Miro-Cens was a multicenter, randomized, open-label, parallel group, interventional study in 35 centers in Japan between March 2022 and April 2024 (Fig. 1). A complete list of investigators and institutions is shown in the Acknowledgements section. Eligible patients were randomly assigned in a 1:1 ratio to each treatment group by a registration system using the permuted block method based on the numerical rating scale (NRS) score (< 6 or ≥ 6) for pain at study enrollment (baseline) as an allocation factor.

Study design. 1Dose for patients with CrCL of ≥ 60 mL/min. Patients with CrCL of 30 to < 60 mL/min receive half the dose of mirogabalin. BID twice daily, CrCL creatinine clearance, CSR cervical spondylotic radiculopathy, NRS numerical rating scale, NSAIDs non-steroidal anti-inflammatory drugs, VAS visual analog scale

The study was approved by Tokyo Medical and Dental University (current name, Institution of Science Tokyo) Certified Review Board (CRB No. 3240003). The study was conducted in accordance with the Clinical Trials Act in Japan, as well as the ethical principles and relevant notifications stipulated in the Declaration of Helsinki (as revised in 2013). This study was registered in the Japan Registry of Clinical Trials under the identifier, jRCTs031210629. All patients provided informed consent to participate in the study.

NSAIDs were given according to their Japanese package inserts, and no dose changes were permitted during the study period in either treatment group. Mirogabalin was administered based on patient’s baseline renal function according to its Japanese package insert [12]: for patients with creatinine clearance (CrCL) ≥ 60 mL/min, mirogabalin was administered at 5 mg twice daily (BID) for the first 2 weeks, 10 mg BID for the next 2 weeks, and 15 mg BID or 10 mg BID after Week 5; and for patients with CrCL 30 to < 60 mL/min, the dose of mirogabalin was half that for those with CrCL ≥ 60 mL/min.

The use of concomitant medications, such as gabapentin, pregabalin, duloxetine, opioids, probenecid, cimetidine, and lorazepam, nerve blocks and any surgical analgesic treatments that might affect the efficacy assessment of upper extremity pain caused by CSR were prohibited during study participation.

After informed consent was obtained, patients were screened for study eligibility in steps. Inclusion criteria were: at the first screening, age ≥ 20 years at the time of consent; a diagnosis of CSR with continuous upper limb pain for ≥ 3 months before enrollment and upper limb pain [visual analog scale (VAS) rating ≥ 40 mm] at enrollment; at the second screening, continuously received NSAIDs with no dosage change (≥ 28 days before enrollment); and with mean NRS score for upper limb pain ≥ 4 points in the 7 days before enrollment. A diagnosis of CSR was based on the following criteria: (1) patients with pain/numbness of a unilateral arm or hand; (2) positive on the Spurling test [14] or the Jackson test; (3) cervical spondylotic changes on X-ray, CT, or MRI; and (4) a clear diagnosis of radiculopathy pain based on other neurological findings even if the Spurling test or the Jackson test was negative.

Key exclusion criteria were: patients with severe pain caused by other than CSR for whom pain assessment was difficult; cervical spondylotic myelopathy; history of cervical spine surgery; CrCL < 30 mL/min; history of mirogabalin use within 28 days prior to obtaining consent; Brief Scale for Psychiatric Problems in Orthopaedic Patients score ≥ 11 for physicians or both ≥ 10 for physicians and ≥ 15 for patients at enrollment [15]; received prohibited concomitant drugs and therapies in the 14 days before enrollment; and having serious underlying disease and complications.

The primary endpoint was the change in the weekly average NRS score for upper limb pain from baseline at Week 12 [16]. Patients were asked to rate the pain they had experienced over the previous 7 days on an 11-point NRS ranging from 0 (“no pain”) to 10 (“worst pain possible”).

Secondary endpoints included: responder rate ≥ 30% or ≥ 50% reduction in the NRS score of the primary endpoint; change from baseline in the VAS score for upper limb pain (range 0–100 mm, with 0 mm meaning no pain and 100 mm meaning the worst pain imaginable); change from baseline in the EQ-5D-5L score [17]; Patient Global Impression of Change (PGIC) score at Week 12 (1 = “very much improved” to 7 = “very much worse”); and change from baseline in the weekly average NRS score for sleep disturbance (0 = “no sleep disturbance” to 10 = “sleep completely disturbed by pain”).

The safety endpoints included the incidences of treatment-emergent adverse events (TEAEs) and of treatment-emergent adverse drug reactions (TEADRs).

Based on the results of previous phase 3 clinical trials of mirogabalin [8, 9], the difference between two groups in the mean weekly change in the NRS score from baseline at Week 12 was estimated to be 1.2, with a standard deviation (SD) of 2.4. Under this assumption, the number of patients needed to ensure 80% power at a two-sided significance level of 5% was 128 (64 for each group). Given possible dropouts, the target sample size was set at 140 (70 for each group) patients.

The primary efficacy analyses used the modified intention-to-treat (mITT) population, which was defined as all randomized patients who received at least one dose of the study drug. A linear mixed model for repeated measures (MMRM) was used to calculate the estimated differences in adjusted means between the mirogabalin add-on group and the NSAIDs alone group, 95% confidence intervals (CIs), and P values for the primary endpoint data. The MMRM included treatment group, treatment duration (week), and treatment group–week interactions as fixed effects, the weekly mean NRS score for upper limb pain at baseline as a covariate, and patients as a random effect. A similar MMRM was used for the secondary efficacy endpoints. As a sensitivity analysis, the last observation carried forward (LOCF) and baseline observation carried forward (BOCF) methods were used to impute the weekly mean NRS scores at Week 12 in the mITT population. Then, analysis of covariance (ANCOVA) was used to compare estimated differences in adjusted means between treatment groups, 95% CIs, and P values. The same analysis was also carried out in the per-protocol set (PPS), which was defined as all patients in the mITT population who adhered to the study protocol and the package inserts.

For the safety analyses, the safety analysis set was used, which was defined as all patients enrolled in the study who received at least one dose of the study drug. TEAEs were coded using the Japanese Medical Dictionary for Regulatory Activities version 27.0. A TEADR was defined as a TEAE judged by the attending physician to have a causal relationship with the study drug.

The significance level for hypothesis testing was set at 5% (two-sided). Statistical analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC, USA).