Abstract
Background
Real-world data on calcitonin gene-related peptide (CGRP) monoclonal antibodies (mAbs) are essential to determine whether a class effect exists and to assess the benefit of switching between treatments. Available data remain limited as most studies focused exclusively on patients who failed previous CGRP mAbs and evidence regarding newer agents such as eptinezumab are still scarce.
Methods
This prospective real-world study included all adult patients enrolled in the FHU InovPain registry who received intravenous eptinezumab 100 mg and after prior treatment with one or more subcutaneous CGRP mAbs, regardless of their response to previous CGRP mAbs. According to the 50% response rate (in terms of monthly migraine days) after 6 months of treatment, a descriptive analysis of switching from subcutaneous CGRP mAbs to eptinezumab was performed. Patients were classified into three subgroups: cross-effectiveness (all CGRP mAbs used were effective), cross-ineffectiveness (all CGRP mAbs used were ineffective), and no cross-response (different responses across CGRP mAbs used). Factors associated with response to CGRP mAbs were investigated by comparing patients with cross-ineffectiveness (with the CGRP mAbs used) to those who responded to at least one CGRP mAb (cross-effectiveness and no cross-response), followed by multivariate logistic regression.
Results
A total of 190 patients (83.7% women; mean age 52.2 ± 13.7 years) were included. The 50% responder rate to eptinezumab was 76.0% (95% CI: 67.3-83.1) in patients who had responded to at least one previously used CGRP mAb, compared with 30.4% (95% CI: 20.2-42.8) in patients with no prior response to CGRP mAbs. Cross-effectiveness, cross-ineffectiveness, and no cross-response were observed in 46.8% (95% CI: 39.6-54.2), 28.9%, (95% CI: 22.7-36.0) and 24.2% (95% CI: 18.4-31.7) of patients, respectively. Only two factors were associated with response to at least one CGRP mAb: a lower helplessness score on the Pain Catastrophizing Scale (AdjOR 0.91, 95% CI: 0.86-0.97, p=0.004) and a lower allodynia score on the ASC-12 (AdjOR 0.91, 95% CI: 0.84-0.98, p=0.010).
Conclusion
This real-world study confirms the clinical benefit of switching to eptinezumab in nearly one-third of patients who did not respond to previous subcutaneous CGRP mAbs. It also demonstrates a class effect that is not absolute, as nearly one-quarter of patients showed no cross-response between CGRP mAbs.
Clinical trial number
Not applicable.
Keywords: CGRP monoclonal antibodies, Real-world, Refractory migraine, Preventive treatment, Eptinezumab, Treatment switching, Cross-response
Background
Since 2018, several monoclonal antibodies (mAbs) targeting the calcitonin gene-related peptide (CGRP) pathway have been approved by the Food and Drug Administration and the European Medicines Agency for the preventive treatment of migraine. These CGRP mAbs include eptinezumab, erenumab, fremanezumab, and galcanezumab. Pivotal randomized controlled trials (RCTs) demonstrated similar efficacy and safety/tolerability among all four anti-CGRP mAbs [1, 2] and suggested a pharmacological class effect. However, differences in therapeutic targets (receptor for erenumab versus ligand for eptinezumab, fremanezumab, and galcanezumab), immunoglobulin subclasses (IgG1 for eptinezumab, IgG2 for erenumab and fremanezumab, and IgG4 for galcanezumab), immunogenicity (fully human antibody for erenumab; humanized antibodies for eptinezumab, fremanezumab, and galcanezumab), and routes of administration (intravenous for eptinezumab; subcutaneous for erenumab, fremanezumab, and galcanezumab) may influence efficacy and safety/tolerability differently [3]. Given these differences, there is a rationale for switching between CGRP mAbs and such switching is one of the answers to the question, “What to do with non-responders to CGRP mAbs?” in clinical practice [4]. Although no RCTs have specifically addressed this question, observational studies support switching between CGRP mAbs, with ≥ 50% responder rates ranging from 10% to 67% in patients who failed one or more previous antibodies [5]. However, most available studies are retrospective and therefore carry a high risk of bias [6]. Moreover, these studies focused exclusively on the effectiveness of switching in non-responders and do not provide data on an overall cross-response to CGRP mAbs.
The specific situation in France provided a unique opportunity to evaluate the sequential use of different CGRP mAbs in individual patients regardless of prior response. Outside the clinical research setting, access to CGRP mAbs in France occurred in three successive phases. Initially, only erenumab was available through a compassionate use program that allowed patients with migraine who had failed oral prophylaxis to receive this antibody for one year. Since the end of this program set up by Novartis, erenumab has no longer been available in France. Galcanezumab and fremanezumab were then marketed but without reimbursement, requiring patients to bear the full cost (at least €245 to €270 per month). Finally, eptinezumab became available on the French market, for hospital use only, in selected healthcare facilities, at no cost to patients [7].
We therefore identified all patients included in the French Fédération Hospitalo-Universitaire (FHU) InovPain registry who received eptinezumab after prior treatment with one or more subcutaneous CGRP mAbs. Because patients had been enrolled in the FHU InovPain registry before initiation of eptinezumab and previous antibodies, prospective evaluation of all CGRP mAbs used in each patient was possible. Before initiation of eptinezumab, the reasons for discontinuing previous CGRP mAbs were not limited to ineffectiveness. In the case of erenumab, treatment discontinuation was related to the end of the compassionate use program, irrespective of treatment response. For fremanezumab and galcanezumab, discontinuation could result either from lack of efficacy or, in responders, from inability to continue paying the full cost of treatment.
The objectives of this prospective real-world study were first to evaluate the response to eptinezumab in patients previously exposed to subcutaneous CGRP mAbs, and second to assess the overall cross-response to CGRP mAbs and its predictive factors.
Methods
Research context
This study was conducted using data from the French FHU InovPain patient registry, previously described elsewhere [8]. The main advantage of this registry lies in the prospective collection of clinical data throughout patient follow-up. The data in this registry are collected during consultations with a headache specialist, initially scheduled every 3 months, then every 6 months, using an assessment form that includes a paper headache diary. The data are entered into the registry by clinical research assistants. The quality control procedure involves meetings between the clinical research assistants and the center’s headache specialists to verify that the data entered the registry are consistent with the patients’ source records.
Study design
This study is an analysis of data from a prospective monocentric cohort of patients treated with eptinezumab after prior exposure to subcutaneous CGRP mAbs.
Study objectives
The primary objective of this study was to evaluate the overall response to eptinezumab in patients previously exposed to subcutaneous CGRP mAbs, and to assess this response according to prior exposure to each individual CGRP mAb. The secondary objective was to evaluate overall cross-response to CGRP mAbs and to identify factors associated with response to CGRP mAbs.
Participants included in the cross-response analysis
The inclusion criteria for the analysis were as follows: (i) patients enrolled in the FHU InovPain registry before initiation of CGRP mAbs; (ii) age ≥ 18 years; (iii) patients with migraine refractory to non-specific migraine prophylaxis, defined by at least three treatment failures among beta-blockers, topiramate, candesartan, tricyclic antidepressants, and onabotulinumtoxinA; (iv) patients treated with eptinezumab (100 mg IV once every 3 months) with prospective assessment of effectiveness after 6 months of treatment; (v) patients previously treated (before eptinezumab treatment) with one or more subcutaneous CGRP mAbs (erenumab 70 or 140 mg SC once monthly / fremanezumab 225 mg SC once monthly / galcanezumab 240 mg SC loading dose, followed by 120 mg SC once monthly) with prospective assessment of effectiveness after 6 months of treatment for each of them; (vi) a washout period of at least 6 months between treatment with two different CGRP mAbs, (vii) patients with at least 8 monthly migraine days (MMDs) before initiation of each CGRP mAb.
The exclusion criteria for the analysis were: (i) unavailable data for one or more antibodies used; (ii) discontinuation of a CGRP mAb because of an adverse event.
Data collected in the FHU InovPain registry and used for the cross-response analysis
For all CGRP mAbs, prospectively collected headache diary data from the month preceding treatment initiation and during the 6-month follow-up period were analyzed.
For each individual CGRP mAb, treatment was considered effective when the number of MMDs decreased by at least 50% after 6 months of treatment without adverse events leading to treatment discontinuation. Conversely, treatment was considered ineffective when the reduction in MMDs was less than 50%.
Patients were classified into three groups: (i) cross-effectiveness when all CGRP mAbs used were effective (these patients showed a response only to the specific CGRP mAbs to which they had been exposed, and not necessarily to all available CGRP mAbs); (ii) cross-ineffectiveness when all the CGRP mAbs used were ineffective (these patients showed a lack of response only to the specific CGRP mAbs to which they had been exposed, and not necessarily to all available CGRP mAbs); and (iii) no cross-response when different CGRP mAbs induced different responses (positive or negative).
To identify potential factors associated with response to CGRP mAbs, we considered, for each patient, data collected in the FHU InovPain registry before initiation of eptinezumab treatment (the last antibody used in all patients). These data included socio-demographic characteristics, migraine clinical features, individual burden induced by migraine, comorbidities, acute migraine treatments and patients’ psychological profile.
Socio-demographic variables included age, gender, and professional status (employed, on sick leave and/or disability leave, job seeker, unemployed, retired, or student).
Clinical migraine characteristics included monthly headache days (MHDs), MMDs, migraine type (high-frequency episodic migraine [HFEM] or chronic migraine [CM]), presence of aura, migraine attack characteristics (unilateral pain, photophobia, nausea, parasympathetic symptoms), continuous interictal pain, and interictal allodynia and its severity assessed using the Allodynia Symptom Checklist (ASC-12) [9].
Migraine-related burden was assessed through functional impact measured by the Headache Impact Test (HIT-6) [10], emotional impact (anxiety and depression) assessed by the Hospital Anxiety and Depression (HAD) scale [11], disability assessed by the Migraine Disability Assessment Scale (MIDAS) [12], burden between attacks assessed by the Migraine Interictal Burden Scale (MIBS-4) [13], and health-related quality of life assessed by the EuroQol instrument (EQ-5D) [14].
The number of comorbidities was collected, including psychiatric comorbidities (with a focus on anxiety and depression) and non-psychiatric comorbidities (with a focus on obesity).
Data collected on acute antimigraine treatment included: monthly days with acute medication use; resistance to triptans and non-steroidal anti-inflammatory drugs (NSAIDs); medication overuse with related medication (triptans, and opioids); and dependence assessed by the Severity Dependence Scale (SDS) [15].
Psychological profile data included pain beliefs and perceptions (4 dimensions: mystery, pain permanence, pain constancy, and self-blame) assessed by the Pain Beliefs and Perceptions Inventory (PBPI) [16], and pain catastrophizing (3 dimensions: helplessness, rumination, and magnification) assessed by the Pain Catastrophizing Scale (PCS) [17].
Statistical analysis
Continuous variables are presented as mean ± standard deviation (SD), and categorical variables as numbers and percentages. Missing data were not recoded prior to analysis.
Eptinezumab responder rates were reported descriptively for the overall population; for patients previously exposed to erenumab, fremanezumab, or galcanezumab, for patients previously exposed to a single subcutaneous CGRP mAb, for patients previously exposed to 2 subcutaneous CGRP mAbs and for patients previously exposed to 3 subcutaneous CGRP mAbs. In all cases, eptinezumab responder rates were analyzed according to response or non-response to previously used subcutaneous CGRP mAbs.
Based on the response to eptinezumab and to previously used subcutaneous CGRP mAbs, patients were classified into three cross-response subgroups: cross-effectiveness; cross-ineffectiveness, and no cross-response.
The search for factors associated with response to CGRP mAbs was based on comparisons between patients with cross-ineffectiveness and patients who responded to at least one CGRP mAb (cross-effectiveness and no cross-response groups combined). Comparisons were performed using baseline data collected before eptinezumab treatment initiation (corresponding to the last CGRP mAb used), using the Wilcoxon–Mann–Whitney test for continuous variables and the χ² test or Fisher’s exact test for categorical variables, as appropriate. Benjamini-Hochberg FDR correction for multiple testing was applied. The variables that were statistically significant (p < 0.05) in univariate analysis were used in a multivariate logistic regression analysis to identify factors associated with response to CGRP mAbs. A stepwise backward selection was performed. Only variables with a p-value < 0.05 were included in the final multivariate logistic model. The factors associated with response to CGRP mAbs are presented as adjusted odds ratios with 95% confidence intervals. We assessed multicollinearity among factors associated with response to CGRP mAbs using the Variance Inflation Factor (VIF). Variables with a VIF > 5 were considered highly collinear. To evaluate model discrimination, the Area Under the Receiver Operating Curve (AUC) was calculated using the model-predicted probabilities.
Statistical analyses were performed using R statistical software (version R-4.5.1).
Ethical considerations
This study was part of the FHU InovPain registry clinical research program, which was authorized by the French Data Protection Authority (Commission Nationale de l’Informatique et des Libertés [CNIL]; authorization no.278 dated 11/09/2017). All patients included provided written informed consent allowing their data recorded in the FHU InovPain registry database to be used for clinical research purposes. This analysis did not require further authorization from an ethics committee or from the CNIL.
This study was conducted in accordance with the declaration of Helsinki.
Results
Characteristics of the study population
A total of 190 patients were included in this analysis: 159 (83.7%) women and 31 (16.3%) men, with a mean age of 52.2 (± 13.7) years. Data regarding socio-demographic characteristics, migraine clinical features, individual burden induced by migraine, comorbidities, acute migraine treatment and patients’ psychological profile are detailed in Table 1.
Table 1.
Population included in cross-response analysis - data collected before initiation of eptinezumab treatment (CI: contraindication / n = number of patients for whom data were collected / SD standard deviation)
| Data collected | n | |
|---|---|---|
| Socio-demographic characteristics | ||
| Age in years (mean ± SD) | 190 | 52.2 ± 13.7 |
| Gender | 190 | |
| women (%) | 159 (83.7) | |
| men (%) | 31 (16.3) | |
| Clinical features of migraine | ||
| Monthly headache days (mean ± SD) | 190 | 19.0 ± 8.1 |
| Monthly migraine days (mean ± SD) | 190 | 17.5 ± 8.3 |
| Migraine form | 190 | |
| episodic migraine (%) | 71 (37.4) | |
| chronic migraine (%) | 119 (62.6) | |
| Aura | 190 | 18 (9.5) |
| Characteristics of migraine attack | 190 | |
| unilaterality (%) | 124 (65.3) | |
| photophobia (%) | 150 (78.9) | |
| nausea (%) | 124 (70.5) | |
| parasympathetic signs | 13 (6.8) | |
| Continuous interictal pain | 190 | 70 (36.8) |
| Allodynia score / ASC-12 (mean ± SD) | 183 | 5.5 ± 4.6 |
| Individual burden induced by migraine | ||
| Functional impact / HIT-6 (mean ± SD) | 190 | 64.6 ± 7.0 |
| Emotional impact / HAD | 189 | |
| anxious impact (mean ± SD) | 8.8 ± 4.0 | |
| depressive impact (mean ± SD) | 6.7 ± 4.7 | |
| Disability / MIDAS (mean ± SD) | 177 | 64.6 ± 70.9 |
| Interictal burden / MIBS-4 (mean ± SD) | 186 | 5.6 ± 3.9 |
| Health-related quality of life / EQ-5D | 189 | |
| index (mean ± SD) | 0.6 ± 0.3 | |
| vas (mean ± SD) | 59.0 ± 20.9 | |
| Patients’ comorbidities | ||
| Number of comorbidities (mean ± SD) | 190 | 2.8 ± 2.7 |
| Number of psychiatric comorbidities (mean ± SD) | 190 | 0.9 ± 1.0 |
| anxiety (%) | 190 | 68 (35.8) |
| depression (%) | 190 | 62 (32.6) |
| Number of somatic comorbidities (mean ± SD) | 190 | 1.9 ± 2.1 |
| BMI (mean ± SD) | 190 | 23.3 ± 3.6 |
| Acute migraine treatment | ||
| Monthly days with acute treatment use (mean ± SD) | 190 | 16.3 ± 9.7 |
| Triptans resistance (%) | 187 (3 CI) | 34 (18.2) |
| Medication overuse - MO (%) | 190 | 122 (64.2) |
| MO involving triptans | 93 (48.9) | |
| MO involving opioids | 36 (18.9) | |
| Dependence score (mean ± SD) | 185 | 8.9 ± 3.7 |
| Patients’ psychological profile | ||
| Mental representation / PBPI scale | ||
| mystery dimension (mean ± SD) | 178 | -0.2 ± 1.2 |
| pain permanence dimension (mean ± SD) | 174 | 0.4 ± 0.9 |
| pain constancy dimension (mean ± SD) | 181 | -0.1 ± 1.3 |
| self-blame dimension (mean ± SD) | 183 | -1.5 ± 0.8 |
| Catastrophism / PCS | ||
| helplessness dimension (mean ± SD) | 185 | 13.7 ± 6.2 |
| rumination dimension (mean ± SD) | 185 | 8.8 ± 4.5 |
| magnification dimension (mean ± SD) | 184 | 4.5 ± 3.2 |
| total score (mean ± SD) | 184 | 27.2 ± 12.0 |
Treatment sequences involving CGRP mAbs
As required by the inclusion criteria, all patients were treated with eptinezumab. Prior exposure included galcanezumab in 132 patients, erenumab in 102 patients, and fremanezumab in 26 patients.
Erenumab was used as part of a one-year compassionate use program established by Novartis, allowing patients who responded to erenumab to continue treatment for up to one year, whereas non-responders discontinued treatment after 6 months. Galcanezumab and fremanezumab were self-funded and, regardless of the response, all patients in our cohort discontinued treatment after 6 months because they could no longer afford the financial burden. Eptinezumab was provided at our hospital at no cost to patients, meaning that responders continued treatment beyond 6 months, and non-responders stopped it.
The sequences of CGRP mAbs used are presented in Fig. 1. Overall, 129 patients received two CGRP mAbs, 52 patients received three CGRP mAbs, and 9 received all four available CGRP mAbs.
Fig. 1.
Sequences of CGRP mAbs used (EPT: eptinezumab, ERE: erenumab, FRE: fremanezumab, GAL: galcanezumab)
Response to eptinezumab
The 50% responder rate to eptinezumab in the overall population and across patient subgroups is shown in Figs. 2 and 3.
Fig. 2.
50% responder rate (in terms of monthly migraine days) to eptinezumab following switching from subcutaneous CGRP mAbs to eptinezumab. From left to right: overall population (patients who responded to at least one CGRP mAb in blue, and those who did not respond to any CGRP mAb in orange); patients who had previously used galcanezumab (responders in blue and non-responders in orange); patients who had previously used erenumab (responders in blue and non-responders in orange); and those who had used fremanezumab (responders in blue and non-responders in orange)
Fig. 3.
50% response rate (in terms of monthly migraine days) to eptinezumab following switching from subcutaneous CGRP mAbs to eptinezumab. From left to right: patients who had used a single CGRP mAb before eptinezumab (responders in blue, non-responders in orange); patients who had used two CGRP mAbs before eptinezumab (responders to at least one CGRP mAb in blue, non-responders to any CGRP mAb used in orange); patients who had used three CGRP mAbs before eptinezumab (patients with a response to at least one CGRP mAb in blue, patients with no response to any CGRP mAb in orange)
Among the 190 patients previously treated with one or more subcutaneous CGRP mAbs, eptinezumab induced a ≥ 50% reduction in the number of MMDs after 6 months of treatment in 111 patients (58.4%, 95%CI: 51.0-65.4), regardless of the response to previous antibodies. Among the 121 patients who had responded to at least one previously used subcutaneous CGRP mAb, 92 (76.0%, 95%CI: 67.3–83.1) achieved a 50% response to eptinezumab. Among the 69 patients who had not responded to any previously used subcutaneous CGRP mAbs, 21 (30.4%, 95%CI: 20.2–42.8) achieved a 50% response to eptinezumab (p < 0.001).
Among patients previously exposed to galcanezumab (n = 132), 50% response rate to eptinezumab was 77.0% (95%CI: 65.5–85.7) in patients who had responded to galcanezumab and 25.9%, (95%CI: 15.6–39.3) in patients who had not responded to galcanezumab (p < 0.001). Among patients previously exposed to erenumab (n = 102), 50% response rate to eptinezumab was 86.0% (95%CI: 73.7–93.3) in patients who had responded to erenumab and 26.7% (95%CI: 15.1–42.2) in patients who had not responded to erenumab (p < 0.001). Among patients previously exposed to fremanezumab (n = 26), 50% response rate to eptinezumab 75.0%, (95%CI: 21.9–98.7) in patients who had responded to fremanezumab and 22.7% (95%CI: 8.9–45.8) in patients who had not responded to fremanezumab (p = 0.072) (Fig. 2).
Among patients previously treated with a single CGRP mAb (n = 117), 50% response rate to eptinezumab was 81.6%, (95%CI: 72.6–88.8) in patients who had responded to that previous antibody and 42.5% (95%CI: 27.4–59.0) who had not responded (p < 0.001). Among patients previously treated with two CGRP mAbs (n = 54), 50% response rate to eptinezumab was 72.0%, (95%CI: 50.4–87.1) in patients who had responded to at least one antibody and 13.8% (95%CI: 4.5–32.6) in patients who had not responded to any previous CGRP mAb (p < 0.001). Finally, among patients previously treated with three monoclonal antibodies (n = 9), all (100.0%, 95%CI: 31.0-100.0) who had responded to at least one previous CGRP mAb achieved a 50% response to eptinezumab, whereas none (0.0%, 95%CI: 0.0-48.3) without prior response achieved such a response (p = 0.012) (Fig. 3).
Overall cross-response to CGRP mAbs and factors associated with CGRP mAbs response
In the overall study population, cross-effectiveness was observed in 83 patients (46.8%), cross-ineffectiveness in 56 (28.9%), and no cross-response in 46 (24.2%). A comparison of patient subgroups who had used 2, 3, and 4 CGRP mAbs revealed statistically significant differences in this distribution (p < 0.001). Cross-effectiveness was observed in 55%, 32.7%, and 22.2% of patients who had used 2, 3, and 4 CGRP mAbs, respectively. Cross-ineffectiveness was observed in 18.6%, 28.1%, and 66.7% of patients who had used 2, 3, and 4 CGRP mAbs, respectively. No cross-response was observed in 26.4%, 19.2%, and 22.2% of patients who had used 2, 3, and 4 CGRP mAbs, respectively.
The proportions of cross-effectiveness, cross-ineffectiveness, and no cross-response for each CGRP mAb sequence are detailed in Table 2. Similar distributions were observed when comparing treatment sequences involving only ligand-targeting CGRP mAbs (eptinezumab, fremanezumab, and galcanezumab) with treatment sequences combining both ligand-targeting CGRP mAbs and erenumab, which targets the CGRP receptor (data not shown).
Table 2.
Cross-response to anti-CGRP mAbs (Ep: eptinezumab; Er: erenumab; F: fremanezumab, G: galcanezumab). Percentages and confidence intervals are shown only for sequences involving more than 10 patients
| Anti-CGRP mAbs sequence | n | Cross-effectiveness n (%, 95%CI) |
Cross-ineffectiveness n (%, 95%CI) |
No cross-response n (%, 95%CI) |
|---|---|---|---|---|
| G-Ep | 72 | 39 (54.2, 42.1–65.8) | 12 (16.7, 9.3–27.7) | 21 (29.2, 19.3–41.2) |
| Er-Ep | 55 | 30 (54.5, 40.7–67.8) | 12 (21.8, 12.2–35.4) | 13 (23.6, 13.7–37.3) |
| Er-G-Ep | 37 | 17 (45.9, 29.8–62.9) | 16 (43.2, 29.6–81.2) | 5 (13.5, 5.1–29.6) |
| F-G-Ep | 14 | 0 (0.0, 0.0-26.8) | 8 (61.5, 32.3–84.9) | 5 (35.7, 14.0-64.4) |
| Er-G-F-Ep | 9 | 1 | 6 | 2 |
| F-Ep | 2 | 2 | 0 | 0 |
| Er-F-Ep | 1 | 0 | 1 | 0 |
| All patients | 190 | 89 (46.8, 39.6–54.2) | 55 (28.9, 22.7–36.0) | 46 (24.2, 18.4–31.1) |
CI: Confidence interval
Among the 46 patients with no cross-response, 25/46 patients (56.3%) treated with eptinezumab were responders, compared with 14/33 patients (42.4%) treated with galcanezumab, 8/20 (40%) treated with erenumab, and 1/7 patients (14.3%) treated with fremanezumab.
Compared with patients with cross-ineffectiveness, patients who responded to at least one CGRP mAb (cross-effectiveness and no cross-response groups combined) showed several significant differences (Table 3). In the multivariate analysis using significant differences found in univariate analaysis, only two risk factors associated with the response to a least one CGRP mAb: a lower helplessness score on Pain Catastrophizing Scale (AdjOR 0.91, 95%CI: 0.86–0.97, p = 0.004) and a lower allodynia score on ASC-12 (AdjOR 0.91, 95%CI: 0.84–0.98, p = 0.010). Eleven patients were excluded from the multivariate logistic regression analysis because of missing data. The Variance Inflation Factor (VIF) values were < 5 for the 2 variables included in the final model. Area Under the Receiver Operating (AUC) was 0.694.
Table 3.
Comparison between patients with cross-ineffectiveness and patients with a response to at least one CGRP mAb (cross-effectiveness and no cross-response groups combined)
| Absence of CGRP mAb response (n = 55) |
At least one CGRP mAb Response (n = 135) |
|||
|---|---|---|---|---|
| p-value | Corrected p-value* | |||
| Socio-demographic characteristics | ||||
| Age in years (mean ± SD) | 56.4 ± 14.5 | 50.5 ± 13.0 | 0.006 | 0.017 |
| Gender | 0.657 | 0.657 | ||
| women (%) | 45 (81.8) | 114 (84.4) | ||
| men (%) | 10 (18.2) | 21 (15.6) | ||
| Clinical features of migraine | ||||
| Monthly headache days (mean ± SD) | 24.4 ± 5.9 | 16.8 ± 7.9 | < 0.001 | < 0.001 |
| Monthly migraine days (mean ± SD) | 22.7 ± 7.8 | 15.3 ± 7.6 | < 0.001 | < 0.001 |
| Characteristics of migraine attack | ||||
| unilaterality (%) | 29 (52.7) | 95 (70.4) | 0.021 | 0.038 |
| photophobia (%) | 38 (69.1) | 112 (83.0) | 0.033 | 0.052 |
| nausea (%) | 32 (58.2) | 102 (75.6) | 0.017 | 0.035 |
| parasympathetic signs (%) | 6 (10.9) | 7 (5.2) | 0.204 | 0.235 |
| Continuous interictal pain (%) | 34 (61.8) | 36 (26.7) | < 0.001 | < 0.001 |
| Allodynia score / ASC-12 (mean ± SD) | 7.1 ± 5.2 | 4.9 ± 4.2 | 0.009 | 0.022 |
| Individual burden induced by migraine | ||||
| Functional impact / HIT-6 (mean ± SD) | 67.3 ± 6.5 | 63.5 ± 7.0 | < 0.001 | 0.002 |
| Emotional impact / HAD | ||||
| anxious impact (mean ± SD) | 9.8 ± 4.6 | 8.4 ± 3.8 | 0.063 | 0.086 |
| depressive impact (mean ± SD) | 8.0 ± 5.1 | 6.1 ± 4.4 | 0.018 | 0.035 |
| Disability / MIDAS (mean ± SD) | 98.1 ± 79.7 | 51.1 ± 62.3 | < 0.001 | < 0.001 |
| Interictal burden / MIBS-4 (mean ± SD) | 7.3 ± 3.7 | 5.0 ± 3.8 | < 0.001 | 0.002 |
| Health-related quality of life / EQ-5D | ||||
| index (mean ± SD) | 0.6 ± 0.3 | 0.7 ± 0.3 | 0.010 | 0.022 |
| vas (mean ± SD) | 52.2 ± 22.4 | 61.7 ± 19.6 | 0.011 | 0.024 |
| Patients’ comorbidities | ||||
| Number of comorbidities (mean ± SD) | 3.5 ± 3.0 | 2.6 ± 2.5 | 0.044 | 0.063 |
| Number of psychiatric comorbidities (mean ± SD) | 1.3 ± 1.0 | 0.8 ± 1.0 | 0.003 | 0.010 |
| anxiety (%) | 26 (47.3) | 42 (31.1) | 0.035 | 0.052 |
| depression (%) | 22 (40.0) | 40 (29.6) | 0.167 | 0.213 |
| Number of somatic comorbidities (mean ± SD) | 2.3 ± 2.3 | 1.8 ± 2.0 | 0.189 | 0.229 |
| BMI (mean ± SD) | 23.8 ± 3.7 | 23.0 ± 3.5 | 0.242 | 0.271 |
| Acute migraine treatment | ||||
| Monthly days with acute treatment use (mean ± SD) | 21.6 ± 9.3 | 14.1 ± 9.0 | < 0.001 | < 0.001 |
| Triptans resistance (%) | 13 (24.5) | 21 (15.7) | 0.157 | 0.208 |
| Medication overuse - MO (%) | 45 (81.8) | 77 (57.0) | 0.001 | 0.004 |
| MO involving triptans (%) | 31 (56.4) | 62 (45.9) | 0.192 | 0.229 |
| MO involving opioids (%) | 19 (34.5) | 17 (12.6) | < 0.001 | 0.002 |
| Dependence score (mean ± SD) | 9.3 ± 3.6 | 8.8 ± 3.7 | 0.317 | 0.335 |
| Patients’ psychological profile | ||||
| Mental representation / PBPI scale | ||||
| mystery dimension (mean ± SD) | 0.1 ± 1.1 | -0.3 ± 1.2 | 0.032 | 0.052 |
| pain permanence dimension (mean ± SD) | 0.6 ± 0.8 | 0.3 ± 0.9 | 0.032 | 0.052 |
| pain constancy dimension (mean ± SD) | 0.5 ± 1.3 | -0.3 ± 1.2 | < 0.001 | 0.001 |
| self-blame dimension (mean ± SD) | -1.4 ± 0.9 | -1.5 ± 0.7 | 0.566 | 0.582 |
| Catastrophism / PCS | ||||
| helplessness dimension. (mean ± SD) | 16.2 ± 5.8 | 12.7 ± 6.0 | < 0.001 | 0.002 |
| rumination dimension (mean ± SD) | 9.5 ± 4.2 | 8.6 ± 4.5 | 0.267 | 0.291 |
| magnification dimension (mean ± SD) | 5.3 ± 3.2 | 4.2 ± 3.1 | 0.032 | 0.052 |
| total score (mean ± SD) | 30.9 ± 11.5 | 25.6 ± 11.9 | 0.006 | 0.017 |
*Benjamini & Hochberg FDR correction for multiple testing
Discussion
The objectives of this observational study were to evaluate the response to eptinezumab in migraine patients previously exposed to subcutaneous CGRP mAbs and, more broadly, to assess cross-response between these CGRP mAbs in real-world settings. Our assessment of cross-response was not limited to evaluating the effect of switching in patients who failed one or more CGRP mAbs. Instead, it was based on evaluating CGRP mAbs used sequentially, regardless of whether previous treatments had been effective. This was made possible by the unique situation regarding access to CGRP mAbs in France. Patients of our cohort previously exposed to erenumab could not use it for more than one year—the duration of the compassionate use program implemented by Novartis—as this CGRP mAb is no longer available in France. Patients previously exposed to galcanezumab or fremanezumab used these CGRP mAbs out-of-pocket, but the financial burden was too high, forcing them to discontinue treatment. Ultimately, all patients previously exposed to one, two, or three subcutaneous CGRP mAbs (irrespective of their response) were treated with eptinezumab, which is now available in selected French hospitals, including ours. Therefore, the sequential use of different CGRP mAbs in our cohort was driven primarily by drug availability rather than by treatment response. The aim of our study was to take advantage of the unique situation in France, with these patients who sequentially used different CGRP mAbs, to provide new data and investigate whether CGRP mAbs exhibit a class effect.
We showed that eptinezumab induced a ≥ 50% reduction in the number of MMDs after 6 months of treatment in 58.4% (95% CI: 51.0-65.4) of patients previously exposed to subcutaneous CGRP mAbs, irrespective of prior response to these treatments. Response to eptinezumab was strongly associated with prior response to subcutaneous CGRP mAbs, with 50% responder rates of 76.0% (95% CI: 67.3–83.1) in previous responders versus 30.4% (95% CI: 20.2–42.8) in previous non-responders.
This 50% response rate in patients with a history of CGRP mAb failure is lower than that observed in prospective observational studies evaluating the effectiveness of eptinezumab in mixed populations of both CGRP mAb-naïve patients and patients with prior CGRP mAb failure: 58% in the multicentric Italian study performed by Barbanti et al. [18], 53.2% in the multicentric Italian study performed by Iannone et al. [19], and 58.3% for EM and 57.8% for CM (according to MHDs) in the multicentric Greek study performed by Dermitzakis et al. [20], with the three studies involving 24, 58, and 68 patients who did not respond to subcutaneous CGRP mAbs and switched to eptinezumab (100 mg per infusion for the vast majority of patients), respectively. Nevertheless, in the monocentric study performed by Andreou et al. in the United Kingdom, which included 80 patients who had failed treatment with subcutaneous CGRP mAbs, a response rate of 32.5% (with a 30% threshold for MMDs) was observed after 6 months of treatment with eptinezumab 100 mg [21]. Interpretation of differences in response rates between these studies and our study is difficult because, despite their prospective design, these studies were not specifically designed to investigate the effect of switching in patients who did not respond to one or more CGRP mAbs.
To date, among studies focusing on switching between CGRP mAbs [5], only two have evaluated switching from subcutaneous CGRP mAbs to eptinezumab [22, 23].
In a retrospective monocentric study performed in the United Arab Emirates, Suliman et al. reported 18 patients (10 with EM and 8 with CM) who did not respond to galcanezumab, including 12 (66.7%) who experienced a reduction of more than 50% in MMDs after 6 months of treatment with eptinezumab (dose not specified in the publication) [22]. This response rate is much higher than the rate (25.9%) observed among patients who did not respond to galcanezumab in our study, but differences in data collection methods (retrospective vs. prospective) and sample size (18 vs. 58) should be noted.
In a retrospective multicentric study performed in Germany, Scheffler et al. described the effectiveness of eptinezumab (used at 100 mg) in 79 patients (19 with EM and 60 with CM), including 65 patients with a history of subcutaneous CGRP mAb failure: 20 with one failed subcutaneous CGRP mAb, 28 with two failed subcutaneous CGRP mAbs, and 17 with three failed subcutaneous CGRP mAbs [23]. The authors reported a 50% response rate of 25% among the 20 patients who had failed treatment with a single subcutaneous CGRP mAb, but they did not report this rate for patients who had failed treatment with more than one subcutaneous CGRP mAb [23]. Nevertheless, they reported the 30% response rate (in terms of MMDs) for all patients without showing any statistically significant difference according to the number of prior ineffective subcutaneous CGRP mAbs (45.1%, 32.1%, and 33.3% for one, two, and three subcutaneous CGRP mAbs, respectively) [23]. The lack of a significant association between the decrease in response to eptinezumab (in terms of MMDs) and the number of prior ineffective subcutaneous CGRP mAbs observed in the German study [23] contrasts with our results, which show that the 50% response rate decreases as this number increases (42.5%, 13.9%, and 0% for one, two, and three subcutaneous CGRP mAbs, respectively). However, the German team demonstrated a significant association between the number of prior ineffective subcutaneous CGRP mAbs and the decrease in the 30% response rate to eptinezumab in terms of MHDs (45%, 32.1%, and 23.5% for one, two, and three subcutaneous CGRP mAbs, respectively) [23], suggesting, as in our results, that the proportion of patients who respond after switching decreases with each successive switch. Such a decrease in switching efficacy with repeated CGRP mAb switching had already been demonstrated in the observational study of the German NeuroTransData registry that did not involve eptinezumab [24].
Real-world data on response rates following switching between CGRP mAbs, according to the number of prior CGRP mAb failures, are important for assessing a potential class effect of CGRP mAbs. Our study, which examined sequential use of different CGRP mAbs—regardless of whether they were effective or ineffective—provides additional data for discussing this class effect. We identified three subgroups of patients according to the effect of the different CGRP mAb sequentially used: (i) cross-effectiveness if all the CGRP mAbs used were effective; (ii) cross-ineffectiveness if all the CGRP mAbs used were ineffective; and (iii) no cross-response if the CGRP mAbs used induced different responses (positive or negative). When considering the entire population included in our study, cross-effectiveness, cross-ineffectiveness, and no cross-response were observed in 46.8% (95% CI; 39.6–54.2), 28.9% (95% CI: 22.7–36.0), and 24.2% (95% CI: 18.4–31.1%) of patients, respectively. Overall, a class effect was observed in approximately three-quarters of our patients, who either responded to all CGRP mAbs used, or failed all CGRP mAbs used. However, this class effect was not absolute, as one-quarter of patients exhibited heterogeneous responses across different CGRP mAbs.
Using multivariate logistic regression, we identified two factors that distinguished patients who exhibited cross-ineffectiveness to CGRP mAbs from those who responded to at least one CGRP mAb. We identified interictal allodynia (assessed using the ASC-12) and helplessness (assessed using the Pain Catastrophizing Scale) as independent factors associated with poor response to CGRP mAbs. In a meta-analysis focused on predictors of response to CGRP mAbs, interictal allodynia had already been identified as a negative predictor of response to CGRP mAbs [25]. Like its detrimental effect on the triptan response [26], the negative influence of allodynia on the response to CGRP mAbs is likely driven by the central sensitization that it clinically reflects. The negative predictive effect of psychiatric comorbidities—and particularly depressive comorbidities—on the response to CGRP mAbs is well established [25], but few studies have evaluated personality traits. Prior to our study, Silvestro et al. found that catastrophizing behavior was a negative predictor of response to erenumab, with an association between the total PCS score and erenumab failure, without association with individual dimensions of catastrophizing (helplessness, rumination, magnification) [27]. Similarly, Bottiroli et al., using the structured clinical interview for DSM-5, demonstrated that a Cluster C personality profile (avoidant, dependent) was also a negative predictor of response to erenumab [28]. In patients with severe migraine, identifying personality profiles that influence treatment response is essential for developing behavioral programs [29] that should be combined with pharmacological treatment and are best implemented as part of regular outpatient care, as required by the intravenous administration of eptinezumab.
The main strength of our study lies in the analysis of prospectively collected data within the FHU InovPain registry. Furthermore, this is the first study to evaluate switching between CGRP mAbs regardless of prior response to the initial CGRP mAb, and it provides data to discuss cross-response and the potential class effect of CGRP mAbs in real-world settings.
Our study has several limitations inherent to real-world studies that must be considered when interpreting the results.
First, the uncontrolled real-world design precludes exclusion of placebo and nocebo effects, which may have differed according to treatment context and route of administration [30]. Thus, a placebo effect may have been enhanced by inclusion in a selective compassionate use program such as the one involving erenumab, by the use of an expensive treatment for which the patient bore the financial burden, as was the case for fremanezumab and galcanezumab, or by the use of a CGRP mAb administered intravenously, such as eptinezumab. In contrast, a nocebo effect could have been induced in anxious patients who had previously responded to a CGRP mAb and who may have feared losing the beneficial effect when using another CGRP mAb.
The small size of certain subgroups is also a limitation. This is particularly true for the subgroup of patients previously treated with fremanezumab (n = 26), for whom the response rate to eptinezumab should be interpreted with caution.
Another limitation relates to the definition of cross-effectiveness and cross-ineffectiveness to CGRP mAbs. In our observational study, cross-effectiveness and cross-ineffectiveness were related to the specific CGRP mAbs to which they had been exposed, and not necessarily to all available CGRP mAbs. For patients who have used only two or three anti-CGRP monoclonal antibodies, there is no way to confirm whether they would have retained their cross-effectiveness or cross-ineffectiveness status if they had used a third or fourth anti-CGRP mAbs. Subgroup analysis based on the number of CGRP mAbs used showed a significantly different distribution of response statuses. However, the proportion of patients with no cross-response appeared to be similar regardless of the number of CGRP mAbs used by the patients.
Moreover, due to the sequential use of CGRP mAbs, the pre-treatment clinical status was not identical depending on the CGRP mAb used. However, each CGRP mAb treatment was initiated only in patients with at least 8 MMDs.
The multivariate analysis included in this study focused on data collected immediately prior to the initiation of eptinezumab, as these were the most extensive data available, particularly regarding patients’ personality traits. Consequently, our study could only identify factors associated with the response to CGRP mAbs rather than true predictive factors because, strictly speaking, prediction of the response to CGRP mAbs would require the analysis of data collected before exposure to the first CGRP mAb.
It is also important to note that our study focused solely on the 100 mg dose of eptinezumab because, due to financial constraints, escalation to the 300 mg dose is not permitted in our center when no clinically sufficient response is observed after the first infusion. This limitation may have led to an underestimation of the effectiveness of eptinezumab in our cohort. Finally, it should be noted that our results were obtained in migraine patients who had failed non-specific preventive treatments, so these findings may not be generalizable to patients receiving CGRP mAbs as first-line preventive therapy, as permitted by their approval.
Conclusions
This French real-world study provides several important findings. First, it confirms the benefit of switching to eptinezumab, which induced a response in nearly one-third of patients who had previously failed subcutaneous CGRP mAbs. Second, by evaluating switching from CGRP mAbs irrespective of response to the initial CGRP mAb, the study supports the existence of a class effect, with either cross-effectiveness or cross-ineffectiveness observed in most patients, although nearly one-quarter of them did not show any cross-response across the different CGRP mAbs. Third, by investigating factors associated with response, it appears that non-response to CGRP mAbs may be related to the presence of central sensitization, as evidenced by allodynia, but also to patient-specific psychological factors such as catastrophizing. Overall, these findings may help refine therapeutic strategies for patients with severe migraine and further highlight the value of real-world clinical research.
Author contributions
SFM and MLM conceived the study. TJ, SFM and MLM collected the data. RF performed statistical analysis. SFM and MLM drafted the manuscript. All authors revised and approved the final manuscript.
Funding
This study was supported by FHU InovPain.
Data availability
The materials used for this work are available from the corresponding author upon reasonable request.
Declarations
Ethics approval and consent to participate
This observational study did not require approval from an ethics committee. It was conducted as part of the FHU InovPain registry clinical research program, which received authorization from the French National Commission on Informatics and Liberty (Commission Nationale de l’Informatique et des Libertés [CNIL]; CIL register no. 278, dated 11/09/2017). Patients were included only after providing written informed consent allowing their data recorded in the FHU InovPain registry database to be used for clinical research purposes.
Consent for publication
Not applicable.
Competing interests
SFM has no conflict of interest. RF had no conflict of interest. TJ had no conflict of interest. MLM received personal fees for consultancy activities from: Abbvie/Allergan, Amgen, Eli Lilly, IPSEN, Lundbeck, Medtronic, Novartis, Orion Pharma, Perfood, Pfizer, Reckitt-Benckiser, Savia, Bioelectronics, Teva, UPSA. MLM is also an Associate Editor of the journal and did not take part in, any review or editorial decision-making process related to this paper.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Haghdoost F, Puledda F, Garcia-Azorin D, Huessler EM, Messina R, Pozo-Rosich P (2023) Evaluating the efficacy of CGRP mAbs and gepants for the preventive treatment of migraine: a systematic review and network meta-analysis of phase 3 randomised controlled trials. Cephalalgia 43(4):3331024231159366. 10.1177/03331024231159366. Erratum in: Cephalalgia. 2024;44(6):3331024241255216. https://doi.org/10.1177/03331024241255216. PMID: 36855951 [DOI] [PubMed]
- 2.Messina R, Huessler EM, Puledda F, Haghdoost F, Lebedeva ER, Diener HC (2023) Safety and tolerability of monoclonal antibodies targeting the CGRP pathway and gepants in migraine prevention: a systematic review and network meta-analysis. Cephalalgia 43(3):3331024231152169. 10.1177/03331024231152169. PMID: 36786548 [DOI] [PubMed]
- 3.Al-Hassany L, Goadsby PJ, Danser AHJ, MaassenVanDenBrink A (2022) Calcitonin gene-related peptide-targeting drugs for migraine: how pharmacology might inform treatment decisions. Lancet Neurol 21(3):284–294. 10.1016/S1474-4422(21)00409-9. Epub 2022 Jan 31. Erratum in: Lancet Neurol. 2022;21(10):e9. https://doi.org/10.1016/S1474-4422(22)00376-3. PMID: 35093196 [DOI] [PubMed]
- 4.Waliszewska-Prosół M, Vuralli D, Martelletti P (2023) What to do with non-responders to CGRP(r) monoclonal antibodies: switch to another or move to gepants? J Headache Pain 24(1):163. 10.1186/s10194-023-01698-8. PMID: 38053061; PMCID: PMC10696682 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Romozzi M, Munafò A, Burgalassi A, De Cesaris F, Vigani G, Altamura C, Rivi V, Guerzoni S, Calabresi P, Raffaelli B, Iannone LF (2025) Pharmacological differences and switching among anti-CGRP monoclonal antibodies: A narrative review. Headache 65(2):342–352. 10.1111/head.14903. Epub 2025 Jan 17. PMID: 39825578 [DOI] [PubMed] [Google Scholar]
- 6.Vikelis M, Rikos D, Argyriou AA, Dermitzakis EV, Andreou AP, Russo A (2025) Switching between anti-CGRP monoclonal antibodies in migraine prophylaxis. Expert Rev Neurother 1–16. 10.1080/14737175.2025.2461766. Epub ahead of print. PMID: 39884968 [DOI] [PubMed]
- 7.Moisset X, Demarquay G, de Gaalon S, Roos C, Donnet A, Giraud P, Guégan-Massardier E, Lucas C, Mawet J, Valade D, Corand V, Gollion C, Moreau N, Grangeon L, Lantéri-Minet M, Ducros A (2024) Migraine treatment: Position paper of the French Headache Society. Rev Neurol (Paris) 180(10):1087–1099 Epub 2024 Oct 15. PMID: 39406556 [DOI] [PubMed] [Google Scholar]
- 8.Lanteri-Minet M, Fabre R, Martin C, Pradat K, Alchaar A, Bozzolo E, Duchene ML, Van Obberghen EK, Donnet A, Fontaine D (2023) One-year prospective real-world assessment of effectiveness and safety of erenumab in migraine prevention: results of the French FHU INOVPAIN registry study. J Headache Pain 24(1):152. 10.1186/s10194-023-01680-4. PMID: 37940860; PMCID: PMC10633983 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Lipton RB, Bigal ME, Ashina S, Burstein R, Silberstein S, Reed ML, Serrano D, Stewart WF, American Migraine Prevalence Prevention Advisory Group (2008) Cutaneous allodynia in the migraine population. Ann Neurol 63(2):148–158. 10.1002/ana.21211. PMID: 18059010; PMCID: PMC2729495 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Kosinski M, Bayliss MS, Bjorner JB, Ware JE Jr, Garber WH, Batenhorst A, Cady R, Dahlöf CG, Dowson A, Tepper S (2003) A six-item short-form survey for measuring headache impact: the HIT-6. Qual Life Res 12(8):963 – 74. 10.1023/a:1026119331193. PMID: 14651415 [DOI] [PubMed]
- 11.Zigmond AS, Snaith RP (1983) The hospital anxiety and depression scale. Acta Psychiatr Scand 67(6):361 – 70. 10.1111/j.1600-0447.1983.tb09716.x. PMID: 6880820 [DOI] [PubMed]
- 12.Stewart WF, Lipton RB, Kolodner KB, Sawyer J, Lee C, Liberman JN (2000) Validity of the Migraine Disability Assessment (MIDAS) score in comparison to a diary-based measure in a population sample of migraine sufferers. Pain 88(1):41–52. 10.1016/S0304-3959(00)00305-5. PMID: 11098098 [DOI] [PubMed]
- 13.Buse DC, Rupnow MF, Lipton RB (2009) Assessing and managing all aspects of migraine: migraine attacks, migraine-related functional impairment, common comorbidities, and quality of life. Mayo Clin Proc 84(5):422 – 35. 10.1016/S0025-6196(11)60561-2. PMID: 19411439; PMCID: PMC2676125 [DOI] [PMC free article] [PubMed]
- 14.Rabin R, de Charro F (2001) EQ-5D: a measure of health status from the EuroQol Group. Ann Med 33(5):337 – 43. 10.3109/07853890109002087. PMID: 11491192 [DOI] [PubMed]
- 15.Lundqvist C, Gossop M, Russell MB, Straand J, Kristoffersen ES (2019) Severity of Analgesic Dependence and Medication-overuse Headache. J Addict Med 13(5):346–353 PMID: 30724760; PMCID: PMC6791494 [DOI] [PMC free article] [PubMed]
- 16.Williams DA, Robinson ME, Geisser ME (1994) Pain beliefs: assessment and utility. Pain 59(1):71–78. 10.1016/0304-3959(94)90049-3. PMID: 7854805 [DOI] [PubMed]
- 17.Sullivan MJL, Bishop SR, Pivik J (1995) The Pain Catastrophizing Scale: Development and validation. Psychol Assess 7(4):524–532. 10.1037/1040-3590.7.4.524 [DOI] [Google Scholar]
- 18.Barbanti P, Aurilia C, Egeo G, Doretti A, d’Onofrio F, Scatena P, Rinalduzzi S, Vinciguerra L, Sansone M, Vecchio R, Drago V, Viticchi G, Bartolini M, Ranieri A, Bandettini di Poggio M, Baldisseri F, Mascarella D, Brusaferri F, Caputi L, Messina S, Autunno M, Valenza A, Orlando B, Distefano M, Borrello L, Pistoia F, Camarda C, Saporito G, Querzola G, Torelli P, Salerno A, Gragnani F, Petolicchio B, Carnevale A, Messina R, Filippi M, Tavani S, Fiorentini G, Bonassi S, Cevoli S, Mannocci A (2025) Italian Migraine Registry (I-GRAINE) study group. A 24-week prospective, multicenter, real-world study on eptinezumab’s effectiveness and safety in migraine prevention (EMBRACE II). J Neurol 272(6):382. 10.1007/s00415-025-13095-z. PMID: 40332560; PMCID: PMC12058817 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Iannone LF, Piella EM, Montisano DA, Fasano C, Sebastianelli G, Coppola G, Ferrandi D, Lanni C, Prudenzano MP, de Tommaso M, Merlo P, De Cesaris F, Chiarugi A, Munafò A, Pistoia F, Ornello R, Doretti A, Grazzi L, Lo Castro F, De Icco R, Vaghi G, Avino G, Romozzi M, Calabresi P, Battistini S, Rufa A, Albanese M, Trimboli M, Carlucci G, Silvestro M, Russo A, Rainero I, Valente MR, Fofi L, Marcosano M, Geppetti P, Altamura C, Vernieri F, Tassorelli C, Sacco S, Guerzoni S (2026) Italian Headache Registry (RICe) Study Group. Levels of migraine controls following International Headache Society (IHS) recommendations with eptinezumab: Effectiveness and tolerability in a 24-week, prospective multicenter study (the TACHIS study). Cephalalgia 46(2):3331024251414659. 10.1177/03331024251414659. Epub 2026 Feb 12. PMID: 41684106 [DOI] [PubMed] [Google Scholar]
- 20.Dermitzakis EV, Argyriou AA, Chondrogianni M, Foska A, Rikos D, Athanasopoulos D, Tsironis C, Soldatos P, Koutsokera M, Tsitsaras N, Litsardopoulos P, Mavraki E, Vikelis M, Greek Research Alliance for the Study of Headache and Pain (GRASP) Study Group (2026) Eptinezumab to prevent difficult-to-treat migraine: prospective, six-month, real-world multicenter evidence from the GRASP study group. J Headache Pain 27(1):48. 10.1186/s10194-026-02275-5. PMID: 41578184; PMCID: PMC12918022 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Andreou AP, Hill B, Al-Rawi R, Murphy M, Soares I, Briscoe J, Kilner R, Lambru G (2025) A prospective real-world analysis of intravenous eptinezumab in migraine management: the first UK experience. J Headache Pain 26(1):235. 10.1186/s10194-025-02180-3. PMID: 41168777; PMCID: PMC12573807 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Suliman R, Santos V, Al Qaisi I, Aldaher B, Al Fardan A, Al Barrawy H, Bader Y, Supena JL, Alejandro K, Alsaadi T (2024) Effectiveness of Switching CGRP Monoclonal Antibodies in Non-Responder Patients in the UAE: A Retrospective Study. Neurol Int 16(1):274–288. 10.3390/neurolint16010019. PMID: 38392960; PMCID: PMC10893254 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Scheffler A, Wenzel P, Bendig M, Gendolla A, Basten J, Kleinschnitz C, Nsaka M, Lindner D, Naegel S, Burow P, Fleischmann R, Holle D (2024) Effectiveness and tolerability of eptinezumab in treating patients with migraine resistant to conventional preventive medications and CGRP (receptor) antibodies: a multicentre retrospective real-world analysis from Germany. J Headache Pain 25(1):79. 10.1186/s10194-024-01788-1. PMID: 38755541; PMCID: PMC11097519 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Hong JB, Israel-Willner H, Peikert A, Schanbacher P, Tozzi V, Köchling M, Reuter U, Raffaelli B, NTD Study Group (2024) Therapeutic patterns and migraine disease burden in switchers of CGRP-targeted monoclonal antibodies - insights from the German NeuroTransData registry. J Headache Pain 25(1):90. 10.1186/s10194-024-01790-7. PMID: 38825722; PMCID: PMC11145812 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Hong JB, Lange KS, Overeem LH, Triller P, Raffaelli B, Reuter U (2023) A Scoping Review and Meta-Analysis of Anti-CGRP Monoclonal Antibodies: Predicting Response. Pharmaceuticals (Basel) 16(7):934. 10.3390/ph16070934. PMID: 37513846; PMCID: PMC10385131 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Burstein R, Collins B, Jakubowski M (2004) Defeating migraine pain with triptans: a race against the development of cutaneous allodynia. Ann Neurol 55(1):19–26. 10.1002/ana.10786. PMID: 14705108 [DOI] [PubMed]
- 27.Silvestro M, Tessitore A, Scotto di Clemente F, Battista G, Tedeschi G, Russo A (2021) Refractory migraine profile in CGRP-monoclonal antibodies scenario. Acta Neurol Scand 144(3):325–333. 10.1111/ane.13472. Epub 2021 May 21. PMID: 34019304; PMCID: PMC8453754 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Bottiroli S, De Icco R, Vaghi G, Pazzi S, Guaschino E, Allena M, Ghiotto N, Martinelli D, Tassorelli C, Sances G (2021) Psychological predictors of negative treatment outcome with Erenumab in chronic migraine: data from an open label long-term prospective study. J Headache Pain 22(1):114. 10.1186/s10194-021-01333-4. Erratum in: J Headache Pain. 2021;22(1):131. https://doi.org/10.1186/s10194-021-01342-3. PMID: 34600468; PMCID: PMC8487575 [DOI] [PMC free article] [PubMed]
- 29.Zamir O, Yarns BC, Lagman-Bartolome AM, Jobanputra L, Lawler V, Lay C (2023) Understanding the gaps in headache and migraine treatment with psychological and behavioral interventions: A narrative review. Headache 63(8):1031–1039. 10.1111/head.14624. Epub 2023 Aug 28. PMID: 37638484 [DOI] [PubMed] [Google Scholar]
- 30.Basedau H, Ornello R, Matteis E, Davaasuren B, Kadyrova B, Vuralli D, Bozhenko M, Azizova I, Bitsadze N, Eralieva E, Ashina M, Mitsikostas D, Puledda F (2023) Placebo and nocebo in the treatment of migraine: How much does real world effectiveness depend on contextual effects? Cephalalgia 43(12):3331024231218392. 10.1177/03331024231218392. PMID: 38041833 [DOI] [PubMed]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The materials used for this work are available from the corresponding author upon reasonable request.