Resumption of migraine preventive treatment with CGRP(-receptor) antibodies after a 3-month drug holiday: a real-world experience

Three monoclonal antibodies (mAbs) targeting the Calcitonin Gene-Related Peptide (CGRP) pathway are currently available in Germany as specific migraine preventive treatments: erenumab binds to the CGRP-receptor (CGRP-R), while galcanezumab and fremanezumab target CGRP directly [1].

The efficacy and tolerability of CGRP(-R) mAbs have been demonstrated in large trial programs and confirmed in numerous real-world studies [2‐13]. However, several questions remain open in the management of these novel drugs.

One key issue is the optimal treatment duration and the need for treatment discontinuation after a period of successful therapy. The expert consensus of the European Headache Federation (EHF) recommends a discontinuation attempt after 6–12 months to reevaluate the need for preventive therapy [14]. Current real-world studies have shown a progressive worsening of migraine frequency during such a discontinuation attempt [15‐17]. In most patients, migraine frequency rapidly returned to the levels before the start of prophylactic mAb therapy [15‐17].

Based on our experience, approximately 90% of patients resume treatment within three months after discontinuation [17]. However, the course of migraine after mAb therapy reinitiation remains unknown.

From a clinical point of view, treatment resumption would ideally lead to an improvement of migraine frequency similar to the initial treatment. A predictable good response after restart would facilitate the decision to temporarily stop treatment and begin again in case of disease deterioration. Alternatively, a second treatment cycle after a long pause could potentially be less effective than the first one, for example due to habituation effects. The evaluation of headache parameters during a second treatment cycle is therefore a crucial step towards improved care of patients with migraine. We have recently published data on migraine frequency and quality of life after mAbs discontinuation [17]. We now report migraine characteristics following the reinitiaton of CGRP(-R) mAb therapy after a three-month discontinuation attempt.

The majority of patients with migraine who resumed preventive treatment with the same CGRP(-R) mAb after a three-month drug holiday experienced a significant reduction of migraine frequency and acute medication use. HIT-6 scores improved, indicating a reduction of headache impact on everyday life. Migraine frequency after three months of mAb therapy following the drug holiday was similar to the frequency that was reported at the end of the first treatment period.

This data is relevant for clinical practice as most patients treated with a CGRP(-R) mAb in Europe undergo a drug holiday after 6–12 months as recommended by national and international guidelines, or even forced by the healthcare systems of some countries [14, 15, 19]. Recent studies have focused on migraine progression after treatment discontinuation. Patients with episodic migraine who received galcanezumab for six months in the EVOLVE-1 and -2 trials showed a progressive deterioration of migraine frequency within four months after treatment cessation but did not return to pretreatment baseline levels [20]. Worsening of migraine in a real-world setting appears more pronounced. De Matteis et al. reported a significant deterioration of migraine frequency as early as in weeks 1–4 after erenumab discontinuation [16]. In the study by Gantenbein et al., half of patients reached after three months of treatment pause a number of MMD comparable to the baseline phase [15]. Vernieri et al. also reported a gradual increase of MMD following cessation of treatment with erenumab and galcanezumab: After three months of drug holiday the 50% responder rates decreased from > 70% to < 30% [19]. Similarly, in the parent study of this analysis, we showed a significant increase of migraine frequency over time after CGRP(-R) mAbs treatment cessation [17], which is also observed in this subgroup analysis.

Headache data after a mAb drug holiday is only available for two small Italian cohorts: In the study by De Matteis et al., n  = 10 patients restarted erenumab treatment after only one month of treatment pause and showed a significant improvement of MMD and AMD in the first month after restart [16]. In the study by Iannone et al., n = 32 patients reported significantly improved MMD, AMD and HIT-6 scores in the first month of retreatment with erenumab or galcanezumab compared to the third month of treatment discontinuation [21]. Our results confirmed these preliminary findings and expand them to a longer time period.

Our findings are reassuring for patients who are advised to stop treatment with CGRP(-R) mAbs. The confirmation that almost 75% of patients have a favorable response after treatment resumption with the same mAb medication might reduce the fear of a drug holiday.

A lack of improvement after a drug holiday was described for other diseases. Multiple reports described a decreased effectiveness of lithium in patients with bipolar disorders after a treatment interruption, a phenomenon called “lithium-discontinuation-induced refractoriness” [22‐25]. A possible neurobiological explanation is based on the phenomenon of episode sensitization: New and more severe episodes occurring during the drug holiday may cause neurobiological alterations resulting in a greater likelihood of recurrence [26]. This observation may apply also for some patients with migraine. Of note, about 25% of patients in this analysis did not benefit from treatment resumption with the mAb they previously responded to. These patients responded particularly well to the first treatment with mAbs prior to the drug holiday. The cause for the poorer response during the second treatment period remains to be determined. Anti-drug antibodies (ADA) may play a role in these patients. So far, ADAs against CGRP(-R) mAbs did not have any impact on their efficacy in randomized-controlled trials [27] albeit in a scenario without treatment interruption. The development of ADAs after treatment interruption or treatment restart has not been investigated yet.

The general recommendation of a drug holiday is based on the treatment with oral migraine preventatives. Oral drugs are commonly discontinued due to lack of efficacy or side effects, only 20% of patients continue treatment for one year [28]. Data from insurance companies’ databases reveals that only 10% of patients restart treatment with the same oral preventive drug after a discontinuation attempt [29]. As opposed to oral medications, a large proportion of our patients wanted to start again with the same mAb medication after three months at the latest [17]. Moreover, mAbs long-term data over several years demonstrated a consistent good efficacy and tolerability [30]. Given the recent evidence on migraine deterioration during discontinuation attempts [15‐17, 19], the need for periodical treatment interruptions remains a matter of discussion.

This is the longest prospective analysis on treatment resumption with CGRP(-R) mAbs after a drug holiday. The inclusion and exclusion criteria allow for a very homogenous population. The small sample size constitutes a limitation and enables only exploratory analyses for the subgroups of patients with CGRP and CGRP-R mAbs. Given the real-world character of our investigation, we cannot control for placebo or nocebo effects during discontinuation or after restart.

In conclusion, we showed a significant reduction of migraine frequency after re-initiation of treatment with CGRP(-R) mAbs. While over 70% of patients returned to the same migraine frequency comparable to the end of the first treatment cycle, in 30% of patients the response was not sufficient. Further research should aim to better characterize these patients and design personalized treatment regimens.