Abstract
Aim
Incorporating patient-reported outcomes (PROs) into clinical practice and real-world research is essential to enable a more holistic assessment of treatment effectiveness, beyond migraine frequency alone. Given the number of calcitonin gene-related peptide (CGRP)-targeted treatments available, switching between CGRP-targeted preventive therapies is a common practice. Real-world evidence on switching to intravenous eptinezumab in adults with migraine for whom prior subcutaneous or oral CGRP-targeted preventive therapies failed is limited. The ongoing INFUSE study was designed to evaluate the real-world effectiveness of switching to eptinezumab in US clinical practice among adults with migraine for whom ≥1 prior CGRP-targeted preventive therapy had failed.
Methods
INFUSE, a real-world, non-interventional study evaluating the effectiveness of eptinezumab in US clinical practice, included adults with migraine for whom ≥1 prior CGRP-targeted preventive therapy had failed and who were newly initiating eptinezumab treatment. PROs were collected electronically. To provide an overall assessment of the change in migraine disease, Patient Global Impression of Change (PGIC) was utilized. Other outcomes allowed for the evaluation of novel end points, such as brain fog and participant-defined good days.
Results
The baseline population included 190 participants (female, 88.9%; 35–54 years of age, 53.7%). At month 6, 44.1% (95% confidence interval (CI): 35.3–53.4%) of participants reported “much/very much improved” on PGIC, and 75.7% (95% CI: 66.9–82.7%) reported any improvement. Mean good days increased by 6.3 per month at month 6 (P < 0.001). Among participants reporting brain fog at baseline (93.7%), 55.8% (95% CI: 46.2–64.9%) reported improvement at month 6.
Conclusion
This research provides insights into capturing the broader migraine burden and the holistic impact of preventive treatment and highlights the potential for achieving ambitious goals. In this population for whom ≥1 prior CGRP-targeted preventive therapy had failed, switching to eptinezumab was associated with participant-reported, clinically meaningful improvements in PROs, including frequency, disability, disease status, and brain fog.
This is a visual representation of the abstract.
Introduction
Migraine is a highly individualized neurological condition characterized by fluctuating symptoms, disability, and quality-of-life impairment; thus, patient-reported outcomes (PROs) provide important insights into how individuals with migraine perceive headache intensity, frequency, and the broader consequences on daily activities and emotional well-being. 1 Incorporating PROs into both clinical practice and real-world research enables a person-centered approach by holistically assessing treatment effectiveness and aligning outcome evaluation with what matters most to individuals with migraine. Furthermore, measuring therapeutic benefits through meaningful, real-life improvements that can be used in clinical practice can potentially improve shared decision-making. Despite being among the most frequently reported symptoms by people living with migraine, cognition and brain fog have received limited systematic assessment in clinical and real-world studies, and at the study initiation, had no brief, validated, migraine-specific measures for use in observational settings.
In recent years, preventive treatment options for migraine have expanded to include therapies targeting the calcitonin gene-related peptide (CGRP), including four CGRP-targeted monoclonal antibodies (mAbs) and two oral CGRP-targeted treatments (gepants) approved by the US Food and Drug Administration.2–7 Given the availability of several CGRP-targeted preventive therapies and no treatment algorithm, switching between CGRP-targeted treatments is a relatively common practice. In a US-based retrospective, observational cohort study, 58% of individuals discontinuing their initial CGRP-targeted mAbs switched to another preventive treatment after an average of approximately 6 months, with 44% of those switching to a different anti-CGRP mAb. 8 Several real-world studies outside the United States have evaluated outcomes in adults with migraine who switched from one subcutaneous (SC) CGRP-targeted treatment to another because of insufficient effectiveness or poor tolerability; however, to date, most of those studies were small; included differing geographies, healthcare systems, and practices; and varied considerably in the reported clinical outcomes following switch. A first switch from one SC CGRP-targeted mAb to another was shown to yield a lower response than in treatment-naive individuals and cycling through multiple CGRP-targeted preventive treatment switches yielded decreasing proportions of responders.9–12 While the current literature has begun to provide some insights into switching patterns, particularly for those switching between SC CGRP-targeted mAbs, more research is needed to inform clinical decision-making in the United States when considering a switch from a SC or oral CGRP-targeted therapy to intravenous (IV) eptinezumab.
Eptinezumab, the only IV-administered CGRP-targeted preventive treatment, was approved in 2020 by the US Food and Drug Administration. 7 The efficacy and safety of eptinezumab have been demonstrated versus placebo for the preventive treatment of episodic migraine (EM) and chronic migraine (CM) in CGRP-targeted treatment-naive adults across multiple clinical trials,13–15 and long-term safety and improvements in PROs have been observed in an open-label trial. 16 The effectiveness and safety of eptinezumab have also been evaluated in a number of real-world studies.17–24 However, data on the real-world effectiveness of eptinezumab in participants with prior CGRP-targeted preventive therapy experience remain limited, especially in the United States.21–24 The multicenter, observational, cross-sectional US-based REVIEW (Real-world EVidence and Insights into Experiences With eptinezumab) study evaluated the real-world effectiveness of eptinezumab in a treatment-experienced population (n = 94), 89% of whom had received prior SC CGRP-targeted mAbs. 17 Participants in REVIEW were recruited after being on eptinezumab for ≥6 months, and a single survey assessed outcomes of interest before and after starting eptinezumab treatment. Approximately 70% of participants were on eptinezumab treatment for a year or more, creating a limitation of long recall periods for the “before eptinezumab” assessments. Despite a history of prior treatment use, after initiating eptinezumab treatment, participants reported improvements in good days, brain fog, migraine symptoms, and elements of daily living and reductions in acute medication use. 17
The ongoing INFUSE study was designed to evaluate the real-world effectiveness of switching to eptinezumab in US clinical practice among adults with migraine for whom ≥1 prior CGRP-targeted preventive therapy had failed. This study aims to expand on the prior real-world REVIEW study 17 by prospectively evaluating similar outcomes in participants newly initiating eptinezumab over time. Outcomes included in this study are intended to capture the multifaceted nature of the disease and descriptively assess the impact of eptinezumab preventive treatment on the overall burden of migraine. This manuscript presents 6-month results from the INFUSE study.
Methods
Study design and population
The INFUSE study is a real-world, prospective, non-interventional effectiveness study evaluating eptinezumab for migraine prevention in US clinical practice. The study period is 6 months, with participant follow-up for an additional 6 months. Findings from this study should be interpreted as exploratory, given the study design and the use of several non-validated outcome measures. All study procedures were conducted in accordance with the principles of the Declaration of Helsinki, Good Pharmacoepidemiology Practices guidelines of the International Society for Pharmacoepidemiology Good Pharmacovigilance Practice, and local rules and regulations. All participants completed electronic informed consent on the study's web-based platform. This study is reported in accordance with the STROBE guidelines for a real-world observational study.
The use of an electronic diary (eDiary) provides the most rigorous method for daily assessment of outcomes and is the standard in randomized clinical trials. However, in routine clinical practice, daily diaries can be burdensome and are often not used by individuals with migraine. Only 31% (58/191) of recruited participants in the INFUSE study reported use of a migraine diary prior to entry. Accordingly, given the observational nature of this study, it leveraged established recall periods rather than requiring daily diary use, balancing data completeness with real-world feasibility.
Participants were recruited across national infusion network partners in the Vyepti® (eptinezumab-jjmr) Infusion Network, which together operate >100 individual infusion centers throughout the United States, striving to gather a geographically and clinically representative sample of participants. Participants who had been newly prescribed eptinezumab for migraine prevention by their treating physician as part of routine clinical care were informed of and invited to participate in the study voluntarily by the infusion center. Data collection was decentralized where participants could access the study's web-based platform and complete the screening questions without encouragement from their prescribing physician. In this non-interventional study, the treating physician independently selected the starting dose of eptinezumab (100 or 300 mg) and determined any subsequent dose adjustments. All other treatment decisions were also at the discretion of the treating physician, and there were no restrictions on the use of concomitant medications. Eligible participants were adults (≥18 years of age) with a diagnosis of migraine (defined as an International Classification of Diseases, 10th revision code starting with G43) who self-reported ≥1 of the following prior CGRP-targeted preventive therapies that had failed: erenumab, fremanezumab, galcanezumab, atogepant, or rimegepant (prescribed on a preventive schedule; every other day). Given the direct-to-patient recruitment approach, a CGRP-targeted preventive therapy was considered as having failed if a participant self-reported failure based on lack of efficacy, undesirable side effects, or both. Participants had to be willing to bring their own device to access and complete the study questionnaires. Participants who were pregnant or planned to become pregnant within the subsequent 12 months, participants who had previously received eptinezumab, and participants who were enrolled in a clinical trial related to headache or migraine were excluded.
The infusion center reported the following data: participant name, date of birth, status (new to eptinezumab therapy), next treatment date, prescribed eptinezumab dose at baseline and subsequent visits, and diagnosis code associated with migraine. Participant-reported baseline characteristics and outcomes were evaluated using surveys delivered electronically. The first (baseline) survey was completed at any time within ≤30 days prior to the first eptinezumab infusion. Follow-up surveys were sent at day 7 and months 3, 6, 9, and 12, with automated prompts to complete the survey. The prompts were triggered by infusion visits or by elapsed time (i.e. 90 days) since last infusion.
Outcomes
The survey included, but was not limited to, the following participant-reported baseline characteristics: age (range), gender, race/ethnicity, height, weight, body mass index (BMI), length of time diagnosed with migraine, previous and current preventive migraine therapies, diagnosis of CM, rationale for switching from prior CGRP-targeted preventive therapy, and bothersome symptoms before starting preventive treatment with eptinezumab.
Patient Global Impression of Change (PGIC), a single-item measure of perceived change in a participant's overall migraine status since starting eptinezumab treatment, was intentionally chosen to capture holistic improvement of disease in people living with migraine. PGIC captures the participant's integrative assessment of the change across all migraine symptoms, pain-related and non-pain-related, such as sensory sensitivities, function, disability, interictal, and affective burden, domains that extend beyond frequency of headache and migraine days. PGIC has been used in numerous eptinezumab clinical trials16,25–27 and real-world CGRP-targeted preventive therapy studies.18–21 In fact, in one study, PGIC showed a stronger correlation with discontinuation (0.541) than monthly migraine day reduction (0.470), highlighting the importance and relevance of this end point in the real world. 28 This study assessed the percentage of participants reaching “much improved” and “very much improved” status on PGIC. Participants responded to the question, “Since your first Vyepti infusion, how would you describe the change (if any) in activity limitations, symptoms, emotions, and overall quality of life, as related to your migraine?” using a 7-point scale (“very much improved,” “much improved,” “minimally improved,” “no change,” “minimally worse,” “much worse,” or “very much worse”) at day 7, month 3, and month 6. Additionally, a subgroup analysis was conducted to evaluate the changes in PGIC through 6 months in participants reporting 1, 2, and ≥3 CGRP-targeted treatment failures as distinct groups of interest.
The Migraine Disability Assessment (MIDAS) questionnaire is a validated and well-established PRO used in several research efforts.16,21,22,24,29 The questionnaire is a 7-item instrument designed to quantify migraine-related disability over the preceding 3 months that includes five questions assessing the impact of migraine on functionality across work/school, household work, and social/leisure activities, plus two additional questions assessing the frequency and severity. 30 The total MIDAS score was calculated by summing the responses to the first five questions for a possible score of 0–270, with higher scores indicating greater disability. The severity of headache pain was assessed using the following question: “On a scale of 0–10, on average how painful were these headaches? (where 0 = no pain at all, and 10 = pain as bad as it can be).”
Headache frequency remains a fundamental component of migraine care, and in other research, MIDAS has demonstrated reliability and concordance with diary-based measures of headache-related burden.29–33 Although prospective headache diaries (including eDiary-based reporting) represent the gold standard for precise quantification of monthly headache days (MHDs) or monthly migraine days in clinical trials, this recall-based approach was selected as a pragmatic alternative for a real-world observational study. The use of a daily diary was not common among INFUSE participants, and sustained daily diary completion can be burdensome, which has the potential to impact participation and data completeness. Therefore, in this study, headache frequency was assessed using the MIDAS headache day item (“On how many days in the past 3 months did you have a headache?”). To report results in a format consistent with commonly used monthly headache metrics, the total number of headache days reported over the 3-month recall period was divided by 3 to estimate an average MHD frequency. Similar to PGIC, a subgroup analysis was conducted to evaluate the changes in MIDAS-derived MHDs through 6 months in participants reporting 1, 2, and ≥3 CGRP-targeted treatment failures as distinct groups of interest. The proportion of participants with ≥50% and ≥75% reductions in MIDAS-derived MHDs from baseline was also evaluated to incorporate both an established threshold of clinical response and a higher goal for clinical response.
Cognition and brain fog were included as exploratory end points because they are among the most commonly reported symptoms by individuals living with migraine, yet remain substantially understudied in both clinical trials and real-world studies, with no brief validated migraine-specific measure available at the time of study design. The inclusion of novel, non-validated measures around cognition and brain fog reflects a deliberate effort to address this gap and align outcome measures with patient-prioritized disease burden. The presence of brain fog was assessed at baseline using the study-specific question, “As part of your migraine disease, have you experienced brain fog (having difficulty concentrating or focusing, trouble finding the right words or speaking, feelings of mental cloudiness)?” Changes in brain fog after eptinezumab initiation (at day 7, months 3 and 6) were only assessed if participants reported brain fog at baseline. The improvement or deterioration was assessed using the question, “If yes, rate the extent your symptoms have changed since starting Vyepti.” Changes were rated as “very much improved,” “much improved,” “minimally improved,” “no change,” “minimally worse,” “much worse,” or “very much worse.”
Participant-defined good days were included as a positive-framing counterpart to traditional burden-focused end points, such as headache and migraine days. Good days were assessed at baseline for the 4 weeks preceding the first survey and at months 3 and 6 for the preceding 3 months. For each observational period visit, participants were asked to respond to the question, “How many good days did you have in the past 3 months?” using a response of 0 to 90 days. For months 3 and 6, the number of participant-defined good days per month was derived by dividing the survey response by 3. For the baseline, the reported number of good days was based on the 1 month preceding study entry to report monthly good days.
Given that eptinezumab is the only IV-administered CGRP-targeted treatment, understanding participants’ perspectives of the infusion can inform shared decision-making. Participants’ infusion-related experience with eptinezumab was assessed, specifically concerns about receiving an infusion at baseline and early infusion experience at day 7. At baseline, participants rated their concerns about receiving an infusion as “not at all concerned,” “slightly concerned,” “moderately concerned,” “somewhat concerned,” or “concerned.” At day 7 after their first infusion, participants rated their perception of their first infusion experience by responding to the question, “Overall, I feel the infusion experience was positive” (response options: “strongly agree,” “agree,” “undecided,” “disagree,” or “strongly disagree”). The list of outcomes collected can be found in Supplemental Table S1.
Statistical analysis
The baseline analysis population comprised all participants who received at least the baseline infusion and completed the baseline assessment. The effectiveness analysis population, or full analysis set (FAS), at month 6 only included participants from the baseline analysis population who received their second eptinezumab infusion on schedule at month 3 (±15 days) and completed the month 6 survey within the predefined time windows (i.e. 90 days after last infusion; Supplemental Figure S1). The inclusion of only participants who received their second dose on time was to ensure that eptinezumab dosing and administration followed the recommendations in the US prescribing information, and the timing of the survey was predefined to ensure that the real-world effectiveness of the first two doses could be properly assessed. Furthermore, assessments conducted 90 days (rather than 30 or 60 days) after an eptinezumab dose take a conservative approach to evaluating efficacy, as outcomes are captured at the end of the dosing window. For all analyses, participants were pooled into a single eptinezumab-treated cohort, regardless of initial dose (100 or 300 mg) or any subsequent dose adjustments.
Baseline continuous variables were summarized using counts, mean, standard deviation (SD), median, and interquartile range, and baseline categorical variables were summarized using counts and percentages. Where appropriate, 95% confidence intervals (CIs) were provided for continuous or categorical data. For the key outcome of PGIC (i.e. proportion of participants who reached “much improved” and “very much improved” status), 95% CIs were calculated using the Wilson score method for binomial distribution. Similarly, for the binary outcomes of the proportion of participants achieving ≥50% and ≥75% reductions in MHDs from baseline at each time point, binomial proportion 95% CIs were calculated. Mean change from baseline was calculated for all post-baseline visits (day 7, month 3, and month 6) for the following outcome measures: good days, MIDAS total scores, and MIDAS-derived MHDs. The 95% CIs for mean changes from baseline were calculated using the Student's t distribution with degrees of freedom equal to N − 1, where N is the sample size at each time point. Participants with missing outcome assessments at any given visit were excluded from the analysis for that specific time point but retained for other available time points (available case analysis). No imputation methods were applied to the primary analyses (i.e. the primary analyses only included complete cases). The reasons for not completing the surveys were not collected.
Due to the decentralized nature of the data collection, reasons for treatment discontinuation were not systematically collected. As a consequence, the results could be subject to survivor bias. To assess the potential impact of missing data and to model within-participant changes over time, a number of analyses were conducted. Baseline participant characteristics (age, gender, BMI, time since migraine diagnosis, and migraine classification) and disease severity (measured by baseline good days, MIDAS, MIDAS-derived MHDs, and MIDAS-derived severity) were compared between participants in the FAS and those who did not meet the criteria for the FAS population to assess whether disease characteristics differed between these groups. A mixed model for repeated measures (MMRM) was applied to continuous outcomes, including good days, MIDAS total score, and MIDAS-derived MHDs. The model included time (month) as a categorical fixed effect, baseline value as a covariate, and a time-by-baseline interaction term. An unstructured covariance matrix was used to model within-participant correlations across time points, and the Kenward–Roger adjustment was applied to estimate denominator degrees of freedom. The impact of missing data was only assessed for continuous outcomes. Finally, a sensitivity analysis was performed that included all available surveys, regardless of whether participants received an infusion or a survey on time or not.
The significance level was set at P < 0.05 for all tests of statistical significance. No adjustments were made to account for multiplicity testing. The design and outcomes assessed were predefined in a protocol finalized prior to study initiation, and the analyses presented here were listed in the statistical analysis plan that was finalized prior to database lock. Analyses were conducted using R software, version 4.5.1.
Results
Participant characteristics
The baseline analysis population included 190 participants, of whom 86.8% (165/190) received a second infusion at month 3 and 73.2% (139/190) received a third infusion at month 6. The effectiveness analysis population (i.e. FAS) at month 6 consisted of 111 participants who had completed their second infusion at month 3 and had completed the survey at month 6 on schedule (Supplemental Figure S1).
Most participants were female (88.9% [169/190]) and White (83.7% [159/190]), with a mean (SD) BMI of 28.7 (8.1) kg/m2; over half of the participants were between 35 and 54 years of age (53.7% [102/190]; Table 1). More than half (55.3% [105/190]) of participants had a migraine history exceeding 15 years, and 81.1% (154/190) reported being given a diagnosis of CM by their treating physician. The most commonly reported comorbidities at baseline were anxiety, allergies, depression, and insomnia. All participants had previously taken ≥1 CGRP-targeted preventive therapy that had failed. The mean (SD) number of prior CGRP-targeted preventive therapies that participants reported as failed was 2.6 (1.3); 24.7% (47/190) of participants had reported one CGRP-targeted preventive therapy that had failed, 20.0% (38/190) reported two CGRP-targeted preventive therapies that failed, and 55.3% (105/190) reported ≥3 CGRP-targeted preventive therapies that failed. Most participants reported their last CGRP-targeted preventive therapy was used more than 3 months prior to study entry (62.6% [119/190]). Participants reported having used acute prescription medications for a mean (SD) of 10.5 (5.2) days and over-the-counter medications for 7.7 (6.3) days during the prior month. At baseline, the mean (SD) total MIDAS score was 108.5 (70.0), and the severity score was 7.0 (1.6). Additionally, the baseline mean (SD) number of MIDAS-derived MHDs was 20.6 (8.8). Brain fog was reported by 93.7% (178/190) of participants and frequently occurred alongside other cognitive symptoms, such as difficulty reading, making decisions, and completing tasks. Taken together, these baseline characteristics indicate a treatment-experienced population with substantial migraine-related disability and burden.
Baseline characteristics, treatment history, and disease burden.
BMI: body mass index; CM: chronic migraine; CGRP: calcitonin gene-related peptide; IQR: interquartile range; MHD: monthly headache day; MIDAS: Migraine Disability Assessment; SD: standard deviation.
Data were missing for 5 (2.6%) participants.
Additional options included American Indian or Alaska Native, Asian, and Native Hawaiian or Other Pacific Islander.
Individual comorbidities reported by ≥10 participants at baseline. Percentages were not calculated because responding to comorbidity questions on the survey was optional; thus, skipping a question did not necessarily mean that a participant did not have that comorbidity.
3-month reporting time frame. The total MIDAS score was derived by summing the individual scores from the first 5 MIDAS questions assessing the impact of headache on work/school, household work, and family/social/leisure activities.
MIDAS-derived MHDs were based on the MIDAS question: “On how many days in the past 3 months did you have a headache? (If a headache lasted more than 1 day, count each day).” Values were divided by 3 to obtain MHDs.
MIDAS severity score ranges between 0 and 10; a higher score indicates greater severity.
Number of days participants reported having a good day in the month prior to the baseline survey. A definition of a good day was not provided to participants.
Change in overall disease status on the PGIC
Improvement in PGIC was evident early; 21.5% (23/107 [95% CI: 14.8–30.2%]) of participants reported that they were “much improved” or “very much improved” at day 7, increasing to 44.1% (49/111 [95% CI: 35.3–53.4%]) of participants in the effectiveness analysis population at month 6 (Figure 1(a)). A total of 59.8% (64/107 [95% CI: 50.3–68.6%]) of participants reported any improvement (“minimally improved,” “much improved,” or “very much improved”) at day 7 and 75.7% (84/111 [95% CI: 66.9–82.7%]) reported any improvement at month 6 (Figure 1(b)).
Proportion of participants reporting (a) “much/very much” improved and (b) any improvementa on the PGICb at day 7, month 3, and month 6. CI: confidence interval; PGIC: Patient Global Impression of Change. aAny improvement includes the responses, “very much improved,” “much improved,” and “minimally improved.” bPGIC was assessed using the question: “Since your first (eptinezumab) infusion how would you describe the change (if any) in ACTIVITY LIMITATIONS, SYMPTOMS, EMOTIONS, and OVERALL QUALITY OF LIFE, as related to your migraine? Choose ONE. (‘very much improved,’ ‘much improved,’ ‘minimally improved,’ ‘no change,’ ‘minimally worse,’ ‘much worse,’ ‘very much worse’).”
Change in MIDAS-derived MHDs
The MIDAS-derived mean (SD) number of MHDs decreased from 20.0 (9.1) at baseline to 14.2 (10.3) at month 3 and 13.2 (10.3) at month 6. This represented mean (95% CI) reductions of 5.6 (3.7–7.5) MHDs per month at month 3 (P < 0.001) and 6.8 (5.2–8.3) MHDs at month 6 (P < 0.001; Figure 2).
Mean change from baseline in MIDAS-derived MHDs at month 3 and month 6.a CI: confidence interval; MHD: monthly headache day; MIDAS: Migraine Disability Assessment; SD, standard deviation. aMIDAS-derived MHDs were based on the MIDAS question: “On how many days in the past 3 months did you have a headache? (If a headache lasted more than 1 day, count each day).” Values were divided by 3 to obtain MHDs. bP <0.001 versus baseline.
A ≥50% reduction from baseline in MIDAS-derived MHDs was reported by 35.8% (34/95 [95% CI: 26.9–45.8%]) of participants at month 3 and 44.1% (49/111 [95% CI: 35.3–53.4%]) at month 6 (Figure 3(a)), while a MIDAS-derived ≥75% MHD reduction was reported by 17.9% (17/95 [95% CI: 11.5–26.8%]) of participants at month 3 and 26.1% (29/111 [95% CI: 18.9–35.0%]) at month 6 (Figure 3(b)).
Participants with (a) ≥50% and (b) ≥75% MIDAS-derived MHDa reduction at months 3 and 6. CI: confidence interval; MHD: monthly headache day; MIDAS: Migraine Disability Assessment. aMIDAS-derived MHDs were based on the MIDAS question: “On how many days in the past 3 months did you have a headache? (If a headache lasted more than 1 day, count each day).” Values were divided by 3 to obtain MHDs.
Change in PGIC and MIDAS-derived MHDs in select subgroups
In the subgroup analysis, 40.7% (95% CI: 24.5–59.3%), 45.0% (25.8–65.8%), and 45.3% (33.7–57.4%) of participants reported that they were “much improved” and “very much improved” at month 6 in the subgroups with 1, 2, and ≥3 previous CGRP-targeted treatments that had failed (Supplemental Figure S2). Additionally, the MIDAS-derived mean number of MHDs decreased from 18.2, 19.9, and 20.7 at baseline to 12.9, 13.5, and 13.2 at month 6 in the subgroups with 1, 2, and ≥3 previous CGRP-targeted treatments that had failed (Supplemental Figure S3). Findings for other outcomes evaluated in these subgroups paralleled the trends seen for PGIC and MIDAS-derived MHDs.
Change in the total MIDAS score and MIDAS-derived headache severity
The mean (SD) total MIDAS score at baseline was 102.3 (67.3), which decreased to 73.0 (66.6) by month 3 and 77.0 (71.6) by month 6, corresponding to mean (95% CI) reductions from baseline of 31.1 (95% CI: 18.6–43.6; P < 0.001) and 25.2 (95% CI: 13.6–36.9; P < 0.001) at months 3 and 6, respectively (Supplemental Figure S4).
Mean (SD) MIDAS-derived headache severity was rated as 7.0 (1.6) at baseline, 6.3 (1.8) at month 3, and 6.3 (1.9) at month 6, for mean (95% CI) reductions from baseline of 0.7 (0.4–1.0) at month 3 and 0.7 (0.4–1.0) at month 6 (both P < 0.001).
Change in brain fog
A total of 93.7% (104/111) of participants in the effectiveness analysis population reported brain fog at baseline. Among these participants who completed the follow-up question at the relevant follow-up visit, improvement was seen early in 39.0% (39/100 [95% CI: 30.0–48.8%]) by day 7, increasing to 44.4% (40/90 [95% CI: 34.6–54.7%]) by month 3, and 55.8% (58/104 [95% CI: 46.2–64.9%]) by month 6 (Figure 4).
Improvement in brain fog at day 7, month 3, and month 6.a,b aBrain fog was assessed using the following question for participants who reported brain fog (i.e. having difficulty concentrating or focusing, trouble finding the right words or speaking, feelings of mental cloudiness) as part of their migraine disease at baseline: “If yes, rate the extent your symptoms have changed since starting eptinezumab (‘very much improved,’ ‘much improved,’ ‘minimally improved,’ ‘no change,’ ‘minimally worse,’ ‘much worse,’ ‘very much worse’).” bThe participants in this analysis are those from the 6-month effectiveness analysis set (n = 111) who responded “yes” to having brain fog at baseline and completed the improvement survey question at follow-up for brain fog.
Change in participant-defined good days
The mean (SD) number of participant-defined good days per month was 10.0 (6.7) days at baseline, 15.3 (9.3) at month 3, and 16.3 (8.6) at month 6. These changes represented mean (95% CI) increases from baseline of 5.3 (3.2–7.3) days per month at month 3 and 6.3 (4.7–7.9) at month 6 (both P < 0.001; Figure 5).
Mean change from baseline in participant-reported good days per month.a CI: confidence interval; SD: standard deviation. aAt baseline, good days were assessed using the question: “How many good days did you have in the past month? (0–31).” At subsequent assessments, the assessment was based on the question: “How many good days did you have in the past 3 months? (0–90 days).” bP <0.001 versus baseline.
Change in acute medication use
At baseline, 62% (54–73%; 69/111) and 45% (36–54%; 50/111) of participants reported 10 or more days per month of prescription acute medication use and over-the-counter acute medication use. At 6 months, the proportion of participants reporting 10 or more days per month of prescription acute medication use and over-the-counter acute medication use decreased to 40% (31–49%: 44/111) and 37% (29–46%; 42/111).
Eptinezumab infusion experience (expectation [prior to infusion] and actual [day 7])
Prior to their first infusion, most participants (72.1%) reported being “not at all concerned” (35.3% [67/190]) or only “slightly concerned” (36.8% [70/190]) about the infusion (Figure 6(a)). Among those expressing concern before the first infusion, the most common reason was about the infusion process itself (47.2% [58/123]). At day 7, 80.6% (141/175) of participants reported a positive experience with their first infusion (Figure 6(b)).
Infusion experience: (a) baseline level of concerna (N = 190); (b) day 7 positive experienceb,c (n = 175). IV: intravenous. aBaseline level of concern about infusion was assessed using the question “How concerned are you about receiving an infusion?” bIV treatment experience at day 7 after the first infusion was assessed using the question “Now that you have had your first eptinezumab infusion, please rate how much you agree/disagree with the following statement: Overall, I felt the infusion experience was positive.” cOf the 190 participants included in baseline analyses, 15 did not have data available for the day 7 assessment of infusion experience.
Sensitivity analyses
Out of the 190 patients with baseline data, 111 patients were included in the FAS and 79 were not (due to discontinuation of treatment, missing month 6 assessments, or completion of assessments outside the protocol-defined time window). Baseline characteristics, treatment history, and disease burden were generally similar for participants who were and were not included in the FAS (Supplemental Table S2a). The baseline severity of disease for the FAS and non-FAS participants was comparable (P > 0.05), with the exception of good days: the participants who were not in the FAS had a lower number of good days at baseline (P = 0.01), indicating a potentially higher severity level (Supplemental Table S2b).
In sensitivity analyses accounting for missing data via MMRM, results for change from baseline in good days, MIDAS total score, and MIDAS severity score were generally comparable to those from the primary analyses of these outcomes (Supplemental Table S3). At month 3 and month 6, respectively, 166 and 144 participants received an infusion and 142 and 123 participants filled out a survey, without any requirement for the survey or infusions to be on time. This compares to 95 participants at month 3 and 111 at month 6 in the FAS. The results between all available surveys and the FAS are comparable (Supplemental Table S2c).
Discussion
In this real-world, non-interventional INFUSE study, participants in whom ≥1 prior CGRP-targeted preventive therapy had failed and who switched to IV eptinezumab reported clinically meaningful improvements in multiple PROs over 6 months. Improvements were noticeable as early as day 7 and increased through months 3 and 6. The results presented here were supported by sensitivity analyses accounting for missing data (including dropouts).
Participants in this 6-month study entered with substantial self-reported disease burden despite reporting prior use of CGRP-targeted preventive therapies, with the majority of participants reporting CM, severe migraine-related disability (based on the total MIDAS score), a high number of MHDs, and a high prevalence of brain fog at baseline. For individuals with high migraine burden, IV eptinezumab offers a pharmacologically distinct option relative to SC or oral CGRP-targeted therapies. Its rapid onset of action and 100% bioavailability with IV administration 34 ensure rapid systemic exposure to the full dose, which may allow for capture of a greater amount of circulating CGRP. This is particularly relevant given the prior data showing substantial residual, free CGRP at the end of dosing intervals for SC anti-CGRP mAbs. 35 Further, the mean BMI (28.7 kg/m2) indicates that this was an overweight population; unlike other routes of administration, the absorption and bioavailability of IV-administered treatments, like eptinezumab, 7 are not affected by BMI and may be favorable for this type of population.
MHDs in this study were derived from MIDAS, introducing the potential for recall bias compared with daily headache diary-based assessments. Nonetheless, these findings contribute to the growing body of real-world evidence by evaluating the benefits of switching from SC or oral CGRP-targeted preventive therapies to IV eptinezumab. Although many participants in the INFUSE study population were severely impacted by migraine, despite prior CGRP-targeted preventive treatment, 44% reported a ≥50% reduction in MIDAS-derived MHDs after switching to eptinezumab. A higher threshold of clinical response, a ≥75% reduction in MHDs, was reported in 26% of participants at month 6.
Ongoing assessment of migraine and headache frequency is an essential component of effective migraine management. Reductions in headache frequency were reported in INFUSE, with a decrease in the number of MIDAS-derived MHDs of 6.8 at month 6. Alongside a reduction in MIDAS-derived MHDs, the proportion of participants who reported 10 or more days per month of prescription and over-the-counter acute medication use decreased by 22 and 8 percentage points, respectively. Given that people living with migraine often require multiple therapies to manage their disease and can have other comorbidities that require medical management, overall medication burden can be an important consideration in their care plan. Furthermore, in clinical practice, novel outcomes should be explored to not only capture migraine frequency and medication use but also the holistic burden of disease.
As migraine is a multifaceted disease and individuals living with migraine can have variable lived experiences, it is important to assess both the overall burden of migraine and the holistic impact of available preventive treatments. Within 1 week of initiating eptinezumab treatment, more than half of participants reported improvements on the PGIC, with that proportion increasing to approximately three-quarters at month 6. Previous findings in studies exploring the impact of CGRP-targeted treatment switches have shown diminishing treatment benefits after multiple SC CGRP-targeted treatment switches, which are contrasted against the consistent PGIC improvements regardless of the prior number of CGRP-targeted treatments in INFUSE. 12 These findings may point to something unique in the transition to eptinezumab treatment; however, additional research is needed. A structured comparison of INFUSE findings to this and other switching studies was not pursued, given the differing study designs, different clinical outcomes that were utilized, and the lack of studies conducted in the United States (Supplemental Table S4).
MIDAS, which is often used in clinical trials and practice, provides insight into the functional impairments associated with migraine, and evaluating change provides an opportunity to assess the impact of preventive treatment on migraine-related disability. In INFUSE, participants reported a reduction in the total MIDAS score of approximately 25 points, indicating a reduction in migraine-related disability.
Furthermore, the concept of brain fog and cognitive impairment in migraine has evolved over the past several years, with brain fog emerging as a frequently reported symptom. 36 At the time of the study design, no validated migraine-specific measure was available for such concepts; therefore, exploratory measures were used. In the current study, 94% of participants reported brain fog at baseline, of whom 56% reported improvements in brain fog by month 6, with improvements evident as early as 7 days after the first infusion. Although the question assessing brain fog was not a validated measure, given the high prevalence of brain fog in adults with migraine, these findings support the importance of evaluating and managing brain fog and cognitive impairments in a clinical setting.
In the INFUSE study, the increase in participant-defined good days was also evaluated, with the definition of a good day independently defined by each participant. Although not a validated assessment, the premise of a good day as a measure of preventive treatment success was to allow participants to consider a variety of elements important to the individual, such as a day in which the participant could engage in their life in a meaningful way, rather than strictly as a day free of headache or migraine. Additionally, it could allow for the capture of burden in multiple phases of the disease (e.g. ictal and interictal, prodrome through postdrome) and the capture of other elements of migraine management (e.g. side effects of medications). At month 6, a 6.3-day increase in the number of participant-defined good days was reported. Although reductions in MHDs and increases in good days are not direct inverses, their concordance may offer valuable insights and support the potential of an individualized approach for assessing treatment effectiveness in real-world settings.
Lastly, the administration of eptinezumab on a quarterly basis under supervision in an infusion center could support improved adherence, contrasted with self-administered monthly SC injections or daily oral dosing at home, as demonstrated by the 73% treatment persistence observed at month 6 in this study. In a retrospective claims database analysis including >1200 participants, persistence for preventive SC anti-CGRP mAbs at 6 months ranged from 44% to 50%. 37 These findings are consistent with a separate claims database analysis including >42,000 participants on CGRP-targeted preventive treatments that showed an approximately 30% to 60% higher risk of discontinuation for SC anti-CGRP mAbs compared with eptinezumab, based on a 15-day treatment gap. 38 The number of surveys is reportedly less than the number of infusions in the current study, as participants could have skipped surveys even if they received an infusion.
The efficacy and tolerability of eptinezumab in participants with EM and CM who were naive to CGRP-targeted preventive therapy have been established in multiple phase 3 clinical trials,13–15,39 while data on effectiveness in adults with prior CGRP-targeted therapy experience is relatively limited. The REVIEW study, in which 89% of participants reported prior exposure to SC CGRP-targeted mAbs, 17 demonstrated improvements in self-reported good days and brain fog, decreases in acute medication use, and improvements in the elements of daily living with eptinezumab treatment. 17 Across studies, despite differences in populations and outcome assessment methods (e.g. diary-based end points in clinical trials vs. recall-based PROs in INFUSE), the direction and magnitude of treatment effects were generally consistent, with clinically meaningful improvements observed in migraine frequency, PROs, and functional measures. The current INFUSE study expands on these findings by providing longitudinal, real-world data on the impact of switching to IV eptinezumab in participants with prior CGRP-targeted treatments that had failed, using a broad range of PROs at prespecified time points. Furthermore, consistent with prior clinical trials,13–15,17,39 eptinezumab demonstrated early effectiveness in INFUSE that appeared to increase over time, with improvements becoming more pronounced by month 6. The observed progressive improvement across evaluated measures suggests the value of allowing sufficient treatment duration to realize the full benefit of eptinezumab.
Limitations
This study was subject to the general limitations associated with an observational, non-interventional design, including the absence of randomization and a control group, which limits causal inference. In particular, participants received an offer to join the study when they called to schedule their first infusion of eptinezumab. Therefore, the potential infusion concerns for this population might not be reflective of the broader migraine population, who may choose not to pursue a treatment given through an infusion. Additionally, data were collected from participants receiving infusion through the Vyepti® (eptinezumab-jjmr) Infusion Network, and therefore the experience may not be generalizable to all individuals receiving infusions at other sites of care. Furthermore, eptinezumab is available in two doses in the United States (100 and 300 mg). The findings of this study should therefore be interpreted in the context of dose selection and potential dose escalation, both of which may influence treatment outcomes with eptinezumab. As a result, these findings may not be generalizable to countries where only the 100-mg dose of eptinezumab is available.
Although some of the items in the survey used to collect participant-reported data in the current study were based on validated PROs (e.g. MIDAS, PGIC), others were novel, study-specific questions designed to assess the selected outcomes of interest, such as cognitive impairment/brain fog, participant-defined good days, and infusion experience. Thus, until validated instruments for those concepts are available, the results of the INFUSE study outcomes should be considered exploratory in nature.
Headache frequency was assessed using the MIDAS 3-month recall item expressed as an average monthly estimate. While this approach provides a pragmatic and validated measure of headache burden in real-world settings, it may be subject to recall bias and averaging effects and does not capture day-to-day variability as precisely as prospective diary-based methods. Accordingly, these findings should be interpreted as recall-based estimates of MHD burden rather than diary-confirmed MHDs or monthly migraine days. Nevertheless, strong concordance between MIDAS-derived MHDs and daily diary-derived monthly migraine days has also been reported in other migraine studies where both tools were administered, which supports the potential validity of using MIDAS-based headache frequency measures in real-world settings.29,40
Given the reliance of the study on participant-reported data, missing data may have impacted outcomes; however, results from the MMRM-based sensitivity analysis accounting for missing data aligned with those from the primary analysis. However, survivor bias cannot be definitively excluded as reasons for discontinuation were not collected. In addition, the length of the survey was limited to allow participants to complete the survey in a relatively short amount of time (approximately 25–30 min for the baseline survey and 15–20 min for follow-up surveys). Thus, some relevant clinical factors (e.g. medication-overuse headache) were not captured, which may limit the characterization of the study population and the interpretation or generalizability of the findings. Furthermore, concomitant medication use and treatment changes over time were not controlled for, consistent with the observational study design. These factors should be considered when interpreting treatment effectiveness in a real-world setting where individuals with migraine are likely using multiple preventive and acute therapies.
Lastly, treatment failures were reported by participants and were not obtained from physicians and/or chart review. Most participants reported that their last CGRP-targeted preventive therapy was used >3 months prior to study entry; however, data were not collected on the actual duration of those prior treatments before discontinuation, which limits interpretation of whether those treatments were given an adequate trial before being deemed ineffective. The use of participant-reported failure rather than uniformly applied objective criteria (e.g. response rate) to define treatment failure likely led to the inclusion of a heterogeneous population, limiting comparability with other studies in preventive-treatment-experienced individuals. Despite these potential limitations, the INFUSE study provides valuable real-world insights into the effectiveness of switching to eptinezumab in adults for whom prior CGRP-targeted preventive therapies had failed.
Conclusions
In this descriptive real-world study, treatment of eptinezumab after ≥1 prior CGRP-targeted preventive therapies had failed was associated with reductions in migraine burden and clinically meaningful improvements in well-established and novel PROs. Participants who switched to IV eptinezumab reported improvements in overall migraine status, reductions in MHDs, improvements in brain fog, an increase in the number of good days, and a reduction in migraine-related disability with the first infusion, continuing through the subsequent infusion. As a survey-based, single-arm non-interventional study, the findings of this study should be considered descriptive, and additional research is needed to confirm the findings.
Clinical implications
The real-world INFUSE study is evaluating the effectiveness of IV eptinezumab for the preventive treatment of migraine in participants in whom ≥1 CGRP-targeted preventive therapy had previously failed.
Switching to eptinezumab treatment was associated with reduced migraine burden and improvements in multiple participant-reported migraine outcomes, including headache frequency, disability, disease status, and brain fog, despite substantial baseline disease burden.
Supplemental Material
sj-docx-1-rep-10.1177_25158163261455673 - Supplemental material for Patient-reported outcomes from a real-world evidence study in individuals with migraine initiating eptinezumab after ≥1 prior calcitonin gene-related peptide (CGRP)-targeted preventive therapy
Supplemental material, sj-docx-1-rep-10.1177_25158163261455673 for Patient-reported outcomes from a real-world evidence study in individuals with migraine initiating eptinezumab after ≥1 prior calcitonin gene-related peptide (CGRP)-targeted preventive therapy by Amaal J. Starling, Stephane A. Regnier, Seema Soni-Brahmbhatt, Susanne F. Awad, S. Wald Grossman, Damian Fiore, Sandeep Sharma and Emad Estemalik in Cephalalgia Reports
Footnotes
Acknowledgments
The authors wish to thank Michelle Townshend for her valuable contributions and assistance in the preparation of this manuscript. Anna Douyon, PhD, and Kiley Margolis, PharmD, of Lumanity Communications Inc. provided medical writing support, which was funded by Lundbeck LLC (Deerfield, IL, USA) and in accordance with Good Publication Practice guidelines.
Ethical considerations
All study procedures were conducted in accordance with the principles of the Declaration of Helsinki, Good Pharmacoepidemiology Practices guidelines of the International Society for Pharmacoepidemiology Good Pharmacovigilance Practice, and local rules and regulations. A US-based Central Institutional Review Board (IRB), Advarra IRB (IRB Registration Number: 00000971), was consulted to determine whether full protocol submission was required or if IRB notification was sufficient, and all necessary regulatory submissions or notifications were completed in accordance with applicable requirements.
Consent to participate
All participants completed electronic informed consent on the study's web-based platform.
Consent for publication
The authors agree to publish with Cephalalgia Reports, if the manuscript is accepted.
Author contributions
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: These analyses were funded by Lundbeck LLC (Deerfield, IL, USA).
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: AJS received consulting fees from AbbVie, Allergan, Amgen, Amneal, Axsome Therapeutics, Eli Lilly, eNeura, Everyday Health, Impel, Lundbeck, Mediq Medscape, Miller Medical, Neurolief, Novartis, Pfizer, Salvia, Satsuma, Teva, Theranica, WebMD, and Woodberry Associates. SAR, SS-B, SFA, SWG, DF, and SS are full-time employees of Lundbeck. EE served on advisory boards and/or speakers bureaus for AbbVie, Axsome Therapeutics, and Lundbeck, and as a faculty advisor for Eli Lilly.
Data availability statement
Data are available from the corresponding author upon reasonable request.
Supplemental material
Supplemental material for this article is available online.
References
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.
