Eptinezumab in treatment-resistant chronic migraine: Four-year real-world effectiveness and treatment persistence

Abstract

Aim

To evaluate 48-month real-world outcomes of eptinezumab users in chronic migraine (CM), including longitudinal effectiveness, treatment persistence, reasons for discontinuation, and post-discontinuation anti-calcitonin gene-related peptide (CGRP) treatment patterns.

Methods

We conducted a retrospective cohort study of 112 CM subjects initiating eptinezumab at a tertiary headache center between 2019 and 2021. Monthly headache days (MHD), migraine disability assessment (MIDAS), and average pain intensity (API) were extracted at baseline and every 6 months for up to 48 months using a predefined outcome form at each visit. Longitudinal changes were analyzed using generalized estimating equations (GEE) with an unstructured correlation matrix, adjusting for age, sex, and body mass index (BMI). Analyses were performed in four predefined analytic cohorts: all initiators, on-treatment, continuation, and discontinuation. Estimated marginal means (EMMs) quantified adjusted mean changes from baseline at each timepoint. No imputation or censoring for treatment discontinuation was applied. Treatment persistence, reasons for discontinuation, factors associated with discontinuation, and post-discontinuation anti-CGRP treatment patterns were summarized descriptively.

Results

Among 112 patients (mean age 50.2 ± 13.7; 77.7% female; baseline MHD median 30 [IQR 25–30]), most had failed prior anti-CGRP therapies (90.2% ≥ 1 failure; 70.6% ≥ 2 failures). In all-initiators, significant reductions in MHD were observed at 6, 18–42, corresponding to model-estimated reductions of 1.5–1.9 MHD. MIDAS increased at 6 months, corresponding to a 44-point increase, but declined over time. API improved significantly at 24, 36 and 48 months, corresponding to a 0.4–0.6-point reduction. Treatment persistence declined to 55.4%, 39.2%, 35.7%, and 29.5% at 1, 2, 3, and 4 years, respectively; the median time to discontinuation was 8 months. Ineffectiveness (34.2%), loss of effectiveness (19.0%), and insurance-related barriers (16.5%) were the most common reasons for discontinuation. After discontinuation, 54.4% initiated another anti-CGRP therapy, 26.6% stopped anti-CGRP therapy entirely, and 20.3% were lost to follow-up.

Conclusions

In this treatment-resistant CM cohort, eptinezumab produced modest but durable benefit, particularly among patients who continued treatment. Persistence declined substantially over time, and post-discontinuation care followed diverse and often complex pathways. Regular reassessment and timely adjustment of preventive strategies remain essential as clinical responses evolve. Larger prospective studies will help clarify long-term effectiveness and treatment trajectories.

Graphical abstract

This is a visual representation of the abstract.

Keywords

chronic migraine eptinezumab CGRP monoclonal antibody treatment persistence longitudinal follow-up

Introduction

Eptinezumab has been approved for the preventive treatment of chronic migraine (CM). Nevertheless, eptinezumab access and usage duration vary across healthcare systems, and evidence from long-term real-world practice is limited. Randomized trials have established short-term efficacy: PROMISE-2 demonstrated significant reductions in monthly migraine days in patients with CM,¹ and DELIVER confirmed benefit in patients with 2–4 prior preventive failures.² Open-label extensions have demonstrated eptinezumab's safety and sustained improvements for up to 2 years.³ Yet these data arose from structured environments with high retention, limiting applicability to headache clinic populations with high treatment refractoriness. For clinicians managing resistant CM, where discontinuation, partial response, and sequencing decisions drive clinical care, the central unanswered question is how eptinezumab behaves across multi-year, routine-practice trajectories.

Existing real-world studies provide important but short-horizon insights. Scheffler et al., in a multicenter analysis, reported a median 4-day reduction in MHD after 3 months in patients who had failed prior calcitonin gene-related peptide (CGRP)-pathway monoclonal antibodies (mAbs).⁴ Similarly, retrospective switch cohorts of 24 weeks reported 93% treatment persistence and 29.7% ≥ 30% response, but no information on post-discontinuation care.⁵ A prospective UK cohort demonstrated clinically meaningful improvements (43% ≥ 50% reduction) over 6 months with intravenous eptinezumab in a treatment-resistant population.⁶ Other prospective evaluations, including EMBRACE II,⁷ focus on early effectiveness and tolerability but do not follow patients long enough to address discontinuation behavior, relapse patterns, or switching decisions. Methodologically speaking, limiting analyses to treatment continuers can introduce survivor bias and overestimate benefit,^8,9 underscoring the need for real-world evidence aligned with intention-to-treat principles and a complete accounting of attrition.^10,11

A growing body of real-world data shows that treatment persistence with migraine preventives declines substantially over time, reinforcing the need to evaluate long-term treatment continuity. In a national US cohort of 66,576 patients, Charleston et al. (2023) reported that subcutaneous anti-CGRP mAbs had 32–58% higher discontinuation hazards than eptinezumab, while eptinezumab persistence was statistically comparable to onabotulinumtoxinA.¹² In a Canadian national claims analysis, only 24.9% of prophylactic-naïve patients remained persistent at 24 months, with 27% switching and 50% fully discontinuing therapy.¹³ Among CM patients treated with anti-CGRP mAbs, de Dios et al. reported a median persistence of 267 days and a steep drop-off in one-year persistence across successive mAb trials (66.7% → 49.8% → 37.2%).¹⁴ Long-term data from Spain showed that after three years, only 37% remained on their initial mAb, though class-level persistence was 66% when allowing switching, with discontinuation driven primarily by inefficacy (12%) and adverse events (5%).¹⁵ Together, these findings illustrate that migraine preventive persistence is generally low and strongly influenced by switching opportunities and treatment refractoriness, reflecting the need to understand long-term, real-world persistence specifically for eptinezumab.

To better understand eptinezumab's long-term effectiveness and persistence, we conducted a retrospective longitudinal cohort study at a single tertiary headache center, with provider-verified outcomes collected over 48 months across four predefined analytic cohorts (all-initiators, on-treatment, continuation, and discontinuation). This infrastructure enables evaluation of clinical outcomes, dose-escalation patterns, and full treatment trajectories in a population characterized by high disease burden and extensive prior treatment failure. Our objective was to define the long-term real-world course of eptinezumab in CM. Specifically, we aimed to: (1) quantify changes in eptinezumab doses, monthly headache days (MHD), disability (MIDAS), and average pain intensity (API) across four years; (2) characterize treatment persistence and reasons for discontinuation; (3) describe dose escalation and concomitant preventive use; and (4) map post-discontinuation pathways, including treatment gaps and subsequent anti-CGRP therapy use.

Methods

Study design

This single-center, retrospective, longitudinal cohort study was conducted at the Jefferson Headache Center (JHC), Philadelphia, PA, USA. The study was approved by the Thomas Jefferson University Institutional Review Board (No. 21E.1141) with a waiver of informed consent and was conducted in accordance with the Declaration of Helsinki and applicable regulatory requirements. This study is reported in accordance with the STROBE guidelines for retrospective, longitudinal cohort.

Patient selection

Eligible patients were adults (≥18 years) with CM diagnosed according to the International Classification of Headache Disorders, 3rd edition (ICHD-3) criteria who initiated their first eptinezumab treatment at the JHC between January 1, 2019, and June 30, 2021. Patients were identified through the infusion schedule. No exclusion criteria were applied; all treated patients were enrolled by convenience and analyzed. Of the 112 patients included, 110 initiated treatments at 100 mg, while 2 received 300 mg as their initial dose. In our region (Pennsylvania, USA), access to eptinezumab is governed by insurance payer criteria, which typically require a diagnosis of CM and documented failure or intolerance of at least two to three preventive therapies, and some may require CGRP mAb failure.

Data collection and outcome measures

Clinical data were obtained through a structured, manual, retrospective review of the electronic medical record, incorporating both patient-reported outcomes and provider-documented clinical information, and were securely stored in a HIPAA-compliant, web-based REDCap electronic data capture system hosted at Thomas Jefferson University.¹⁶ Baseline was defined as the date of the first eptinezumab infusion at JHC. Patient demographics (age, sex, BMI, race/ethnicity, working status), comorbidities, migraine characteristics (age of headache onset, migraine duration, presence of daily headache), prior preventive use, baseline preventive and acute medication use were extracted. At baseline and each follow-up encounter, patients completed a standardized outcome form (data verified by the treating provider during each visit) that captured monthly headache days (MHD; 0–30 days), average pain intensity (API; 0–10 scale), and migraine disability assessment (MIDAS) scores. Monthly migraine days were not routinely documented in the electronic medical record.

Longitudinal data were collected at baseline and approximately every 6 months through 48 months from all patients. For each predefined timepoint, the visit closest to that interval (±1 month) was used, with absent data coded as missing and no imputation applied. The primary outcomes were MHD, API, MIDAS, and ≥30% responder rate (reduction in MHD from baseline). Secondary outcomes included treatment persistence (time from first infusion to provider-documented discontinuation), dose escalation from 100 mg to 300 mg, concomitant use of onabotulinumtoxinA, reasons for treatment discontinuation (ineffectiveness, loss of effectiveness, insurance barriers, adverse events), and post-discontinuation anti-CGRP treatment patterns (initiation of another anti-CGRP therapy, complete cessation of anti-CGRP therapy, or loss to follow-up).

Statistical analysis

All statistical analyses were performed using IBM SPSS Statistics (version 28, IBM Corp., Armonk, NY). Descriptive statistics were used to summarize baseline patient demographics, comorbidities, migraine characteristics, and medication use. Longitudinal changes in MHD, MIDAS, and API were analyzed using generalized estimating equations (GEE) to account for within-subject correlation across repeated measures (unstructured correlation matrix). Timepoint was modeled as a categorical variable, with baseline specified as the reference, allowing estimation of outcome changes at each follow-up interval relative to baseline. Models were adjusted for age, sex, and BMI. To characterize outcome trajectories under these conditions while acknowledging survivor bias, longitudinal analyses were conducted across four predefined cohorts: an all-initiator cohort including all available data following treatment initiation; an on-treatment cohort restricted to periods of active treatment; a continuation cohort including patients who remained on therapy throughout the 48-month follow-up; and a discontinued cohort including patients who stopped therapy at any point during follow-up, with outcomes modeled across their entire observation period regardless of therapy cessation or subsequent switching to alternative medications. While this approach does not eliminate bias related to informative discontinuation, analyzing multiple exposure-defined cohorts allows for explicit visualization of how attrition and treatment persistence influence observed outcomes, avoiding overreliance on on-treatment-only analyses that may overestimate long-term effectiveness. Estimated marginal means (EMMs) and corresponding 95% confidence intervals (95% CI) were calculated for all analyses using the same GEE modeling approach, without multiplicity adjustment. Treatment persistence was illustrated descriptively using Kaplan–Meier survival curves based on observed time to treatment discontinuation. No inferential comparisons were conducted. Loss to follow-up was treated as treatment discontinuation. Factors associated with treatment discontinuation due to lack of effectiveness or adverse events were examined using multivariable logistic regression. Based on clinical relevance and the limited number of discontinuation events, the final model included age, sex, BMI, baseline MHD, and the number of prior anti-CGRP therapy failures as covariates. Age and BMI were entered as mean-centered continuous variables and scaled per 5-unit increase (5 years for age and 5 kg/m² for BMI) to improve interpretability. Two-sided p-values <0.05 were considered statistically significant.

Results

Study population and baseline characteristics

A total of 112 patients initiated eptinezumab during the study period. The cohort was predominantly female (77.7%) with a mean age of 50.2 ± 13.7 and a median BMI of 28.3 (IQR 23.8–32.7). At baseline, patients had a median of 30 MHD (IQR 25–30), a median MIDAS of 90 (IQR 30–172.5), and a mean API of 6.3 ± 1.9; 77 patients (68.8%) reported daily headache. Most patients (90.2%) had failed at least one prior anti-CGRP preventive, and 70.6% had failed at least two (Tables 1 and 2). Continuation and discontinuation groups were grossly similar in demographics, comorbidities, headache characteristics, and medication use.

Table 1.

Patient demographic and clinical characteristics.

	Total cohort n = 112	Continuation group n = 33	Discontinuation group n = 79
Patient characteristics
Age	50.2 ± 13.7; 51 (40.3–61.8)	52.7 ± 13.7; 56 (40.5–62)	49.2 ± 13.7; 49 (40–60)
Female	87 (77.7)	27 (81.8)	60 (75.9)
BMI	29.4 ± 7.5; 28.3 (23.8–32.7)	29.8 ± 9.3; 27.3 (23.7–33.2)	29.2 ± 6.6; 28.3 (23.9–32.3)
Race
Caucasian	92 (82.1)	27 (81.8)	65 (82.3)
African American	12 (10.7)	4 (12.1)	8 (10.1)
Hispanic/Latino	4 (3.6)	0 (0)	4 (5.1)
Asian	1 (0.9)	1 (3.0)	0 (0)
Unanswered	3 (2.7)	1 (3.0)	2 (2.5)
Working status
Working	41 (36.6)	11 (33.3)	30 (38.0)
Not working	22 (19.6)	7 (21.2)	15 (19.0)
Disability	49 (43.8)	15 (45.5)	34 (43.0)
Comorbidities
Psychiatric conditions
Anxiety	72 (64.3)	20 (60.6)	52 (65.8)
Depression	63 (56.3)	17 (51.5)	46 (58.2)
Bipolar disorder	3 (2.7)	1 (3.0)	2 (2.5)
PTSD	1 (0.9)	0 (0)	1 (1.3)
Chronic pain syndromes
Fibromyalgia	16 (14.3)	4 (12.1)	12 (15.2)
CRPS	2 (1.8)	1 (3.0)	1 (1.3)
Endocrine conditions
DM	11 (9.8)	4 (12.1)	7 (8.9)
Sleep disorders
Insomnia	57 (20.9)	19 (57.6)	38 (48.1)
OSA	24 (21.4)	5 (15.2)	19 (24.1)
Migraine characteristics
Age of headache onset	26.4 ± 17.2; 20 (13–38.3)	29.0 ± 16.5; 28.5 (13.8–46.3)	25.4 ± 17.5; 19 (12.8–37.5)
Migraine duration, years	14.2 ± 12.2; 9 (4–22)	16.7 ± 12.4; 14 (4–24)	13.3 ± 12.2; 8.5 (4–21.8)
Migraine with aura	33 (29.5)	11 (33.3)	22 (27.8)
Baseline MHD	26.4 ± 6.7; 30 (25–30)	24.8 ± 7.5; 30 (19–30)	27.1 ± 6.2; 30 (28–30)
Baseline daily headache	77 (68.8)	19 (57.6)	58 (73.4)
Baseline MIDAS	116.8 ± 109.8; 90 (30–172.5)	128.5 ± 116.5; 90 (47.5–192.5)	112.1 ± 107.9; 90 (25.5–175.0)
Baseline API	6.3 ± 1.9; 6.3 (5–8)	6.2 ± 1.5; 6 (5.3–7.0)	6.3 ± 2.0; 6.5 (4.5–8.0)

Continuous data are presented as mean ± SD; median (IQR); frequency data are presented as count (%).

BMI, body mass index; PTSD, post-traumatic stress disorder; CRPS, complex regional pain syndrome; DM, diabetes mellitus; OSA, obstructive sleep apnea; MHD, monthly headache days; MIDAS, migraine disability assessment; API, average pain intensity; SD, standard deviation; IQR, interquartile range.

Table 2.

Summary of medication use.

	Total cohort n = 112	Continuation group n = 33	Discontinuation group n = 79
Prior preventive use
No. of prior oral preventives	6.7 ± 3.5; 6 (3.3–9)	6.8 ± 3.9; 6 (3.5–10)	6.7 ± 3.4; 6 (3–9)
Antiepileptics	104 (92.9)	31 (93.9)	73 (92.4)
Antidepressants	103 (92.0)	28 (84.8)	75 (94.9)
Beta-blockers	70 (62.5)	18 (54.5)	52 (65.8)
ACEI/ARB	64 (57.1)	15 (45.5)	49 (62.0)
Calcium channel blockers	23 (20.5)	7 (0.2)	16 (20.3)
OnabotulinumtoxinA	91 (81.3)	26 (78.8)	65 (82.3)
No. of prior anti-CGRP therapy^a
0	2 (1.8)	1 (3.0)	1 (1.3)
1	20 (17.9)	5 (15.2)	15 (19.0)
2	57 (50.9)	18 (54.5)	39 (49.4)
3	32 (28.6)	9 (27.3)	23 (29.1)
4	1 (0.9)	0 (0)	1 (1.3)
No. of prior anti-CGRP therapy failures (lack of effectiveness and adverse events)
0	11 (9.8%)	3 (9.1)	8 (10.1)
1	22 (19.6%)	7 (21.2)	15 (19.0)
2	52 (46.4%)	15 (45.5)	37 (46.8)
3	26 (23.2%)	8 (24.2)	18 (22.8)
4	1 (0.89%)	0 (0)	1 (1.3)
Baseline preventive medication use
No. of baseline oral preventives	1.9 ± 1.2; 2 (1–3)	2.1 ± 1.2; 2 (1–3)	1.9 ± 1.2; 2 (1–3)
Antiepileptics	58 (51.8)	20 (60.6)	38 (48.1)
Antidepressants	66 (58.9)	20 (60.6)	46 (58.2)
Beta-blockers	32 (28.6)	9 (27.3)	23 (29.1)
ACEI/ARB	27 (24.1)	7 (21.2)	20 (25.3)
Calcium channel blockers	5 (4.5)	4 (12.1)	1 (1.3)
OnabotulinumtoxinA	45 (40.2)	16 (48.5)	29 (36.7)
Concurrent onabotulinumtoxinA throughout eptinezumab using period
Concurrent users	57 (50.9)	24 (72.7)	33 (41.8)
Overlap rate	87.8 ± 26.9; 100 (99.1–100)	88.5 ± 24.9; 100 (89.6–100)	87.3 ± 28.6; 100 (100–100)
Baseline acute medication use
NSAIDs	78 (69.6)	26 (0.79)	52 (65.8)
Gepants	60 (53.6)	19 (57.6)	41 (51.9)
Triptans	44 (35.5)	14 (42.4)	30 (38.0)
DHE	39 (31.5)	11 (33.3)	28 (35.4)
Opioids	18 (16.1)	8 (24.2)	10 (12.7)
Ditans	16 (14.3)	3 (9.1)	13 (16.5)
Acetaminophen	11 (9.8)	2 (6.1)	9 (11.4)
Other^b	43 (34.7)	17 (51.5)	26 (32.9)

Continuous data are presented as mean ± SD; median (IQR); frequency data are presented as count (%).

The patient with 4 prior therapies had been treated sequentially with erenumab, fremanezumab, galcanezumab, and rimegepant, whereas all other patients (0–3 prior therapies) had received only subcutaneous anti-CGRP mAbs. ^bOther include DHE/lidocaine nasal spray, ketamine nasal spray, lidocaine nasal spray, oxytocin nasal spray.

ACEI, angiotensin-converting enzyme inhibitors; ARB, angiotensin II receptor blockers; CGRP, calcitonin gene-related peptide; NSAID, nonsteroid anti-inflammatory drugs; DHE, dihydroergotamine; SD, standard deviation.

Effectiveness

In all-initiators, MHD decreased significantly from baseline at multiple time points (6 and 18–42 months), corresponding to a model-estimated absolute reduction of 1.5–1.9 MHD. Similar significant reductions were observed in the on-treatment group (1.9–3.5 MHD) and the continuation group (3.0–4.6 MHD), but not in the discontinuation group ( Table 3 ; Figure 1A, B; Supplementary Table 1 ).

Figure 1.

A-F. Change in clinical outcomes over 48 months after initiating eptinezumab. Panels A and B show changes in MHD, C and D show changes in MIDAS, and E and F show changes in API. Values are presented as estimated marginal means (EMMs); error bars are omitted for visual clarity. Symbols indicate patient groups: all-initiator cohort (square), on-treatment cohort (circle), continuation group (triangle), and discontinuation group (rhomboid). MHD, monthly headache days; MIDAS, Migraine Disability Assessment; API, average pain intensity; EMM, estimated marginal mean.

Table 3.

Back-transformed estimated marginal means of monthly headache days.

Month	0	6	12	18	24	30	36	42	48
All-initiator cohort (N = 112)
n	112	101	93	93	92	82	83	79	82
EMM (95% CI)	27.1 [25.5–28.9]	25.6 [23.8–27.6]*	25.5 [23.6–27.6]	25.4 [23.8–27.1]*	25.5 [23.7–27.4]*	25.3 [23.4–27.4]*	25.3 [23.4–27.3]*	25.5 [23.6–27.6]*	26.0 [24.1–28.0]
On-treatment cohort
N	112	82	63	50	46	44	40	36	33
n	112	76	55	48	44	41	36	33	31
EMM (95% CI)	25.2 [21.5–29.5]	22.9 [19.2–27.4]*	23.3 [19.4–27.9]*	23.4 [20.4–26.8]	23.7 [20.2–27.7]	22.8 [19.6–26.7]*	22.8 [19.5–26.6]*	21.7 [18.6–25.3]*	22.1 [19.0–25.7]
>30% reduction		11	10	8	8	9	7	9	8
Continuation group (N = 33)
n	33	30	31	31	32	31	30	31	31
EMM (95% CI)	27.0 [23.2–31.5]	23.3 [19.4–28.0]*	23.3 [19.4–28.1]*	24.4 [20.8–28.6]	24.0 [19.8–29.1]*	22.4 [18.8–26.7]*	24.0 [20.5–28.0]*	22.6 [19.0–26.8]*	22.7 [19.0–27.1]*
>30% reduction		6	8	6	8	9	7	9	8
Discontinuation group (N = 79)
On eptinezumab^a	79	49	30	17	13	11	7	3	0
n	79	71	62	62	60	51	53	48	51
EMM (95% CI)	25.5 [22.8–28.6]	24.6 [21.8–27.7]	25.0 [22.1–28.3]	24.3 [21.4–27.6]	25.1 [22.3–28.2]	25.4 [22.6–28.6]	24.3 [21.5–27.6]	25.6 [22.7–28.9]	25.9 [22.9–29.2]

Estimated marginal means were obtained from generalized estimating equation models with a log link (adjusted for age, sex, BMI). Model estimates were calculated on the log scale and exponentiated to yield marginal mean estimates and 95% confidence intervals on the original scale of monthly headache days. N denotes the total number of patients in each cohort. n denotes the number of patients contributing data at each time point. Missing values were not imputed. *indicates p < 0.05 versus baseline. MHD, monthly headache days; GEE, generalized estimating equation; EMM, estimated marginal mean; BMI, body mass index; CI, confidence interval.

Among patients receiving eptinezumab, the ≥30% MHD responder rate ranged from 14.5% to 25.8% over 48 months of follow-up ( Table 3 ).

MIDAS increased initially across all groups but declined over time; however, changes at individual timepoints did not reach statistical significance in either all-initiators or discontinuation group. In the on-treatment group, progressive improvement was observed during late follow-up, reaching statistical significance at 30, 36 and 48 months (score reduction of 34.5–59). A similar trend was observed in the continuation group, with a significant reduction at 48 months (51.5-point reduction) ( Table 4 ; Figure 1C, D; Supplementary Table 2 ).

Table 4.

Back-transformed estimated marginal means of migraine disability assessment score.

Month	0	6	12	18	24	30	36	42	48
All-initiator cohort (N = 112)
n	112	101	93	93	92	82	83	79	82
EMM (95% CI)	129.5 [101.9–164.5]	173.5 [131.0–229.9]*	137.2 [107.4–175.1]	136.6 [109.5–170.3]	123.9 [98.6–155.6]	121.4 [98.4–149.9]	119.8 [96.7–148.2]	120.2 [95.9–150.7]	100.2 [78.8–127.3]
On-treatment cohort
N	112	82	63	50	46	44	40	36	33
n	112	42	34	34	29	34	30	25	23
EMM (95% CI)	135.8 [108.4–170.3]	165.0 [128.7–211.4]	118.7 [80.0–176.0]	111.7 [74.0–168.7]	113.7 [84.2–153.8]	93.0 [62.2–139.0]*	101.3 [72.7–141.0]*	100.9 [74.6–136.4]	76.8 [45.6–129.2]*
Continuation group (N = 33)
n	33	30	31	31	32	31	30	31	31
EMM (95% CI)	118.9 [74.7–189.1]	133.2 [80.2–221.1]	119.5 [67.5–211.4]	78.1 [51.7–117.8]	93.4 [63.7–136.9]	89.4 [55.1–145.2]	79.1 [53.1–118.0]	76.5 [48.5–120.7]	67.4 [42.4–107.1]*
Discontinuation group (N = 79)
On eptinezumab^a	79	49	30	17	13	11	7	3	0
n	79	71	62	62	60	51	53	48	51
EMM (95% CI)	124.6 [94.3–164.7]	159.7 [116.8–218.4]*	132.8 [95.4–184.9]	132.8 [95.4–184.9]	116.4 [89.3–151.6]	125.7 [98.2–161.0]	114.0 [86.1–150.9]	118.5 [90.4–155.2]	108.0 [79.0–147.6]

Estimated marginal means were obtained from generalized estimating equation models with a log link (adjusted for age, sex, BMI). Model estimates were calculated on the log scale and exponentiated to yield marginal mean estimates and 95% confidence intervals on the original scale of monthly headache days. N denotes the total number of patients in each cohort. n denotes the number of patients contributing data at each time point. Missing values were not imputed. *indicates p < 0.05 versus baseline. MIDAS, migraine disability assessment; GEE, generalized estimating equation; EMM, estimated marginal mean; BMI, body mass index; CI, confidence interval.

In all-initiators, API showed significant reductions at 24, 36 and 48 months, corresponding to a reduction of 0.4–0.6 points. The continuation group showed a more consistent reduction at 12 and 24–48 months, with decreases of 0.6–1.2 points. A similar trend was observed in the on-treatment group, reaching significance at 24–36 and 48 months (0.5–1.0-point reduction). Discontinuers did not demonstrate meaningful or consistent change ( Table 5 ; Figure 1E, F; Supplementary Table 3 ).

Table 5.

Back-transformed estimated marginal means of average pain intensity.

Month	0	6	12	18	24	30	36	42	48
All-initiator cohort (N = 112)
n	112	101	93	93	92	82	83	79	82
EMM (95% CI)	6.2 [5.7–6.7]	6.0 [5.4–6.5]	5.9 [5.4–6.4]	6.0 [5.5–6.6]	5.6 [5.0–6.2]*	5.9 [5.4–6.4]	5.8 [5.2–6.3]*	6.0 [5.4–6.6]	5.7 [5.1–6.3]*
On-treatment cohort
N	112	82	63	50	46	44	40	36	33
n	112	76	55	48	44	41	36	33	31
EMM (95% CI)	6.0 [5.4–6.7]	6.0 [5.2–6.7]	5.8 [5.1–6.5]	5.8 [5.0–6.6]	5.5 [4.7–6.2]*	5.5 [4.7–6.3]*	5.8 [4.9–6.7]*	5.5 [4.4–6.6]	5.0 [4.0–6.1]*
Continuation group (N = 33)
n	33	30	31	31	32	31	30	31	31
EMM (95% CI)	6.7 [5.8–7.5]	6.2 [5.2–7.2]	6.0 [5.1–6.9]*	6.5 [5.3–7.7]	5.6 [4.6–6.5]*	5.9 [4.8–6.9]*	6.1 [4.9–7.2]*	5.7 [4.5–6.8]*	5.5 [4.3–6.7]*
Discontinuation group (N = 79)
On eptinezumab^a	79	49	30	17	13	11	7	3	0
n	79	71	62	62	60	51	53	48	51
EMM (95% CI)	6.2 [5.6–6.7]	6.1 [5.5–6.6]	5.9 [5.4–6.4]	6.0 [5.5–6.6]	5.9 [5.4–6.5]	6.1 [5.6–6.5]	6.1 [5.6–6.7]	6.3 [5.8–6.9]	6.0 [5.3–6.6]

Estimated marginal means were obtained from generalized estimating equation models with a log link (adjusted for age, sex, BMI). Model estimates were calculated on the log scale and exponentiated to yield marginal mean estimates and 95% confidence intervals on the original scale of monthly headache days. N denotes the total number of patients in each cohort. n denotes the number of patients contributing data at each time point. Missing values were not imputed. *indicates p < 0.05 versus baseline. API, average pain intensity; GEE, generalized estimating equation; EMM, estimated marginal mean; BMI, body mass index; CI, confidence interval.

Dose escalation

Of the 112 patients, two initiated treatments at 300 mg. One patient achieved a sustained 30% reduction in headache frequency and continued treatment for the full 48 months, while the other discontinued after 34 months due to loss of effectiveness, never having achieved a 30% reduction. Dose escalation occurred in 87 patients (77.7%), all of whom started at 100 mg. Of these, 86 escalated to 300 mg, and one increased to 150 mg due to intolerance. Escalation generally occurred early, with a mean of 7.7 ± 7.4 months, typically at the third infusion.

Concurrent onabotulinumtoxinA use

During follow-up, onabotulinumtoxinA was the only concurrent preventive therapy tracked. Overall, 57 of 112 patients (50.9%) received onabotulinumtoxinA, including 24 of 33 (72.7%) in the continuation group and 33 of 79 (41.8%) in the discontinuation group. Among these concurrent users, the mean overlap rate with eptinezumab was high and similar (87.8 ± 26.9%). Subgroup analyses were not performed due to the limited number of patients receiving eptinezumab alone, which precluded statistically meaningful comparisons.

Treatment persistence and discontinuation

Persistence declined steadily over time. Kaplan–Meier estimates showed continuation rates of 55.4% at one year, 39.3% at two years, 35.7% at three years, and 29.5% at four years ( Figure 2 ). The treatment trajectory throughout the 48-month is illustrated in Table 6 . The most common reasons for discontinuation were ineffectiveness (34.2%), loss of effectiveness (19.0%), insurance or cost barriers (16.5%), and loss to follow-up (15.2%). Adverse events accounted for 6.3% of discontinuations ( Table 6 ). In a multivariable analysis, younger age was the only factor associated with discontinuation; BMI, baseline MHD, and the number of prior anti-CGRP failures were not associated (Tables 7 and 8).

Figure 2.

Kaplan–Meier curve of eptinezumab treatment persistence. The curve illustrates the probability of treatment persistence over 48 months in the total cohort (N = 112). Loss to follow-up was treated as treatment discontinuation.

Table 6.

Treatment trajectory and discontinuation reasons.

Month after eptinezumab	0	6	12	18	24	30	36	42	48
Continue using eptinezumab	112	78	63	50	44	42	40	35	33
100 mg	110	33	11	7	7	5	1	1	1
300 mg	2	45	52	43	37	37	39*	34	32
Cumulative discontinuations	0	34	49	62	68	70	72	77	79
Discontinuations in current interval	0	34	15	13	6	2	2	5	2
Discontinuation reason
Lost to follow-up		4	3	3				2
Ineffectiveness		13	8	3	3
Loss of effectiveness				7	2	2	2	1	1
Insurance and cost		8	4		1
Adverse events^a		4						1
Transportation difficulty		2							1
Pregnancy		1
Disease stabilization^b								1
Patient request^c		1
Deceased		1

*One shift from 100 mg to 150 mg. ^aAdverse events were reported in five patients: two cases of severe constipation; one case each of nasal stuffiness with respiratory problems, angioedema with bronchospasm, and wheezing with bronchospasm and cough. ^bPatient discontinued treatment after 41 months due to sustained migraine control. ^cPatient requested return to prior anti-CGRP therapy (galcanezumab); reason not documented. Two patients have combined reasons: ineffectiveness and adverse events, ineffectiveness and insurance.

Table 7.

Baseline comparison between treatment continuation group and discontinuation group due to lack of effectiveness or adverse events.

	Continuation group	Discontinuation group	p-value
n	33	47
Age	52.7 ± 13.7; 56 (40.5–62)	46.3 ± 12.5; 50 (39–54)	0.033
Female	27 (81.8)	36 (76.6)	0.782
BMI	29.8 ± 9.3; 27.3 (23.7–33.2)	29.4 ± 5.9; 29.1 (24.6–32.1)	0.516
Migraine duration, years	16.7 ± 12.4; 14 (4–24)	13.44 ± 13.1; 7.5 (4.3–20.3)	0.301
No. of prior anti-CGRP therapy failures^a	1.8 ± 0.9; 2 (1–2.5)	2.1 ± 0.9; 2 (2–3)	0.170
Baseline MHD	24.8 ± 7.5; 30 (19–30)	26.5 ± 7.0; 30 (25–30)	0.137
Baseline MIDAS	128.5 ± 116.5; 90 (47.5–192.5)	110.1 ± 100.1; 87.5 (29–180)	0.639
Baseline API	6.2 ± 1.5; 6 (5.3–7.0)	6.2 ± 2.1; 6.5 (4.5–8.0)	0.944

Continuous data are presented as mean ± SD; median (IQR); frequency data are presented as count (%). P-values comparing continuation and discontinuation groups were calculated using the chi-square test for categorical variables, and the non-parametric or independent t-test for continuous variables, as appropriate. ^aNo. of prior anti-CGRP therapy failures (lack of effectiveness and adverse events)

BMI, body mass index; CGRP, calcitonin gene-related peptide; MHD, monthly headache days; MIDAS, migraine disability assessment; API, average pain intensity; SD, standard deviation; IQR, interquartile range.

Table 8.

Multivariate analyses of treatment discontinuation.

	Adjusted OR (95% CI)	p-value
Age (per 5 years)^a	0.82 (0.68–0.99)	0.039
Female	0.68 (0.21–2.18)	0.528
BMI (per 5 kg/m²)^a	0.96 (0.72–1.28)	0.805
Baseline MHD	1.04 (0.97–1.11)	0.293
No. of prior anti-CGRP therapy failures^b	1.24 (0.72–2.13)	0.433

Age and BMI were modeled as mean-centered continuous variables and scaled per 5 units (age per 5 years; BMI per 5 kg/m²). ^bNo. of prior anti-CGRP therapy failures (lack of effectiveness and adverse events)

OR, odds ratio; CI, confidence interval; BMI, body mass index, MHD, monthly headache days; CGRP, calcitonin gene-related peptide.

Post-discontinuation trajectories

Among the 79 discontinuers, 54.4% initiated another anti-CGRP therapy during follow-up, with a mean gap of 13.78 months before re-entry into treatment. Switching was common: 29 patients switched once, 8 switched twice, 4 switched three times, and 2 switched four times. An additional 26.6% stopped all anti-CGRP therapies, and 20.3% were lost to follow-up. Treatment pathways after discontinuation were heterogeneous and frequently involved multiple transitions (Figure 3).

Figure 3.

Anti-CGRP switching patterns after discontinuation of eptinezumab. This flowchart illustrates the therapeutic pathways for patients (n = 79) following the discontinuation of eptinezumab. Categories include switching to other anti-CGRP preventives, stopping anti-CGRP preventives, and loss to follow-up (LTF).

Discussion

This 48-month real-world analysis shows that eptinezumab can yield sustained, multidomain benefit in a subset of treatment-resistant CM patients, while also clarifying long-term treatment behaviors relevant to clinical care. Patients who remained on therapy demonstrated durable reductions in headache frequency and later improvements in disability and pain intensity, whereas those who discontinued did not maintain benefit. Despite declining persistence over time, more than half of discontinuers ultimately returned to anti-CGRP therapy via heterogeneous pathways, reflecting the clinical complexity of managing CM in real-world practice. These patterns help clinicians set realistic expectations: even after failing multiple prior CGRP mAbs, some patients may still improve with eptinezumab, though responses are modest and inconsistent across individuals. The observed dissociation between MHD, MIDAS, and API also reflects the complexity of disease burden in treatment-resistant CM, particularly in a cohort where most patients required dose escalation to 300 mg.

The longitudinal outcome patterns in this cohort illustrate substantial heterogeneity in treatment response, a feature consistently observed in re CM populations with multiple prior preventive failures.^17,18 Headache frequency, disability, and pain intensity followed nonparallel trajectories, with MHD and API showing gradual improvement while MIDAS initially worsened before later declining. This dissociation is consistent with prior work demonstrating that disability measures, such as MIDAS, reflect a broader functional burden that may improve more slowly than headache frequency alone.^19,20 Fluctuation in ≥30% MHD responder status among continuers (16.7–29% across intervals), with only two patients maintaining responder status at all timepoints, illustrating that fixed responder thresholds such as the ≥30% MHD reduction benchmark can fluctuate substantially in treatment-resistant CM, and mirroring real-world findings that responder rates decline progressively with increasing numbers of prior anti-CGRP mAb failures.^4,14,21 It is likely that clinically meaningful benefit may manifest intermittently or incompletely, necessitating longitudinal rather than single-timepoint assessment when evaluating treatment benefit durability.

Long-term persistence with eptinezumab was modest, with 55.4% of patients remaining on treatment at 12 months and fewer than 40% continuing beyond 24 months, consistent with the wide persistence range (27.8%–66.7%) reported across real-world anti-CGRP mAb studies in similarly treatment-resistant populations.^8,14,22,23 In our cohort, discontinuation was driven primarily by inadequate response or loss of effectiveness, followed by insurance-related barriers. These patterns align with the dominant reasons for discontinuation reported in other real-world studies. Logistic regression identified younger age as the only factor associated with discontinuation, whereas baseline headache frequency, and number of prior anti-CGRP failures were not significant. This age effect is consistent with prior Japanese literature suggesting better treatment outcomes among older patients.¹⁸ It would also be interesting to examine whether younger patients are more psychologically eager to change medication. For example, younger age groups may find it more challenging to take time off work because they are earlier in their careers. Similarly, perhaps they have less ability to travel to an infusion center if it is not conveniently located and are therefore more likely to discontinue eptinezumab if not overtly helpful. We would also be curious to determine whether insurance coverage, for example, Medicare vs. private insurance, influences discontinuation.

Beyond age, prior studies have linked higher attack frequency, multiple preventive failures, psychiatric comorbidities, and immune-rheumatologic burden to poorer treatment durability,^17,18,24,25 the lack of such associations in our analysis likely reflects the cohort's uniformly high refractoriness, which limits variability in baseline characteristics and may obscure factors detectable in more heterogeneous populations. Additionally, persistence in real-world treatment-resistant CM is often influenced by time-varying variables, such as fluctuating clinical benefit, evolving disability, changes in insurance coverage, and the cumulative burden of repeated treatment attempts, that are not captured by baseline phenotyping alone in retrospective datasets. Together, these findings indicate that long-term eptinezumab use in treatment-resistant CM is more strongly shaped by dynamic clinical and access-related pressures than by static patient characteristics at initiation. Clinically, this underscores the need for early and periodic reassessment, attention to partial or delayed improvement, and a proactive approach to dose escalation or within-class switching to support long-term preventive management in highly treatment-resistant patients.²⁶

Post-discontinuation treatment trajectories following eptinezumab were heterogeneous, but continued engagement with anti-CGRP therapy remained common despite prior discontinuation. In this heavily pretreated cohort, 54.4% of discontinuers pursued another anti-CGRP agent, whereas 26.6% stopped anti-CGRP therapy entirely, and 20.3% were lost to follow-up. This re-engagement suggests that discontinuation of eptinezumab often reflects therapeutic adjustment rather than abandonment of the class, likely influenced by the limited availability of alternative preventive options in treatment-resistant CM. Post-switch effectiveness was not evaluated in our study because the timing of treatment discontinuation and subsequent initiation varied across individuals, precluding standardized comparisons. Prior real-world studies also provide limited and inconsistent data on outcomes after multiple treatment switches, and incomplete follow-up further constrains interpretation.^11,27,28

This study has several notable strengths. The 48-month observation period provides one of the longest real-world assessments of eptinezumab to date, and the intention-to-treat design, incorporating both continuers and discontinuers, reduces survivor bias and more accurately reflects real-world treatment trajectories. Provider-verified outcome forms and GEE statistics enabled longitudinal modeling of headache frequency, disability, and pain intensity despite irregular follow-up intervals, offering insight into multidomain change over time beyond what categorical responder thresholds can capture. Nonetheless, important limitations must be acknowledged. The retrospective, single-center design depends on routine clinical documentation rather than standardized research protocols, which may introduce variability in how outcomes, discontinuation reasons, and follow-up status were recorded. Attrition over time resulted from treatment discontinuation and evolving clinical care, and incomplete outcome forms at later timepoints are an expected feature of long-term EMR-based longitudinal studies. Moreover, treatment courses were dynamic, with multiple medication changes and evolving management strategies occurring over extended monitoring periods. The lack of systematic, continuous collection of contextual factors, such as medication overuse headache, infusion-related reactions, medication adherence, behavioral interventions, and detailed tracking of concomitant preventive and acute medication use, limits the ability to attribute outcome trajectories solely to eptinezumab. The cohort's uniformly high refractoriness, while representative of tertiary headache practice, also reduces external validity; patients with fewer prior treatment failures or lower baseline disability may exhibit different persistence patterns or response trajectories than those observed here. Finally, several methodological limitations warrant consideration. First, comparisons between treatment continuers and discontinuers may be subject to survivor bias, as patients who remain on therapy could differ in clinically meaningful ways from those who discontinue, potentially affecting the observed outcomes. Second, the presence of concomitant onabotulinumtoxinA therapy in some participants represents a potential confounding factor, as improvements in outcomes may not be solely attributable to the study intervention.

Conclusions

In this 48-month real-world cohort of treatment-resistant CM, eptinezumab was associated with gradual, multidomain improvement among continuers, while overall treatment trajectories showed considerable heterogeneity and early discontinuation. Larger prospective, multi-center studies with standardized outcome collection are needed to confirm these patterns and further guide long-term preventive strategies in treatment-resistant CM.

Highlights

Multiple exposure-defined cohorts (all-initiator, on-treatment, continuation, discontinuation) were used to address the complexity of long-term real-world treatment trajectories.

Symptom improvement was modest and most apparent among patients who remained on eptinezumab, while overall treatment persistence declined over time.

Post-discontinuation pathways were heterogeneous, with frequent switching and re-engagement with anti-CGRP therapy, reflecting iterative preventive management in chronic migraine.

Supplemental Material

sj-docx-1-cep-10.1177_03331024261451427 - Supplemental material for Eptinezumab in treatment-resistant chronic migraine: Four-year real-world effectiveness and treatment persistence

Supplemental material, sj-docx-1-cep-10.1177_03331024261451427 for Eptinezumab in treatment-resistant chronic migraine: Four-year real-world effectiveness and treatment persistence by Wan-Jen Hsieh, Kelly Nicol, Chai Ching Ng, Tsung-Wei Hou, Phillip Phan, Scott W. Keith, Brittany Heckel, Michael J. Marmura and Hsiangkuo Yuan in Cephalalgia

Footnotes

Acknowledgements

During the preparation of this manuscript, the authors used Grammarly (Grammarly, Inc., San Francisco, CA) and Microsoft 365 Copilot (Microsoft Corporation, Redmond, WA), accessed via Thomas Jefferson University's enterprise license, solely for proofreading and grammar refinement. These tools did not generate any content, ideas, or substantive text. All intellectual content and writing were developed entirely by the authors, who take full responsibility for the final version of this manuscript. The authors also thank John Waters, MD, Adam Mermelstein, MD, Qiu Zi Yang, MD, and Emily Casaletto, MD for their valuable contributions to this work.

ORCID iDs

Wan-Jen Hsieh

Chai Ching Ng

Michael J. Marmura

‌Hsiangkuo Yuan

Ethical considerations

The study was approved by the Thomas Jefferson University Institutional Review Board (No. 21E.1141).

Consent to participate

IRB approval with a waiver of informed consent.

Consent for publication

All authors have reviewed the final version of the manuscript and agree to its publication in Cephalalgia.

Author contributions

Wan-Jen Hsieh contributed to conceptualization, data curation, formal analysis, methodology, software, writing – original draft, and writing – review & editing. Kelly Nicol, Tsung-Wei Hou, Chai Ching Ng, and Brittany Heckel contributed to writing – review & editing. Phillip Phan contributed to data curation, methodology, and validation. Scott Keith contributed to methodology and validation. Michael J. Marmura contributed to writing – review & editing and supervision. Hsiangkuo Yuan contributed to conceptualization, methodology, validation, writing – review & editing, and supervision.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Miles for Migraine, Patient Philanthropy.

Declaration of conflicting interests

The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Wan-Jen Hsieh, Kelly Nicol, Chai Ching Ng, Tsung- Wei Hou, Phillip Phan, Scott Keith and Brittany Heckel report no conflict of interest to declare. Michael J. Marmura has received honoraria for consultations from Lundbeck, Theranica, Axsome, and AbbVie; research funding for his role as principal investigator from Teva and AbbVie. Hsiangkuo Yuan has received funding from AHS Early-Stage Investigator Research Award, Pfizer-Jefferson Collaborative Research; institutional support for serving as an investigator from Teva, AbbVie, Ipsen, Rehaler, Pfizer, Shiratronics, Lundbeck, Johnson & Johnson; consultant/advisory fees from Salvia, AbbVie, Pfizer, Cerenovus; and royalties from Cambridge University Press and MedLink Neurology.

Data availability statement

Data are available upon reasonable request. De-identified data may be made available upon reasonable request to the corresponding author, subject to institutional review and data use agreements.

Open practices

Not applicable

Supplemental material

Supplemental material for this article is available online.

References

Lipton

Goadsby

Smith

, et al. Efficacy and safety of eptinezumab in patients with chronic migraine: PROMISE-2. Neurology 2020 Mar 31; 94: e1365–e1377.

Ashina

Tepper

Gendolla

, et al. Long-term effectiveness of eptinezumab in patients with migraine and prior preventive treatment failures: extension of a randomized controlled trial. J Headache Pain 2023 Nov 20; 24: 155.

Kudrow

Cady

Allan

, et al. Long-term safety and tolerability of eptinezumab in patients with chronic migraine: a 2-year, open-label, phase 3 trial. BMC Neurol 2021 Mar 19; 21: 126.

Scheffler

Wenzel

Bendig

, et al. 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 2024 May 16; 25: 79.

Triller

Blessing

Overeem

, et al. Efficacy of eptinezumab in non-responders to subcutaneous monoclonal antibodies against CGRP and the CGRP receptor: a retrospective cohort study. Cephalalgia 2024 Oct; 44: 3331024241288875.

Andreou

Hill

Al-Rawi

, et al. A prospective real-world analysis of intravenous eptinezumab in migraine management: the first UK experience. J Headache Pain 2025 Oct 30; 26: 235.

Barbanti

Aurilia

Egeo

, et al. A 24-week prospective, multicenter, real-world study on eptinezumab's effectiveness and safety in migraine prevention (EMBRACE II). J Neurol 2025 May 7; 272: 382.

Ray

Virk

Regmi

, et al. Patient adherence and long-term tolerability of anti-calcitonin gene-related peptide (CGRP) monoclonal antibodies in migraine prevention: a systematic review. Cureus 2025 Aug 31; 17: e91347.

Vaghi

Corrado

Pocora

, et al. Real world effectiveness of anti-CGRP monoclonal antibodies over three consecutive one-year treatment cycles: an intention-to-treat analysis. Cephalalgia 2025 Aug; 45: 3331024251353421.

10.

Pozo-Rosich

van Veelen

Caronna

, et al. Guidelines of the International Headache Society for Real-World Evidence studies in migraine and cluster headache. Cephalalgia 2025 Mar; 45: 3331024251318016.

11.

Vandenbussche

Pisarek

Paemeleire

. Methodological considerations on real-world evidence studies of monoclonal antibodies against the CGRP-pathway for migraine: a systematic review. J Headache Pain 2023 Jun 22; 24: 75.

12.

Charleston

Talon

Sullivan

, et al. Persistence to anti-CGRP monoclonal antibodies and onabotulinumtoxinA among patients with migraine: a retrospective cohort study. J Headache Pain 2023 Aug2; 24: 101.

13.

Ojo

Zhang

Bleibdrey

, et al. Persistence and switching patterns of migraine prophylactic medications in Canada: a retrospective claims analysis comparing adherence and evaluating the economic burden of illness. J Pharm Pharm Sci 2023; 25: 402–417.

14.

de Dios

Pages-Puigdemont

Ojeda

, et al. Persistence, effectiveness, and tolerability of anti-calcitonin gene-related peptide monoclonal antibodies in patients with chronic migraine. Headache 2025 Jan; 65: 24–34.

15.

Garrido Hernández

Andrade Zumarraga

Perea Justicia

, et al. Long-term persistence of anti-calcitonin gene-related peptide monoclonal antibodies in treatment-resistant migraine: a three-year follow-up study in Spain. Cephalalgia Reports 2025; 8: 25158163251388920.

16.

Harris

Taylor

Minor

, et al. The REDCap consortium: building an international community of software platform partners. J Biomed Inform 2019 Jul; 95: 103208.

17.

Caronna

Gallardo

Egeo

, et al. Redefining migraine prevention: early treatment with anti-CGRP monoclonal antibodies enhances response in the real world. J Neurol Neurosurg Psychiatry 2024 Sep 17; 95: 927–937.

18.

Ihara

Ohtani

Watanabe

, et al. Predicting response to CGRP-monoclonal antibodies in patients with migraine in Japan: a single-centre retrospective observational study. J Headache Pain 2023 Mar 9; 24: 23.

19.

Stewart

Lipton

Kolodner

. Migraine disability assessment (MIDAS) score: relation to headache frequency, pain intensity, and headache symptoms. Headache 2003 Mar; 43: 258–265.

20.

Carvalho

Luedtke

Braun

. Minimal important change and responsiveness of the migraine disability assessment score (MIDAS) questionnaire. J Headache Pain 2021 Oct 21; 22: 126.

21.

Scheffler

Schenk

Wurthmann

, et al. CGRP Antibody therapy in patients with drug resistant migraine and chronic daily headache: a real-world experience. J Headache Pain 2021 Sep 20; 22: 111.

22.

Varnado

Manjelievskaia

, et al. Treatment patterns for calcitonin gene-related peptide monoclonal antibodies including galcanezumab versus conventional preventive treatments for migraine: a retrospective US claims study. Patient Prefer Adherence 2022 Mar 29; 16: 821–839.

23.

Schwedt

Lee

Knievel

, et al. Real-world persistence and costs among patients with chronic migraine treated with onabotulinumtoxinA or calcitonin gene-related peptide monoclonal antibodies. J Manag Care Spec Pharm 2023 Oct; 29: 1119–1128.

24.

Sanchez-Rodriguez

Gago-Veiga

Garcia-Azorin

, et al. Potential predictors of response to CGRP monoclonal antibodies in chronic migraine: real-world data. Curr Pain Headache Rep 2024 Dec; 28: 1265–1272.

25.

Caronna

Mas-de-Les-Valls

Egeo

, et al. Long-Term effectiveness and persistence factors of anti-CGRP monoclonal antibodies in migraine: 2-year results from the EUREkA cohort. Neurology 2026 Feb 24; 106: e214659.

26.

Iannone

Burgalassi

Vigani

, et al. Switching anti-CGRP(R) monoclonal antibodies in multi-assessed non-responder patients and implications for ineffectiveness criteria: a retrospective cohort study. Cephalalgia 2023 Apr; 43: 3331024231160519.

27.

Waliszewska-Prosół

Montisano

Antolak

, et al. The impact of primary headaches on disability outcomes: a literature review and meta-analysis to inform future iterations of the Global Burden of Disease study. J Headache Pain 2024 Mar 4; 25: 27.

28.

Burgalassi

Romozzi

Vigani

, et al. Outcomes, unmet needs, and challenges in the management of patients who withdraw from anti-CGRP monoclonal antibodies: a prospective cohort study. Cephalalgia 2024 Nov; 44: 3331024241273968.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.04 MB

0.00 MB