Evaluating the Complementary Nature of Randomized Controlled Trials and Real-World Evidence Using Over 5 Years of Experience with Fremanezumab: A Narrative Review

Abstract

Introduction

Since the pivotal phase 3 randomized controlled trials (RCTs) and regulatory approval of fremanezumab, a calcitonin gene-related peptide (CGRP) monoclonal antibody (mAb) for migraine prevention, several real-world evidence (RWE) studies have been conducted. In this narrative review, we highlight the complementary value of RCTs and RWE studies, by examining > 6 years of fremanezumab efficacy and effectiveness data.

Methods

Embase, MEDLINE, MEDLINE In-Process, and Cochrane databases, bibliographies, and conference materials were searched for published RCT data and RWE on efficacy and effectiveness outcomes for fremanezumab. Electronic database and conference abstract searches were limited to articles published between January 2017 and September 2023, and January 2022 and September 2023, respectively.

Results

Of 48 publications identified for inclusion, 14 reported data from 10 RCTs and 34 reported data from 23 RWE studies. RCTs were of 8–12 weeks duration (with open-label extensions up to 52 weeks) and generally excluded individuals receiving preventive migraine treatment at baseline, and those with previous preventive treatment failures, previous exposure to CGRP pathway mAbs, or severe comorbidities. Exceptions were the UNITE study, which included participants with migraine and major depressive disorder, and the FOCUS studies, which included those with ≥ 2 prior preventive therapies. Most RWE studies included a broader population than RCTs, and were ≥ 3 months long, with some following participants for up to 2 years. Across RCTs, RWE studies, and population subgroup analyses, fremanezumab was consistently associated with improvements in monthly migraine and headache days, from as early as 4 weeks and sustained for up to 52 weeks.

Conclusion

While RCTs and RWE studies have differing strengths and weaknesses, these evidence-generation approaches have together highlighted the potential benefit of fremanezumab for preventive migraine treatment among broad, heterogenous patient populations.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12325-026-03647-8.

Key Summary Points

Why carry out this study?

Randomized controlled trials (RCTs) and real-world evidence (RWE) studies both allow for the generation of valuable data and insights; RCTs are the gold standard approach to establishing treatment efficacy and safety, while RWE studies typically include more heterogenous patient populations and can demonstrate effectiveness in routine clinical practice with a potentially longer follow-up

Integrating RCT and RWE findings provides a more comprehensive understanding of the use of treatments in clinical practice. We conducted this narrative review of RCTs and RWE studies using over 5 years of efficacy and effectiveness data for fremanezumab

What was learned from the study?

Across the RCT and RWE publications identified for inclusion, treatment with fremanezumab was consistently associated with reductions in monthly migraine and headache days

Despite their differing strengths and weakness, RCTs and RWE studies have provided invaluable data on the efficacy and effectiveness of fremanezumab for the preventive treatment of migraine across broad, heterogeneous patient populations

Introduction

Since the approval of fremanezumab as a preventive migraine treatment for adults in 2018 [1, 2], the evidence base for the efficacy and safety of the calcitonin gene-related peptide (CGRP) monoclonal antibody (mAb) has expanded, and now encompasses more than 5 years of randomized controlled trial (RCT) data and real-world evidence (RWE).

Phase 3 RCTs remain the gold standard for generating evidence on the efficacy and safety of a medicinal product [3] owing to their high internal validity (the extent to which the observed difference between groups can be correctly attributed to the intervention under investigation) and trial design that minimizes the impact of confounding variables [4]. Importantly, placebo-controlled RCTs can quantify placebo responses, an essential step in establishing the efficacy and safety of a medical intervention. RCTs are generally conducted among restricted, homogeneous populations under tightly controlled, experimental conditions [5]. Consequentially, RCTs provide highly reliable results; however, these conditions can limit the generalizability of RCT data to the real-world setting [4]. Additionally, the population sizes in RCTs tend to be smaller, and follow-up periods are often shorter compared with RWE studies [6].

RWE is derived from the analysis of data collected in routine clinical practice settings, and often relates to patient health status and/or the delivery of healthcare (e.g., electronic health record data, medical claims data, and data from product or disease registries) [7]. RWE studies often include large, unrestricted, heterogeneous patient populations and longer durations of follow-up compared with RCTs [8]. Additionally, RWE study designs are less restrictive than RCTs and can assess a range of outcomes, including patient preferences, natural history, compliance, and medication adherence [8]. Once the efficacy and safety of a medical intervention has been demonstrated in appropriate RCTs, RWE can complement these data, providing insights on disease epidemiology, infrequently occurring adverse events (AEs), special populations, and the effectiveness of medical interventions in routine clinical practice [9]. However, interpretations of RWE must be made with caution because of potential bias (e.g., due to selection and/or confounding) and lack of a comparative placebo-control group, as this can affect valid inference [4]. Other limitations include the availability, quality, and completeness of real-world data and comparability of outcomes collected in the real world versus those more rigorously assessed in RCTs [9].

The relative strengths and weaknesses of RCTs and RWE studies make these two approaches highly complementary, with each allowing for the generation of valuable data and insights within the respective limitations of their design (Fig. 1) [4, 8, 10–15]. As such, the collective and critical assessment of both RCT findings and RWE should ultimately inform clinical decision-making.

Fig. 1.

Strengths and weaknesses of RCTs and RWE studies. RCT randomized controlled trial, RWE real-world evidence

In this report, we aim to highlight the complementary value of RCTs and RWE using, as an example, evidence of efficacy and effectiveness accumulated for fremanezumab for the preventive treatment of migraine in adults.

Methods

This article is based on previously conducted studies and does not contain any new studies with human participants or animals involving any of the authors.

To investigate the breadth and depth of current fremanezumab efficacy and effectiveness data, we searched medical databases (including Embase, MEDLINE, MEDLINE In-Process, and Cochrane), bibliographies, and conference materials (e.g., from the 2022 The Deutsche Gesellschaft für Neurologie congress, the 2022 Migraine Trust International Symposium, and the 2023 American Academy of Neurology annual meeting) for publications summarizing the results of RCTs and RWE studies. Electronic database searches were limited to English-language publications between January 2017 and September 2023, and conference abstract searches were limited to abstracts published between January 2022 and September 2023.

The study screening process was performed in two phases. During the first level of screening, titles and abstracts of studies identified from the electronic databases and the internet searches were screened by one reviewer per record, with a 10% quality check of decisions by an independent researcher, with eligibility determined according to predefined inclusion and exclusion criteria. This process was repeated for full texts of studies selected during the first phase.

Data extraction was limited to study characteristics, participant characteristics, and standard efficacy and effectiveness outcomes including monthly migraine days (MMD; the number of days in a month where a person experiences a migraine, typically defined by established diagnosis criteria [16, 17], e.g., a qualified migraine headache lasting for > 30 min and having at least two of the following features: unilateral location, throbbing, moderate to severe intensity, exacerbated by exercise/physical activity or at least one of the following symptoms: nausea and/or vomiting, photophobia, or phonophobia; or any day during which a subject takes acute migraine medications, regardless of the duration of pain), monthly headache days (MHD; the number of days in a month during which a person experienced a qualified migraine or non-migraine headache that lasts continuously for ≥ 4 h or a headache of any duration for which acute treatment is administered), responder rates, migraine-related disability (evaluated by the Migraine Disability Assessment [MIDAS], and/or the Headache Impact Test-6 [HIT-6]), days of acute medication use, pain severity, and participant quality of life (evaluated by the Migraine-Specific Quality of Life [MSQoL] questionnaire). Where available, efficacy and effectiveness data were extracted for subgroups of particular interest: participants with episodic migraine (EM, defined as headaches that occur on ≤ 14 days per month) [18, 19]; chronic migraine (CM, defined as experiencing headaches on ≥ 15 days per month for ≥ 3 months, with features of migraine occurring on > 8 days per month) [18–20]; > 2 prior migraine preventive treatment failures and/or prior CGRP pathway mAb exposure; and participants with psychiatric comorbidities.

This report will focus on the complementary value of the data extracted from these RCT and RWE publications, with a focus on MMD and MHD as key efficacy and effectiveness outcomes. While MMD is often considered a more clinically targeted endpoint for evaluating the efficacy of preventive migraine treatments [16, 17, 21], individuals with CM typically experience frequent headache days with variable symptom profiles, and not all headache days meet full migraine criteria, making classification of individual days as migraine versus non-migraine less distinct [22]. In light of this, studies that include individuals with CM typically assess the change from baseline in the number of MMD and/or MHD of moderate/severe intensity and/or requiring use of an acute migraine medication as the primary efficacy endpoint [16].

Results

A total of 48 publications were identified for inclusion: 14 publications from 10 RCTs and 34 publications from 23 RWE studies (Supplementary Material). Nine of the 10 RCTs were supported by peer-reviewed journal articles, while one was represented by conference abstracts/presentations (UNITE). Of the 23 RWE studies, 20 were supported by peer-reviewed journal articles, with the remaining three being supported by conference abstracts/presentations (Cheng et al. [2022]; Afridi et al. [2022]; PEARL).

Study Characteristics

Information on study duration (RCTs)/duration of follow-up (RWE studies) and participant numbers are displayed in Fig. 2.

Fig. 2.

Study durations/durations of follow-up and participant numbers in the 10 RCTs and 23 RWE studies. RCT randomized controlled trial, RWE real-world evidence

RCTs were 8–52 weeks in duration [23–34], whereas RWE studies were 1–24 months in duration [35–58]. RCTs included 50–1890 participants [23–34], while RWE studies included 10–1003 participants [35–58].

Participant Characteristics

Migraine Diagnoses

Information on migraine diagnoses in the RCTs and RWE studies are displayed in Fig. 3. Of the 10 RCTs, two included only participants with EM [24, 27], two included only participants with CM [25, 26], and six included participants with either EM or CM [23, 28–34]. Among the 23 RWE studies, nine studies included participants with CM only [39, 40, 44, 47, 50, 51, 55, 56, 58] and the remaining 14 studies included participants with both EM and CM [35–38, 41–43, 45, 46, 48, 49, 52–54, 57].

Fig. 3.

Migraine diagnoses in the 10 RCTs and 23 RWE studies. ^aEM, defined as headaches that occur on ≤ 14 days per month; ^bCM, defined as experiencing headaches on ≥ 15 days per month for ≥ 3 months, with features of migraine occurring on > 8 days per month. CM chronic migraine, EM episodic migraine, RCT randomized controlled trial, RWE real-world evidence

Prior Migraine Preventive Treatment and Concomitant Migraine Preventive Medication Use

In the FOCUS RCT, all participants had received at least two prior migraine preventive therapies, with almost 20% having received four prior migraine preventive therapies [23]. In contrast, the 12-week HALO-EM [24, 59], HALO-CM [25, 60], and phase 2b/3 RCTs in Japan and Korea [26, 27] excluded individuals who had previously failed (as a result of lack of efficacy) two or more of the following medication groups: (A) divalproex sodium and sodium valproate; (B) flunarizine and pizotifen; (C) amitriptyline, nortriptyline, venlafaxine, and duloxetine; and (D) atenolol, nadolol, metoprolol, propranolol, and timolol. Additionally, in the majority of RCTs, individuals who had used onabotulinumtoxinA within 3–4 months prior to screening and/or used opioid- or barbiturate-containing treatments on ≥ 4 days per month were excluded [23–28, 30, 32, 33].

Seventeen of the RWE studies presented data on prior and concomitant migraine preventive therapies, but the amount of prior exposure varied considerably and included other CGRP pathway mAbs, beta blockers, anticonvulsants, and therapies for psychiatric comorbidities (e.g., antidepressants) [35, 38, 40–44, 47–50, 52, 55–58, 61–65].

Prior CGRP Pathway mAb Exposure

The HALO-EM, HALO-CM, FOCUS, FOCUS open-label extension (OLE), and phase 2b/3 RCTs in Japan and Korea excluded individuals with any prior exposure to a CGRP pathway mAb [23–27, 32, 59, 60]. The UNITE RCT excluded individuals with prior CGRP pathway mAb failure or exposure to a CGRP pathway mAb within 5 half-lives prior to screening [29, 30]. Six RWE studies reported data for participants who had switched from or had been previously exposed to another CGRP pathway mAb [42, 47, 48, 64–66].

Psychiatric Comorbidities

The phase 4 UNITE RCT assessed fremanezumab for participants with comorbid major depressive disorder (MDD) [30]. However, in most other RCTs, participants who had evidence or medical history of clinically significant psychiatric issues were excluded [23–27]. Twelve of the 23 RWE studies reported baseline comorbidities with a focus on psychiatric comorbidities [38, 41, 45, 49–52, 55–57, 63, 67], with anxiety disorders and depression reported in six of these studies [45, 49, 50, 52, 55, 63].

Efficacy and Effectiveness Outcomes

Efficacy and effectiveness outcomes assessed in the 10 RCTs and 23 RWEs are shown in Fig. 4. MMD was reported for all 10 RCTs, while the next most common efficacy outcomes, days of acute medication use, MHD, responder rates, and HIT-6, were reported by 8, 7, 7, and 7 RCTs, respectively. Only 2 RCTs presented results for MSQoL, and none identified pain severity as an outcome, and any mention of pain was associated with AEs such as injection-site pain. The most common effectiveness outcomes for the 23 RWE studies were MMD (16 studies), MHD (16 studies), and responder rates (13 studies). No RWE studies reported on MSQoL, because it is not routinely collected in clinical practice.

Fig. 4.

Standard efficacy and effectiveness outcomes and participant subpopulations assessed in RCTs and RWE studies of fremanezumab. An endpoint analysis was counted for a population subgroup if patients falling within that subgroup were explicitly stated as being included in the total analysis population for which the endpoint was measured or if the endpoint was measured separately for the specific subgroup. CGRP calcitonin gene-related peptide, CM chronic migraine, EM episodic migraine, HIT-6 Headache Impact Test-6, mAb monoclonal antibody, MHD mean headache days, MIDAS Migraine Disability Assessment, MMD monthly migraine days, MSQoL Migraine-Specific Quality of Life, RCT randomized controlled trial, RWE real-world evidence

Overall, RWE studies tended to report results for participant subgroups, including those with prior and concomitant preventive treatment use, prior CGRP pathway mAb exposure, and psychiatric comorbidities, more often than RCT studies.

MMD and MHD

Across all RCT and RWE studies that assessed MMD and/or MHD as efficacy endpoints, fremanezumab treatment was associated with improvements in these measures. See Supplementary Material for a full breakdown of MMD and MHD results for the RCTs and RWE studies. In general, substantial reductions in MMD and/or MHD were seen as early as 4 weeks after initiation of fremanezumab treatment.

In the HALO-EM/CM long-term extension (LTE) study, where MMD at baseline was 16.4 days, the mean change from baseline in MMD over 12 months was − 8.0 days and − 7.2 days among participants with CM treated with monthly and quarterly fremanezumab, respectively [33]. For participants with EM enrolled in the HALO-EM/CM LTE study, the mean change from baseline in MMD over 12 months for monthly and quarterly fremanezumab was − 5.1 days (baseline MMD 9.1 days) and − 5.2 days (baseline MMD 9.2 days), respectively [33]. Among RWE studies, the largest improvement in MMD was seen in a small (n = 10), retrospective, observational, independent study among participants with CM, where change from baseline in MMD over 3 months was − 16.7 days; MMD at baseline in this group was 21.8 days [40]. Regarding participants with EM, substantial reductions in MMD were reported for this population in a 6-month, retrospective, panel-based chart review [64], where the mean change from baseline at Month 6 of fremanezumab treatment was − 7.7 days (baseline MMD 10.0 days).

In general, there was greater improvement in MHD with fremanezumab in RWE studies compared with RCTs; the highest mean reduction in an RCT (HALO-EM/CM LTE) was − 7.8 days over 12 months among participants with CM who received monthly fremanezumab (baseline MHD 16.3 days) [33], while the highest mean reduction in a RWE study, the 24-week prospective, multicenter FRIEND2 study in participants with CM, was − 14.7 days over 12 weeks (baseline MHD 22.7 days) [38]. A similar trend was observed for the EM subpopulation; the highest mean reductions in MHD among RCTs for participants with EM (HALO-EM/CM LTE) were − 4.8 days (baseline MHD 8.5 days) and − 4.9 days (baseline MHD 8.6 days) over 12 months of monthly and quarterly fremanezumab, respectively [33], whereas higher reductions were observed in the RWE publications. For example, in a 3-month, retrospective, panel-based chart review [43], MHD were reduced by − 6.0 days over 3 months of treatment (baseline MHD 10.4 days).

Participants with Prior Migraine Preventive Treatment Failures

Among participants with 2–4 documented migraine preventive treatment failures enrolled in the FOCUS OLE study, fremanezumab was efficacious in reducing MMD and MHD over 24 weeks. The mean change from baseline in MMD was − 5.1 days, − 5.5 days, and − 4.7 days, respectively, for quarterly fremanezumab, monthly fremanezumab, and placebo. For MHD, the mean change from baseline was − 4.8 days for quarterly fremanezumab, − 5.2 days for monthly fremanezumab, and − 4.5 days for placebo. Participants with < 2 prior migraine preventive treatment failures and those with ≥ 2 prior migraine preventive treatment failures could not be compared because a subgroup with < 2 prior migraine preventive treatment failures was not included.

In a 6-month, retrospective, panel-based online chart review reported by Driessen et al. [63], participants with < 2 prior migraine preventive treatment failures had lower MMD at baseline (10.5 days) compared with participants with ≥ 2 prior migraine preventive treatment failures (13.2 days); participants in both groups achieved a percentage reduction in MMD at 6 months of greater than 70%. In the same study, participants with < 2 prior migraine preventive treatment failures had lower MHD at baseline (11.0 days) compared with participants with ≥ 2 prior migraine preventive treatment failures (14.6 days), suggesting a less severe population [63].

Participants with Prior CGRP Pathway mAb Exposure

In the aforementioned real-world Driessen et al. (2022) study, participants naive to another CGRP pathway mAb had similar reductions in MMD at 6 months after fremanezumab initiation compared with those with prior CGRP pathway mAb exposure (72.4% and 68.7%, respectively).

When comparing the proportion of participants reaching a ≥ 50% reduction in MMD, findings varied between the 24-month, observational, prospective PEARL and FINESSE studies. In the PEARL study, the percentage of overall participants and participants with either EM or CM with previous CGRP pathway mAb exposure reaching a ≥ 50% reduction in MMD after 9 months was 48.9%, 46.2%, and 50.0%, respectively [65]. It is important to note that these PEARL data were based on a conference abstract/presentation, not a peer-reviewed publication. In the FINESSE study, only 67 participants (EM: n = 38; CM; n = 29) had prior exposure to a CGRP mAb and results were variable [66].

In real-world publications that evaluated MHD, participants naive to another CGRP pathway mAb exhibited a larger reduction in MHD than participants previously exposed to a CGRP pathway mAb [42, 47]. In the real-world Driessen et al. (2022) study, the prior exposure subgroup had 58.9% reduction in MHD compared with a 70.7% reduction in MHD for the no prior exposure group at 6 months after fremanezumab initiation. In a 24-month, retrospective chart review by Cohen et al. (2022), the mean change from baseline in MHD following 6 months of treatment with fremanezumab was − 5.6 days in CGRP pathway mAb-naive participants with CM versus − 0.6 days in the prior exposure group.

Participants with Psychiatric Comorbidities

Participants with migraine and MDD enrolled in the UNITE RCT experienced a significant reduction in MMD versus placebo (− 5.1 days for fremanezumab from 15.1 days at baseline and − 2.9 days from 14.4 days at baseline for placebo; P < 0.0001) following treatment with fremanezumab during the 12-week double-blind period (DBP) [30]. It should be noted that the aforementioned UNITE RCT data are from a conference abstract/presentation, rather than a full-length publication. These results were similar to those seen in the 12-week HALO-CM RCT among participants without psychiatric comorbidities; baseline MMD were 16.2 days (quarterly fremanezumab), 16.0 days (monthly fremanezumab), and 16.4 days (placebo) and reductions from baseline were − 4.9 days, − 5.0 days, and − 3.2 days, respectively [25].

In the real-world Driessen et al. study (2022), similar baseline MMD and changes from baseline in MMD after 6 months of fremanezumab treatment were seen for participants with migraine and comorbid MDD (mean reduction of − 9.9 days from 14.5 days), participants with migraine and comorbid generalized anxiety disorder (GAD; mean reduction of − 9.5 days from 14.3 days), and the overall participant population (mean reduction of − 9.2 days from 12.7 days). For both the UNITE RCT and Driessen RWE study, participants with and without psychiatric comorbidities could not be compared because a subgroup without psychiatric comorbidities was not reported.

A direct comparison of changes in MHD between participants with difficult-to-treat CM with or without psychiatric comorbidities was made in a post hoc, subgroup analysis of a 3-month, prospective, open-label, multicenter independent Greek cohort study [68]. In this study, participants with and without psychiatric comorbidities had similar mean baseline MHD (22.4 days and 23.9 days, respectively) and similar reductions in MHD between baseline and Months 3 to 18 (− 9.8 days and − 8.9 days, respectively) [68].

Discussion

This report aimed to highlight the complementary value of RCT data and RWE by reviewing results of fremanezumab efficacy and effectiveness studies published over 5 years or more. Our findings illustrate how together RCTs and RWE studies have contributed to the breadth (characterized by different subpopulations of individuals with migraine) and depth (varied efficacy and effectiveness outcomes) of our knowledge of fremanezumab for preventive migraine treatment. This approach to evidence generation (i.e., combination of randomized and non-randomized methods) that includes a variety of data sources, with data prospectively or retrospectively collected, has been successful in highlighting the potential benefit of fremanezumab for preventive migraine management across broad, heterogenous patient populations.

When assessing efficacy and effectiveness data from RCTs or RWE studies, it is important to consider the complementary value of differing elements in study design, including study duration/duration of follow-up, sample size, and participant inclusion/exclusion criteria, and how these may affect the interpretation of results. For instance, the double-blind periods of fremanezumab RCTs were all 12 weeks in duration apart from one LTE study where participants were randomized for up to 52 weeks of treatment. The follow-up periods of fremanezumab RWE studies were between 4 and 12 months, which supported findings of sustained, long-term improvements in MMD and MHD observed in the RCT LTEs. Interestingly, a higher relative proportion of RCTs (6/10) had > 500 participants compared with RWE studies (3/23). While larger participant populations tend to correlate with more robust and precise data findings [69], smaller, independent studies are important in providing data from different regional settings, often among individuals with more severe or difficult-to-treat migraine.

Across RCTs and RWE publications, fremanezumab treatment was consistently associated with improvements in MMD and MHD. In contrast to the patterns seen in other disease areas, in many instances, the improvements reported in MMD and MHD in fremanezumab RWE publications were greater than those seen in RCT publications. This may reflect that patients with more heterogenous and/or complex clinical profiles are better represented in RWE studies, including those with more severe or refractory migraines, prior therapeutic failure, higher migraine frequency, and/or comorbidities (e.g., depression). These factors result in higher baseline scores and greater room for improvement across various outcome measures. It should also be noted that the generally shorter treatment durations assessed in RCTs versus RWE studies may impact the extent of response to fremanezumab, as RCTs would have captured data among participants who only responded to treatment within 12 weeks. As such, RWE studies may have the potential to identify those individuals who take longer to respond to treatment, which was estimated to make up 55.1% of individuals who did not initially respond to CGRP pathway mAb treatment within 12 weeks, in a multicenter, prospective, observational study published in 2023 [70].

It is important to note that these observations are descriptive in nature and intended to be hypothesis-generating rather than confirmatory. Additionally, any interpretations should consider the increased susceptibility of RWE studies (compared with RCTs) to certain analytical biases that can impact treatment effect estimation, including selection bias, regression to the mean, responder bias, performance bias, and confounding bias. For example, the outcomes of patients referred to headache specialists for treatment as part of routine clinical practice may be more favorable than those of the underlying population as a result of the systematic exclusion of individuals who do not have access to such services, due to selection bias [71, 72]. Regarding regression to the mean, RWE studies may be more susceptible to this, including in the context of preventive migraine treatment. In routine clinical practice, patients often initiate preventive treatment during periods of particularly high disease burden, such as a flare in migraine frequency. As a result of the fluctuating nature of migraine, these high baseline values may be followed by lower values at subsequent assessments, even in the absence of a true treatment effect [73, 74]. Conversely, RCTs are less susceptible to regression to the mean owing to pre-defined eligibility criteria and study design features, such as baseline run-in periods and placebo control arms, that account for natural variability in disease activity [75]. Responder bias may be introduced in longer-term, prospective RWE studies where participant numbers decrease over time as participants are lost to follow-up, have missing data entries, or discontinue treatment because of lack of efficacy, adverse events, non-compliance with study procedures, etc. In these scenarios, the interpretation of data at later timepoints may be biased as participants who persevere with treatment are more likely to be responders. As a result, outcomes may appear more favorable than would be seen in the full study population. Performance bias may also influence the estimation of treatment effects in open-label, single-arm RWE studies, as the absence of blinding can alter participant behavior, care provision, and treatment adherence, potentially leading to an overestimation of the observed benefit [76]. Confounding bias can also distort treatment–outcome associations in RWE studies, leading to over- or underestimation of effectiveness if differences in patient characteristics are not adequately controlled [77]. Regarding fremanezumab specifically, access-related factors may influence the heterogeneity of RWE studies. In many healthcare systems, CGRP pathway mAbs are reimbursed only after failure of a specified number of prior preventive treatments or are restricted to particular populations (e.g., high-frequency EM or CM) [78]. In addition, their higher cost compared with traditional oral preventive therapies can further limit access [79]. Together, these factors may result in more selected and/or severe patient populations being represented in RWE studies. With respect to participant subgroups of particular interest, RCT and RWE studies have provided efficacy and effectiveness data across a broad spectrum of individuals with migraine, including those with multiple prior migraine preventive treatment failures, individuals with prior exposure to a CGRP pathway mAb, and/or individuals with comorbid psychiatric conditions. These findings are particularly valuable as they suggest that fremanezumab has the potential to benefit those patients who are traditionally more difficult to treat.

Results of both the FOCUS DBP and OLE study and a RWE retrospective chart review demonstrated the efficacy and effectiveness of fremanezumab among participants with ≥ 2 prior migraine preventive treatment failures. Considering that the large majority of reimbursement authorities in the USA [80], EU, and other regions [81] require individuals to have tried and experienced failure of ≥ 2 prior preventive treatments in order to qualify for access to or reimbursement of CGRP pathway mAbs, most individuals who might benefit from CGRP pathway mAbs in the real world would fall within this category. Evidence of the effectiveness of fremanezumab among individuals with ≥ 2 prior preventive treatment failures is therefore imperative.

In the RWE Driessen et al. (2022) and PEARL studies, responder rates to fremanezumab (based on reductions in MMD) were similar among participants with and without prior exposure to another CGRP pathway mAb. The other RWE publications which reported MMD response rates in participants with prior exposure to another CGRP pathway mAb had low sample sizes [35, 41, 66]. As such, the findings of these publications should be interpreted with caution. With regard to improvements in MHD, findings from both the Driessen et al. (2022) and Cohen et al. (2022) RWE studies indicated that participants naive to CGRP pathway mAbs had greater reductions in MHD from baseline versus those previously exposed to another CGRP pathway mAb. The Cohen et al. (2022) study showed less benefit within the naive exposure group and even less in the prior exposure group compared with other RWE studies, which is likely due to the concomitant use of onabotulinumtoxinA. Although responder rates to CGRP pathway mAbs are generally high, with most RCTs showing rates between 40% and 70% [82], a substantial proportion of recipients do not respond to treatment within 12 weeks [70, 82, 83].

As a result of differing therapeutic targets (CGRP ligand versus receptor), it is reasonable to question whether different mechanisms of the CGRP pathway mAbs may play a relevant role in the variability of efficacy found among participants. However, evidence regarding switching between CGRP mAbs is lacking and clinical practice guidelines have not provided recommendations in this regard [84]. Studies that have assessed the effectiveness of CGRP pathway mAbs among individuals with previous exposure to another CGRP pathway mAb are therefore important in elucidating potential benefits of switching, especially for individuals with limited alternative options.

In publications that included analyses for individuals with migraine and comorbid MDD or GAD, fremanezumab was shown to be effective in reducing MMD from baseline in this population. Despite the absence of direct comparator subgroups of individuals with migraine without MDD or GAD, reductions in MMD demonstrated in the UNITE RCT and real-world Driessen et al. (2022) study were similar to those seen among the comparable HALO-CM RCT participant group (where participants with severe psychiatric comorbidities were excluded) and the overall population of participants with or without MDD/GAD in the real-world Driessen et al. (2022) study, respectively. In line with these findings, direct comparison of MHD improvements with fremanezumab among individuals with migraine with and without comorbid psychiatric conditions demonstrated similar results for both groups in a RWE Greek cohort study [68].

Depression is one of the most common psychiatric diseases comorbid with migraine and individuals with migraine are 2–4 times more likely to develop depression than individuals without migraine [85]. Although this association has been well established, limited data are available on the efficacy of migraine preventive therapy in participants with migraine and psychiatric comorbidities. Considering this, the evidence of CGRP pathway mAb efficacy and effectiveness in individuals with migraine and psychiatric conditions reported here is extremely valuable in informing treatment management strategies for this population.

It is important to acknowledge that the purpose of this narrative review was to describe and contextualize the existing evidence of fremanezumab efficacy and effectiveness from RCTs and RWE studies, rather than to quantitatively compare or pool results across studies. A formal meta-analysis was not feasible because of the substantial heterogeneity of the RCTs and RWE studies. As such, findings should be interpreted qualitatively and within the context of the individual studies.

Although the efficacy and safety profile of fremanezumab has been well established in RCTs and RWE studies, evidence in certain areas remains limited. For instance, a recent systematic review reported that fremanezumab was effective in post hoc analyses of RCTs and RWE studies in adults with migraine complicated by medication overuse [86]; however, the available data are currently limited. Further robust, standardized RWE studies and long-term safety and effectiveness data are needed to better define the role of fremanezumab in this population. To date, no head-to-head studies directly comparing fremanezumab with other CGRP pathway mAbs, or with other preventive migraine treatments, have been conducted. As such, comparisons are limited to indirect meta-analyses, small observational studies, and switch cohorts (participants who switched to fremanezumab from another CGRP pathway mAb). Regarding the RCT and RWE study patterns discussed for fremanezumab, it should be noted that comparisons with other CGRP pathway mAbs have not been and are beyond the scope of this narrative review. Furthermore, it was not possible to determine if these observations are specific to fremanezumab or represent a class effect; such determination was also beyond the scope of this study.

Every effort was made to include all relevant publications within the specified timeframe of this narrative review. However, subsequent to this, in a fast-moving area such as migraine, new studies have been published since our searches were conducted. One such study, the SPACE EM study, is of particular interest, as it reports the efficacy and safety of fremanezumab in children and adolescents with EM, a population for which CGRP pathway mAb use has not been evaluated in a randomized clinical trial before [87]. These data, together with real-world experience, will be essential in informing best practice and optimal treatment management for children and adolescents with migraine.

Conclusion

The complementary value of data generated from RCTs and RWE studies was demonstrated through this narrative review of over 5 years of data for fremanezumab. Across over 30 studies of varying design, size, duration, and patient populations, fremanezumab showed consistent efficacy and effectiveness in improving MMD and MHD. Thus, with differing strengths and weaknesses, RCTs and RWE studies have together highlighted the potential benefit of fremanezumab for preventive migraine treatment across broad, heterogenous patient populations. Further high-quality RCTs and RWE studies will likely continue to provide invaluable data and insights into the optimal use of preventive migraine therapies such as fremanezumab.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

Piero Barbanti, Timothy Smith, Richard James Stark, Alejandra Elizabeth Guevara Morel, and Verena Ramirez Campos contributed to the study concept and design, interpretation, and critical revision of the manuscript; Caroline Ling, Preet Bhogal, and Caoimhe Leonard contributed to the study design, data acquisition, analysis, and critical revision of the manuscript.

Funding

This study was funded by Teva Pharmaceuticals. The journal’s Rapid Service Fee and Open Access Fees were also funded by Teva Pharmaceuticals.

Data Availability

The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Conflicts of Interest

Piero Barbanti has received consulting fees from AbbVie, Allergan, Angelini, Assosalute, Bayer, Eli Lilly, Fondazione Ricerca e Salute, Lundbeck, Orion Pharma, Organon, Pfizer, and Teva Pharmaceuticals; has received payment or honoraria from AbbVie, Allergan, Angelini, Bayer, DOC Pharma, Eli Lilly, Lundbeck, Organon, Orion Pharma, Teva Pharmaceuticals, and Viatris; has received support for attending meetings and/or travel from AbbVie, Lundbeck, Teva Pharmaceuticals, and Viatris; has participated on a data safety monitoring board or advisory board for AbbVie, Angelini, Eli Lilly, Lundbeck, Organon, Orion Pharma, Pfizer, and Teva Pharmaceuticals; has received other financial or non-financial interests from AbbVie, Amgen, Alder, Allergan, Biohaven, ElectroCore, Eli Lilly, GSK, Lundbeck, New Penta, Noema Pharma, Novartis, and Teva Pharmaceuticals. Timothy Smith received fees for speaker’s bureau from Amgen, AbbVie, Biohaven, Novartis, and Eli Lilly, and for advisory boards or consultancies from AbbVie, Biohaven, Amgen, Eli Lilly, Lundbeck, Impel Neuropharma, Neurolief, Nocira, Nuvie Bio, Theranica, Teva Pharmaceuticals, and Vorso/Nesos. He has received research support from AbbVie, Aeon, Amgen, BioAge Labs, Biohaven, Charleston Labs, ElectroCore, Eli Lilly, Impel Neuropharma, Lundbeck, Nocira, Novartis, Novo Nordisk, Pfizer, Structure Therapeutics, Teva Pharmaceuticals, Theranica, and Vorso/Nesos. Richard James Stark has received advisory board and/or lecture fees from AbbVie, Eli Lilly, Lundbeck, Pfizer, Teva Pharmaceuticals, and Viatris. Verena Ramirez Campos, Maurice Theodoor Driessen, and Alejandra Elizabeth Guevara Morel are employees of Teva Pharmaceuticals. Caroline Ling and Caoimhe Leonard are employees of RTI Health Solutions, which conducted the data collection and analyses that were funded by Teva Pharmaceuticals. Preet Bhogal is currently an employee of WEP Clinical and was an employee of RTI Health Solutions at the time the study was undertaken, which conducted the data collection and analyses that were funded by Teva Pharmaceuticals.

Ethical Approval

This article is based on previously conducted studies and does not contain any new studies with human participants or animals involving any of the authors.