Over the past few years, monoclonal anti-Calcitonin Gene-Related Peptide-(Receptor-)monoclonal Antibodies (anti-CGRP(R)-mAbs) have emerged as a promising prophylactic treatment for migraine (1–3). These antibodies bind to either the CGRP ligand or its receptor, thereby blocking its actions. While the efficacy of anti-CGRP(R)-mAbs in preventing migraines has been established, it is crucial to consider the potential physiological implications associated with their prolonged usage. This necessity arises from the multifaceted role of CGRP in various biological processes within the human body (4).

CGRP, a potent systemic vasodilator capable of reducing blood pressure (BP), and exerting both chronotropic and inotropic effects on the heart, plays a critical role in maintaining the cerebro-and cardiovascular (CV) homeostasis (5). In theory, anti-CGRP(R)-mAbs may diminish the protective effect of CGRP on CV-infarct size and increase the risk of heart failure (5–7). Notably, migraine itself is an independent CV risk factor, further emphasizing the need for a comprehensive understanding of the CV implications of anti-CGRP(R)-mAbs (8–10). Moreover, as CGRP can be found in the digestive tract, lungs, kidney, liver and adipose tissue as well, other regulatory systems may be affected by the neuropeptide, although less is known about the exact sites of action and underlying mechanisms (4, 11).

In vitro studies demonstrated that erenumab inhibited CGRP-mediated vasodilation, while not interacting with other vasoactive compounds (12). Furthermore, work in mice demonstrated that treatment with a CGRP receptor antagonist worsened cerebral ischemic outcomes (6), and rat studies demonstrated that cardiac ischemic outcome is worsened by the blocking of the CGRP receptor (7). Thus, these animal studies indicate an important role of CGRP in preserving tissue during ischemic conditions. A handful of studies have evaluated the effect of erenumab in vivo and yielded mixed results, with one study that did not find any alterations in vasomotor reactivity or flow-mediated dilation in patients with migraine, while another demonstrated that it did affect trigeminovascular reactivity by a decrease in capsaicin-induced dermal blood flow (13, 14).

Recent studies, including a meta-analysis of 19 randomized controlled trials and the analysis of the US FDA Adverse Event Reporting System (FAERS) database, aimed to evaluate the safety and tolerability of monoclonal antibodies and gepants targeting the CGRP pathway (15, 16). While these RCTs did not reveal differences in serious adverse events between active treatments and placebo, they were limited by their short-term nature of 3 to 6 months, their focus on solely SAEs, or lack of investigation into laboratory or ECG findings (16). Furthermore, the FAERS database, while reporting also low frequency SAEs, has limitations due to a short 6-month follow-up period and reliance on only self-reported adverse events, which likely leads to underreporting (15).

Earlier, a case report highlighted a CV-event that occurred 5 months after starting erenumab (17). Furthermore, during an open-label extension study a CV related death was reported (18). These CV-related AEs are particularly noteworthy since clinical trials typically focus on enrolling patients with migraine who need to be generally in good health. Therefore, other reviews have emphasized the importance of incorporating real-world data and post-marketing surveillance studies to validate and expand upon these trial results (19).

Of particular concern are post-marketing case reports of elevated BP associated with erenumab, raising questions about CV safety (20). An independent study found an average increase of 5.2 mm Hg in systolic BP and 3.5 mm Hg in diastolic BP after starting anti-CGRP(R)-mAbs erenumab or fremanezumab in patients with migraine (21). Studies with clinical data collected from independent researchers and with longer follow-up time seem important to obtain a more comprehensive insight and understanding of individual responses and side-effects.

In this observational cohort study we assessed the safety of erenumab and fremanezumab regarding CV safety by assessing CV events and ECGs during a period of at least 1 year of treatment with anti-CGRP(R)-mAbs in a real world setting.

All patients with migraine who received treatment with either erenumab or fremanezumab at the Leiden Headache Center were considered eligible for inclusion. Exclusion criteria were established based on elevated risks of cardiovascular disease (CVD), and included hypertension at baseline, a medical history of hypertension and a prior CV event. At the time of this study we defined hypertension as a systolic BP ≥140 mmHg and/or a diastolic BP ≥90 mmHg in accordance with the 2018 ESC/ESH guidelines (22). In addition, we excluded patients that were previously treated with an anti-CGRP(R)-mAb and thus had no distinct baseline period. After one baseline month patients started treatment with either erenumab or fremanezumab. Due to the restricted availability of anti-CGRP(R)-mAbs in the Netherlands at the time of inclusion, all patients had to have at least 6 monthly migraine days (MMD) and failed on ≥4 migraine prophylactic treatments, including at least candesartan, beta-blockers, valproate and topiramate. Treatment failure was defined as ineffective treatment, discontinuation because of side effects or ineligibility because of contra-indications. Migraine diagnosis was made by a neurology resident in consultation with a neurologist with headache expertise, or by neurologists themselves, based on the International Classification of Headache Disorders (ICHD)-3 criteria (23). Patients with a second headache diagnosis other than tension-type headache were excluded. Approval for this study was obtained from the LUMC Medical Ethical Committee who declared no ethical concerns.

Over a span of 3 years, from mid-November 2018 to mid-November 2021, a comprehensive assessment was carried out to establish study eligibility, involving a total of 270 patients. The study’s inclusion process is visually presented in Figure 1. Among the patients considered for participation, a total of n = 77 were excluded. Primary reasons for exclusion were previous treatment with anti-CGRP(R)-mAbs (n = 43) or an increased CV risk due to hypertension at baseline (n = 30), a history of hypertension (n = 3) or a history of CV events (n = 1).

A total of 193 patients with migraine started treatment with either erenumab (n = 101) or fremanezumab (n = 92). Among the patients treated with erenumab, 82/101 (81%) were female, the average age was 43 years and 43/101 (43%) patients had migraine with aura. Among the patients treated with fremanezumab, 76/92 (83%) were women, the average age was 43 years and 32/92 (35%) patients had migraine with aura. None of the patients had hypertension or were smoking at baseline. Baseline characteristics for both treatment groups are presented in Table 1.

This study investigated the long-term safety of anti-CGRP(R)-mAbs as migraine treatment. Among all patients, 3.11% developed abnormal ECG or CV adverse events during treatment with erenumab (n = 3) or fremanezumab (n = 3). Within this group, 1.55% developed moderate to severe CV adverse events that led to treatment discontinuation. These adverse events included cerebellar stroke, SCAD, and pericarditis. The remaining 1.55% developed non-threatening ECG abnormalities without physical complaints. It is noteworthy that these events occurred in patients with no prior hypertension and no prior CV complaints. No clinically meaningful changes were observed in liver, kidney function, lipid, electrolyte, or glucose levels over time.

Migraine itself is associated with an elevated risk of myocardial infarction (OR 2.2, 95% CI 1.7–2.8) and stroke (OR 1.5, 95% CI 1.2–2.1) (25). It is important to note that the absolute risk of CV disease in patients with migraine remains generally low, and is influenced by various factors, including migraine subtype, age, sex and the presence of additional risk factors. In a large population-based study spanning 11.9 years, 697 CV events consisting of both stroke or myocardial infarction were reported among 3,577 women with migraine, resulting in an annual incidence of 1.64% (26). Another population-based study among people with migraine reported cumulative incidences over 19 years for myocardial infarction and ischemic stroke separately. Specifically, for myocardial infarction, there were 25 cases per 1,000 people, which translates to an annual incidence of 0.13%. For ischemic stroke, the study found 45 cases per 1,000 people, resulting in an annual incidence of 0.24% (27).

We recognize the inherent complexity in directly comparing our cohort study with population-based studies. Our selection process may introduce bias by favoring patients with fewer CV risk factors or more severe migraine cases. Additionally, our study encompasses non-ischemic CV events, including pericarditis. Nonetheless, our findings align with these population-based studies. We identified one individual with an ischemic stroke within a 1.5-year follow-up period, resulting in an annual incidence rate of 0.34% for ischemic stroke. To our knowledge, there is no study investigating the association between pericarditis and migraine. Furthermore, although the lifetime prevalence of migraine seems to be elevated in individuals with SCAD compared to the general population (28), there are no studies investigating the incidence of SCAD specifically among patients with migraine. However, a history of migraine is correlated with an increased risk of cervical artery dissection, with an odds ratio of 1.74 (95% CI 1.38–2.19) (29). The precise biological mechanisms underlying this association remain elusive, although genetic analyses have hinted at the existence of common genetic factors influencing vascular structure and function in both migraine and cervical artery dissection (29).

While our findings suggest a comparable annual incidence of 0.34% for ischemic stroke among patients with migraine, it is essential to acknowledge that we cannot definitively exclude the possibility that treatment with anti-CGRP(R)-mAbs might influence the risk of all CV events. Notably, our cohort consisted of healthy individuals with no cardiovascular risk factors at baseline. In contrast, the aforementioned population-based study reporting a one-year incidence of 1.64% included approximately 10% of subjects who were smokers, some had a history of diabetes, and the mean age was approximately 10 years higher compared to our cohort (26). Additionally, the other population-based study included individuals with diabetes (1.37%), hypercholesterolemia (0.73%), hypertension (3.03%), and even other comorbidities such as cancer or liver/renal disease (27). Given these differences, further studies are warranted, particularly those that include control groups matched for CV risk factors and migraine severity. For now, it remains crucial for practitioners to be vigilant, as anti-CGRP(R)-mAbs may potentially increase the severity or impact of such events if they do occur (6). During myocardial infarction (MI) or stroke, there is a release of CGRP, which possesses vasodilatory properties (30, 31). Vasodilation improves the blood flow and oxygen delivery to affected tissue and therefore plays a protective role. Consequently, CGRP may mitigate the extent of tissue damage. Inhibiting CGRP, either through blocking its receptors or capturing the released CGRP before it reaches the receptor, has the potential to exacerbate the deleterious effects of MI or TIA/stroke (5).

The above is especially important since real-world data (RWD) studies revealed elevated BP after exposure to anti-CGRP(R)-mAbs, which was not previously observed in the pivotal randomized clinical trials (20, 21, 32). This discrepancy is most likely due to the dichotomization of the BP outcome variable used in the clinical trials, which focused solely on the occurrence of hypertension. This approach may mask the absolute impact of anti-CGRP(R)-mAbs on BP and potentially downplay the significance of the data (32). Treatment with anti-CGRP(R)-mAbs is associated with an average increase of 5.2 mmHg in systolic BP (21). While some clinicians may perceive this increase as mild, it is crucial to recognize the clinical significance. In fact, a mere 5 mmHg increase in systolic BP will raise the risk of non-fatal CV events by approximately 10% (33). Due to the possibility of many years of treatment, this relative risk increase is clinically important even in participants aged 55 years or younger, which is the age group most represented among patients with migraine (33). Hence, these seemingly modest elevations in BP associated with anti-CGRP(R)-mAbs warrant careful consideration and monitoring to ensure optimal CV health outcomes. Interestingly, patients that developed CV events did not have hypertension, indicating that this factor likely did not contribute to the incidence of CV adverse events. However, as previously emphasized in our earlier publication (21), monitoring of BP remains extremely important. According to the American College of Cardiology and the American Heart Association Task Force, recommended long-term BP targets are ≤130/80 mmHg in patients younger than 65 years and ≤140/90 mmHg in patients 65 years or older (34, 35). Following this stricter blood pressure target of ≤130/80 mmHg, there was one individual that developed pericarditis that had a baseline BP of 126/84 mmHg and would have been excluded from treatment under this guideline. However, all other cases of cardiovascular adverse events and the observed developments of ECG abnormalities in this study had baseline blood pressure readings below the ≤130/80 mmHg threshold. Additionally, it is worth noting that in total, 68 out of 193 patients (35%) in this study had a baseline blood pressure that was higher than 130/80 mmHg but still within the range of ≤140/90 mmHg. According to our current blood pressure guidelines, these patients would not have been eligible for anti-CGRP(R)-mAb treatment unless their blood pressure was actively managed.

A limitation of this study may be some missing data toward the end of the follow-up period, primarily caused by patients discontinuing the treatment. It is important to acknowledge that the occurrence of missing data was not random, as discontinuation may have resulted from adverse events and was therefore possibly associated with an outcome. Therefore, a complete case analysis was added, and the results demonstrated no changes in ASAT, ALAT, GGT, lipid panel, electrolytes, and glucose levels over time and the previously observed changes in ALP and creatinine serum levels were no longer present. In addition, we chose not to include a control group in our study, a decision justified by a single key reason. The inclusion of a control group would not have replicated the findings of pre-existing large population-based studies that already established annual incidence rates of CV events in patients with migraine (26, 27).

A major strength of this study is the utilization of RWD. RWD mirrors the precision, consistency, and verifiability inherent in prospective clinical trials. Unlike experimental studies, RWD-driven observational research dispenses with control groups or manipulated interventions. It relies solely on observed data, capturing patient health statuses and healthcare specifics from diverse sources. The associated Real-World Evidence (RWE) holds great importance, offering a contextual perspective for randomized trials, filling in their data gaps, and addressing broader questions regarding intervention impact in real-life settings. Even in the absence of control groups, RWD and RWE cast light upon healthcare practices, patient outcomes, and the real-world implications of medical interventions. Another strength is that this RWD study excels in incorporating a baseline period and establishing an initial health and CV risk profile. By contrasting treatment period safety outcomes against baseline data, we aimed to identify any notable deviations or adverse effects. The longitudinal design, with over 12 months of follow-up, extends the assessment over time. Nevertheless, we advocate for lengthier RWD studies with similar data collection on CV adverse events for a more comprehensive understanding. For now, we advise to establish the CV risk for each patient prior to the start of treatment with anti-CGRP(R)-mAbs by blood tests, BP measurement and an ECG. After the start of treatment, regular blood tests or ECG measurements may not be indicated unless the patient develops symptomatic complaints. Regular BP monitoring during treatment with CGRP monoclonal antibodies is advised (21).

In conclusion, we identified CV events in 1.55% of patients receiving anti-CGRP(R)-mAbs with 1.5-year follow-up, despite having no CV risk factors or hypertension at baseline. We advise awareness regarding CV events in patients with migraine undergoing CGRP-targeted treatment, not as a confirmation of increased risk but as a proactive measure to address potential multifactorial influences.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving humans were approved by nWMO-committee division 3 Leiden University Medical Center. The studies were conducted in accordance with the local legislation and institutional requirements. The ethics committee/institutional review board waived the requirement of written informed consent for participation from the participants or the participants’ legal guardians/next of kin because of the retrospective non-WMO design with data of patients who previously gave written informed consent through protocol P17.232 and the Biobank regulation of the LUMC.

BA: Conceptualization, Data curation, Formal analysis, Methodology, Project administration, Writing – original draft, Writing – review & editing. NV: Conceptualization, Formal analysis, Methodology, Writing – original draft, Writing – review & editing. JR: Data curation, Formal analysis, Writing – original draft, Writing – review & editing. MO: Writing – review & editing. AM: Conceptualization, Investigation, Methodology, Supervision, Writing – review & editing. GT: Conceptualization, Investigation, Methodology, Supervision, Writing – review & editing.