Erenumab for Migraine Prevention in a 1-Year Compassionate Use Program: Efficacy, Tolerability, and Differences Between Clinical Phenotypes

Abstract

During a 1-year compassionate use program, 156 patients with migraine self-administered a monthly dose of erenumab 140 mg with a subcutaneous autoinjector. Main inclusion criteria were: ≥ 4 migraine days/month and ≥two prior prophylactic treatment failures. The patients covered the migraine severity spectrum from episodic migraine (EM) (n = 80) to chronic migraine (CM) (n = 76). During the 3rd month of treatment, monthly headache days decreased by 45.7% in EM and 35.5% in CM. The 50% responder rate for reduction in monthly headache days was significantly higher in EM (55%) than in CM (43%) (p = 0.05). In both the migraine subgroups, the clinical improvement vs. baseline was already significant during the 1st month of treatment (p < 0.001). There were also significant reductions in mean headache severity, duration, and monthly days with acute drug intake. The 30% responder rate at 3 months was 60% in CM and 54.1% of patients reversed from CM to EM. The therapeutic effect was maintained at 12 months when 50% responder rates, considering discontinuation for lack of efficacy or adverse effects as 0% response, still were 51% in EM and 41% in CM. A total of 10 patients with EM (12.5%) and 23 patients with CM (30.3%) had discontinued treatment, considering the treatment as ineffective. At 3 months, 48% of patients reported non-serious adverse events among which the most frequent was constipation (20.5%); corresponding figures at 12 months were 30 and 15%. Discontinuation due to an adverse effect for the entire 12 month period was rare (3.8%). The lower efficacy in CM than in EM was mainly due to a very low 50% responder rate in patients with CM with continuous pain (13%) as compared to CM with pain-free periods (58%) (p < 0.001). Similarly, the 50% responder rate was lower in patients with ≥two prior prophylactic treatment failures (40.5%) compared to those with two failures (70%) (p < 0.05). There was no significant efficacy difference between low (4-7 migraine days/month, n = 22) and high frequency (8-14 days, n = 59) EM nor between patients with CM with (n = 50) or without (n = 26) acute medication overuse. Erenumab had no effect on the frequency of auras. Taken together, erenumab 140 mg monthly was highly effective for migraine prophylaxis over the whole severity spectrum of the disease, except in patients with continuous headaches. Its effect is significant after the first injection, quasi-maximal after the second injection, and does not wear off after 12 months. The most frequent adverse effect was constipation. These results are compared to those published for erenumab in the pivotal randomized placebo-controlled trials and to those reported in several recent real-world studies.

Keywords: compassionate use; erenumab; migraine prophylaxis; monoclonal antibodies blocking CGRP transmission; outcome predictors.

Conflict of interest statement

JS is an advisor for Novartis Benelux, Teva Pharma, Lundbeck and Man & Science; he has been a principal investigator in clinical trials sponsored by Novartis, Teva and Eli Lilly. GT has been an investigator in clinical trials sponsored by Novartis, Teva and Eli Lilly. RN and PG have been data managers in clinical trials sponsored by Novartis, Teva and Eli Lilly. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2021 Schoenen, Timmermans, Nonis, Manise, Fumal and Gérard.

Figures

Figure 1

Change in monthly headache days after 3 months of erenumab treatment in patients with episodic migraine (EM) (dashed line) and chronic migraine (CM) (continuous line). ***p < 0.001 vs. baseline (Friedman's ANOVA, the post-hoc Dunn's test).

Figure 2

Proportion of patients with ≥50% reduction in monthly headache days during 3 months of treatment with 140 mg erenumab per month. At 3 months, 54.1% of patients with CM had reverted to EM. #p = 0.05 chi-squared test “EM vs. CM”.

Figure 3

Proportions of different levels of decrease in monthly headache days during 3 months of erenumab treatment in EM (light bars) and CM (dark bars). *p = 0.014 chi-squared test “EM vs. CM”.

Figure 4

Mean monthly headache severity (three-point scale from 1-mild to 3-severe headache) and number of days with acute drug intake (mean ± SEM) during the first 3 months of treatment. *p < 0.05; ***p < 0.001 (Friedman's ANOVA, the Dunn's post-hoc test).

Figure 5

Proportion of patients reporting no (52%) or various adverse effects (48%) after 3 months of treatment with erenumab.

Figure 6

Monthly headache days during 12 months of erenumab treatment (mean ± SEM) and their percentage decrease during 12 months compared to the pretreatment month in EM (dashed line) and CM (continuous line). The number of patients who stopped treatment because of inefficacy or adverse effects is shown at 6, 9, and 12 months. p < 0.0001 vs. baseline at all the timepoints; p < 0.05 between EM and CM.

Figure 7

Proportion of patients with CM with 50 or 30% decrease in headache days during the 3rd month of erenumab treatment: on the left, patients with CM with pain-free periods (ICHD3 A1.3.1); on the right, patients with CM with continuous pain (ICHD3 A1.3.2). Insert: monthly change in headache days during the first 3 months of treatment and percentage decrease during 3 months (pain-free group: dashed line; continuous pain: continuous line). ***p < 0.001 chi-squared test “pain-free” vs. “continuous pain”.

Figure 8

Fifty and 30% responder rates for monthly headache days at 3 months of treatment in patients with two prior treatment failures (on the left) and those with >two prior failures (on the right). *p < 0.05, **p < 0.01 chi-squared test “two prior failures” vs. “>two prior failures”.

Figure 9

Change in monthly numbers of headache days (dashed line) and auras (continuous line) during the first 3 months of treatment with erenumab in patients having both migraine with and without aura attacks. ***p < 0.001 vs. baseline (Friedman's ANOVA, the Dunn's post-hoc test).