2023 Impact Factor
2023 Impact Factor
Migraine is a primary headache disorder with the second-highest number of cases among disabling diseases globally [1]. It is characterized by headaches that can be of varying intensity, throbbing, and accompanied by associated symptoms (nausea, vomiting, photophobia, and phonophobia), which have been shown to have a negative impact on the quality of life [1,2]. Chronic migraine (CM) is characterized by headaches occurring at least 15 days per month, with at least eight days of migraine per month [2]. Albeit various pharmacological interventions are available for the treatment and prevention of migraine, the lack of efficacy and side effect profile of the available standard preventive therapies warrants the development of newer and better preventive modalities. Due to the unmet needs in the pharmacotherapeutic management of migraine, the past decade has seen focused research on novel preventive therapeutics for migraine to improve the quality of life [3]. Studies have discovered that the calcitonin gene-related peptide (CGRP), a multifunctional neuropeptide that plays an important role in the pathogenesis of migraine, is a promising target for migraine treatment and prevention as well [4-6].
CGRP, a 37-aminoacid peptide, is a potent vasodilator released from trigeminal neurons and plays a key role in migraine pathophysiology [7]. The potential therapeutic role of the CGRP blockade was assessed three decades ago in the form of CGRP receptor antagonists, but their short half-lives have restricted their usage to acute migraine only [8]. Concurrently, certain schools of thought have focussed their research on monoclonal antibodies (mAbs) blocking the CGRP pathways. The utility of mAbs as therapeutic modalities lies in part with their target-specificity, longer elimination half-lives in humans, as well as decreased severe adverse events such as drug-drug interactions and hepatotoxicity [9].
A literature search has shown that presently there are four mAbs targeting the CGRP, namely, eptinezumab, fremanezumab, galcanezumab, and erenumab [10-12]. The former three are humanized mAbs which selectively bind to CGRP and are potent agents, while erenumab is the only mAb that targets the CGRP receptor complex instead of the CGRP ligand [13,14].
Although several research articles have individually discussed the efficacy of various mAbs targeting CGRP in chronic migraine prevention [15-18], there are no studies providing a robust conclusion regarding the relative clinical efficacy of these mAbs when compared to each other in the prevention of chronic migraine. Thus, the present network meta-analysis (NMA) study was conducted to analyze various mAbs in chronic migraine and thus further enable the use of these agents in a clinical setting and improve migraine management in clinical practice.
The study protocol was developed according to the Preferred Reporting Items for Systematic Review Protocols (PRISMA-P) statement guidelines. It was registered with the International Prospective Register of Systematic Re-views (PROSPERO) database (CRD42021232659).
Adult patients of both sexes with a diagnosis of chronic migraine based on The International Classification of Headache Disorders (ICHD-3) criteria [2] were included in the NMA.
Types of interventionsAny anti-CGRP mAb, irrespective of the dose, frequency, and duration of administration, was considered as the intervention. Patients receiving standard medication, other anti-CGRP mAb, or a placebo were regarded as the control.
OutcomesOutcomes analyzed in this NMA are as follows:
Primary:• Comparison of change in mean monthly migraine days in patients with chronic migraine at the end of the 12-week follow-up period between the intervention and control groups.
Secondary:• Comparison of responder rate (percentage of patients having ≥ 50% reduction in monthly migraine days) at the end of the 12-week follow-up period between the intervention and control groups.
• Comparison of treatment-emergent adverse events between the intervention and control groups.
The literature search was done on PubMed, MEDLINE, the Cochrane database, and the International Clinical Trial Registry Platform (ICTRP). The search strategy and the “Medical Subject Headings” terms used for the literature search have been provided in the Supplementary Material (available online). A randomized controlled trial published in English comparing the mAbs for migraine prophylaxis, with each other, with the standard of care, or with a placebo was considered eligible to be included in the NMA. Two authors (HMP and AT) independently screened the titles and abstracts of all the studies obtained following the search. The full texts of relevant studies were retrieved and assessed for eligibility criteria for inclusion in the NMA. In case of any disagreement, it was resolved through discussion with another author (RM), and reasons for the exclusion of studies, if any, were mentioned.
Following the screening of research articles, three authors evaluated all of the included papers before the data extraction process. The author AS confirmed the data, a clinical pharmacologist who also served as a statistical advisor.
The information which was gathered from each rando-mized controlled trial included authorship, year of publication, the arms of the interventions (experimental and control), and the sample size associated with each trial. The data pertaining to the endpoints, such as mean migraine days, percentage responders, and various adverse events, were also collected.
The Risk of Bias 2 (ROB2) tool was used to assess the quality of each randomized controlled trial [12]. This tool attempts to evaluate the many methods through which bias could be introduced into a randomized controlled trial (RCT) across six domains. The domains included the bias arising from the randomization process, bias due to deviations from intended interventions, bias due to missing outcome data, bias in the measurement of the outcome, bias in the selection of the reported result, and overall bias. The risk of bias judgment under each domain was scored as either “low risk of bias,” “some concerns,” or “high risk of bias.” Based on the responses given for the individual domains, the overall risk of bias for each RCT was also decided and reported.
Estimates of the pooled mean difference (MD) in migraine days were used to analyze the primary outcome. According to the recommendation made by Cochrane, the PlotDigitizer was used to extract data from graphs when the studies did not include numerical values in the published publication [19]. If the data was reported as the median and interquartile range (IQR), it was transformed into the mean and standard deviation using the formula provided by Wan et al. [20]. The odds ratio (OR) was used to assess the responder rate and treatment-emergent side events between the intervention and control groups in this study.
When doing a NMA, data is gathered to make direct pairwise comparisons of interventions and to calculate indirect evidence between them and between them and any other interventions being evaluated in the NMA. A random-effects model built on the Bayesian framework was used to conduct this NMA, which was carried out with the help of the “gemtc” package and R v4.0.3 [21]. A network graph was constructed individually for mean migraine days, responder rate, and adverse events showing the amount of direct evidence available for each of these measures. It was decided to show all indirect or mixed evidence calculated from all potential combinations of the various interventions in a league table format. The data were presented as the MD in the case of mean migraine days and the OR for responder rate and adverse events. Credible intervals (95%CrI) were also determined for all endpoints.
The various therapies were ranked from best to worst based on their efficacy (change in mean migraine days and responder rate) and safety (as measured by the number of people who experienced migraine treatment-emergent adverse effects) based on the surface under the cumulative ranking (SUCRA) curve analysis. SUCRA analysis would serve as a valuable tool to decide on the overall efficacy of various mAbs against CGRP ligands and receptor complexes in migraine. We also estimated the inconsistency concerning the direct and indirect comparison results, if any, using node splitting analysis and overall model diagnostics.
Based on the SUCRA rankings obtained for change in mean monthly migraine days and the OR for treatment- emergent adverse effects, we plotted the efficacy against the safety of these mAbs to have an idea about the risk- benefit analysis of these drugs while using them in the prevention of chronic migraine.
We initially identified 319 potentially relevant clinical trials on mAbs against CGRP approved for migraine in adults and children till January 2023. The full texts of 29 articles were assessed. Among them, 19 studies were deemed ineligible for the reasons listed (Fig. 1). Finally, ten studies were included for NMA [12,16,22-29].
The characteristics of all the included studies have been described in Supplementary Table 1 (available online). The risk of bias for each study and the overall risk of bias was assessed using the tool ROB2. The risk of bias chart and risk of bias for individual studies are described in Supplementary Figure 1 (available online) and Supplementary Table 2 (available online), respectively.
The network graph shows the available direct comparisons of mean migraine days (Fig. 2A) among various mAbs used in the prevention of migraine. Mean migraine days were reported as an outcome measure in 10 RCTs. As shown in the network graph, the majority of the direct comparisons have been studied between placebo and fremanezumab (monthly), placebo and fremanezumab (quarterly), fremanezumab (monthly), and fremanezumab (quarterly) for all the endpoints (Fig. 2).
Inconsistency in an NMA refers to the differences in treatment effects of a pair of interventions across different trials. It is analyzed using node split analysis. In the node splitting analysis run on the direct comparisons for mean migraine days, responder rate, and adverse event profile, the inconsistency factor was found to be non-significant for fremanezumab (quarterly) compared with placebo (p = 0.56). The node-splitting model could not be generated for other treatment options. Therefore, inconsistency for the overall model for all three outcomes was assessed by the goodness of fit.
The model fit statistics for all three endpoints show residual deviance (27.7) and are comparable to the number of data points (27). Thus, the consistency of the model can be assured. Residual deviance per arm has been mentioned in Supplementary Table 3 (available online).
The interventions were compared based on mean migraine headache days. Also, the OR for responder rate and adverse events was calculated and compared between the intervention. The efficacy may be evaluated using SUCRA values, which show the relative ranking of each intervention in lowering migraine headache days and improving response rates.
Galacanezumab 120 mg (MD, −2.7; 95%CrI, −4.8 to −0.83) followed by eptinezumab 300 mg (MD, −2.6; 95%CrI, −5.3 to −0.0034) had shown the highest efficacy in reduction of mean migraine days as compared to the other interventions and placebo. Amongst the various dose routines and dosages of fremanezumab, the monthly dose showed the highest efficacy as compared to others (MD, −1.9; 95%CrI, −3.1 to −0.79) (Fig. 3A). In the case of response to these mAbs, in terms of responder rate (> 50% reduction in migraine headache days), the fremanezumab group of mAbs had the highest efficacy. The fremanezumab quarterly dose regimen showed the best response (OR, 2.9; 95%CrI, 1.9−4.6) followed by fremanezumab monthly dose administration (OR, 2.9; 95%CrI, 1.9−4.3) (Fig. 3B).
The safety was assessed by comparing the odds of treatment-emergent adverse events associated with each intervention and their comparator (Fig. 3C). Eptinezumab was found to be safer and have fewer adverse effects than other therapies (mAbs) and even a placebo in terms of safety (OR, 0.88; 95%CrI, 0.61−1.3) (Fig. 3C).
Pooled MDs along with 95%CrI of the various direct and indirect pairwise evidence built from the model have been represented in the league tables for the difference in mean migraine days (Supplementary Table 4; available online), the OR of responder rate (Supplementary Table 5; available online), and adverse events (Supplementary Table 6; available online) associated with mAbs. A rankogram is produced for all the interventions that show the relative efficacy of the drugs in reducing migraine days (Fig. 4), response rate (> 50% reduction in mean migraine days) (Fig. 5), and the rate of adverse events based on the SUCRA score of each. Higher the SUCRA score, the better the intervention in the prophylaxis of chronic migraine.
Galcanezumab 120 mg has shown the most considerable reduction in mean migraine days, followed by eptinezumab and erenumumab. SUCRA ranking for change in mean migraine days showed the following results (Fig. 4A).
Among the various mAbs compared in response rate, fremanezumab quarterly, followed by fremanezumab monthly, had the best response. SUCRA ranking based on highest responder to lowest responder rate drug is as follows (Fig. 4B).
Safety was assessed by the number of adverse events associated with each intervention, where eptinezumab 100 mg showed a favorable safety profile (Fig. 4C). A scatter plot was plotted to show the relative association between safety and efficacy (Fig. 5). Among the various mAbs assessed above, a scatter plot was developed, which showed the comparative assessment between safety and efficacy, and it was found that eptinezumab 100 mg had a favorable safety and efficacy profile. The funnel plot was not constructed as all the comparisons had fewer than 10 studies [30].
The present NMA compared various pharmacological interventions for the prophylactic treatment of chronic migraine using a Bayesian approach. Although there have been many meta-analyses [31,32] that have been conducted to compare the efficacy of various mAbs in the prophylactic management of migraine, this is the first NMA using a Bayesian approach that has been undertaken to assess and compare various mAbs against CGRP in the prevention of chronic migraine in adults.
Since migraine is a debilitating disorder affecting the quality of life of people around the globe, effective management consists of not only slowing the important pathophysiology associated with migraine but also preventing further attacks of migraine. Thus, the present NMA was carried out to find the relative efficacy and safety of various mAbs approved by the US FDA in preventing a migraine attack.
This NMA focused on four approved mAbs used as preventive medications for chronic migraine. It was reported in various trials that all four mAbs, their different dosage forms, and their schedules used in various studies had shown similar efficacy in terms of reduction in mean migraine days, which was also the primary outcome variable in our study. The change in mean migraine days was common in all the selected RCTs and was the primary outcome variable in all the studies as well. However, in our study, galcanezumab was found to be the most effective mAb in reducing migraine headache days, followed by Eptinezumab. Response rate (> 50% reduction in mean mi-graine days) was also analyzed in our study. Fremanezumab (quarterly) had better efficacy in terms of responder rate. These results are consistent with the meta-analyses done on various mAbs against CGRP for the treatment of episodic and chronic migraine [33,34], even though these studies focussed on treatment rather than the prevention of migraine which is the focus of this study.
Treatment-emergent adverse events were assessed to analyze the safety factor with these novel agents against CGRP for prophylaxis. Eptinezumab showed a favorable safety profile in comparison to other mAbs, and when compared to a placebo, galcanezumab was found to be much more likely to elicit TEAEs and major adverse events. These findings are also consistent with the safety analysis from the meta-analysis by Messina et al. [35], even though our study focused on efficacy, and safety being the secondary evaluation. Current research has demonstrated that several of these innovative medications are efficacious and well-tolerated over the long term; however, caution should be exercised in light of the possible dangers of inducing hypertension and worsening ischemic stroke, which has been documented by the most recent trials [24,36].
We were able to compare pharmaceutical classes among each other and with placebo both directly and indirectly using NMA, which often provides a more exact estimate of the relative efficacy and safety of a medicine class than paired comparisons. We were able to combine direct and indirect comparisons among several types of CGRP mAbs to rank their efficacy and safety for migraine therapy through network analysis.
The results of the current study revealed that fremanezumab was more effective than eptinezumab in terms of enhancing 50% response rates; nevertheless, the difference was just a marginal improvement. The findings of this study should be confirmed by more head-to-head comparisons.
According to the present American Headache Society/ American Academy of Neurology guidelines, there have been no recommendations for mabs as preventive medications for migraine [37]. Even though they have comparable efficacies and safeties when compared to standard treatment. Only erenumab has been recommended by US FDA as an alternative for preventive migraine ther-apy. Several mAbs against calcitonin gene-related peptides or receptors, such as eptinezumab, erenumab, fremanezumab, and galcanezumab, are recommended in the European Headache Federation (EHF) Guidelines for migraine therapy [38]. They did not, however, offer any recommendations as to which medicine was the most successful [38,39].
The present NMA suggests that fremanezumab (quarterly) has the best efficacy in terms of response rate compared to other CGRP mAbs, and Eptinezumab has the best safety profile in terms of the lowest treatment-related adverse events and also a better efficacy profile. In the absence of head-to-head trials, these findings are important for developing guidelines and assisting doctors in making judgments about which medicine to use.
Through NMA, we were able to, directly and indirectly, compare the efficacy and safety of all CGRP mAbs with placebo treatment, which helped in better estimation of their efficacy than pairwise comparisons. The utilization of a network method to investigate the relative effectiveness of different CGRP mAbs for migraine is a fundamental strength of our study. The Cost-effectiveness of various mAbs was not assessed in any of the studies. Therefore, for better application of these mAbs in clinical practice, studies focussing on those aspects were needed as well.
Migraine is a very common neurological disorder that has a major effect on the quality of life. This NMA suggests all the mAbs approved for migraine have similar efficacy and safety profiles. Fremanezumab (quarterly) ranked better in terms of response, and eptinezumab fared well with respect to both efficacy and safety in the prophylactic treatment of migraine. Future large multi-centric clinical trials may be planned (especially as active comparison randomized clinical trials) to confirm the efficacy and safety results of the meta-analysis.
No potential conflict of interest relevant to this article was reported.
Conceptualization: Mundot Puliappadamb Haridas, Anand Srinivasan. Methodology: Anand Srinivasan, Amruta Tripathy, Rituparna Maiti. Data acquisition and curation: Mundot Puliappadamb Haridas, Rituparna Maiti, Amruta Tripathy. Analysis using R Programming Language, Writing—Reviewing and Editing: Anand Srinivasan, Mundot Puliappadamb Haridas. Validation, Visualiza-tion, Investigation and Interpretation: Rituparna Maiti, Amruta Tripathy. All authors participated in writing the draft of the manuscript or revising it for intellectual content, and all authors participated in the final approval of the version to be published. All authors reviewed and approved the manuscript before it was submitted for pub-lication.
Fig. 1. PRISMA flowchart depicting the literature search for the included studies.
Fig. 2. (A) Net graph of mean migraine days, (B) Net graph of responder rate, and (C) Net graph of adverse events. In each network graph, the size of each node is proportional to the total number of randomized participants (sample size) allocated to the corresponding treatment across all trials, and the width of each line is proportional to the total number of randomized controlled trials evaluating the corresponding treatment comparison.
Fig. 3. Forest plot illustrating the (A) mean difference (95% credible interval, 95%Crl) in mean migraine days of various pharmacotherapeutic management vs. placebo, (B) odds ratio (95%Crl) of the responder rate of various pharmacological management against placebo, (C) odds ratio (95%Crl) of the rate of adverse drug reactions of various drug regimens against placebo.
Fig. 4. SUCRA (surface under cumu-lative ranking) score with ranking for (A) difference in mean migraine days, (B) odds ratio of responder rate, and (C) odds ratio of adverse events for various pharmacotherapeutic regi-mens (the x-axis denotes treatment, and the y-axis denotes the SUCRA score).
Fig. 5. A scatter plot comparing the SUCRA (surface under cumulative ranking) scores of differences in mean migraine days against the odds ratio of the adverse events of the various pharmacotherapy regimens.
CrossRef (1) Scopus (1) Download (484)
Author ORCID Information Mundot Puliappadamb Haridas Amruta Tripathy Rituparna Maiti Anand Srinivasan |
 |
