Comparative effectiveness and tolerability of calcitonin gene‐related peptide (CGRP) monoclonal antibodies and onabotulinumtoxinA in chronic migraine: A multicenter, real‐world study in Taiwan

Abstract

Objective

To compare the real‐world effectiveness and tolerability of calcitonin gene‐related peptide (CGRP) monoclonal antibodies (mAbs) and onabotulinumtoxinA in chronic migraine (CM) patients.

Methods

This multicenter study involved retrospective analysis of prospectively collected data of CM patients treated with CGRP mAbs or onabotulinumtoxinA, including difficult‐to‐treat (DTT) patients (i.e., ≥3 preventive failures). Treatment outcomes were determined at 6 months based on prospective headache diaries and Migraine Disability Assessment (MIDAS).

Results

The study included 316 (55 M/261F, mean age 44.4 ± 13.5 years) and 333 (61 M/272F, mean age 47.9 ± 13.4 years) CM patients treated with CGRP mAbs or onabotulinbumtoxinA, respectively. At 6 months, CGRP mAb treatment was associated with a greater decrease in monthly migraine days (MMDs) (−13.0 vs. −8.7 days/month, p < 0.001) and a higher ≥50% responder rate (RR) (74.7% vs. 50.7%, p < 0.001) compared with onabotulinumtoxinA injections. The findings were consistent in DTT patients (−13.0 vs. −9.1 MMDs, p < 0.001; ≥50% RR: 73.9% vs. 50.3%, p < 0.001) or those with medication‐overuse headache (MOH) (−13.3 vs. −9.0 MMDs, p < 0.001; ≥50% RR: 79.0% vs. 51.6%, p < 0.001). Besides, patients receiving CGRP mAbs had greater improvement (−42.2 vs. −11.8, p < 0.001) and a higher ≥50% RR (62.0% vs. 40.0%, p = 0.001) in MIDAS scores and a lower rate of adverse events (AEs) (6.0% vs. 21.0%, p < 0.001). However, none of the patients discontinued treatment due to AEs.

Conclusions

In this multicenter, real‐world study, CGRP mAbs were more effective than onabotulinumtoxinA in CM patients, even in DTT or MOH patients. All of these injectables were well tolerated. Further prospective studies are needed to verify these findings.

INTRODUCTION

Chronic migraine (CM) is characterized by frequent migraine attacks over a prolonged period of time, and is associated with greater disability and poorer quality of life when compared with episodic migraine (EM) [1, 2]. Preventive treatment is usually challenging in patients with CM, and is complicated by frequent coexistence of medication overuse (MO) or medication‐overuse headache (MOH) [1, 3, 4, 5]. In fact, only a minority of patients could overcome major barriers to appropriate care [6]. Even in those who receive preventive treatment, the adherence rate to oral preventive medications is frequently disappointing [7].

Two classes of injectable preventive medications became available for clinical use in recent years, namely onabotulinumtoxinA and calcitonin gene‐related peptide (CGRP) monoclonal antibodies (mAbs). In the pivotal trials, onabotulinumtoxinA, galcanezumab, and fremanezumab were all more effective than placebo in reducing headache frequency and headache‐related disability in patients with CM [8, 9, 10, 11, 12], even in the presence of MO [13, 14, 15], Besides, galcanezumab and fremanezumab, which are the only two CGRP mAbs currently available in Taiwan, were of proven efficacy in patients who were difficult to treat (DTT), that is, those who had failed multiple migraine preventive medications [16, 17]. Treatment with these two classes of injectables was well tolerated, and discontinuations due to adverse events (AEs) were uncommon [8, 9, 18]. These findings were confirmed in open‐label or real‐world studies [18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29].

The comparative effectiveness of CGRP mAbs and onabotulinumtoxinA should be an issue of clinical significance and importance. However, studies involving head‐to‐head comparisons are lacking. According to a report by the Institute for Clinical and Economic Review (ICER), the treatment efficacies of CGRP mAbs and onabotulinumtoxinA were comparable based on network meta‐analysis [30]. Conversely, other meta‐analyses indicated that CGRP mAbs were associated with numerically greater treatment efficacies in certain outcome measures when compared with onabotulinumtoxinA [31, 32]. However, these reports involved indirect comparisons based on data from clinical trials of individual medications, and interpretation of the findings could be difficult due to heterogeneities in the study design, patient population, outcome measure, and so on.

The reimbursement regulations of Taiwan National Health Insurance (NHI) [33] provide a unique opportunity to understand the treatment effectiveness of these two classes of injectables at 6 months in a real‐world setting. CGRP mAbs and onabotulinumtoxinA are reimbursed only in DTT CM patients, defined as those who have failed ≥3 categories of preventive medications recommended in the guidelines of the Taiwan Headache Society [34], among which topiramate should be included, and concomitant use of CGRP mAbs and onabutulinumtoxinA is not covered. The reimbursement consists of two stages, and the injectables should be administered in a clinic or hospital setting [33]. Patients with granted approval typically have a 3‐month or 6‐month treatment course of CGRP mAbs or onabotulinumtoxinA, respectively. Those who have a ≥50% reduction in monthly migraine days (MMDs) after treatment are qualified to apply for the second stage of reimbursement, which consists of another 3‐month treatment of CGRP mAbs or 12‐month treatment of onabotulinumtoxinA. After completion of the treatment courses, there is a 6‐month mandatory break, after which re‐application is feasible for patients who fulfill the reimbursement regulations. Both medical records and prospective headache diaries are needed for the initial and subsequent applications for reimbursement. By taking advantage of these regulations, the present study aimed to compare the effectiveness and tolerability of CGRP mAbs and onabotulinumtoxinA in CM patients in a real‐world clinical setting.

METHODS

Patients

This was a retrospective analysis of prospectively collected data for routine clinical use in CM patients treated with galcanezumab, fremanezumab, or onabotulinumtoxinA in the headache clinics of five tertiary medical centers (Taipei Veterans General Hospital [TVGH], Tri‐Service General Hospital [TSGH], MacKay Memorial Hospital [MMH], Chang Gung Memorial Hospital Linkou Medical Center [CGMHLMC], and National Cheng Kung University Hospital [NCKUH]) and three regional hospitals (Taipei Municipal Gan‐Dau Hospital [TMGDH], Kuang Tien General Hospital [KTGH], and An Nan Hospital [ANH]) across different regions of Taiwan between April 2020 and March 2023. Headache diagnoses were based on the Third Edition of the International Classification of Headache Disorders (ICHD‐3) [35]. Migraine Disability Assessment (MIDAS) [36], Hospital Anxiety and Depression Scale (HADS) [37], and Pittsburgh Sleep Quality Index (PSQI) [38] were included as part of the routine practice in certain centers. Acute and preventive treatments were determined by the treating physician, and the patients were asked to keep headache diaries. Patients were included in the analysis if (1) they were at least 20 years old, (2) they were diagnosed as CM according to medical records, (3) they received at least one dose of galcanezumab, fremanezumab, or onabotulinumtoxinA for the first time, and (4) galcanezumab, fremanezumab, or onabotulinumtoxinA was administered according to standard protocols, that is, a 240 mg loading dose followed by 120 mg monthly doses for galcanezuamb, 225 mg monthly or 675 mg quarterly for fremanezumab, and 155 ± 5 U at intervals of 12 ± 2 weeks according to the PREEMPT paradigm [8, 9] for onabotulinumtoxinA. Patients were excluded if (1) the diagnosis of CM was not fulfilled according to headache diaries before the first dose of galcanezumab, fremanezumab, or onabotulinumtoxinA; (2) there was no reliable information on the post‐treatment efficacy; (3) there were overlaps between treatment periods with galcanezumab or fremanezumab and onabotulinumtoxinA in the same patients who had exposure to both of the two classes of agents, that is, onabotulinumtoxinA was used within 1 month of the last dose of galcanezumab treatment or fremanezumab monthly treatment or within 3 months of fremanezumab quarterly treatment, or galcanezumab or fremanezumab was used within 12 weeks of the last onabotulinumtoxinA injection cycle; or (4) there was a switch among these three agents during the follow‐up period. The number of prior preventive failures was recorded, and DTT patients were defined as those who had failed ≥3 categories of preventive treatment. The analysis included reimbursed and non‐reimbursed treatment courses. The study protocols were approved by the Institutional Review Boards of TVGH (2023–09‐013 AC), TSGH (A202305181), MMH (23MMHIS095e), CGMHLMC (202301858B0), NCKUH (A‐ER‐112‐113), TMGDH (1120003132), KTGH (KTGH 11227), and ANH (TMANH112‐REC034).

Outcome measures and tolerability

The primary outcome was the mean change from baseline in the number of monthly (28‐day) migraine days (MMDs) at 6 months, which was based on prospective headache diaries. The secondary outcome measures included the mean change from baseline in the number of monthly moderate‐to‐severe migraine days (MMSMDs) and the proportions of patients with ≥50% or ≥ 75% reduction in the number of MMDs and MMSMDs at 6 months. The pre‐ and post‐treatment headache frequencies were calculated directly based on data in the electronic diary system (TVGH and TMGDH) or paper diaries (all other hospitals), which were kept for reimbursement purposes or routine practice. When the original diaries were not available, data were extracted from electronic medical records. Comparisons were made between the two treatment groups, and subgroup analysis was carried out for DTT or MOH patients. Besides, MIDAS scores and patients' overall subjective ratings on effectiveness at follow‐up, which were available in a subset of patients, were included in the analysis. The percentages of patients who reported AEs or who discontinued treatment prematurely were determined.

Statistical analysis

Baseline characteristics were compared between the two treatments using Student's t‐test for continuous variables, or the Mann–Whitney U test for non‐normally distributed variables, and using the chi‐square test for categorical variables. Only patients with at least one post‐treatment efficacy measure were analyzed. The least‐squares‐mean (LSM) changes from baseline in the number of MMDs, MMSMDs, or MIDAS scores at each of the time points were compared between the two categories of injectables. Mixed‐effect models with repeated measures without imputation of missing data were used in the analysis, and included the fixed categorical effects of sex, treatment, month, and treatment‐by‐month interaction, as well as age and baseline value as continuous covariates. The responder rates (RR) (i.e., the proportions of patients with ≥50% or ≥ 75% reductions in MMDs, MMSMDs, or MIDAS scores, at 6 months in the per‐protocol population) were compared using the chi‐square test. Logistic regression modeling was used to estimate the odds ratios (ORs) and the 95% confidence intervals (CIs) for having a treatment response at 6 months between the two treatment groups. Statistical analysis was carried out using IBM SPSS Statistics for Windows, version 24.0 (IBM Corp., Armonk, NY, USA). Statistical significance was set at a two‐sided p of < 0.05.

RESULTS

Patient selection and baseline characteristics

During the study period, 899 CM patients were screened (154 M/745F, mean age 46.3 ± 13.8 years), and patients without reliable baseline frequency in the medical records (n = 131), those with a baseline headache frequency not fulfilling the criteria for CM (n = 41), and those who were lost to follow‐up (n = 78) were excluded. In total, 316 (55 M/261F, mean age 44.4 ± 13.5 years) patients treated with CGRP mAbs and 333 (61 M/272F, mean age 47.9 ± 13.4 years) receiving onabotulinumtoxinA were analyzed (Figure 1). Patients treated with onabotulinumtoxinA were older (47.9 ± 13.4 vs. 44.4 ± 13.5 years, p = 0.001) and had fewer preventive failures (3.2 ± 1.4 vs. 3.6 ± 1.2, p < 0.001) compared with those receiving CGRP mAb injections. However, other demographic and clinical profiles at baseline were comparable (Table 1).

FIGURE 1.

Patient selection. CGRP, calcitonin gene‐related peptide; CM, chronic migraine; mAb, monoclonal antibody; OnabotA, onabotulinumtoxinA.

TABLE 1.

Baseline characteristics in chronic migraine patients receiving calcitonin gene‐related peptide monoclonal antibodies and onabotulinumtoxinA.

Characteristic	CGRP mAbs (n = 316)	OnabotA (n = 333)	P value
Age (years)	44.4 ± 13.5	47.9 ± 13.4	0.001
Female	261 (82.6%)	272 (81.7%)	0.761
MOH	238 (75.3%)	236 (70.9%)	0.202
Migraine with aura	43 (13.6%)	30 (9.0%)	0.064
Monthly migraine days	22.3 ± 5.4	22.9 ± 5.5	0.180
Monthly moderate‐to‐severe migraine days	19.0 ± 7.3	18.0 ± 7.4	0.117
Monthly days with acute medication use	17.3 ± 9.0	17.1 ± 10.1	0.423
Prior preventive failures (categories)	3.6 ± 1.2	3.2 ± 1.4	<0.001
Beta‐blockers	238 (75.3%)	196 (58.9%)
Flunarizine	261 (82.6%)	209 (62.8%)
Topiramate	298 (94.3%)	256 (76.9%)
Valproate	43 (13.6%)	42 (12.6%)
Tricyclics	119 (37.7%)	100 (30.0%)

	CGRP mAbs (n = 238)	OnabotA (n = 254)	P value
HADS‐Total	17.7 ± 9.2	17.2 ± 9.2	0.498
HADS‐A	9.4 ± 4.9	9.1 ± 4.8	0.491
HADS‐D	8.4 ± 5.0	8.1 ± 5.1	0.456
MIDAS	52.1 ± 48.7	59.3 ± 55.8	0.572
PSQI	11.9 ± 4.3	11.9 ± 4.2	0.766

Abbreviations: CGRP, calcitonin gene‐related peptide; CM, chronic migraine; HADS, Hospital Anxiety and Depression Scale (A, Anxiety subscale; D, Depression subscale); mAb, monoclonal antibody; MIDAS, Migraine Disability Assessment; MOH, medication‐overuse headache; OnabotA, onabotulinumtoxinA; PSQI, Pittsburgh Sleep Quality Index.

Treatment effectiveness in CM patients

During the study period, 245 (43 M/202F, mean age 43.9 ± 13.4 years) and 71 patients (12 M/59F, mean age 46.2 ± 14.0 years) were treated with galcanezumab (mean total dose 679.3 ± 206.8 mg) or fremanezumab (mean total dose 969.7 ± 377.6 mg), respectively, and the mean total dose for onabotulinumtoxinA was 296.7 ± 44.3 U. In the CGRP mAb group, the average treatment duration was 4.6 ± 1.7 months, and 144 patients (45.6%) did not complete the 6‐month treatment course (lack of efficacy in 54, financial issues in 66, travel issues in 16, and loss to follow‐up in 8 patients). In patients receiving onabotulinumtoxinA injections, the average treatment duration was 5.7 ± 0.8 months, and 28 patients (8.4%) dropped out (lack of efficacy in 6, financial issues in 20, and loss to follow‐up in 2 patients).

At 6 months, treatment with CGRP mAbs was associated with greater reductions from baseline in the numbers of MMDs (−13.0 vs. −8.7 days/month, p < 0.001) (Figure 2a) and MMSMDs (−11.7 vs. −8.7 days/month, p < 0.001) (Figure 2b) when compared with onabotulinumtoxinA injections. The between‐group differences started at the first month and remained significant throughout the 6‐month treatment period (Figure 2a,b). Besides, treatment with CGRP mAbs was associated with higher ≥50% (74.7% vs. 50.7%, p < 0.001) (OR = 1.91 [95% CI = 1.39–2.63], p < 0.001) and ≥ 75% RRs (43.8% vs. 23.8%, p < 0.001) (OR = 1.88 [95% CI = 1.34 to 2.64], p < 0.001) in MMDs (Figure 2c,d), and the findings were consistent after controlling for age, gender, the number of prior preventive failures, and coexisting MOH (≥50% RR: OR = 2.07 [95% CI = 1.49 to 2.87], p < 0.001; ≥75% RR: OR = 2.02 [95% CI = 1.42 to 2.86], p < 0.001). The results were similar for ≥50% RR (71.1% vs. 57.2%, p = 0.009) and ≥ 75% RR (48.4% vs. 31.3%, p = 0.002) in MMSMDs. The trends were consistent in the intention‐to‐treat analysis (Figure S1).

FIGURE 2.

Treatment outcomes of chronic migraine patients treated with calcitonin gene‐related peptide monoclonal antibodies or onabotulinumtoxinA at 6 months. Least‐squares mean change (standard error) from baseline in the number of monthly migraine days (MMDs) (a) and monthly moderate‐to‐severe migraine days (MMSMDs) (b) and proportions of patients with ≥50% (c) or ≥ 75% (d) reduction in MMDs and MMSMDs at 6 months. CGRP, calcitonin gene‐related peptide; DTT, difficult‐to‐treat; mAb, monoclonal antibody; MMD, monthly migraine day; MMSMD, monthly moderate‐to‐severe migraine day; MOH, medication‐overuse headache; onabotA, onabotulinumtoxinA. ***p ≤ 0.001, **p ≤ 0.01.

Subgroup analysis

Among the study participants, 544 (96 M/448F, mean age 46.0 ± 13.1 years) were categorized as DTT. The CGRP mAb group (n = 292) (49 M/243F, mean age 44.1 ± 13.4 years) had greater improvement in MMDs (−13.0 vs. −9.1 days/month, p < 0.001) than the onabotulinumtoxinA group (n = 252) (47 M/205F, mean age 48.3 ± 12.5 years) at 6 months (Figure 3a). Besides, patients treated with CGRP mAb had greater ≥50% (73.9% vs. 50.3%, p < 0.001) or ≥ 75% RRs (43.8% vs. 25.2%, p < 0.001) in MMDs at 6 months (Figure 3b). Conversely, in patients with MOH (n = 474) (91 M/383F, mean age 47.5 ± 13.5 years), 238 were treated with CGRP mAbs (41 M/197F, mean age 45.4 ± 13.4 years) and 236 received onabotulinumtoxinA injections (50 M/186F, mean age 49.7 ± 13.2 years). Patients in the CGRP mAb group had greater decrease in MMDs (−13.3 vs. −9.0 days/month, p < 0.001) than those receiving onabotulinumtoxinA (n = 236) (Figure 3c), and were more likely to have ≥50% (79.0% vs. 51.6%, p < 0.001) and ≥ 75% reduction in MMDs (45.2% vs. 19.9%, p < 0.001) at 6 months (Figure 3d). The trends were consistent in the intention‐to‐treat analysis (Figure S1).

FIGURE 3.

Subgroup analysis in difficult‐to‐treat (DTT) and medication‐overuse headache (MOH) patients. Least‐squares mean change (standard error) from baseline in the number of monthly migraine days (MMDs) (a) at 6 months and ≥ 50% and ≥ 75% responder rates for MMDs (b) in patients who were DTT and those with MOH (c, d). CGRP, calcitonin gene‐related peptide; DTT, difficult‐to‐treat; mAb, monoclonal antibody; MMD, monthly migraine day; MOH, medication‐overuse headache; onabotA, onabotulinumtoxinA. ***p ≤ 0.001, **p ≤ 0.01, *p < 0.05.

Conversely, there were 57 patients (14 M/43F, mean age 47.5 ± 13.6 years) in the CGRP mAb group that had prior exposure to onabotulinumtoxinA and 32 patients (5 M/27F, mean age 43.2 ± 15.5 years) in the onabotulinumtoxinA group that had prior exposure to CGRP mAbs, although there were no overlaps between the treatment periods of onabotulinumtoxinA and CGRP mAbs in the same patients. The findings were consistent after excluding these patients. Patients receiving CGRP mAbs had greater reduction in MMDs (−14.4 vs. −8.8 days/month, p < 0.001), and higher ≥50% (85.4% vs. 51.2%, p < 0.001) and ≥ 75% RRs (50.8% vs. 24.1%, p < 0.001) in MMDs at 6 months when compared to those in the onabotulinumtoxinA group.

Patient‐reported outcomes (PROs)

At follow‐up, MIDAS scores were available in 269 patients (217F/52M, mean age 47.2 ± 13.4 years) and subjective ratings on overall effectiveness in 343 patients (272F/71M, mean age 47.1 ± 13.5 years). The CGRP mAb group (n = 79) had greater decreases in MIDAS scores (−42.2 vs. ‐11.8, p < 0.001) (Figure 4a), and had higher ≥50% (62.0% vs. 40.0%, p = 0.001) and ≥ 75% RR (45.6% vs. 29.5%, p = 0.011) (Figure 4b) at 6 months when compared with the onabotulinumtoxinA group (n = 190). Conversely, in the CGRP mAb group (n = 122) the proportions of patients reporting no improvement, partial improvement, significant improvement, and marked improvement were 11.5%, 40.2%, 32.0%, and 16.4%, respectively. The corresponding figures for onabotulinumtoxinA (n = 221) were 16.3%, 43.0%, 27.1%, and 13.6%, respectively (Figure 4c).

FIGURE 4.

Patient‐reported outcomes in chronic migraine patients. (a) Least‐squares mean change from baseline in Migraine Disability Assessment (MIDAS) score, (b) proportions of patients with ≥50% and ≥ 75% response, and (c) patient's subjective ratings on overall effectiveness. ***p ≤ 0.001, ** p ≤ 0.01.

Safety

Overall, a lower proportion of patients in the CGRP mAb group reported AEs (6.0% vs. 21.0%, p < 0.001) when compared with those in the onabotulinumtoxinA group. The most common AE in patients treated with CGRP mAbs was injection site pain or reaction (5.7%), followed by cold symptoms (0.3%). AEs reported by patients treated with onabotulinumtoxinA included lateral eyebrow elevation (9.3%), eyelid ptosis (7.5%), neck tightness (7.2%), injection site pain or reaction (2.4%), and cold symptoms (1.2%). All of the AEs were mild and transient, and none of the patients discontinued treatment because of AEs.

DISCUSSION

In this multicenter, real‐world study, it was found that treatment with CGRP mAbs in CM patients was associated with greater reductions in MMDs, MMSMDs, and MIDAS, and higher RRs at 6 months, when compared with onabotulinumtoxinA, which was suggestive of greater effectiveness of CGRP mAbs, even in DTT or MOH patients. However, all of these injectable preventive medications were effective and well tolerated. None of the patients discontinued treatment because of AEs, although AEs were more common in those treated with onabotulinumtoxinA. However, further studies are needed to confirm these findings.

One of the major strengths of the present study is the sample size (>300 patients in each arm), which could give more accurate estimates for the treatment effectiveness. Second, the study involved direct comparisons between these two classes of injectables at standard doses in CM patients. In comparison, prior meta‐analyses comparing the efficacies of CGRP mAbs and onabotulinumtoxinA were based on indirect comparisons derived from data in clinical trials of individual injectables, including those using non‐standard doses [30, 31, 32, 39]. In addition, data on DTT patients were lacking for onabotulinumtoxinA. Third, the data were of high quality. Although this was not a prospective study, the analysis was based on data collected prospectively for routine clinical use. The numbers of MMDs and MMSMDs were based on prospective headache diaries, which are mandatory for patients who apply for reimbursement from the Taiwan NHI [33]. Besides, patients without reliable diaries were excluded from the analysis. Finally, unlike clinical trials, which only recruited a highly selected group of patients, this was a real‐world study recruiting patients from both tertiary medical centers and regional hospitals across different regions of our country. Therefore, the concerns about selection bias could be minimized, and the findings could be more generalizable to patients encountered in routine practice.

It was found that CGRP mAbs were associated with greater effectiveness compared with onabotulinumtoxinA in CM patients, even in those with multiple treatment failures or MOH. In fact, the finding could be in line with post hoc analyses of double‐blind randomized placebo‐controlled trials and certain real‐world studies, in which CM patients with limited or no response to onabotulinumtoxinA could still benefit from CGRP mAbs [40, 41, 42, 43, 44, 45, 46, 47]. Besides, in a retrospective claims‐based analysis, it was found that patients treated with erenumab had greater reductions in both acute medication use and healthcare resource utilization when compared with those receiving onabotulinumtoxinA injections [48]. However, further studies are needed to clarify this issue.

The findings of the present study could have a more practical impact. According to practice guidelines and regulations, the clinical use of CGRP mAbs and onabotulinumtoxinA is frequently limited to patients who have failed multiple preventives [22, 33, 34, 49, 50]. Although the effectiveness of these two classes of injectables in DTT patients were supported by open‐label or real‐world studies [19, 20, 24, 25, 27], only CGRP mAbs were tested in double‐blind randomized placebo‐controlled trials [16, 17]. Therefore, the comparative effectiveness of CGRP mAbs and onabotulinumtoxinA in this patient population could not be inferred from published meta‐analyses [30, 31, 32]. In the current study, the reductions in headache frequency and disability were greater in patients treated with CGRP mAbs compared with those receiving onabotulinumtoxinA injections. However, all of these agents were effective in CM patients encountered in real‐world practice.

The safety profiles of the CGRP mAbs and onabotulinumtoxinA in the current study were consistent with prior reports. Although AEs were not uncommon, the majority were mild, local, and self‐limited. More importantly, none of the patients discontinued the treatment due to AEs. The proportions of patients reporting AEs were not very different from those observed in real‐world studies [19, 20, 21, 24, 25]. However, in the current study, whether a particular AE was present was specifically asked about based on those reported in prior studies, including our own [51], and therefore the percentages of certain AEs were slightly higher than those in clinical trials. Of note, lateral eyebrow elevation was seen in 9.3% of patients, and it was probably categorized as facial paresis in prior studies, which occurred in 1.2%–2.2% [22]. The discrepancy might also be attributed to cultural differences in esthetic perception. Lateral eyebrow lift is one of the commonly performed plastic surgeries in Western populations [52]. However, elevation of the lateral part of the eyebrow is often perceived as aggressive and intimidating in East Asian cultures, and is therefore considered as an AE. In fact, such an AE could be managed with additional injections of 2.5 U just above the lateral part of the eyebrow on each side for cosmetic purposes [22].

The present study has some limitations. First, as this was a retrospective study, the results of MIDAS, HADS, and PSQI at baseline, and certain outcome measures, especially PROs, were incomplete or unavailable. However, the present study included those that are the most important and most commonly used, such as the change from baseline in MMDs and ≥50% RRs. Second, patients were enrolled from routine practice without a predefined treatment algorithm. There could be concerns about confounding bias. In particular, patients in the CGRP mAb group had more preventive failures. For non‐reimbursed treatment courses, especially for patients with <3 preventive failures, onabotulinumtoxinA injection was preferred because of the costs. For reimbursed treatment courses during COVID outbreaks, CGRP mAbs were favored to avoid close contact of the head and face regions. Also, the completion rates were lower in the CGRP mAb group. In addition to financial issues, travel is another factor that could have contributed to the duration of treatment. Since administration of injectables should take place in clinics or hospitals according to the reimbursement regulations, monthly visits are necessary for monthly injections. Therefore, premature termination could occur despite significant improvement, since patients could have prescriptions of a duration of up to 3 months if they chose to take oral preventives only. However, the baseline clinical profiles of these two treatment groups were generally comparable. Although patients in the CGRP mAb group appeared to be more difficult to treat, they still had better treatment response compared with those in the onabotulinumtoxinA group. The results were consistent in subgroup analysis based on the presence of multiple prior treatment failures or MOH. In addition, the treatment responses to CGRP mAb were not very different between DTT and non‐DTT patients in the clinical trials [10, 11, 16, 17]. Third, the RRs were higher than those reported in some studies [8, 9, 19, 23, 26]. There could be concerns as to whether patients preferred to report a positive response in order to obtain approval for subsequent reimbursement. However, the responder rates were, in fact, close to some real‐world studies [24, 29, 53]. The variations in the RRs across different studies could be related to differences in patient characteristics, study design, reimbursement regulations or financial issues, methodology of outcome measure (intention‐to‐treat vs. per‐protocol analysis), and so on. Finally, CGRP mAbs in the current study were limited to galcanezumab and fremanezumab, which are the only two options available in our country. It is uncertain whether the findings could be applied to erenumab or eptinezumab.

In conclusion, CGRP mAbs were more effective in reducing migraine frequency and disabilities in CM patients than onabotulinumtoxinA, even in DTT or MOH patients. However, these two categories of injectables were effective and well tolerated, and should both be reasonable options in the treatment of CM patients. Prospective studies are needed to verify these findings.

AUTHOR CONTRIBUTIONS

Yen‐Feng Wang: Conceptualization; methodology; investigation; validation; formal analysis; funding acquisition; project administration; resources; writing – review and editing; writing – original draft; data curation. Fu‐Chi Yang: Data curation; resources. Lu‐An Chen: Data curation; resources. Ting‐Yu Chang: Data curation; resources. Hui‐Chen Su: Data curation; resources. Chun‐Pai Yang: Data curation; resources. Yi‐Hsien Tu: Data curation; resources. Yi‐Shiang Tzeng: Data curation; formal analysis; visualization; software. Shih‐Pin Chen: Data curation; resources; writing – review and editing. Jong‐Ling Fuh: Data curation; resources. Kuan‐Lin Lai: Data curation; resources; writing – review and editing. Yu‐Hsiang Ling: Data curation; resources; writing – review and editing. Wei‐Ta Chen: Data curation; resources; writing – review and editing. Shuu‐Jiun Wang: Conceptualization; methodology; data curation; validation; investigation; supervision; funding acquisition; resources; writing – original draft; writing – review and editing.

FUNDING INFORMATION

This study was funded in part by Taiwan National Science and Technology Council (109‐2314‐B‐075‐054 and 110‐2314‐B‐075‐041‐MY3 [to Y.F.W.], and 110‐2321‐B‐010‐005, 111–2321‐B‐A49‐004, 111–2314‐B‐075‐086‐MY3, 111–2321‐B‐A49‐011, and 112‐2321‐B‐075‐007 [to S.J.W.]); Taiwan Ministry of Health and Welfare (MOHW112‐TDU‐B‐211‐144,001 [to S.J.W.]); and Taipei Veterans General Hospital (V108C‐092, V109C‐096, V110C‐111, V111C‐161, V112C‐078, and V112D67‐003‐MY3 (to Y.F.W.]). This work was also supported by the Brain Research Center, National Yang Ming Chiao Tung University, from the Featured Areas Research Center Program within the framework of the Higher Education Sprout Project by the Ministry of Education of Taiwan. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

CONFLICT OF INTEREST STATEMENT

Y.F.W. has received personal fees as an advisor or a speaker from Allergan/AbbVie, Boehringer Ingelheim, Chugai, Eli Lilly, Hava Bio‐Pharma, Lundbeck, Novartis, Orient EuroPharma, Pfizer, Sanofi, Teva, UCB, and Viatris. He has received research grants from the Taiwan National Science and Technology Council, and Taipei Veterans General Hospital. W.T.C. has received honoraria as a speaker from Allergan/AbbVie, Hava Bio‐Pharma, Orient EuroPharma Pfizer, and Viatris. He has received research grants from the Taiwan National Science and Technology Council, and Taipei Veterans General Hospital. S.J.W. has received personal fees as an advisor or speaker from AbbVie, Orient EuroPharma, Pfizer, and Percept; and has been the PI in trials sponsored by Allergan/AbbVie, Lundbeck, Novartis, and Orient EuroPharma. He has received research grants from the Taiwan National Science and Technology Council and Taipei Veterans General Hospital. F.C.Y., L.A.C., T.Y.C., H.C.S., C.P.Y., Y.H.T., Y.S.T., S.P.C., J.L.F., K.L.L., and Y.H.L. report no relevant conflict of interest.

ETHICS STATEMENT

The study protocols were approved by the Institutional Review Board of Taipei Veterans General Hospital (VGH IRB No. 2022‐11‐003 BC), Tri‐Service General Hospital (A202305181), MacKay Memorial Hospital (23MMHIS095e), Chang Gung Memorial Hospital Linkou Medical Center (202301858B0), National Cheng Kung University Hospital (A‐ER‐112‐113), Taipei Municipal Gan‐Dau Hospital (1120003132), Kuang Tien General Hospital (KTGH 11227), and An Nan Hospital (TMANH112‐REC034).

Supporting information

Figure S1. Responder rates in the intention‐to‐treat populations. Proportions of patients with ≥50% and ≥ 75% response in monthly migraine days (MMDs) in the entire study population (a), and difficult‐to‐treat (DTT) (b) and medication‐overuse headache (MOH) (c) patients. ***p ≤ 0.001, **p ≤ 0.01, *p < 0.05.

Wang Y‐F, Yang F‐C, Chen L‐A, et al. Comparative effectiveness and tolerability of calcitonin gene‐related peptide (CGRP) monoclonal antibodies and onabotulinumtoxinA in chronic migraine: A multicenter, real‐world study in Taiwan. Eur J Neurol. 2024;31:e16372. doi: 10.1111/ene.16372

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.