Medication overuse headache generally refers to monthly headache days of ≥15 per month resulting from acute “medication overuse” (MO) defined as the frequent consumption of acute pain medications above a certain threshold (e.g., 10 or 15 days per month depending on the medication type) for ≥3 months. Diagnostic criteria for MOH were introduced in the International Classification of Headache Disorders, 2nd edition (ICHD-2) (1) without a the requirement of the diagnosis of chronic migraine (CM; see Appendix for details on MOH and CM). In the third edition beta version (ICHD-3β), updated in 2013, concurrent CM and MOH are diagnosed if both criteria are met (2). These criteria remain the same in the ICHD-3 criteria.

The prevalence of MOH is still uncertain (3) and varies from 2% in a recent population-based study (4) to 60% in the Global Burden of Disease Study (5). Whether MOH is a cause or consequence of frequent acute medication intake has been debated for years (6, 7). Nonetheless, MO and MOH management are important in clinical care. MO is associated with an increased risk of chronic headache (8) and chronic migraine (9, 10).

Prognosis of MOH due to different type of chronic headache is probably variable. Although there is evidence that chronic tension-type headache (CTTH) and CM may have partly distinct pathophysiological mechanisms, clinical experience suggests that these two clinical entities may be at opposite poles of a single disease spectrum (11, 12). Of note, CM with MO represents a more severe subgroup of patients and can be difficult to treat (13). In a preplanned subgroup analysis that evaluated efficacy and safety of erenumab in patients with CM and MO, the medication overuse subgroup had a higher proportion of failing to respond to preventive treatment (14).

Effective education for people with MOH plays a major role in effective withdrawal therapy. Successful educational interventions rely largely on a productive doctor-patient alliance (15) and patients' motivation for adherence. In addition, it is often recommended to add a preventive therapy or optimize existing preventive regimens to reduce the number of monthly headache days, improve sense of self efficacy and as a result reduce both the physical and psychological need for acute medication for migraine. Preventive treatments may be pharmacologic, non-pharmacologic, or both.

Moreover, the addition of behavioral therapies to medication withdrawal has been shown to significantly improve outcomes and the maintenance of benefits when measured at 3 years post treatment) (16, 17). Results from a previous systematic review (18) and a newly published randomized clinical trial (19) support a program of prophylaxis treatment together with abrupt withdrawal of acute analgesic intake for treating MOH. Evidence from randomized controlled trials confirms complete withdrawal of the offending acute medication(s) is an effective treatment for MOH (20–22), but withdrawal symptoms may lead to treatment failure or rebound reactions. In the past two decades, it has been suggested that prednisone may be effective in reducing withdrawal symptoms (23, 24). However, results from a well-designed controlled trial confirmed no therapeutic effect of prednisolone on MOH (25). Therefore, an acceptable level of evidence in the recommending use of any of prophylactic medications or treatments other than behavioral therapies is still undetermined up to now (26). Hence the utility in bridging therapy using a preventive drug during the withdrawn phase needs to be determined. Topiramate is the first choice for people with MOH despite frequent adverse events (27). Data from our ongoing network meta-analysis (PROSPERO, CRD 42021193370) indicate that despite lower safety and greater tolerability issues, topiramate has large beneficial effects on reducing headache frequency and monthly acute medication intake frequency. Onabotulinumtoxin A and the humanized monoclonal calcitonin gene-related peptide (CGRP) antagonists eptinezumab and fremanezumab, are also helpful for reversion to no medication overuse; however, until CGRP antagonists are readily available around the world there are challenges with cost-effectiveness and accessibility.

Rhizoma Gastrodiae, a kind of traditional Chinese herb medicine, was firstly recorded for treating primary headache in a famous compendium of Chinese medicines called Sheng Nong's Herbal Classic in 25 AD to 220 AD. Gastrodin (4-[β-D-glucopyranosyloxy] benzyl alcohol) is the main bioactive constituent of Rhizoma Gastrodiae and has been applied clinically as an analgesic to treat migraine and other types of headache for more than 30 years in China (28). Several potential anti-migraine mechanisms of gastrodin have been discovered including inducing a dose-dependent reduction of CGRP- mRNA expression similar to flunarizine (29), preventing the release of CGRP from pre-synaptic central projections, inhibiting the post-synaptic effects of the ERK1/2 downstream signaling pathway to inhibit the firing of second-order trigeminal nociceptive neurons (30), altering acid-evoked membrane excitability of rat dorsal root ganglion neurons and decreasing the amplitude of the depolarization (31). Gastrodin has determined to be safe for use in toxicity studies in animals and occasionally reported adverse drug reactions (ADR) or events (ADE) in clinical practice (32). Recommended dosage of oral gastrodin for treating episodic migraine is 300 mg/day. Although there is no evidence that gastrodin is superior to other analgesics in treating primary headache, low quality evidence obtained from the tremendous amount of case series and observational studies indicates that gastrodin is beneficial for reducing headache frequency and headache intensity with increased dosage. This may largely be explained by the phenomenon that gastrodin is not brain penetrant with a cerebrospinal fluid /plasma ratio of only 5.5% (33). So theoretically, increasing the dosage of gastrodin maybe an ideal option for better efficacy. Previously, the safety of intravenous 40 mg/kg (up to 2,400 mg/day for a 60 kg human) gastrodin had been determined in a clinical trial (34). For efficacy and safety concern, the dosage of oral gastrodin in the present trial will be increased to 600 mg/day.

We will test the hypothesis that gastrodin may serve as a candidate therapy drug during the detoxification phase of MOH with advantages of improving chances of treatment adherence and success. The present study aims to determine the efficacy and tolerability of oral gastrodin as a prophylaxis treatment in MOH.

This is an investigator-initiated, multicenter, 1:1 ratio randomized, double-blind, parallel-group, placebo-controlled trial which consisted of a 28-day screening period, and a 12-week intervention period, with a final evaluation at week 24. A schematic diagram of study design is provided in Figure 1. The protocol of the present study has been registered in the Chinese Clinical Trial Registry (ChiCTR2200063719).

Eligible patients will be screened at neurological outpatient clinic from three study sites at different provinces in China: Affiliated hospital of Yunnan University, Kunming, Yunnan; Xiangya Changde Hospital, Changde, Hunan; and Affiliated hospital of Chengdu University, Chengdu, Sichuan; and Puer People's Hospital, Puer, Yunnan. These sites are all tertiary hospitals which may provide extensive healthcare service for persons with headaches.

All patients will receive full individualized brief intervention (BI) education according to Kristoffersen et al.'s strategy (36) before randomization.

In the gastrodin group, patients will receive gastrodin 600 mg/day (200 mg tid) from baseline to the end of week 12. In the placebo group, matching placebo made of starch will also be prescribed three times a day. If a patient experience unacceptable adverse events as recorded in the instructions for gastrodin such as dry mouth, dizziness and upper abdominal discomfort, the target dose will be tapered down to 300 mg/day.

During the 24-week study period, their usual acute medication intake will be restricted to ≤ 2 days per week. Participants will be asked to return the experimental drug (or placebo) packaging at week 4, and week 8 in the clinic in order to get drugs for the next 4 weeks.

As there is a lack of guidelines for reporting results from clinical trials on MOH, we will adopt Hagen et al.'s suggestion on endpoints for MOH studies in follow-up including headache days/month, acute medication use days/month, headache intensity, cured MOH, and responder rate (37).

Primary efficacy endpoint is reduction in the mean number of monthly headache days from baseline at 24-week follow-up.

Secondary efficacy endpoints include:

All subjective adverse events (41) will be collected from the time of allocation through the last visit (week 24). The investigators are responsible for assessment of the severity and causal relationship of all AEs to study intervention. Documentation on start date and duration of AE, treatment measures taken for the AE will be recorded.

Tolerability will be assessed by proportion of drop-outs due to AEs throughout double-blind period, with lower dropout rates indicate higher tolerability.

Eligible participants will be enrolled from January, 2023 to December, 2024. During the 28-day screening period, baseline information on demographics, number of monthly headache days, number of monthly acute medication intake days, headache intensity, and sum scores of HIT-6 and SF-36 for the past 4 weeks before randomization will be collected through a patient's headache diary or face to face interview in outpatient clinic. Education on medication overuse using BI strategy (36) will be finished before randomization (Visit 0). Training on the eDiary will be started after a signed consent form obtained from a study participant and prior to the first administration of study intervention. At each visit during the double-blind phase and follow-up (Visit 1–4), outcomes and adverse events will be evaluated correspondingly. Schedule of enrollment, interventions, and assessments following SPIRIT 2013 statement (42) is provided in Table 1.

We will estimate each of the above mentioned variables to calculate endpoints for MOH studies according to a published study (22). Hypothesis test will be 2-tailed with 90% power and a significance level of P < 0.05. Sample size estimations for the four recommended endpoints are provided in Table 2.

Finally, we assume group sample sizes of 77 in gastrodin group and 77 in the control group will achieve 90.188% power to detect a difference between the group proportions of 24%. The proportion in the gastrodin group is assumed to be 22% under the null hypothesis and 46% under the alternative hypothesis. The proportion in control group is 22% based on previously published literature (22). The test statistic used is the two-sided Z-Test with unpooled variance. To allow for 20% dropouts, we will include 186 patients, corresponding to 93 in each group.

The patients, nurses, study investigators and statistician are blinded to the result of the randomization process until data analyses are finished. The blind code could be broken by the principal investigator only for safety concerns. Gastrodin tablets and matching placebo will be provided in identical blister cards to maintain blinding of the study.

Baseline survey and efficacy assessments will be based on information recorded by participants using an eDiary mobile application. Baseline information includes age, educational level (years), any medical history, family history of headache, body mass index, age of headache onset, pre-exiting headache type, monthly headache days, and month medication intake days.

Following variables will be recorded both at baseline survey and each headache attack after randomization: headache characteristics, headache associated symptoms, the headache pain intensity, pain duration of an attack (hours), and acute medication use (type and dosage).

Monthly headache days, month medication intake days, the headache pain intensity, score of HIT-6 and SF-36, treatment failure, drop-outs and any adverse events will be assessed at 4, 8, 12, and 24-week follow-up.

A summary table of eDiary is shown in Table 3. Training for the eDiary will be provided for qualified participants during before randomization.

All the data will be downloaded from a cloud server after each visit and recorded in the CRF and then entered into SPSS software (version 20.0) by an independent data administrator. Data on identification information from participants will be de-identified in SPSS database.

All analyses on efficacy endpoints will be performed using the intention-to-treat (43) principle which includes all randomized patients who had an evaluable baseline period of eDiary data (week-0), received at least 1 dose of gastrodin or placebo, and had at least one evaluable follow-up appointment. Safety evaluations will be performed consisting of all participants who received at least 1 dose of study intervention.

Continuous variables are presented as mean ± standard deviation (44) conforming to the normal distribution and median (interquartile range; IQR) for the skew distribution data. Student's t-test and Mann-Whitney U-test will be used to test differences in continuous variables where appropriate. Categorical variables will be expressed by the number and percentage of occurrences. Differences between two groups will be compared using chi-squared tests or Fisher's exact test where appropriate.

Cox proportional hazard regression (backward stepwise method) analysis and the Kaplan-Meier survival curve will be performed when necessary.

To more completely understand the clinical meaningfulness of the new therapy for MOH results from different headache type, we will conduct a subgroup analysis based on prior headache type (CM/CTTH or CM/ other types of headaches) for all the efficacy endpoints.

Participants who do not complete or are lost to follow-up will be coded as treatment failure, missing values will be handled using multiple imputation method under the missing-at-random (MAR) assumption (45).

An independent data monitoring committee (DMC) will be established to review any spontaneously reported adverse events and identify any safety issues, analyze unintended effects of the study intervention, make recommendations to sponsors, including modification or early termination of the trial when needed. Monitoring on safety trends will be implemented through timely recorded data in eDiary. Other data review on safety issues and pertinent details will be checked by DMC after each visit through source documents. Source documents include a participant's medical records, the eDiary records, and the results of diagnostic tests such as laboratory and neuroimaging tests.

An authorized representative of sponsors will conduct online monitoring to review, audit and download copy of study-related documents on a periodic basis. The representative will meet with the investigator(s) in-site to discuss study-related questions at mutually convenient times when needed.

The present study protocol was approved by Ethic Committee of Affiliated Hospital of Yunnan University (No. 2022047).

All modifications to the protocol will be submitted in consultation with Institutional Review Board approval. Summaries of all relevant modifications will be disseminated to each study investigator via e-mail in time.

Signed consent form including consent provisions for collection and use of participant data and biological specimens will be obtained from participants and authorized surrogates before baseline screening.

Data confidentiality will be implemented by identifying ID numbers for all participants, with all data stored and managed by sponsors in a secure. When data sharing is requested, de-identified data files will be transferred on an encrypted mobile storage by sponsors.

Dr. Fanyi Kong was supported by Yunnan Fundamental Research Project (Grant No. 202001AT070127).

All the investigators, including the project statistician will have access to the final trial database after study results are published. Dataset on clinical information is available with the permission of the principal investigators after all the study results are published.

Compensation to those who suffer harm from trial participation will be covered by commercial insurance provided by KPC Pharmaceuticals, Inc. Free medical services will be provided by Affiliated Hospital of Yunnan University to those who suffer harm from trial participation.

Study findings will be disseminated to healthcare academia through abstracts, posters, or presentations at local, national and international conferences. Full-texts of the present protocol and study results will be published in peer-reviewed journals. Citations of the study findings by professionals are allowed without permission. Participant-level dataset will be provided to professional authors only for conducting individual patient meta-analysis.