Abstract
Background
Calcitonin gene-related peptide-targeted drugs have proven safe and effective for migraine prevention in large randomized-controlled, double-blind trials with an average duration of six months. Open-label studies may provide additional information on the long-term safety and efficacy of these substances.
Methods
We searched PubMed for open-label trials with calcitonin gene-related peptide(-receptor) monoclonal antibodies and calcitonin gene-related peptide-receptor antagonists. We summarized and critically analyzed the literature in a narrative way.
Results
Overall, 13 open-label trials were included in this review (n = 4 for erenumab, n = 4 for galcanezumab, n = 3 for fremanezumab, n = 1 for eptinezumab, n = 1 for atogepant). Open-label trial duration ranged between 12 and 264 weeks. No safety concerns emerged, and the adverse events profile was similar to the double-blind study phase. Discontinuation rates were generally low with >75% of patients remaining in the trials after one year. Efficacy data showed a sustained reduction of migraine frequency throughout the trials, along with a lasting improvement in quality of life.
Conclusions
The open-label study program for calcitonin gene-related peptide-targeted migraine preventives confirms the favorable safety and efficacy profile of these drugs over time. Treatment adherence appears higher than with previous unspecific migraine preventives. Real-world data and post-marketing surveillance studies may corroborate and complement open-label results.
Introduction
Calcitonin Gene-Related Peptide (CGRP) is a potent endogenous vasodilator that plays a crucial role in migraine pathophysiology (1,2). CGRP-targeted therapies in migraine prevention comprise monoclonal antibodies (mAbs) against either the CGRP-receptor (erenumab) or CGRP ligand (galcanezumab, fremanezumab, eptinezumab), and small molecule CGRP-receptor antagonists, the so-called gepants (atogepant and rimegepant). The superiority of these drugs over placebo have been proven in comprehensive, double-blind (DB), placebo-controlled study programs (see Messina et al. in this Special Issue).
The duration of the DB studies was usually twelve weeks and does not allow conclusions on the long-term effects of these substances. Long-term, open-label (OL) studies can therefore provide a meaningful addition to the DB study program (3). OL studies or unblinded studies are defined as clinical trials in which both the subjects and the investigators are aware of the treatment being given. For CGRP-targeted drugs, OL studies were mostly conducted as OL extensions (OLE) of previous DB trials, providing an additional exposure time of one year and, in one case, five years (4).
The key reason for conducting OL trials is to examine the safety and tolerability of a drug over time (3). There can be many benefits to conducting OL trials, but many challenges as well. The benefits in OL trials are that they allow investigators to determine if the incidence rate of adverse events (AEs) seen in the shorter pivotal trials keeps stable, or changes over a longer period. Moreover, OL trials give the opportunity to address the persistence of known AEs over time. For clinicians, OL trials can help us see if a patient does well on a treatment, if they may have a sustained response over time and if the response will continue to improve or if the response will stabilize over time. For patients in a clinical trial, it allows them to continue on a treatment that may not yet be available on the market.
There are also limitations to OL trials that are important to understand when evaluating the data. These trials do not have a placebo arm so listed AEs can be large and can include events that can normally occur over a given year. For example, upper respiratory infection and urinary tract infections may be common occurrences over the course of a 52-week time period in women and may not have anything to do with a medication, and this should be considered when evaluating the data. Any OL trial that continues for 52 weeks or longer is going to lose a number of participants because of the difficulty of continuing long term in a trial setting, so understanding the average drop off rate in a long-term OL trial is important for the reader as well. For example, in previous OL trials testing efficacy of antiepileptics for migraine prevention in adults or adolescents, the discontinuation rate ranged between 29% and 32% (5,6). Those participants who remain in the trial may be doing so because they are responding to the treatment, giving a clear bias to the responder data over time.
If we examine CGRP-targeted treatments, there are several notable concerns that warrant consideration in OL trials. These include possible vascular consequences of blocking the CGRP signaling pathways, as well as concerns over long term gastrointestinal safety (7).
In this narrative review, we present an overview of the published OL and OLE trials for CGRP(-receptor) antibodies and CGRP-receptor antagonists in the preventive treatment of migraine.
Methods
Two independent reviewers (RDI, MC) searched PubMed for original research publications on OL and OLE trials with CGRP(-receptor) antibodies and CGRP-receptor antagonists. We also included selected conference abstracts where publicly available. The combination of the following terms were used for search: “long-term OR long term”, “open-label OR open label”, “extension” AND “erenumab”, “galcanezumab”, “fremanezumab”, “eptinezumab”, “rimegepant”, “atogepant”. For the present narrative review, we did not include: real-world studies, studies with retrospective design, trials involving patients with headaches other than migraine with or without aura or chronic migraine, OL trials with duplication of study population and follow-ups. In case of duplication of study population, we considered the trial reporting data on efficacy in terms of monthly migraine days (MMD) reduction or responder rate. The search was carried out on 30 April 2022. The full text of articles deemed eligible based on title and abstract was then evaluated by two other authors (BR and MT). In the next paragraphs, we will define OLE studies as those trials including patients rolled-over from previous randomized controlled trials (RCTs), and OL studies as trials including patients directly enrolled in an OL study design. We summarized the key efficacy and safety endpoint in a narrative way and in an overview table (Table 1).
Key characteristics of open-label (OL) studies with CGRP-targeted substances for migraine prevention.
AE, adverse events; CGRP, calcitonin gene-related peptide; CM, chronic migraine; DB, double-blind; EM, episodic migraine; MMD, monthly migraine days; n/a, not available; NCT, National Clinical Trial number; OL, open-label; OLE, open-label extension.
Results
Erenumab
The long-term safety and efficacy of erenumab has been assessed in five OLE of prior DB, placebo-controlled studies.
From the Phase II erenumab study in episodic migraine (EM), 383 patients entered an OLE and initially received 70 mg erenumab, which was later increased to 140 mg (4,8,9). After three years, 235 patients (61.3%) were still in the trial and 215 (56.1%) completed the OLE after five years (4,9). Patients reported on average 8.8 MMD at baseline, which progressively decreased to 3.7 MMD after the first OLE year (8). Patients who received placebo or lower erenumab doses in the DB trial phase (DBTP) quickly caught up with the prior 70 mg-group and reached comparable MMD levels already in the first OLE month (8). The reduction of MMD remained stable throughout the OLE with a mean change of −5.3 MMD from baseline after five years (4). The proportions of patients with a ≥50%, ≥75% and 100% response during the last four OLE weeks were 65%, 42%, and 26% at week 64; (4). Response rates increased to 71%, 47%, and 36% at week 268, respectively (4). Accordingly, patients reported a significant reduction of days with acute medication use and a stable improvement in functionality and quality of life (4,8). As this is an OLE, these results are not compared to placebo and are reported compared to the patient’s baseline at entry into the trial.
Tolerability data revealed a similar adverse event (AE) profile as in the DBTP. AEs during the OLE occurred at an even lower exposure-adjusted incidence rate than in the placebo group of the parent study (123.0 AE/100 subjects-years in the OLE vs. 230.5 in the placebo group of the DBTP). Nasopharyngitis, upper respiratory tract infections, and influenza remained the most common AEs after one, three, and five years (4,8,9). Serious AEs occurred in less than 5% of patients and were mostly isolated events without a clear relation to treatment. Only 14 patients (2.5%) discontinued treatment due to AEs in the first year, 17 after three years, and 18 after five years (4,8,9). Two fatal events occurred during the OLE. One study participant died of coronary arteriosclerosis during the first OLE year (8). Due to severe preexisting cardiovascular risk factors and consumption of alcohol and cardiac stimulants, this event was not considered related to the study drug (8). Another patient with a preexisting lupus erythematosus died of unknown causes after five years and one month of erenumab treatment (4). The incidence of cardiovascular events was well in the range of the background incidence in the general population (9). Almost 10% of patients developed anti-drug antibodies (ADA) during the OLE. ADA had no impact on safety or efficacy outcomes (4).
One further long-term analysis for erenumab in patients with EM was the OLE of the LIBERTY trial (10). LIBERTY focused on a difficult-to-treat population with 2–4 prior treatment failures (10). During the 12-week DBTP, 246 patients were randomized to receive placebo or erenumab 140 mg every four weeks (10). The great majority of patients (97.6%) continued trial participation in a three-year OLE, in which they received erenumab 140 mg monthly (11). After two years, 76.4% of patients were still in the trial (12). Half of patients (50.0%) who switched from placebo to erenumab reached a ≥50% reduction of MMD from baseline to weeks 61–64. The 50% responder rate in the group who continued with erenumab was 44.3% after one year (11). In the population who continued the trial over 112 weeks, the ≥50% responder rate increased to 57.2% during the last four weeks (12). MMD were reduced by −3.7 in weeks 61–64 and by −4.2 in weeks 109–112 compared to baseline (11,12). This also reflected in functional improvement and diminished impact of headache on everyday life (11).
AEs occurred at an exposure-adjusted incidence rate of 198.0/100 patient-years over the OLE, which was lower than in the placebo group of the DBTP (377.9/100 patients-years) (12). Patients experienced nasopharyngitis as the most common AE, followed by influenza, and back pain (11,12). AE incidence rates remained stable during the whole OLE. Four patients interrupted treatment due to AEs in the first year of OLE, nine more during the second year. Cardiovascular events were rare (4.2%) and also similar to the placebo arm of the parent trial (12).
Another OLE trial in adult patients with EM treated with erenumab was performed specifically in Japanese patients (13). After the parent study, there was an option to enroll in a 76-week OLE. In the OLE, 459 patients were initially administered erenumab 70 mg, which was increased to 140 mg (13). Over 90% of patients completed the study (13). MMD decreased from 7.9 at baseline to 5.0 MMD at the end of the OLE (13). Similar to the previously reported OLE, AEs occurred at a lower exposure-adjusted rate than in the placebo group of the DBTP. No fatal AE occurred and only six patients discontinued treatment due to AE (13).
Long-term data for erenumab in patients with chronic migraine (CM) is available for the 52-week OLE of a 12-week Phase II, DB, placebo-controlled trial (14). After the DBTP, 609 patients continued the study in the OLE and 470 patients (77.2%) completed OL treatment with 70 mg erenumab monthly, which was increased to 140 mg (15). MMD decreased from 18.1 at baseline to 8.8 at week 52 (15). The proportion of patients reaching a ≥50% reduction of MMD during weeks 49-52 was 59% (15). Over 70% of patients reverted from a chronic to an episodic migraine at the end of the OLE (16). A post-hoc analysis revealed a greater benefit for patients with erenumab 140 mg than those with 70 mg (−8.5 days for the 70 mg dose vs. −10.5 days for the 140 mg dose at week 52).
AEs occurred less frequently than in the DBTP (126.3/100 patient-years in the OLE vs. 202.0 in the placebo group of the DBTP). Similar to the previously mentioned trials, the most frequent AEs were upper respiratory tract infections, sinusitis, and arthralgia (15). Treatment discontinuation due to AEs occurred in 16 patients. ADA were detected in 6% of patients and had no effect on safety or efficacy outcomes (15).
In all mentioned studies, no clinically relevant changes in vital signs or laboratory parameters were observed. Pooled secondary analyses of the OL data confirmed the long-term tolerability of erenumab with a lower AE incidence than in the DBTP and no increases in AE rates over time (17). In particular, no cardiovascular safety concern emerged (18). The safety profile was similar in patients with migraine with and without aura (19).
Galcanezumab
Galcanezumab had an independent OL study that was completed in parallel to its DB placebo-controlled trials (20). The OL trial was completed by 210 patients (77.8%). All patients in this trial received a loading dose of 240 mg. One group continued with 240 mg monthly while the other group received 120 mg monthly after the loading dose. The overall change in MMD over 12 months was −5.6 for the 120 mg-group and −6.5 for the 240 mg-group (20). Most of the reduction occurred in the first month and then stabilized during the further trial period. The proportion of patients who reached a ≥50% response across months 1 to 12 was 65.6% in the 120 mg-group and 73.7% in the 240 mg-group. Most patients (80–85%) reported that they felt much or very much better under galcanezumab treatment. Accordingly, patient-reported outcomes (PRO) showed a significant improvement in functionality and health-related quality of life over the observation period (21).
The discontinuation rate due to AEs was 5% for both groups, most commonly due to local reactions at the injections site (20). Injection site pain was also the most common AE overall, followed by nasopharyngitis, and upper respiratory tract infections. No severe AE was considered related to galcanezumab treatment (20). Four patients had a transient elevation of liver enzymes, which did not lead to treatment discontinuation (20). Elevated systolic or diastolic blood pressure was reported in nine and 22 patients, respectively. However, most cases were only transient and no one persisted for more than two consecutive visits (20). Approximately 10% of patients developed ADA, mostly with very low titers. There was no apparent effect of ADA on safety outcomes (20).
Besides this purely OL trial, further long-term evidence derives from OLE of three of the DB placebo-controlled studies. REGAIN, a Phase III study of galcanezumab in the prevention of chronic migraine had an OLE period of nine months (22). Of the 1,022 patients who continued in the OLE, 80.7% completed the trial (22). The OLE started with a 240 mg dose and continued with 120 or 240 mg galcanezumab monthly (22). During the OLE, a further improvement in MMD was observed with a reduction of −8.0/−9.0 MMD compared to baseline in the last OLE month (22). The 50% responder rate at month 12 was 53.3–56.9% (22). The AE profile remained similar to the DBTP. The most frequent AEs were nasopharyngitis, upper respiratory tract infection and injection-site reaction (22). Only 4.5% of patients withdrew from the OLE due to AEs (23). The positive long-term findings reflected in a reduced healthcare resource utilization with less healthcare professional or emergency room visits, admissions to hospital and overnight hospital stays (24).
The second Phase III study with an OLE is CONQUER, which enrolled patients with 2–4 prior preventive treatment failures (25). CONQUER began with a three-month DBTP and continued with a three-month OLE. In the DBTP, 462 patients with episodic or chronic migraine were randomly assigned to placebo or galcanezumab 120 mg (with a loading dose of 240 mg) (25). Of the 451 patients who completed the DBTP, all but two patients entered the OLE and 96% completed the study (26). The previous placebo group started the OLE with a blinded loading dose of 240 mg and reported a reduction of −3.9 MMD in the first OLE month (26). Both groups reported thereafter similar reductions with 120 mg galcanezumab monthly (26). At the end of the OLE, the mean MMD reduction from baseline was −5.2 for the previous placebo group and −5.6 for the previous galcanezumab group (26). More than half of patients (53.1–53.6%) reached a ≥50% reduction of MMD during the last OLE month. PRO measures showed a decreased disability and an improved quality of life during the trial (26). AEs occurred in similar rates as in the DBTP with nasopharyngitis and injection site pain as the most frequent ones (26). AEs led to treatment discontinuation in five patients, three of them had injection site reactions. There were no treatment-related serious AEs and no fatal AEs during the OLE. A subgroup analysis in elderly patients (65–75 years) did not reveal any safety concerns in this particular population (26).
The last available OL study enrolled only Japanese patients with episodic and chronic migraine (27). Patients with EM entered the 12-month OLE after a six-month DBTP (27). Patients with CM were directly enrolled in the OLE without a previous DBTP (27). All 311 patients received galcanezumab 120 or 240 mg monthly. Safety and tolerability analyses did not reveal any relevant difference between the Japanese population and the previously described trials (27). Treatment discontinuation due to AEs was reported for 19 patients, in three cases due to injections site reactions and two cases due to anxiety. Secondary efficacy analyses showed a significant reduction in MMD (−10.6–10.7 for patients with CM and –3.2–4.3 for patients with EM previously treated with placebo) (27).
Fremanezumab
OLE evidence for fremanezumab is from the pivotal Phase III studies (HALO), the FOCUS study, and an OL Japanese study.
HALO consisted of a 12-week placebo-controlled DBTP for patients with episodic and chronic migraine, followed by a 12-month long-term safety study. The long-term study also allowed the enrollment of new patients who did not participate in the DBTP (28). During the OL trial, patients received either 225 mg fremanezumab monthly or 675 mg fremanezumab quarterly with placebo injections at intervening months (28). Patients coming from the fremanezumab groups of the DBTP continued with their previous treatment regimen, while patients previously on placebo and new patients were randomized 1:1 to the monthly or quarterly group (28). 1,493 completed the 12-month trial. AEs occurred in 84-89% of patients. The most frequent ones were injection-site reactions, followed by upper-respiratory tract infection and nasopharyngitis. AEs led to treatment discontinuation in 4-5% of patients. These were in most cases injection site reactions (28). This event was not considered related to the study drug (28).
The most common cardiovascular AE was hypertension, occurring in 2% of patients. High blood pressure was mostly transient and observed in patients with a known history of hypertension (28). There were 15 serious cardiovascular AEs during the study, including two strokes. One fatal AE was reported: A 44-year-old patient died from a ruptured brain aneurysm 10 months after the last fremanezumab injection. Both the sponsor and the investigators did not consider any serious cardiovascular AE to be related to the study drug (28). ADA occurred in 2.3% of patients without any safety consequence (28).
As for efficacy, MMD were reduced by −7.2 days, −8.0 days, −5.2 days, and −5.1 days for the CM quarterly, CM monthly, EM quarterly, and EM monthly group, respectively, at month 12 (28). An average monthly MMD reduction ≥50% was reached by 53-57% of patients with CM and 66-68% of patients with EM across the OLE (26). Accordingly, headache-related disability was significantly reduced (28). Reductions of MMD were stable in all analyzed weeks without any wearing-off effects at the end of the dosing interval (29).
Data from the FOCUS trial confirmed the positive long-term profile of fremanezumab (30). FOCUS enrolled patients who previously failed 2–4 classes of migraine preventive medication (30). The trial consisted of a 12-week placebo-controlled DBTP followed by a 12-week OLE. In the DBTP, 838 patients were administered either placebo, fremanezumab 225 mg monthly or fremanezumab 375 mg quarterly. Overall, 807 patients finished the DBTP and entered the OLE, which was completed by 772 patients (31). During the OLE, all patients received 225 mg fremanezumab monthly. Compared to baseline, MMD were reduced by −4.7-5.5 at the end of the OLE (31). The 50% responder rates at week 24 ranged between 38% for the previous placebo group and 46% for the previous fremanezumab monthly group. The top-three AEs were nasopharyngitis, injection site erythema, and injection site induration. Only seven patients withdrew from the trial due to AEs (31).
Finally, one OL trial was conducted specifically in a Japanese population: 50 patients with episodic or chronic migraine received fremanezumab 225 mg monthly or 375 mg quarterly for 52 weeks (32). The trial was completed by 44 patients. AEs were the reason for treatment discontinuation in only two patients. Nasopharyngitis was the most common AE also in this study, followed by injection site reactions (32). In the monthly group, MMD were reduced by -2.9 at month 6 and -5.9 at month 12. In the quarterly treatment, the respective changes were -4.7 and -1.6. Headache-related PROs showed a reduced disability and headache impact in both groups (32).
Eptinezumab
The PREVAIL study is the only available study assessing the long-term safety of eptinezumab (33). PREVAIL was a Phase III, OL trial comprising a 48-week treatment phase with eptinezumab 300 mg I.V. every 12 weeks and a 20-week safety follow up. The study population consisted of 128 adults with CM, 78% completed the study protocol (33).
Treatment-emergent AE were experienced by 91 patients (71.1%). The most common AEs included nasopharyngitis, upper respiratory infection, and sinusitis. After I.V. administration, five patients experienced hypersensitivity reactions and one patient suffered an allergic reaction that required intravenous treatment. 10 patients interrupted at least one infusion due to AEs, most commonly infusion-site extravasation. Eight patients withdrew from the study due to AEs, three of them had a hypersensitivity reaction. Overall, 23 patients (18%) developed ADA, all reversed through continued administration of the drug. There was no association between the occurrence of ADA and safety outcomes (33).
In the PREVAIL study, long-term efficacy was only evaluated through PROs. Improvements in health-related quality of life, headache impact, and patients’ global impression were observed after the first eptinezumab administration and remained stable until the end of the trial (33).
Atogepant
Atogepant’s 40-week OL study was conducted alongside its Phase III EM trial (ADVANCE). The OLE was completed by 511 of 685 patients (74.6%). Safety results are currently only available in abstract form (34). AEs occurred in 62.5% of study participants, most commonly upper respiratory tract infections and urinary tract infections (34). Serious AEs were reported by 3.4% of participants. No serious AE was considered related to the study drug (34).
Discussion
In the present review, we summarized the key findings derived from OLE trials of drugs targeting the CGRP pathway, namely monoclonal antibodies (mAbs) and gepants.
Regarding mAbs, OLE studies confirmed a favorable safety and tolerability profile, which was comparable to the corresponding DBTP. The overall rate of patients reporting at least one AE adverse event varied between 43% and 93% among studies. Consistently, the most reported AEs were nasopharyngitis, upper respiratory tract infections, and injection-site reactions (4,8,9,22). The absolute number of SAEs was low, and none of them was considered related to treatment.
Limited OLE safety data are available for atogepant (34). AEs occurred in 62.5% of participants; of these, the most commonly reported were upper respiratory and urinary tract infections.
Since the development of CGRP-targeted therapies, there have been various safety concerns. The cardiovascular concerns are justified by the role of CGRP in vasodilatation and the large expression of its receptor in the vascular system and, as such, patients with major vascular events were excluded from the parent RCTs (35). Moreover, the role of CGRP in inducing gastrointestinal motility disorders has been studied in animal models (36) and gastrointestinal complaints, especially constipation, were among the most frequently reported AEs in RCTs and post-marketing studies (37). Indeed, constipation was reported in a significant higher percentage of patients taking erenumab when comparing to placebo (17). In the OLE trials, all of the study drugs showed good cardiovascular safety and tolerability. The incidence of cardiovascular events did not appear to be higher across all study populations as compared to the background incidence in the general population (9,11,12,18). Constipation occurred in no more than 3% of patients enrolled in OLE studies, with an apparently lower rate when indirectly compared with the parent DBTP, and generally did not lead to treatment discontinuation (12,17).
It is worth noting that the study population of OLE trials is made of patients that meet rigorous inclusion criteria. In this context, patients with significant medical conditions, especially major cardiovascular diseases and serious gastrointestinal pathologies were not included. This was also the case of female patients in their fertile age without secure contraception. In addition, all patients involved in OLE trials, by definition, did not previously experience AEs leading to a discontinuation of treatment during the DBTP, while patients who reported tolerability issues were less likely included in the OLE studies. This of course might have reduced the occurrence of AEs.
Data regarding sustained efficacy must be taken with caution because: i) the OL design does not allow for a direct placebo comparison, ii) participants with a poor response are more likely to drop out of the study, and therefore study completers are likely to represent only responders, iii) at the end of the DBTP, the clinical features of the participants may not be comparable to the original inclusion/exclusion criteria. With these limitations in mind, OLE trials suggested a long-term efficacy for EM as well as CM, with a pattern suggestive for a further reduction of MMD at the end of the OL follow-up when compared to the end of the DBTP. Efficacy data from OL studies are not available for eptinezumab, as the PREVAIL trials only reported PROs measures (33).
Regarding EM, MMD reduction was between 2.9 and 5.2 after an observation period varying from 56 to 76 weeks. Noteworthy, a MMD reduction of 5.3 was confirmed for up to five years with erenumab treatment (4). The 50% responder rate was between 44.5% and 71%. In CM, after 40–52 weeks of treatment, MMD were reduced from 7.2 to 9.3, with a 50% responder rate between 53% and 59%. Of note, response rates were calculated differently across trials: While the erenumab trials based response calculation on the last four weeks vs. baseline, trials with galcanezumab and fremanezumab often considered the monthly average reduction across the whole trial period. Four studies reporting OLE data with galcanezumab and fremanezumab reported cumulative data from EM and CM patients, thus limiting the interpretability of the results and generalizability to the clinical practice (22,26,28,32).
The discontinuation rate is a crucial point in OLE studies; indeed, it may represent a major bias for the observed findings. In these OLE studies, the discontinuation rate varies depending on the duration of the follow-up: it was below 5% in studies with a 12-week observation, 7–26% for studies with a follow-up of maximum 76 weeks, and 25–44% for trials with a follow-up of three to five years. Noteworthy, the discontinuation rate due to AEs varies between less than 1% and 12% (12,13). Other reasons for discontinuing the trial included withdrawal of consent, lack of efficacy, protocol deviations, lost to follow-up, noncompliance with study procedures.
This said, the treatment adherence appears to be satisfactory when compared to oral preventive drugs; indeed, previous studies on propranolol, amitriptyline, and topiramate suggested a discontinuation rate between 21% and 81% at six months and between 45% and 93% at 12 months (38). A recent RCT compared erenumab and topiramate, one of the oral preventive drugs with the strongest evidence of efficacy (39). In this trial (HER-MES), 777 patients with EM or CM were randomized to either topiramate, titrated up to 100 mg/day, or erenumab (70 or 140 mg) (40). The proportion of patients discontinued due to AEs was strikingly more marked in the topiramate (38.9%) that in the erenumab (10.6%) group (40). Similarly, treatment-related AEs were more common in the patients treated with topiramate arm (81.2% vs 55.4% of patients with erenumab). The erenumab group also showed a higher percentage of 50% responders (55.4% against 31.2% in the topiramate arm) (40)
In recent years, real-world studies provided a growing body of data on the use of CGRP-targeted therapies in the clinical practice. These studies are useful to evaluate drugs “effectiveness”, defined as a measure of the “extent to which an intervention does what is intended to do in routine circumstances” (41).
Real-world studies have the advantage to include a broader spectrum of patients, who are more representative of the daily clinical practice. These studies however bear important limitations: i) the open-label, non-randomized, uncontrolled design, ii) the need to cope with the national/regional regulations, and iii) the high heterogeneity of procedures among studies.
Real-world studies including EM patients suggested a reduction in MMD between 2.9 to 6.4; the 50% responder rate ranged between 35-76.5% (42–44).
In CM, real-world studies proved a reduction in MMD from 8.2 to 13.9, while the 50% responder rate was between 34 and 75% (42–45). Noteworthy, erenumab proved effective in refractory CM patients as well (46). The difference among studies is primary ascribable to different study procedures and duration of follow-up, and different baseline features of patients enrolled, in particular the number of MMD and of previous preventive drugs failures.
In the real-world setting, at least one AE was reported by 19% to 48% of patients, with constipation, nausea, flu-like symptoms, and site injection reaction being the most prevalent (>2%) and consistently reported (42–45).
Altogether, the evidence suggests that treatments targeting the CGRP-pathway are characterized by good effectiveness, safety and tolerability in real-world setting, comparable or even slightly better when compared to previous RCTs or OLE trials. This observation is relevant when considering that several real-world studies included patients with difficult-to-treat migraine phenotypes and more comorbidities, as compared with the population of RCTs.
It is also true that OLE and real-world trials are not generally designed and powered to specifically evaluate safety, and rare or unexpected AEs may be missed out (3). Thus, a third step approach, namely post-marketing surveillance and spontaneous AEs reports, is crucial for long-term monitoring. As the most important example, an analysis of the FDA Adverse Event Reporting System revealed an association between erenumab treatment and elevated blood pressure. Consequently, hypertension was included in the warnings paragraph of the erenumab prescribing information (47).
Although well established in the clinical research setting, the role of OLE is still under scrutiny for several considerations. The lack of a control group and blinding to treatment limits the usefulness of these trials to assess efficacy. Moreover, their utility in addressing long-term safety and tolerability concerns has been questioned by several authors (3,48). As a matter of fact, given the limited study population, tardive serious adverse events would be noted only if there was a large increase in their incidence. The roll-over of patients from the DBTP to the OL phase generally selects a population which already tolerated the study drug, which may also partially differ from the originally intended study population. Moreover, the naive patients switched from placebo to study drug during the OLE will generally take part in a study phase with less strict follow-up timepoints. Unblinding the patient at the end of the DBTP has been proposed as a partial solution to compensate these ethical issues. In such cases, patients would have a valid consent and thus be able to choose competently (49). Finally, we should bear in mind that the CGRP(-receptor) mAbs are not available on the market in many countries or reimbursed only under very strict conditions. This results in an ethical dilemma, as the study participants are deprived of an effective medication at the end of the OL trials.
Conclusions
In conclusion, although findings from OLEs trials should be carefully evaluated, there are several key findings worth consideration: i) monoclonal antibodies targeting the CGRP pathway showed a satisfactory safety profile, ii) the rate of drop-outs was generally low, iii) patients switched from placebo to active drug quickly reported a benefit in terms of MMD reduction similar to the active arm of DBTP, and iv) the mean reduction of MMD was generally sustained until the last follow-up, at least in the subset of patients who completed the observation period.
Taken together, OL and OLE trials support the safety and tolerability, clinical improvement, and the satisfactory and the good patients’ compliance of CGRP-targeted mAbs. OL and OLE trials on gepants are needed to confirm their safety and clinical impact.
Article highlights
Open-label studies with CGRP(-receptor) monoclonal antibodies and CGRP-receptor antagonists confirm a favorable tolerability profile of these substances over time. Dropout rates were generally low, indicating a good therapeutic adherence. Treatment response was sustained for up to five years, along with an improvement in functionality and quality of life.
Footnotes
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: BR reports research grants from Novartis, and personal fees from Abbvie/Allergan, Hormosan, Lilly, Novartis, and Teva. RDI received speaker honoraria from Eli Lilly and TEVA. MC has nothing to declare. MT reports personal fees from TEVA. JA reports consulting honoraria from Abbvie, Amgen, Aeon (Data monitoring board), Axsome, Biohaven, BioDeliveryScientificInternational, Eli-Lilly, GlaxoSmithKline, Lundbeck, Impel, Neurolief, Neso, Satsuma, Theranica, Teva. JA received institutional grants from Abbvie, Biohaven, Eli-Lilly, Satsuma, and Zosano.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
