustment.
Demographics and baseline disease characteristics
A total of 217 intent-to-treat (ITT) patients were randomised and received galcanezumab (n=107) or placebo (n=110) in the phase II-a study.24 Baseline demographics and disease state characteristics for galcanezumab and placebo groups, respectively, included an average age of 41±11 and 42±12 years; gender distribution of 88 (82%) and 96 (87%) women, and 19 (18%) and 14 (13%) men; mean number of MHD of 7±2 and 7±2 per 28 days during the baseline period; and MHD or probable MHD of 8±3 and 8±3 per 28 days during the baseline period. The post-hoc analyses for sustained response was performed with the modified ITT population, consisting of the ITT patients who have non-missing values at months 1, 2 and 3, and included 99 patients receiving galcanezumab and 105 patients on placebo. The baseline demographics and disease state characteristics for galcanezumab and placebo groups for patients classified as responders (≥50% reduction in MHD) and non-responders (<50% reduction in MHD) are presented in table 1. The mean number of MHD or probable MHD for galcanezumab and placebo groups for patients classified as responders (≥50% reduction in MHD) was 8±2 and 9±3, respectively, per 28 days during the baseline period, whereas the mean number of MHD or probable MHD for galcanezumab and placebo groups for patients classified as non-responders (<50% reduction in MHD) was 9±3 and 8±3, respectively, per 28 days during the baseline period (table 1).
Onset of efficacy
Analyses of mean change in weekly MHD from baseline showed that onset of efficacy of galcanezumab started at week 1 after the first injection as demonstrated by a statistically significant reduction in mean change in weekly MHD from baseline of −0.89±0.11 (galcanezumab) vs −0.53±0.11 (placebo) at week 1 (p=0.018); there was no statistically significant treatment-by-week interaction. A higher proportion of galcanezumab-treated patients (compared with placebo-treated patients) responded at week 1, with 62% of galcanezumab-treated patients having a ≥50% reduction in the number of weekly MHD compared with 42% of patients on placebo. This difference between galcanezumab-treated patients and placebo-treated patients for ≥50% reduction in the number of weekly MHD at week 1 was statistically significant (p<0.05). The mean reduction in weekly MHD in the galcanezumab-treated group remained statistically significantly greater (p<0.05) compared with the placebo-treated group at all weeks1–5 7 8 10–12 except for week 6 and week 9 (figure 2A).
Analyses of mean change in weekly migraine or probable MHD from baseline showed that the onset of efficacy of galcanezumab also started at week 1 after the first injection as demonstrated by a statistically significant reduction in mean change in weekly MHD or probable MHD from baseline at week 1 (p=0.002); no statistically significant treatment-by-week interaction was observed. The mean change in weekly migraine or probable MHD from baseline in the galcanezumab-treated group was statistically significantly greater (p<0.05) compared with the placebo-treated group at most weeks (figure 2B).
The proportions of galcanezumab-treated versus placebo-treated patients meeting ≥50%, ≥75% and 100% sustained response at months 1, 2 and 3 are noted in figure 3. Forty-seven per cent of galcanezumab-treated patients (n=99) vs 25% of patients on placebo (n=105) (p<0.001) responded (ie, ≥50% response rates) at months 1, 2 and 3. There was no significant difference between galcanezumab-treated versus placebo-treated patients, 22% and 13%, respectively, who met ≥75% response at months 1, 2 and 3 (figure 3). The 11% of galcanezumab-treated patients was statistically significantly greater (p<0.05) than the 2% of placebo-treated patients who met 100% response at months 1, 2 and 3 (figure 3). Among patients who were ≥50% responders at month 1, 27% of galcanezumab-treated vs 45% of placebo-treated patients either did not sustain a ≥50% response at month 2 or 3, or discontinued early.
Subsequent response in initial non-responders
Subsequent response outcomes for non-responders at month 1 and non-responders at months 1 and 2 are reported in figures 4 and 5, respectively. Among month 1 non-responders (ie, patients who did not have at least a 50% reduction in MHD at month 1), 27% (galcanezumab; n=41) vs 20% (placebo; n=61) of patients responded at both months 2 and 3 (figure 4). Seventeen per cent of galcanezumab-treated vs 3% of placebo-treated (p<0.05) patients experienced at least a 75% reduction in MHD at months 2 and 3, and 10% of galcanezumab-treated vs 2% of placebo-treated patients experienced 100% reduction in MHD at months 2 and 3; only the ≥75% response rate was statistically significantly different (figure 4).
Of the patients who did not respond (ie, have ≥50% reduction in MHD) at months 1 and 2, 50% of those who were treated with galcanezumab (n=22) vs 24% of those who were treated with placebo (n=46) (p≤0.05) subsequently experienced at least a 50% reduction in MHD at month 3, 18% of galcanezumab-treated patients vs 4% of those treated with placebo experienced at least a 75% reduction in MHD at month 3, whereas no galcanezumab-treated patients and 4% of placebo-treated patients experienced 100% reduction in MHD at month 3; only the ≥50% response rate was statistically significantly different (figure 5).
This post-hoc analysis suggests the onset of efficacy of galcanezumab appears to begin within 1 week after the first dose in a substantial number of patients, and once improvement is achieved it is sustained during the 3-month treatment period in a substantial proportion of patients. In addition, 47% of patients who responded to galcanezumab at month 1 continued to respond at months 2 and 3. Importantly, there is also a response in a proportion of patients that may occur at months 2 and 3 despite the lack of a response 1 month after initiation of treatment. Further, response may still occur in a subset of patients at month 3 despite the lack of a response 1 or 2 months after initiation of treatment.
Little is known about the onset of efficacy, sustained response and response over time despite an initial non-response in migraine prevention trials. Rapid onset of efficacy is considered to be important for patient compliance, adherence to treatment, and therefore favourable long-term clinical outcomes,36 although any promise for onset needs to be tempered against the consequences of disappointment. Given that these are important questions for both the patient and the physician, these types of analyses could usefully be carried out more frequently in migraine preventive studies. Unlike with acute therapies, there is also a paucity of research evaluating the most important attributes of preventive medications among patients. In one study conducted on patients in Brazil and the USA, overall efficacy and the speed of onset in particular were considered by the majority of patients to be the most important attributes of preventive medications.37 However, the importance of a sustained response was not addressed in that study. In a study in a headache specialty practice, issues such as being involved in care decisions, knowledge of side effects and being prescribed treatments with published efficacy ranked highly.38 Our data are supported by similar onset of action analyses in the CGRP monoclonal antibody class that support separation of active treatment from placebo by 1 week39 40 or indeed earlier.41 The rapidity of onset of effect reinforces the importance of CGRP in the pathophysiology of migraine, its role in the genesis of headache in people with migraine and the rapid target engagement of the antibodies.
Non-planned analyses have important limitations, particularly where the parent studies were not designed to evaluate the time frame of the analyses, here weekly efficacy outcomes. Due to the nature of post-hoc analyses, no multiplicity adjustment was conducted; thus, the results need to be interpreted under the consideration that false positives may occur. More comprehensive data are needed to confirm the findings in these post-hoc analyses. The evaluation of mean changes in migraine frequency over the course of a week in a population of patients may not accurately reflect overall changes in migraine frequency for an individual patient. Migraine frequency is variable, and weekly assessment of migraine frequency is probably not as stable as monthly assessment of migraine frequency. In particular, calculations for weekly change in MHD in patients who have migraine frequency at the lower end of the range (eg, 4 MHDs/month) may have a greater risk for inaccurate frequency interpolations than those with higher frequency MHD. Moreover, by definition episodic migraine frequency limits the frequency of MHD and thus the granularity of measurement of changes in that frequency. Chronic migraine is more suitable with its greater frequency to explore onset of action questions in migraine prevention. In addition, for the non-responders at month 1 or at months 1 and 2, only a subset of the randomised patients are included in the analyses; therefore, statistical comparison between treatment groups is limited due to potential selection bias.
Despite these limitations, the results for responders and non-responders suggest a sustained and cumulative benefit over time in subgroups of patients treated with galcanezumab. These types of data will be essential to build evidence-based clinical guidelines for treatment initiation and duration of a treatment trial with this new class of treatment. Future studies should prospectively examine the earliest onset of a significant treatment effect, as well as sustained and delayed response outcomes as they are important treatment outcomes for patients and helpful in guiding clinical decision making for physicians.
These post-hoc analyses attempt to begin to address clinically important questions about the onset of efficacy, sustained response and probability of a delayed response in patients with migraine, and suggest that galcanezumab has an onset of action 1 week after treatment is started. In a proportion of patients who respond within 1 month, the response is sustained over the subsequent 2 months. In those patients who had less than robust response to galcanezumab in the first month or first 2 months of treatment, a subsequent response in a relevant proportion of patients may be possible.
The authors would like to thank Michael H Ossipov, PhD, for writing, formatting and editing assistance with this manuscript.
Contributors PJG contributed to the conception and interpretation of data of the work, drafting of the manuscript, and critical revision for important intellectual content. DWD contributed to the conception of the work, analysis and interpretation of data, and critical revision of the manuscript for important intellectual content. JMM contributed to the conception of the work, interpretation of data and critical revision of the manuscript for important intellectual content. MBF contributed to the conception and design of the work, interpretation and analysis of data, drafting of the manuscript, and critical revision for important intellectual content. TMO contributed to the conception of the work, interpretation of data and critical revision of the manuscript for important intellectual content. QZ contributed to the conception and design of the work, analysis and interpretation of data, drafting of the manuscript, and critical revision of the manuscript for important intellectual content. VS contributed to the conception of the work, interpretation of data and critical revision of the manuscript for important intellectual content. SKA contributed to the interpretation of data and critical revision of the manuscript for important intellectual content.
Funding This work was supported by Eli Lilly and Company, Indianapolis, Indiana, USA
Competing interests The studies included in the analyses were sponsored and/or supported by Eli Lilly and Company. VS, MBF, JMM, TMO and SKA are full-time employees of Eli Lilly and Company and/or one of its subsidiaries, and are stockholders. QZ was an employee of Eli Lilly and Company at the time the manuscript was written and is presently an employee of Sanofi and is an Eli Lilly and Company stockholder. PJG reports grants and personal fees from Amgen and Eli Lilly and Company; personal fees from Alder BioPharmaceuticals, Allergan, Autonomic Technologies, Dr Reddy’s Laboratories, Biohaven Pharmaceuticals, Electrocore, eNeura, Novartis,Impel Neuropharma, Mundipharma, Teva Pharmaceuticals and Trigemina; and personal fees from MedicoLegal work, UptoDate, Oxford University Press, Massachusetts Medical Society and Wolters Kluwer; and a patent magnetic stimulation for headache assigned to eNeura without fee. Within the last 12 months, DWD reports personal fees from Amgen, lder, Allergan, Autonomic Technologies, Biohaven, Eli Lilly, eNeura, Foresight Capital, Neurolief, Zosano, WL Gore, Vedanta Associates, Promius Pharma, Nocira, Novartis, Electrocore, Teva, Ipsen, Impel, Satsuma, Charleston Laboratories and Theranica; compensation for activities related to data safety monitoring committee from Axsome; compensation related to CME content development: HealthLogiX, Medicom Worldwide, MedLogix Communications, MedNet, Miller Medical Communications, PeerView Operation Services America, WebMD/Medscape, American Academy of Neurology, American Headache Society, PeerView Institute for Medical Education, Chameleon Communications, Academy for Continued Healthcare Learning, Universal Meeting Management, Haymarket Medical Education, Global Scientific Communications, UpToDate and Meeting LogiX; royalties from editorial or book publishing: Oxford University Press, Cambridge University Press, Wiley-Blackwell, Sage and Wolters Kluwer Health; consulting use agreement through employer: Neuro Assessment Systems and Myndshft; holds equity in Aural Analytics, Healint, Theranica, Second Opinion/Mobile Health and Epien; and board of directors position: King-Devick Technologies and Ontologics.
Patient consent for publication Not required.
Ethics approval The study was approved by the appropriate institutional review board for each study site.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.