xRead - Episodic Vertigo (January 2026)

1266 |

15264610, 2024, 10, Downloaded from https://headachejournal.onlinelibrary.wiley.com/doi/10.1111/head.14835 by University Of Michigan Library, Wiley Online Library on [21/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

HEADACHE

of other vestibular disorders other than treated benign paroxys mal positional vertigo, a failure of treatment with > 4 prophylactic medications for VM, prior or current prophylactic treatment with a CGRP medication, or serious medical or psychiatric conditions at the discretion of the investigator. Participants were allowed to continue up to two prophylactic VM medications on a stable dose during the trial. All participants provided written informed consent. They were also allowed to use abortive treatment as needed. The only major change to the protocol during the trial was increasing the age of eli gibility from 65 to 75. This was done very early in the trial, 2months after the first participant was recruited, based on consensus that it would both increase generalizability and recruitment, without signif icant negative trade-offs. Each day of the trial, participants were texted the following: “Over the last 24 hours, what is the highest level of dizziness that you have experienced?” Answer choices were (0) no dizziness, (1) mild dizzi ness, (2) moderate dizziness, and (3) severe dizziness. If the partici pant chose moderate or severe dizziness, they would get a second question asking about usage of medications for relief, and if they answered in the affirmative, a third text provided an opportunity to give the medication names and dosages. Responses were auto matically entered into the REDCap database via Twilio. 2,15,16 Any day with moderate or severe dizziness was considered a DDD. The primary outcome measure was the change in VM-PATHI score, comparing baseline (SV2) to after 3 months of treatment (SV4). VM-PATHI is a validated, disease-specific patient-reported outcome measure for VM. 17 It consists of 25 items, like “dizziness with head or body movements,” which participants rate as (0) no problem, (1) mild problem, (2) moderate problem, (3) severe problem, or (4) prob lem as bad as it can be. VM-PATHI shows good test–retest reliability, discriminant validity, and responsiveness. Scores range from a maxi mum of 100 to a minimum of 0. Secondary outcome measures included the change in the Dizziness Handicap Inventory (DHI), comparing scores at baseline to after 3months of treatment (SV4). 18 The DHI is the most widely used metric of vestibular outcomes. Scores range from a maximum of 100 to a minimum of 0. The minimal clinically important differ ence has been calculated at 11 points; however, a recent study found that the minimal detectable change was 18 points. 19,20 Another sec ondary endpoint was the change in PROMIS SF (Patient-Reported Outcomes Measurement Information System Short Form) v1.2 Global Health score, which generates both a physical and mental subscale score. Reduction in DDDs from baseline to month 4 was a secondary outcome measure. No changes were made to endpoints after the trial commenced. Additional data collected included vital signs, physical exam, assessment of adverse events and serious adverse events, medical history, concomitant medications, Headache Impact Test, Patient Health Questionnaire-9, Generalized Anxiety Disorder-7, complete Outcomes

blood count, comprehensive metabolic panel, a pregnancy test, an electrocardiogram, and an end of study form at the final visit. The end of study form asked whether participants believed they re ceived drug or placebo. It also asked, “How effective was the med ication you received in the trial?” which was rated as: “not at all,” “helped a little,” “it helped a moderate amount,” “helped quite a lot,” or “I felt cured.”

Statistical analysis

In total, 50 participants were planned for randomization. Assuming a dropout rate of 80% and a standard deviation of VM-PATHI of 16, this would provide 80% power to detect a 15-point difference between placebo and galcanezumab arms, with an alpha of 0.05, with a two-tailed t -test. The modified intent to treat (mITT) analy sis included all participants who were randomized and treated with galcanezumab or placebo and who completed the minimum data necessary for analysis. The per protocol analysis included all par ticipants who completed the trial without a major protocol devia tion. The primary outcome measure was a difference in differences analysis comparing placebo to galcanezumab. The difference in VM-­ PATHI score between baseline (SV2) and month 4 (SV4) compared placebo change from baseline to galcanezumab change from base line. This was assessed using an independent groups Student's t -test, both one- and two-tailed tests. Similarly, change in each group (pla cebo and galcanezumab) in DHI between baseline (SV2) and month 4 (SV4) was calculated and compared in an independent groups two-tailed Student's t -test. The change in monthly DDD count from baseline to month 4, comparing placebo to galcanezumab, was also assessed the same way. Given the small size of the trial, no interim analysis was planned. Statistical analyses were carried out in Stata/ SE version 14.2 (StataCorp LLC, College Station, TX). This study was funded by Eli Lilly and Company as an investigator-initiated trial. The study team designed and executed the study and performed all data analysis and manuscript writing. Descriptive statistics were used to summarize all the clinical characteristics and demographic outcomes. These were summa rized overall, by treatment group, by assessment time, and by treat ment group and time. Continuous variables were summarized using means, medians, standard deviations, and ranges. Categorical data were summarized by number and percent. Bivariate analyses com paring groups and time points were summarized using independent groups Student's t -tests and one-way analysis of variance for con tinuous variables that were not extremely skewed and by Wilcoxon rank sum and the Mann–Whitney U test for variables requiring non-­ parametric statistics. Categorical analyses used chi-squared tests of independence, Fisher exact tests, and Z -tests of proportions. Effect sizes were calculated using Cohen's d , with the standard deviation of the control group as the denominator. Significance was assessed using an alpha of 0.05. Missing data were minimized by daily mon itoring of text message data by the study coordinator and an elec tronic survey design that did not allow for skipping questions.