Emerging literature on patients with migraine treated with erenumab and other CGRP monoclonal antibodies outside of clinical trials has demonstrated improved outcomes for a significant proportion of cases [17,18,19,20]. We audited 17–30 months of PROM/QoL data in our cohort of treatment/medically resistant or difficult to treat CM patients to give us insight into the more medium to long-term benefits of anti-CGRP intervention in this type of patient population. During the first year, 68/177 (38.4%) patients stopped treatment mainly due to lack of efficacy (marginal or no improvement) and less frequently due to lack of efficacy and/or side effects. The reported side effects in our patients were generally mild, with the most common being injection site reactions, abdominal bloating and constipation. Two patients stopped erenumab after the first dose of 70 mg, both due to potentially more severe side effects. After 1 year, 109/177 (persistence of 61.6%) of patients continued treatment. By June 2021, 97/177 (persistence of 54.8%) had sustained clinical improvement and had remained on erenumab for 17–30 months (with 54 of these remaining on treatment for at least 25–30 months). This is a very significant proportion of patients showing improvement considering that all patients in this report had failed at least three conventional migraine preventive agents (typically these patients had cycled through 4–8 failed preventive treatments).

Persistence with preventive treatments in migraine in the medium to long term (1–3 years) is a good combined measure of efficacy and side effects. In a recent report from Canada, the persistence of erenumab in the medium term in more than 14,000 pooled EM and CM patients was approximately 71% at 1 year and 63% at 18 months [20]. This is slightly higher than in our patients (persistence of 61.6% at 1 year). This may be due in part to the treatment resistant nature of our patients. In addition, all of our patients had CM and the above study had a combination of EM and CM (with the assumption that EM patients are less treatment resistant). In contrast, it should be noted that the persistence of conventional oral preventive treatments for migraine (such as B-blockers, tricyclics, and anti-convulsant medications) is in the region of 20–30% after 1 year [21]. Given the data from numerous clinical trials, the more recent real-world studies referred to above and our data, it would appear that erenumab and the other CGRP monoclonal antibodies (fremanezumab, galcanezumab and eptinezumab) have superiority over conventional preventive treatments with better efficacy and less side-effects. However, it is early days as few direct head-to-head trials have been done. In this regard, the HERMES trial was the first randomised controlled trial comparing a conventional oral preventive agent (topiramate) to a CGRP antagonist (erenumab) in 777 migraine patients [22]. In support of the argument that CGRP antagonists are superior to conventional migraine preventive treatment, the investigators found that some 10% discontinued erenumab mainly due to adverse events compared to almost 40% in the topiramate group. In addition, significantly more patients in the erenumab group had a ≥ 50% reduction in monthly migraine days when compared with the topiramate group [22].

Erenumab has previously been shown to be beneficial for both treatment resistant EM and CM in short duration clinical trials. For example, the LIBERTY trial evaluated the efficacy of erenumab in patients with episodic migraine who failed 2–4 other prophylactic agents [23]. After 12 weeks, approximately 30% of those receiving erenumab 140 mg had achieved a ≥ 50% reduction in migraine days, compared to 13.7% for the placebo group [8, 9]. Long term outcomes from the LIBERTY study over a 64-week period highlighted persistent efficacy and safety of erenumab. In an open label extension study, sustained efficacy and safety with erenumab over a five-year period has been demonstrated [24]. Given our findings and those of others more recently, further studies looking at longer term outcomes are needed to evaluate more meaningful clinical response in a real-world setting. However, in this report, we provide real-world data for 17–30 months of treatment with erenumab which has not been previously documented in such a group of difficult to treat or treatment resistant CM patients.

In our evaluation, we included the HIT-6, MIDAS and MSQ questionnaires. These Patient Reported Outcome Measures (PROM) are frequently used and internationally accepted instruments for this cohort of patients. For example, a subgroup analysis of a double-blind trial of those with CM and medication overuse headache (MOH) comparing erenumab 70 mg or 140 mg to placebo, utilised MIDAS, HIT-6 and MSQ [11]. In this study, erenumab demonstrated improvement in PROM/QoL and disability using these tools. These findings are consistent with other recently published real-world evidence [18, 19].

A clinically meaningful or minimally important difference on the HIT-6 was considered to be ≥6-point improvement [25]. Clinically meaningful or minimally important difference for the MIDAS and MSQ questionnaires are not as clear cut and there is currently no expert consensus (although individual groups have suggested criteria). However, our data demonstrated a similar reduction in MIDAS and MSQ scores from baseline that mirrors the HIT-6 data. The decision to continue treatment was based on a subjective clinical improvement of at least 30% (as reported by the patient): this decision was reinforced by headache diaries and PROM/QoL questionnaires completed by our patients. Approximately one fifth of patients were deemed to be ‘super-responders’, as they had an overall clinical improvement in the region of 80–90% or more.

There are currently no biomarkers or diagnostic tests available to assess whether a particular migraine treatment is helpful for patients, and therefore QoL questionnaires, diaries and patient consultations are currently the only reliable ways to assess for meaningful clinical improvement. There are a number of known limitations when using QoL questionnaires or PROM in CM. One factor is that none of the QoL questionnaires used in our analysis have been specifically designed for patients with CM. They are “migraine specific questionnaires” and the majority of patients with migraine have EM. This needs to be taken into account as significant differences exist between groups of EM and CM patients. For example, patients with CM have a significantly higher level of disability than patients with EM [26]. Another stumbling block is that the MSQ questionnaire was designed to assess QoL attributable to migraine in patients who were in a non-migrainous state [15]. This is certainly not the case for our cohort of patients as, by definition, CM patients have headache and migraine symptoms on more days per month than not. In fact, many of our patients had daily symptoms and few or no crystal-clear days. In support of our methodology, there are several studies in the last ten years that have demonstrated a reasonable level of legitimacy of use of these QoL questionnaires in CM. For example, Rendas-Baum, Bloudek [27] demonstrated the validity of MSQ in assessing QoL outcomes in CM patients. This study confirmed further credibility by demonstrating strong associations with HIT-6. The recently published erenumab trials included the newly developed Migraine Physical Function Impact Diary (MPFID) scale, a PROM which is designed to assesses the impact of migraine physical function over the previous 24 hours and 7 days [26]. We did not use this questionnaire in our study as it is not currently widely accepted or routinely used in clinical practice. It may have been useful to compare our patient group to studies using this parameter [28]. The most important clinical factors for patients relating to treatment include reducing disability, improving quality of life, increasing workplace productivity, reducing presenteeism, and being able to function in their daily activities. In this regard, PROM/QoL questionnaires can be used as an additional resource to improve our understanding of the impact of certain therapies for our patients with greater disability and to monitor evolution on treatment over time [25].

It is generally accepted that CM cohorts as a subgroup represent a complex and heterogenous group of patients. There are various factors which predispose patients to evolve from EM to CM. Risk factors for chronicity include: age, female sex, obesity, depression, low educational status and medication overuse [7, 29]. There is also clinical evidence that genetic factors make some individuals prone to developing more chronic forms of migraine. From a treatment perspective, there is preliminary data that individual responses to anti-CGRP therapies in migraine may be dependent on epigenetics [30]. From our small cohort and emerging evidence in the literature, it is clear that there is significant individual variation in terms of response to CGRP blocking treatment. Interestingly, certain patients do not experience any improvement whatsoever, while others are ‘super-responders’ and achieve a near complete resolution of their migraine headache and associated neurological symptoms.

All CM patients in this study had failed at least three conventional migraine preventive treatments prior to starting erenumab and are therefore considered to be medically resistant, or difficult to treat migraineurs. This more resistant group is of particular interest for us as anti-CGRP therapies have recently been approved for reimbursement in the Irish healthcare system specifically for those who have failed at least three conventional options. This is consistent with several other healthcare systems internationally.

Many of our 177 erenumab patients opted to remain on various additional migraine preventive therapies for all or part of the 17–30 month treatment period (including onabotulinumtoxin A injections or oral prophylaxis), perhaps partly because they were having some perceived additional benefit with these non-CGRP therapies. We have therefore wondered if there may be some clinical synergy with respect to these different treatments. In a recent review in Headache, it has been proposed that there may be a pathophysiological rational to dual prescribing of onabotulinumtoxinA and anti-CGRP agents, as there is a possibility of clinical synergism [31]. However, this symbiotic relationship has not yet been confirmed at the time of writing, and there is currently only preclinical evidence to support a rational for dual therapy. It may be worthwhile in future studies to address whether patients being treated with additional preventive treatments had incremental improvement in QoL outcomes compared to those on CGRP treatment alone.

The disability associated with migraine should not be underestimated. When comparing migraine to other neurological disorders, it is the second cause of neurological impairment worldwide next to stroke [32]. In a disease burden study of over 11,000 participants for whom preventive treatments have failed, 52% reported impairment in both overall work productivity (absenteeism and presenteeism combined) and daily activities due to migraine [4]. In the same study, 64% of respondents reported that migraine had affected their private life, including relationships with friends, relatives, and partners. It is well documented that loss in work productivity and activity impairment due to migraine is higher in individuals with previously failed preventive treatments, further establishing the need for more effective treatment options in this subpopulation to ensure that they can fully contribute to the workforce and to society [4].