Patient Disposition

A total of 571 patients were enrolled at 105 sites; 15 patients did not meet at least one selection criterion and/or did not document a starting dose for erenumab and were excluded. Therefore, 556 patients were included in the FAS and SAF. Of the 556 patients, 99 (17.9%) completed the study over the course of 2 years. The majority of patients (64.4%; n = 293) did not complete the study because the study was terminated early by the sponsor.

Demographic and Baseline Characteristics

Demographics and baseline characteristics are summarized in Table 1. The mean (standard deviation; SD) age of the patients was 45.0 (12.3) years. The majority of the patients were female (89.0%; n = 495), Caucasians (97.6%; n = 521), and non-smokers (85.1%; n = 458). Migraine developed at a mean age of 25.3 (14.1) years before baseline; and time since migraine diagnosis was 17.0 (13.2) years before baseline. There were 63.3% (n = 352) of patients who had CM according to ICHD-3 criteria [9]. Almost all patients (99.3%; n = 552) were treated with at least one previous preventive therapy at baseline. The most common previous treatments were topiramate (78.2%), beta-blocker (75.5%; metoprolol [50.7%], propranolol [13.8%], bisoprolol [11.0%]) and amitriptyline (69.2%). A total of 97.5% of patients took acute medication in the 3 months preceding baseline. The most common acute medications were triptans (100%) and non-steroidal anti-inflammatory drugs (NSAIDs) (45.3%). Among triptans, the most common were sumatriptan (37.8%) and rizatriptan (27.5%). The most common NSAID used was ibuprofen (28.8%). Patients reported on average 3.7 (1.5) treatment failures (i.e., lack of efficacy of a prior preventive treatment). The disease burden of migraine was high in the study participants at baseline: patients experienced on average 14.6 (6.7) MHD and 10.8 (5.4) MMD, and on average patients took acute medication on 10.3 (5.5) days per month.

Table 1 Demographics and baseline characteristics (full analysis set)Primary OutcomesStarting Dose of Erenumab in the Overall Population

The starting dose of erenumab was 70 mg in 68.5% (n = 381) of patients and 140 mg in 31.5% (n = 175) of patients (Fig. 2).

Fig. 2

Starting dose of erenumab in overall population (FAS)

Starting Dose of Erenumab in the Subgroups of Patients

The proportion of patients with a starting dose of 70 mg or 140 mg erenumab did not differ by < 4 or ≥ 4 treatment failures (Table 2). The proportion of patients with 140 mg as starting dose was highest (43.5%; n = 40) in the patients aged 30 to ≤ 40 years versus other age groups. Erenumab 140 mg was less frequently used as the starting dose in patients with 4 to ≤ 7 MHD (22.6%; n = 14) or with ≥ 15 MHD and < 8 MMD (26.3%; n = 10) versus other subgroups. The proportion of patients with 140 mg as starting dose decreased with increasing time since migraine diagnosis (< 1 year: 44.0% [n = 11], 1 to ≤ 5 years: 37.5% [n = 39], ≥ 5 years: 29.3% [n = 125]). The proportion of patients with erenumab 70 mg or 140 mg as starting dose stratified by HIT-6 showed no meaningful difference between the subgroups. More than 69.5% (n = 214) of patients were prescribed a starting dose of erenumab 70 mg regardless of psychiatric, cardiac, autoimmune, and/or GI disorders. There was no difference observed in choice of starting dose between patients with or without psychiatric, cardiac, autoimmune and/or GI disorders. In all the subgroups, erenumab 70 mg was the commonly prescribed starting dose (Table 2).

Table 2 Starting dose of erenumab in subgroups of patients (full analysis set)Other OutcomesReason for Choice of Starting Dose

The most common reason for the choice of starting dose was “according to prescribing information” (90.3%; n = 344) in patients starting with 70 mg and “high dose because of severity of migraine” (47.4%; n = 83) in patients starting with 140 mg. Other reasons for the selection of the higher or lower starting dose are presented in Table 3. The proportion of patients who took 140 mg of erenumab increased during the observational period and especially within the first 6 months (Table 3).

Table 3 Other outcomes (FAS)Time Until First Dose Change

The median time between treatments was 29.7 days (range 25.0–96.0), and the median number of injections administered during the study was 14.0 (range 1.0–27.0), corresponding to 12.0 (range 0.5–17.7) injections per year per patient. The most common reason for the first dose increase (39.2%; n = 94) or decrease (58.3%; n = 7) was “according to prescribing information.” For 252 (45.3%) patients, at least one dose change (i.e., 70 mg to 140 mg or 140 mg to 70 mg) was performed during the study. The erenumab dose was increased (i.e., from 70 to 140 mg) in 250 patients (45.0%) and decreased (i.e., from 140 to 70 mg) in 43 patients (7.7%) anytime during the study. For 240 patients, the first change was a dose increase and occurred on average after 5 months. The dose was increased in 250 (45.0%) cases (i.e., 70–140 mg; occurred on average after 5.0 months) and decreased in 43 (7.7%) cases (i.e., 140–70 mg; occurred on average after 6.2 months).

Treatment Interruption

During the study, 118 (21.2%) patients attempted at least once to discontinue treatment. On average, the median time until the first omission attempt was 336.0 (range 37–633) days. The mean duration of omission attempt (i.e., period with discontinuation of erenumab and no other antibody treatment) was 75.2 (range 3–447) days. A total of 92 (78.0%) patients restarted erenumab after their omission attempt. There were 90 patients (76.3%) who received erenumab 140 mg versus 29 patients (24.6%) who received erenumab 70 mg as the last dose before the omission attempt. A total of 83 patients (70.3%) were treated with the same dose before and after the omission attempt. In this study, 29 patients (24.6%) discontinued erenumab treatment after the omission attempt.

Drug Discontinuation

During the study, 85 (15.3%) patients discontinued treatment with erenumab after 7.1 months on average. Twenty-two patients who started treatment with another monoclonal antibody during the study received a mean of 8.2 (SD 3.9) injections of erenumab during 10.6 (SD 3.9) months before switching. The therapy was discontinued most frequently because of “no treatment response” (8.3%; n = 46) or “insufficient improvement of migraine” (5.2%; n = 29). The most common reason for omission attempt was “good effectiveness” (9.0%; n = 50) (Table 3).

Change in MHD and MMD

In more than half of the patients (59.2%), 4–5 injections were administered before the first visit with a reduction in MHD by ≥ 50%. The diary data were available for 507 patients (91.2%). Overall, the mean number of MHD and MMD decreased during the study with erenumab treatment. After 6 months, the mean number of MHD and MMD decreased by −7.0 (6.6) and −4.9 days (5.9), respectively (Fig. 3). The mean (SD) number of MHD and MMD decreased by −8.0 (7.2) and −5.7 (6.2) between the first and last visit, respectively.

Fig. 3

MMD and MHD and at selected visits (FAS)

After 6 months (visit 3; V3), more than half of the patients (55.4%; n = 237) showed a reduction in MHD by ≥ 50%. This proportion increased after 1 year (visit 5; V5: 59.7%; n = 197) and 2 years (visit 9; V9: 70.0%; n = 42). In 22.9% (n = 98) of patients, MHD were reduced by ≥ 75% after 6 months (Fig. 3). After 1 and 2 years, 30.9% (n = 102) and 46.7% (n = 28) showed a reduction of MHD by ≥ 75%, respectively.

Change in Non-migraine Symptoms

At baseline, 37.0% (n = 202) and 35.8% (n = 196) of patients suffered from insomnia and fatigue, respectively. In addition, 28.6% (n = 156) of patients had pain in different parts of the body at baseline. During the study, the proportion of patients with non-migraine symptoms decreased for all symptoms (Table 4).

Table 4 Change in proportion of patients with non-migraine symptomsTreatment Response by Comorbidity (Change in MHD)

About three-fourth of the patients (75.9%; n = 422) had at least one comorbidity (Table 1). At month 6, more than 50% of patients with comorbidities (psychiatric disorder [54.3%; n = 88], depression [51.9%; n = 56], cardiac disorder [58.8%; n = 10], autoimmune disorder [51.8%; n = 57]) showed a decrease in MHD by ≥ 50%, except for patients with GI disorders (39.0%; n = 16) (Table 5). The proportion of patients with a ≥ 50% reduction in MHD increased from months 6 to 24 in patients with all types of comorbidities. Similarly, the proportion of patients with a ≥ 75% reduction in MHD increased from months 6 to 24 in patients with all types of comorbidities except in patients with autoimmune disorders (Table 5).

Table 5 Treatment response by comorbidity (change in MHD): FASPatient-Reported Outcomes

Headache Impact Test-6: At baseline, the mean HIT-6 score was 64.5 (range 0–100, with 100 being worst), with lower values for patients previously treated with erenumab versus naïve patients (62.0 [6.9] versus 66.5 [4.7]). The mean HIT-6 scores improved by −6.8 (8.2) after 6 months (visit 3; V3) and by −9.3 (9.2) after 2 years (visit 9; V9) (Fig. 4). At baseline, headaches had a severe impact (HIT-6 score 60) on patients’ lives in 83.1% (previously treated: 71.3%, naïve: 92.5%) and no or little impact (HIT-6 score < 50) in 2.3% of patients. At visit 3 (V3) and visit 9 (V9), a severe impact was reported in 43.5% and 35.6% of patients and no or little impact in 18.7% and 27.1% of patients, respectively.

Fig. 4

Mean headache impact test (HIT-6) scores at selected visits (FAS)

Treatment Satisfaction Questionnaire for Medication: At baseline, mean TSQM scores were 72.4 for “global satisfaction,” 64.9 for “effectiveness,” 74.3 for “side effects,” and 80.4 for “convenience” (Fig. 5). At month 6, the mean TSQM score was 83.5 for “global satisfaction,” 73.8 for “effectiveness,” 89.1 for “side effects,” and 88.4 for “convenience.”

Fig. 5

Mean Treatment Satisfaction Questionnaire for Medication (TSQM) scores at selected visits (FAS)

At baseline, the mean number of days per month incapacitated for work or with sick leave due to migraine was 3.9 (6.8) days. The mean number of days of patients not at work/school (month 6: 1.1 [2.3] days, month 12: 0.9 [2.1] days, and month 24: 0.8 [2.0] days) and with reduced performance at work/school (month 6: 1.6 [2.6] days, month 12: 1.4 [2.4] days, and month 24: 1.3 [2.5] days) decreased from month 6 to month 24.

Overview of Adverse Events

In total, 776 AEs were reported in 52.9% (n = 294) patients. A total of 6.5% (n = 36) of patients experienced at least one serious adverse event (SAE), and in 6.3% (n = 35) of patients, the AE led to the withdrawal of erenumab (Table 6). One fatal SAE was reported in a 63-year-old female patient who suffered from cerebral infarction. This fatal SAE was assessed by the physician as unrelated to erenumab. In 2.5% (n = 14/556) of all patients, an adverse drug reaction (ADR) was observed after a dose increase. The most common AEs were (by System Organ Class [SOC] terms; > 15%) GI disorders (19.8%), nervous system disorders (16.5%), and general disorders and administration site conditions (13.1%) (Table 6). In this study, constipation (12.1%) was the most frequently reported AE, in line with the SmPC of Aimovig® (erenumab) [16]. No new safety signals were identified.

Table 6 Overview of adverse events (SAF)

Of the 776 AEs, 8.8% (n = 68) were serious and 12.3% (n = 95) were severe. More than half of the patients recovered from AEs (59.4%; n = 447). There were 32.7% (n = 252) patients with AEs who showed a causal relationship with erenumab. In 53.5% of patients with AEs which led to a dose reduction (4.0%) or permanent/temporary withdrawal of erenumab (25.1%), an improvement was observed after the reduction/withdrawal. The mean time until the first ADR was 4.1 months (range 0.0–19.6).