Sage Journals: Discover world-class research

Abstract

Background

To evaluate the benefit-risk assessment of atogepant and calcitonin gene–related peptide (CGRP) monoclonal antibodies (mAbs) vs. placebo based on the number needed to treat (NNT) and the number needed to harm (NNH) in a blended episodic migraine and chronic migraine (EM + CM) population.

Methods

The NNT was calculated based on achievement of a ≥ 50% reduction in mean monthly migraine days (MMDs) from baseline across 12 weeks. The NNH was calculated using the proportion of participants reporting a discontinuation due to adverse events (AEs). The primary analysis included data from studies of atogepant, erenumab, galcanezumab, eptinezumab and fremanezumab.

Results

In the primary analysis, the calculated NNT for atogepant 60 mg vs. placebo was 4.2 (95% credible interval (CrI) = 3.1–6.7), which was the lowest relative to the CGRP mAbs in the blended EM + CM population. Participants who received atogepant 60 mg or fremanezumab showed the most favorable NNH values (−1010 (95% Crl = 44 to ∞ to number needed to benefit 80) for atogepant) resulting from lower rates of discontinuation due to AEs compared with those receiving placebo.

Conclusions

Atogepant demonstrated a favorable benefit-risk profile, with NNT and NNH values comparable (not statistically significant) with those of CGRP mAbs across all analyses.

Graphical abstract

This is a visual representation of the abstract.

Keywords

antibodies CGRP receptor antagonists gepant headache disorders migraine disorders monoclonal

Introduction

Migraine is a debilitating neurologic disease that affects more than 1 billion people worldwide (1,2). A migraine attack is characterized by intense headache often accompanied by photophobia, phonophobia and nausea (1). People who have up to 14 monthly headache days (MHDs) are classified as having episodic migraine (EM) and those experiencing ≥15 MHDs, at least eight of which are migraine days, for >3 months are classified as having chronic migraine (CM) (1,3). Functional impairment because of migraine includes substantial negative impacts on professional and financial well-being, relationships, and overall health (4,5).

Circulating calcitonin gene–related peptide (CGRP) in the blood and activation of the neuronal CGRP receptor have long been implicated in migraine pathogenesis (6 –10). Several different preventive therapies targeting the CGRP pathway are available to alleviate migraine. Atogepant is an oral, small-molecule CGRP receptor antagonist approved for the preventive treatment of episodic migraine (EM) and chronic migraine (CM) (11). There are also several monoclonal antibodies (mAbs) that target the CGRP pathway to reduce migraine frequency (12,13). Erenumab is an approved mAb that targets the CGRP receptor (14); galcanezumab (15), eptinezumab (16) and fremanezumab (17) are approved mAbs that target the CGRP ligand. Atogepant is an oral tablet taken daily, whereas the CGRP mAbs are administered via injection, either intravenously (eptinezumab) or subcutaneously (erenumab, fremanezumab, galcanezumab) monthly or quarterly, depending on the antibody and dosage (11,14 –17).

The recently updated American Headache Society position statement recommends CGRP-targeted therapies, including gepants, as first-line options for the preventive treatment of migraine. The statement points to the efficacy, safety and tolerability of CGRP-targeted therapies demonstrated in clinical trials and real-world studies (18). However, there is limited evidence comparing the efficacy of these preventive migraine therapies (19 –23). The disease burden of migraine on the population and the disability experienced by each individual with migraine generate a considerable need for healthcare professionals to make informed decisions when selecting a therapeutic regimen. This clinical choice is largely driven by the overall risk-benefit profile of treatments, which encompasses the appropriate efficacy, as well as safety and tolerability outcomes. Number needed to treat (NNT) and number needed to harm (NNH) are metrics that have been used to quantify the effect sizes of clinically relevant benefits and harms, respectively (24). However, randomized clinical trials simultaneously comparing these treatments are rare, and therefore, indirect comparisons can provide useful estimates for differences in treatments if each has been compared against a common comparator. The goal of this meta-analysis was to evaluate the benefit-risk profiles of atogepant and CGRP mAbs (erenumab, galcanezumab, eptinezumab and fremanezumab) vs. placebo in a blended episodic migraine and chronic migraine (EM + CM) population. The results from the various clinical trials allowed for the determination of NNT and NNH as an indirect treatment comparison to aid in clinical decision-making.

The NNT reflects the number of participants that would need to be treated with a medication for one additional participant to reach a favorable outcome such as a treatment response (24,25). In this meta-analysis, the desired outcome was at least a 50% reduction in monthly migraine days (MMDs) across 12 weeks. Although MMD reduction is a commonly reported endpoint in clinical trials, a 50% response rate is more informative in clinical practice to guide treatment decisions. Thus, a smaller NNT indicates greater efficacy and a more powerful medication (24). The NNH, on the other hand, represents the number of participants that would need to be treated with a medication for one additional patient to reach an unfavorable outcome (24,25). In this meta-analysis, the NNH outcome of interest was an adverse event (AE) leading to treatment discontinuation because this outcome allows for comparison of tolerability across different treatments. A larger NNH value indicates greater tolerability (24). These NNT and NNH values are clinically intuitive and relevant assessments of effect size that can inform real-world treatment management.

Methods

Study design

This study was a post hoc analysis of existing randomized clinical trial data and published, peer-reviewed literature. The analysis was performed on a blended EM + CM population. The study characteristics are outlined in the supplemental material (Table S1). Of the 12 key registrational studies included in the primary analysis, 11 were phase 3, one was phase 2, and all were double-blind, placebo-controlled, randomized clinical trials that lasted 12–48 weeks. Sensitivity analyses were also conducted to encompass a broader scope of data. Each of these sensitivity analyses were performed in conjunction with the primary analysis. The first sensitivity analysis included five dose-relationship studies, which were phase 2, 2b, 2b/3 or 3 double-blind, placebo-controlled, randomized clinical trials that spanned 12–49 weeks. The second sensitivity analysis incorporated seven Asian studies (conducted in China, Japan, India, Russia and other countries), which were phase 2 and 3 double-blind, placebo-controlled, randomized clinical trials that lasted 12–24 weeks. Detailed methodologies have been published previously (see supplemental material, Table S1).

Data sources and outcomes

A comprehensive literature search was performed on 1 September 2022 and relevant studies were identified. The studies meeting the following criteria were included in this analysis: randomized clinical trial design of migraine preventive treatment for EM or CM in a non-treatment failure-specific population; use of atogepant or a CGRP mAb licensed dose approved by the UK/European Medicines Agency or United States Food and Drug Administration; use of placebo as one of the comparators; and evaluation of the outcomes outlined below (19 –23,26 –44). An overview of the studies included in each analysis is presented in Table 1.

Table 1.

Episodic and chronic migraine preventive treatment clinical trials included in the NNT/NNH analysis.

	Primary analysis	Dose–response relationship studies	Asian studies
Number needed to treat	ADVANCE (19)	CGP-MD-01 (33)	Hu 2022 (38)
	PROGRESS (26)	Bigal 2015 EM (34)	Sakai 2021 EM (40)
	EVOLVE 1 and 2 (20,29)	Sun 2016 (36)	Wang 2021 (44)
	REGAIN (30)	Dodick 2019 (37)
	ARISE (27)
	STRIVE (21)
	Tepper 2017 (28)
	HALO EM (22)
	PROMISE 1 and 2 (23,31)
Number needed to harm	ADVANCE (19)	CGP-MD-01 (33)	Hu 2022 (38)
	PROGRESS (26)	Bigal 2015 EM (34)	Sakai 2021 EM (40)
	EVOLVE 1 and 2 (20,29)	Bigal 2015 CM (35)	Sakai 2021 CM (41)
	REGAIN (30)	Sun 2016 (36)	Sakai 2019 (39)
	ARISE (27)	Dodick 2019 (37)	Sakai 2020 (42)
	STRIVE (21)		Takeshima 2021 (43)
	Tepper 2017 (28)		Wang 2021 (44)
	HALO EM (22)
	HALO CM (32)
	PROMISE 1 and 2 (23,31)

CM = chronic migraine; EM = episodic migraine; NNH = number needed to harm; NNT = number needed to treat.

Outcomes

The efficacy outcome for NNT calculation was 50% response rate derived from MMDs; that is the percentage of participants who achieved at least a 50% reduction in their MMDs across 12 weeks from baseline. The NNT represents the number of participants over a given time that need to be treated to achieve one additional 50% clinical response vs. placebo.

The safety outcome for NNH calculation was discontinuation due to AEs. NNH represents the number of participants that would need to be treated before observing one additional safety outcome of interest relative to the placebo cohort. It should be noted that the data for discontinuation due to AE from the different studies were not based on the same time point; study duration is outlined in the supplemental material (Table S1). The approach to comparing these data is described below. For treatments with higher discontinuation due to AE rate than with placebo, NNH was positive, with higher NNH values being more favorable. Treatments with a lower incidence of discontinuation due to AEs relative to placebo resulted in negative NNH values, suggesting that additional discontinuation due to AEs are less likely to occur when participants are treated with treatment of interest relative to placebo. Negative values favor the treatment of interest relative to placebo.

Statistical analysis

This meta-analysis calculated the NNT and NNH of atogepant and CGRP mAbs vs. placebo. Bayesian network meta-analysis methodology was used to produce a single pooled estimate of the placebo rate used to calculate the NNTs or NNHs for atogepant and CGRP mAbs. This approach is consistent with the recommendations from the National Institute for Health and Care Excellence (NICE) Technical Support Document (45). The Bayesian method considers the relative precision of different studies using the exact binomial likelihood with different sizes of Ns. In other words, the relative weighting of each study is dependent on the number of participants in that study, not the classification of EM or CM. Using the binomial likelihood, studies with a larger sample size were weighted more because it is less likely there will be significant outliers in those studies. A random effects (RE) model was preferred over the fixed effect (FE) model to calculate the median 50% response rates, NNTs and 95% credible intervals (CrIs) vs. placebo across 12 weeks. Measure of heterogeneity to compare FE and RE models was performed based on residual deviance and deviance information criterion statistics (45,46). If there were truly significant differences in relative treatment effects in separate EM and CM subgroups, we would expect notable heterogeneity expressed in the RE models’ superior fit to the FE model. However, based on the heterogeneity measures, there was no significantly superior fit for RE models relative to FE models, suggesting that the treatment effects across EM and CM were comparable.

NNT and NNH were calculated as the inverse of the absolute risk reduction or ARR [1/absolute risk reduction]. For NNT, the ARR was the difference between the proportion of participants who achieved a 50% response rate with the active treatment and the proportion of participants who achieved a 50% response rate with placebo. For NNH, the ARR was the difference between the proportion of participants who discontinued due to AEs with the active comparator and the proportion of participants who discontinued due to AEs with placebo. Data for discontinuation due to AEs were reported at different time points and discontinuation can differ depending on which time point it is measured. To account for this, a cloglog link function model was used to estimate median discontinuation rates due to AEs and calculate NNH and 95% CrIs. The cloglog model takes time point of the data into consideration and assumes a constant event rate over time, with estimates reported as hazard ratios. When 95% CrIs for NNH included infinity, the difference was deemed not significant.

Results

Of the 12 studies comprising the primary analysis, seven enrolled participants diagnosed with EM and five enrolled participants with CM. The demographic and baseline clinical characteristics were similar across trials (see supplemental material, Table S2). In the EM trials, the mean MMDs at baseline was 7.4–9.2, the mean age was 39.8–42.9 years and 83.7%–88.8% of participants were female. In the CM trials, the mean MMDs at baseline was 16.0–19.6, the mean age was 39.7–42.9 years and 82.8%–88.2% of participants were female.

In the blended EM + CM primary analysis, all treatments had comparable calculated NNT values (Figure 1), with a median NNT of 4.2 (95% CrI = 3.1–6.7) for atogepant 60 mg once daily vs. placebo, which was numerically lower than the NNTs of all CGRP mAbs. This NNT value indicates that, if approximately four participants were to be treated with atogepant 60 mg once daily across 12 weeks, one extra would reach a 50% reduction in MMDs as a result of the treatment. Among the CGRP mAbs, erenumab 140 mg once monthly, galcanezumab 120 mg once monthly, eptinezumab 300 mg once every 3 months and fremanezumab 225 mg once monthly clustered with NNTs of 4.9–5.1. Across all the sensitivity analyses, fremanezumab 225 mg once monthly had a numerically slightly lower NNT than all other treatments (Table 2). Fremanezumab 225 mg had an NNT of 4.6 (95% CrI = 3.2–7.6) compared with 5.0 (95% CrI = 3.7–7.6) for atogepant 60 mg once daily upon inclusion of the dose-response studies. With inclusion of the Asian studies, fremanezumab 225 mg once monthly had an NNT of 3.9 (95% CrI = 2.8–6.1), followed closely by fremanezumab 675 mg once every 3 months with an NNT of 4.0 (95% CrI = 2.8–6.4) and atogepant 60 mg once daily with an NNT of 4.2 (95% CrI = 3.1–6.7). Across all analyses, atogepant 60 mg once daily NNTs ranged from 4.2–5.0. The NNT range was 4.9–7.8 for erenumab (both doses), 4.8–4.9 for galcanezumab 120 mg once monthly, 5.1–6.8 for eptinezumab (both doses) and 3.9–6.0 for fremanezumab (both doses). None of these results were statistically significant and the CrIs overlapped, indicating comparability between atogepant and the CGRPs. Overall, across all analyses, the NNT for atogepant 60 mg once daily vs. placebo was comparable with the NNTs for CGRP mAbs in the blended EM + CM population.

Figure 1.

Median number needed to treat (NNT) for atogepant and calcitonin gene–related peptide monoclonal antibodies in the primary analysis.^a Response rate outcome is defined as percent of participants who achieved at least a 50% reduction in their monthly migraine days across 12 weeks. ^aPrimary analysis includes ADVANCE (19), PROGRESS (26), EVOLVE 1 and 2 (20,29), REGAIN (30), ARISE (27), STRIVE (21), Tepper et al. (28), HALO EM (22) and PROMISE 1 and 2 (23,31). Crl = credible interval.

Table 2.

NNT for atogepant and CGRP mAbs (blended EM + CM^a) for primary and sensitivity analyses.

	Primary analysis only^b	Primary + dose–response studies^c	Primary + Asian studies^d	All studies
Atogepant 60 mg QD	4.2	5.0	4.2	5.0
95% CrI	3.1–6.7	3.7–7.6	3.1–6.7	3.7–7.6
Erenumab 70 mg QM	7.0	6.8	7.8	7.5
95% CrI	4.7–12.6	4.8–10.7	5.3–13.7	5.4–11.6
Erenumab 140 mg QM	5.0	4.9	5.1	5.0
95% CrI	3.5–8.3	3.6–7.5	3.8–7.8	3.8–7.2
Galcanezumab 120 mg QM	4.9	4.8	4.8	4.8
95% CrI	3.7–7.3	3.7–6.8	3.8–6.7	3.8–6.4
Eptinezumab 100 mg Q3M	6.8	6.8	6.8	6.8
95% CrI	4.4–14.2	4.7–11.8	4.4–14.4	4.7–12.0
Eptinezumab 300 mg Q3M	5.1	5.2	5.1	5.2
95% CrI	3.6–8.5	3.9–7.9	3.6–8.5	3.9–8.0
Fremanezumab 225 mg QM	5.0	4.6	3.9	3.9
95% CrI	3.1–11.2	3.2–7.6	2.8–6.1	3.0–5.6
Fremanezumab 675 mg Q3M	6.0	5.6	4.0	4.0
95% CrI	3.5–17.1	3.6–12.3	2.8–6.4	3.0–6.1

CGRP = calcitonin gene–related peptide; CM = chronic migraine; CrI = credible interval; EM = episodic migraine; FE = fixed effects; mAb = monoclonal antibody; MMD = monthly migraine day; NNT = number needed to treat; QD = once daily; QM = once a month; Q3M = once every 3 months; RE = random effects.

Response rate outcome is defined as percent of participants who achieved at least a 50% reduction in their MMDs across 12 weeks.

Based on heterogeneity measures, there was no significantly superior fit for RE models relative to FE models, suggesting that the treatment effects across EM and CM are comparable.

Primary analysis includes ADVANCE (19), PROGRESS (26), EVOLVE 1 and 2 (20,29), REGAIN (30), ARISE (27), STRIVE (21), Tepper 2017 (28), HALO EM (22), and PROMISE 1 and 2 (23,31).

Dose–response relationship studies include CGP-MD-01 (33), Bigal 2015 (34,35), Sun 2016 (36) and Dodick 2019 (37).

Asian studies include Hu 2022 (38), Sakai 2021 (40,41) and Wang 2021 (44).

The NNH for atogepant 60 mg once daily was −1010 (95% CrI = 44 to ∞ to number needed to benefit [NNB] 80) in the primary analysis (Figure 2), with a negative value indicating lower discontinuation due to AEs among participants who received atogepant compared with those who received placebo. An NNH value of −1010 indicates that for approximately every 1000 participants treated with placebo, one would discontinue due to a treatment-related adverse event. The proportion of participants that discontinued treatment with atogepant in these pooled populations was even lower. Similar findings were observed among participants who received fremanezumab, with lower discontinuation due to AEs compared with those who received placebo, resulting in negative NNH values (fremanezumab 225 mg once monthly: −1124 (95% CrI = 40 to ∞ to NNB 78); fremanezumab 675 mg once every 3 months: −346 (95% CrI = 52 to ∞ to NNB 65)). For the majority of treatments across all NNH analyses, the 95% CrIs indicated nonsignificant values as they included infinity. Atogepant 60 mg once daily and eptinezumab 100 mg once every 3 months had the only negative NNH values upon inclusion of the dose-response studies (both −6667 (95% CrI = 58 to ∞ to NNB 107 for atogepant and 33 to ∞ to NNB 80 for eptinezumab 100 mg) (Table 3). Atogepant 60 mg once daily also had the lowest NNH upon inclusion of the Asian studies (−1261 (95% CrI = 39 to ∞ to NNB 78)) followed by erenumab 140 mg once monthly (−593 (95% CrI = 47 to ∞ to NNB 65)).

Figure 2.

Number needed to harm (NNH) for atogepant and calcitonin gene–related peptide monoclonal antibodies in the primary analysis.^a Bold values indicate that rates are higher for PBO relative to active treatment. ^aPrimary analysis includes ADVANCE (19), PROGRESS (26), EVOLVE 1 and 2 (20,29), REGAIN (30), ARISE (27), STRIVE (21), Tepper et al. (28), HALO EM (22), PROMISE 1 and 2 (23,31). ^b95% confidence interval (CI) indicates nonsignificant values because it includes infinity. Ato = atogepant; Ept = eptinezumab; Ere = erenumab; Fre = fremanezumab; Gal = galcanezumab; NNB = number needed to benefit; PBO = placebo; QD = once daily; QM = once a month; Q3M = once every 3 months; Tx = treatment.

Table 3.

NNH for atogepant and CGRP mAbs (blended EM + CM^a) for primary and sensitivity analyses.

	Primary analysis only^b		Primary + dose–response studies^c		Primary + Asian studies^d		All studies
Atogepant 60 mg QD	−1010		−6667		−1261		1499
95% CrI	44 to ∞ to NNB 80		58 to ∞ to NNB 107		39 to ∞ to NNB 78		55 to ∞ to NNB 153
Risk rate: Placebo \| Atogepant	1.22%	1.12%	1.13%	1.11%	0.88%	0.80%	0.76%	0.83%
Erenumab 70 mg QM	422		265		7042		722
95% CrI	25 to ∞ to NNB 103		26 to ∞ to NNB 149		37 to ∞ to NNB 91		43 to ∞ to NNB 167
Risk rate: Placebo \| Erenumab 70 mg	1.22%	1.46%	1.13%	1.51%	0.88%	0.89%	0.76%	0.90%
Erenumab 140 mg QM	332		296		−593		−1019
95% CrI	19 to ∞ to NNB 98		21 to ∞ to NNB 112		47 to ∞ to NNB 65		58 to ∞ to NNB 84
Risk rate: Placebo \| Erenumab 140 mg	1.22%	1.52%	1.13%	1.47%	0.88%	0.71%	0.76%	0.66%
Galcanezumab 120 mg QM	204		224		88		101
95% CrI	26 to ∞ to NNB 171		28 to ∞ to NNB 199		(12–4742)		(15–1203)
Risk rate: Placebo \| Galcanezumab 120 mg	1.22%	1.71%	1.13%	1.58%	0.88%	2.02%	0.76%	1.76%
Eptinezumab 100 mg Q3M	552		−6667		730		−5714
95% CrI	22 to ∞ to NNB 88		33 to ∞ to NNB 80		20 to ∞ to NNB 91		43 to ∞ to NNB 98
Risk rate: Placebo \| Eptinezumab 100 mg	1.22%	1.40%	1.13%	1.11%	0.88%	1.01%	0.76%	0.74%
Eptinezumab 300 mg Q3M	114		183		152		130
95% CrI	13 to ∞ to NNB 176		19 to ∞ to NNB 150		12 to ∞ to NNB 162		15 to ∞ to NNB 725
Risk rate: Placebo \| Eptinezumab 300 mg	1.22%	2.10%	1.13%	1.67%	0.88%	1.53%	0.76%	1.53%
Fremanezumab 225 mg QM	−1124		159		−448		422
95% CrI	40 to ∞ to NNB 78		21 to ∞ to NNB 275		66 to ∞ to NNB 63		40 to ∞ to NNB 264
Risk rate: Placebo \| Fremanezumab 225 mg	1.22%	1.13%	1.13%	1.76%	0.88%	0.65%	0.76%	1.00%
Fremanezumab 675 mg Q3M	−346		8333		−250		−373
95% CrI	52 to ∞ to NNB 65		35 to ∞ to NNB 89		128 to ∞ to NNB 47		164 to ∞ to NNB 64
Risk rate: Placebo \| Fremanezumab 675 mg	1.22%	0.93%	1.13%	1.14%	0.88%	0.48%	0.76%	0.49%

AE = adverse event; CGRP = calcitonin gene–related peptide; CM = chronic migraine; CrI = credible interval; EM = episodic migraine; FE = fixed effect; mAb = monoclonal antibody; NNB = number needed to benefit; NNH = number needed to harm; QD = once daily; QM = once a month; Q3M = once every 3 months; RE = random effects.

Negative values for NNH indicate higher rates of discontinuation due to AEs with placebo relative to comparator of interest.

Based on heterogeneity measures, there was no significantly superior fit for RE models relative to FE models suggesting that the treatment effects across EM and CM are comparable.

Primary analysis includes ADVANCE (19), PROGRESS (26), EVOLVE 1 and 2 (20,29), REGAIN (30), ARISE (27), STRIVE (21), Tepper 2017 (28), HALO EM (22), HALO CM (32) and PROMISE 1 and 2 (23,31).

Dose-response relationship studies include CGP-MD-01 (33), Bigal 2015 EM (34), Bigal 2015 CM (35), Sun 2016 (36) and Dodick 2019 (37).

Asian studies include Hu 2022 (38), Sakai 2021 EM (40), Sakai 2021 CM (41), Sakai 2019 (39), Sakai 2020 (42), Takeshima 2021 (43) and Wang 2021 (44).

Discussion

This study provides an indirect comparison of the efficacy (NNT) and safety (NNH) of atogepant and CGRP mAbs vs. placebo in the preventive treatment of migraine for a blended EM + CM population. Atogepant 60 mg once daily demonstrated the lowest NNT to achieve a 50% reduction in MMDs compared with CGRP mAbs in the primary analysis. All of the treatments had single-digit NNT values, which indicates effective treatments with at least one in 10 participants experiencing clinically relevant improvement over placebo (24). Notably, in the primary analysis, the negative NNH value of atogepant relative to placebo means that participants receiving atogepant 60 mg once daily had even lower discontinuation rates due to AEs than participants receiving placebo. Fremanezumab 225 mg once monthly and 675 mg once every 3 months also had negative NNH values; however, none of the results were statistically significant. All treatments (atogepant and mAbs) displayed a similar efficacy and higher likelihood to help (low NNT numbers) than harm (higher or negative NNH values).

Previous studies have derived NNT and NNH values for the CGRP mAbs with the 50% response rate and discontinuation due to AEs, respectively. One such study compared CGRP mAbs, topiramate and propranolol, for the prevention of EM and CGRP mAbs, topiramate and onabotulinumtoxinA, for the prevention of CM (47). Overall, the results of their NNT analyses are consistent with the CGRP mAb results of our NNT analysis; however, the results of the NNH analysis were mixed. Another meta-analysis examining topiramate and CGRP mAbs for the preventive treatment of EM found similar NNT efficacies across all treatments. The NNH safety profiles, on the other hand, favored CGRP mAbs over traditional migraine treatments such as topiramate, which had lower NNH values (range 5–50) illuminating higher discontinuation due to AEs (48). A summary of the clinical trial populations included along with each result for our analyses and previous meta-analyses can be found in the supplemental material (Table S3). Differences between previous meta-analyses and the one presented here may be due to the analysis of different studies (e.g. our inclusion of Tepper et al. (28) on erenumab for CM) or differences in treatment center practices or placebo rates. Most notably, the previous analyses focused on separate EM and CM populations, with each treatment NNT/NNH calculated from a single trial, whereas our analyses pooled data from multiple clinical trials into a blended EM + CM population, eliminating biases from individual trials.

The blended EM + CM population reflects the reality of migraine disease, in which MMDs and thus EM/CM definition fluctuates over time (49). A potential limitation of our analyses is that the clinical trials included had different protocols with varying definitions of migraine and primary outcomes and studied different oral versus injectable treatments. Additionally, it is possible that the blended EM + CM analysis introduces differences among study participants. However, based on heterogeneity measures, there was no significantly superior fit for RE models relative to FE models, suggesting that the treatment effects across EM and CM were comparable. To investigate potential differences in treatment efficacy across EM and CM populations, we also performed meta-regressions on baseline MMDs across the included outcomes because EM and CM populations are distinguished by differences in baseline MMDs and MMD definitions can be different in different trials. The meta-regression models confirmed that there are no differences in treatment efficacy because the model did not demonstrate significantly superior fit. Thus, evaluation of the overall migraine population allowed for a comprehensive summary of each of the available therapies.

There are also several limitations inherent to a study of this nature. Predominantly, NNT and NNH calculations are indirect metrics for comparison across distinct clinical trials, each with their own criteria. In terms of safety, injection-site reactions that may occur with injectable treatments are not relevant for oral treatments. Additionally, treating physicians in the later studies benefited from adverse event findings in the earlier studies and were more aware and better prepared to address them, which may have led to a reduced discontinuation rate in later studies. Extrapolation from clinical trial participants to the general population may also be limited. For all analyses, the treatment period was limited to 12 weeks, preventing the evaluation of long-term efficacy and safety of atogepant and CGRP mAbs vs. placebo. As the data available from different studies for discontinuation were not all based on the same time point and it is generally thought that time period affects this outcome, it was necessary to account for this difference in these analyses. A model using the cloglog link function was implemented for the discontinuation endpoint. The cloglog model considers the time point of the data and assumes a constant event rate over time for the treatments; the relative effects estimated from this analysis are given as a hazard ratio. The constant event rate is a strong assumption for the cloglog model. In addition, the cloglog model assumes that all patients will experience the event with sufficient follow-up.

The true contribution of this study lies in its application to clinical decision-making. Atogepant is an oral medication option and can be a preferred route for some individuals with migraine, whereas CGRP mAbs are administered as either a subcutaneous or intravenous therapy (50,51). Understanding how treatments compare from an NNT and NNH perspective and being able to explain this to patients is helpful at the point of care (52). A 50% reduction in MMDs is widely regarded as a clinically significant benchmark for response to a preventive migraine treatment (53). Discontinuation due to AEs is especially relevant, as it reflects the patient experience underlying adherence and persistence rates, which are known to be the major obstacles in treating migraine (54 –56). The markedly favorable NNH value of atogepant indicates the potential for greater treatment adherence. In the absence of direct clinical comparison of preventive migraine treatments in EM and CM, these NNT and NNH derivations provide risk-benefit profiles for clinicians to interpret and apply.

Conclusions

The NNT for atogepant 60 mg once daily vs. placebo was comparable with CGRP mAbs in the blended EM + CM population. Atogepant demonstrated a numerically lower NNT compared with CGRP mAbs in the primary analysis. Participants who received atogepant 60 mg once daily demonstrated lower rates of discontinuation due to AEs compared with those receiving placebo in most of the analyses in this study, resulting in negative NNH values. These strong (but not statistically significant) data demonstrate that atogepant is efficacious and well-tolerated, in line with other CGRP-targeted medications and the American Headache Society's recommendation of these treatments as first-line options for the preventive treatment of migraine. Clinical implications

The NNT for atogepant 60 mg once daily vs. placebo was comparable with CGRP mAbs.

Atogepant demonstrated a numerically lower NNT compared with CGRP mAbs in the primary analysis.

Atogepant demonstrated a favorable benefit-risk profile, supporting the American Headache Society's recommendation of using CGRP-targeted therapies, including gepants, as first-line options for the preventive treatment of migraine

Supplemental Material

sj-docx-1-cep-10.1177_03331024241299377 - Supplemental material for Benefit-risk assessment based on number needed to treat and number needed to harm: Atogepant vs. calcitonin gene–related peptide monoclonal antibodies

Supplemental material, sj-docx-1-cep-10.1177_03331024241299377 for Benefit-risk assessment based on number needed to treat and number needed to harm: Atogepant vs. calcitonin gene–related peptide monoclonal antibodies by Jessica Ailani, Anjana Lalla, Rashmi B Halker Singh, Dagny Holle-Lee, Krisztian Nagy, Kari Kelton, Cristiano Piron, Pranav Gandhi and Patricia Pozo-Rosich in Cephalalgia

Footnotes

Acknowledgments

AbbVie funded this study and participated in the study design; the collection, analysis and interpretation of data; and the review and approval of the final manuscript for publication. All authors had access to relevant data and participated in the drafting, review and approval of this publication. No honoraria or payments were made for authorship. Medical writing support was provided to the authors by Sara Nathan, PhD, of Peloton Advantage, LLC, an OPEN Health company, and was funded by AbbVie. Support for the study design and implementation was provided by James Gahn and David Rowe of Medical Decision Modeling Inc., and was funded by AbbVie.

Author contributions

AL, KN, KK, CP and PG were responsible for the study design. KK and CP were responsible for collection and assembly of data. AL, KK, CP and PG were responsible for data analysis. All authors were responsible for data interpretation. All authors were responsible for manuscript preparation. All authors were responsible for manuscript review and revisions. All authors approved the final version of the manuscript submitted for publication.

Data availability

AbbVie is committed to responsible data sharing regarding the clinical trials we sponsor. This includes access to anonymized, individual and trial-level data (analysis data sets), as well as other information (eg, protocols, clinical study reports or analysis plans), as long as the trials are not part of an ongoing or planned regulatory submission. This includes requests for clinical trial data for unlicensed products and indications. These clinical trial data can be requested by any qualified researchers who engage in rigorous, independent, scientific research, and will be provided following review and approval of a research proposal, Statistical Analysis Plan (SAP), and execution of a Data Sharing Agreement (DSA). Data requests can be submitted at any time after approval in the USA and Europe and after acceptance of this manuscript for publication. The data will be accessible for 12 months, with possible extensions considered. For more information on the process or to submit a request, see: .

Declaration of conflicting interests

The authors declared the following potential conflicts of interest with respect to the research, authorship and/or publication of this article: Jessica Ailani, MD, has served as a consultant for AbbVie, Aeon, Amgen, Axsome, BioDelivery Sciences International, Biohaven, Eli Lilly and Company, GlaxoSmithKline, Gore, Impel, Ipsen, Linpharma, Lundbeck, Merz, Miravo, Nesos, Neurolief, Pfizer, Satsuma, Teva, and Theranica. She has owned stock shares in Ctrl M, has provided editorial services for SELF magazine, and has received clinical trial grants from AbbVie, Biohaven, Eli Lilly and Company, Ipsen, Parema, Satsuma and Zosano. Anjana Lalla, MS, Pranav Gandhi, PhD, and Krisztian Nagy, MD, are employees of AbbVie and may hold AbbVie stock. Rashmi B. Halker Singh, MD, serves on the editorial board of Current Neurology and Neuroscience Reports, serves as deputy editor for Headache, and previously received grants for research support from Amgen and Eli Lilly. In the past 3 years, she has received consulting fees from AbbVie and Pfizer. Dagny Holle-Lee, MD, PhD, has received honoraria for consulting from AbbVie/Allergan, Amgen, Eli Lilly, Hormosan, Lundbeck, Novartis, Teva and Zuellig Pharma. Patricia Pozo-Rosich, MD, PhD, has received, in the past 3 years, personal fees for advisory boards and speaker panels from AbbVie, Amgen, Chiesi, Eli Lilly, Lundbeck, Medscape, Novartis, Pfizer and Teva Pharmaceuticals, and for serving on a scientific advisory or data safety monitoring board for Lilly Foundation Spain; is the principal investigator for clinical trials sponsored by AbbVie, Eli Lilly, Lundbeck, Novartis and Teva; has received (through her group) grants from AbbVie, EraNet Neuron, Instituto de Salud Carlos III, Novartis and Teva; and serves as an associate editor for Cephalalgia, Headache, Neurologia, The Journal of Headache and Pain, and Revista de Neurologia. Kari Kelton, PhD, and Cristiano Piron, MPH, are employees of Medical Decision Modeling Inc., which received payment from AbbVie for the analysis.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and publication of this article: This study was sponsored and funded by AbbVie. Medical writing support was provided to the authors by Sara Nathan, PhD, of Peloton Advantage, LLC, an OPEN Health company, and was funded by AbbVie. Support for the study design and implementation was provided by James Gahn and David Rowe of Medical Decision Modeling Inc., and was funded by AbbVie.

ORCID iDs

Cristiano Piron

Pranav Gandhi

Patricia Pozo-Rosich

Supplemental material

Supplemental material for this article is available online.

References

Headache Classification Committee of the International Headache Society. The international classification of headache disorders, 3rd edition. Cephalalgia 2018; 38: 1–211.

Stovner

Hagen

Linde

, et al. The global prevalence of headache: an update, with analysis of the influences of methodological factors on prevalence estimates. J Headache Pain 2022; 23: 34.

Bigal

Lipton

. Clinical course in migraine: conceptualizing migraine transformation. Neurology 2008; 71: 848–855.

Buse

Yugrakh

Lee

, et al. Burden of illness among people with migraine and ≥ 4 monthly headache days while using acute and/or preventive prescription medications for migraine. J Manag Care Spec Pharm 2020; 26: 1334–1343.

Buse

Fanning

Reed

, et al. Life with migraine: effects on relationships, career, and finances from the chronic migraine epidemiology and outcomes (CaMEO) study. Headache 2019; 59: 1286–1299.

Goadsby

Edvinsson

Ekman

. Vasoactive peptide release in the extracerebral circulation of humans during migraine headache. Ann Neurol 1990; 28: 183–187.

Lassen

Haderslev

Jacobsen

, et al. CGRP May play a causative role in migraine. Cephalalgia 2002; 22: 54–61.

Tepper

. History and review of anti-calcitonin gene-related peptide (CGRP) therapies: from translational research to treatment. Headache 2018; 58: 238–275.

Avona

Burgos-Vega

Burton

, et al. Dural calcitonin gene-related peptide produces female-specific responses in rodent migraine models. J Neurosci 2019; 39: 4323–4331.

10.

Paige

Plasencia-Fernandez

Kume

, et al. A female-specific role for calcitonin gene-related peptide (CGRP) in rodent pain models. J Neurosci 2022; 42: 1930–1944.

11.

Qulipta [package insert]. North Chicago, IL: AbbVie, 2023.

12.

Dodick

Goadsby

Silberstein

, et al. Safety and efficacy of ALD403, an antibody to calcitonin gene-related peptide, for the prevention of frequent episodic migraine: a randomised, double-blind, placebo-controlled, exploratory phase 2 trial. Lancet Neurol 2014; 13: 1100–1107.

13.

Dodick

Goadsby

Spierings

, et al. Safety and efficacy of LY2951742, a monoclonal antibody to calcitonin gene-related peptide, for the prevention of migraine: a phase 2, randomised, double-blind, placebo-controlled study. Lancet Neurol 2014; 13: 885–892.

14.

Aimovig [package insert]. Thousand Oaks, CA, and East Hanover, NJ: Amgen Inc., and Novartis Pharmaceuticals Corporation, 2021.

15.

Emgality [package insert]. Indianapolis, IN: Eli Lilly and Company, 2019.

16.

Vyepti [package insert]. Bothell, WA: Lundbeck Seattle BioPharmaceuticals, Inc, 2021.

17.

Ajovy [package insert]. North Wales, PA: Teva Pharmaceuticals USA, Inc., 2021.

18.

Charles

Digre

Goadsby

, et al. Calcitonin gene-related peptide-targeting therapies are a first-line option for the prevention of migraine: an American Headache Society position statement update. Headache 2024; 64: 333–341.

19.

Ailani

Lipton

Goadsby

, et al. Atogepant for the preventive treatment of migraine. N Engl J Med 2021; 385: 695–706.

20.

Stauffer

Dodick

Zhang

, et al. Evaluation of galcanezumab for the prevention of episodic migraine: the EVOLVE-1 randomized clinical trial. JAMA Neurol 2018; 75: 1080–1088.

21.

Goadsby

Reuter

Hallstrom

, et al. A controlled trial of erenumab for episodic migraine. N Engl J Med 2017; 377: 2123–2132.

22.

Dodick

Silberstein

Bigal

, et al. Effect of fremanezumab compared with placebo for prevention of episodic migraine: a randomized clinical trial. JAMA 2018; 319: 1999–2008.

23.

Ashina

Saper

Cady

, et al. Eptinezumab in episodic migraine: a randomized, double-blind, placebo-controlled study (PROMISE-1). Cephalalgia 2020; 40: 241–254.

24.

Citrome

Ketter

. When does a difference make a difference? Interpretation of number needed to treat, number needed to harm, and likelihood to be helped or harmed. Int J Clin Pract 2013; 67: 407–411.

25.

Andrade

. The numbers needed to treat and harm (NNT, NNH) statistics: what they tell us and what they do not. J Clin Psychiatry 2015; 76: e330–e333.

26.

Pozo-Rosich

Ailani

Ashina

, et al. Atogepant for the preventive treatment of chronic migraine (PROGRESS): a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet 2023; 402: 775–785.

27.

Dodick

Ashina

Brandes

, et al. ARISE: a phase 3 randomized trial of erenumab for episodic migraine. Cephalalgia 2018; 38: 1026–1037.

28.

Tepper

Ashina

Reuter

, et al. Safety and efficacy of erenumab for preventive treatment of chronic migraine: a randomised, double-blind, placebo-controlled phase 2 trial. Lancet Neurol 2017; 16: 425–434.

29.

Skljarevski

Matharu

Millen

, et al. Efficacy and safety of galcanezumab for the prevention of episodic migraine: results of the EVOLVE-2 phase 3 randomized controlled clinical trial. Cephalalgia 2018; 38: 1442–1454.

30.

Detke

Goadsby

Wang

, et al. Galcanezumab in chronic migraine: the randomized, double-blind, placebo-controlled REGAIN study. Neurology 2018; 91: e2211–e2221.

31.

Lipton

Goadsby

Smith

, et al. Efficacy and safety of eptinezumab in patients with chronic migraine: PROMISE-2. Neurology 2020; 94: e1365–e1377.

32.

Silberstein

Dodick

Bigal

, et al. Fremanezumab for the preventive treatment of chronic migraine. N Engl J Med 2017; 377: 2113–2122.

33.

Goadsby

Dodick

Ailani

, et al. Safety, tolerability, and efficacy of orally administered atogepant for the prevention of episodic migraine in adults: a double-blind, randomised phase 2b/3 trial. Lancet Neurol 2020; 19: 727–737.

34.

Bigal

Dodick

Rapoport

, et al. Safety, tolerability, and efficacy of TEV-48125 for preventive treatment of high-frequency episodic migraine: a multicentre, randomised, double-blind, placebo-controlled, phase 2b study. Lancet Neurol 2015; 14: 1081–1090.

35.

Bigal

Edvinsson

Rapoport

, et al. Safety, tolerability, and efficacy of TEV-48125 for preventive treatment of chronic migraine: a multicentre, randomised, double-blind, placebo-controlled, phase 2b study. Lancet Neurol 2015; 14: 1091–1100.

36.

Sun

Dodick

Silberstein

, et al. Safety and efficacy of AMG 334 for prevention of episodic migraine: a randomised, double-blind, placebo-controlled, phase 2 trial. Lancet Neurol 2016; 15: 382–390.

37.

Dodick

Lipton

Silberstein

, et al. Eptinezumab for prevention of chronic migraine: a randomized phase 2b clinical trial. Cephalalgia 2019; 39: 1075–1085.

38.

, et al. Galcanezumab in episodic migraine: the phase 3, randomized, double-blind, placebo-controlled PERSIST study. J Headache Pain 2022; 23: 90.

39.

Sakai

Takeshima

Tatsuoka

, et al. A randomized phase 2 study of erenumab for the prevention of episodic migraine in Japanese adults. Headache 2019; 59: 1731–1742.

40.

Sakai

Suzuki

Kim

, et al. Efficacy and safety of fremanezumab for episodic migraine prevention: multicenter, randomized, double-blind, placebo-controlled, parallel-group trial in Japanese and Korean patients. Headache 2021; 61: 1102–1111.

41.

Sakai

Suzuki

Kim

, et al. Efficacy and safety of fremanezumab for chronic migraine prevention: multicenter, randomized, double-blind, placebo-controlled, parallel-group trial in Japanese and Korean patients. Headache 2021; 61: 1092–1101.

42.

Sakai

Ozeki

Skljarevski

. Efficacy and safety of galcanezumab for prevention of migraine headache in Japanese patients with episodic migraine: a phase 2 randomized controlled clinical trial. Cephalalgia Rep 2020; 3: 2515816320932573.

43.

Takeshima

Sakai

Hirata

, et al. Erenumab treatment for migraine prevention in Japanese patients: efficacy and safety results from a phase 3, randomized, double-blind, placebo-controlled study. Headache 2021; 61: 927–935.

44.

Wang

Roxas

Jr. Saravia

, et al. Randomised, controlled trial of erenumab for the prevention of episodic migraine in patients from Asia, the Middle East, and Latin America: the EMPOwER study. Cephalalgia 2021; 41: 1285–1297.

45.

Dias

Welton

Sutton

, et al. NICE DSU technical support document 2: A Generalised linear modelling framework for pairwise and network meta-analysis of randomised controlled trials. London, United Kingdom: National Institute for Health and Care Excellence, 2016.

46.

Dias

Sutton

Welton

, et al. NICE DSU technical support document 3: Heterogeneity: Subgroups, meta-regression, bias and bias-adjustment. London, United Kingdom: National Institute for Health and Care Excellence, 2012.

47.

Drellia

Kokoti

Deligianni

, et al. Anti-CGRP monoclonal antibodies for migraine prevention: a systematic review and likelihood to help or harm analysis. Cephalalgia 2021; 41: 851–864.

48.

Overeem

Raffaelli

Mecklenburg

, et al. Indirect comparison of topiramate and monoclonal antibodies against CGRP or its receptor for the prophylaxis of episodic migraine: a systematic review with meta-analysis. CNS Drugs 2021; 35: 805–820.

49.

Serrano

Lipton

Scher

, et al. Fluctuations in episodic and chronic migraine status over the course of 1 year: implications for diagnosis, treatment and clinical trial design. J Headache Pain 2017; 18: 101.

50.

Mansfield

Gebben

Sutphin

, et al. Patient preferences for preventive migraine treatments: a discrete-choice experiment. Headache 2019; 59: 715–726.

51.

Hubig

Smith

Chua

, et al. A stated preference survey to explore patient preferences for novel preventive migraine treatments. Headache 2022; 62: 1187–1197.

52.

Urtecho

Wagner

Wang

, et al. A qualitative evidence synthesis of patient perspectives on migraine treatment features and outcomes. Headache 2023; 63: 185–201.

53.

Ailani

Burch

Robbins

. The American Headache Society consensus statement: update on integrating new migraine treatments into clinical practice. Headache 2021; 61: 1021–1039.

54.

Hepp

Bloudek

Varon

. Systematic review of migraine prophylaxis adherence and persistence. J Manag Care Pharm 2014; 20: 22–33.

55.

Orlando

Mucherino

Monetti

, et al. Treatment patterns and medication adherence among newly diagnosed patients with migraine: a drug utilisation study. BMJ open 2020; 10: e038972.

56.

Hepp

Dodick

Varon

, et al. Persistence and switching patterns of oral migraine prophylactic medications among patients with chronic migraine: a retrospective claims analysis. Cephalalgia 2017; 37: 470–485.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.08 MB