The enzyme histone deacetylase type 6 (HDAC6) deacetylates α-tubulin, a component of microtubules that plays an important role in axonal transport. Despite its name, unlike other HDAC subtypes, it acts in the cytosol and not on histones.1,2 Histone deacetylase type 6 has been hypothesized to contribute to the pathophysiology of small fiber neuropathies such as painful diabetic peripheral neuropathy (DPN), chemotherapy-induced peripheral neuropathy (CIPN), and Charcot–Marie–Tooth disease (CMT).5 In rodent models of CIPN, HDAC6 inhibition and HDAC6 knockout are protective against loss of intraepithelial nerve fibers and mechanical allodynia,3,4 and HDAC6 inhibition can also reverse these findings after they have been induced.3 These results suggest that HDAC6 inhibition may have value as a treatment for DPN and CIPN and some subtypes of CMT.

The selective HDAC6 inhibitor ricolinostat is an orally administered investigational inhibitor of HDAC6 with a plasma half-life of approximately 4 hours. After administration of ricolinostat to humans, at exposures associated with the 120-mg dose of ricolinostat in the most current formulation, there is an approximately 4-fold increase in acetylated tubulin in peripheral blood mononuclear cells (PBMCs).6 We hypothesized that administration of ricolinostat to individuals with painful DPN could improve neuropathic pain and possibly other neuropathic signs and symptoms. To test this hypothesis, we conducted a 12-week, phase 2, randomized, double-blind, placebo-controlled study to assess of the efficacy of ricolinostat in individuals with painful DPN.

Eligible subjects were 18 to 80 years with a documented history of type 1 or 2 diabetes for at least 6 months, hemoglobin A1c > 6.5% and <11%, and painful distal symmetric sensorimotor polyneuropathy because of diabetes defined as a Douleur Neuropathique (DN) score >4, moderate to severe pain, and one or more symmetric, distal neuropathic sign(s). Potentially confounding pain conditions were exclusionary, as was severe neuropathy, defined as a Utah Early Neuropathy Scale (UENS) score >24. Subjects taking a stable dose of duloxetine, gabapentin, pregabalin, or carbamazepine for >3 months could continue during the study. Other pain medications were not permitted, except acetaminophen 500 mg with a maximum of 2000 mg/d as needed.

Individuals meeting all above criteria and reporting average daily pain >4 and <9 on the 0- to 10-point numerical pain rating scale (NRS) with an SD <1.5 during the second week of a two-week lead-in observation phase were eligible to enter a 12-week, randomized, double-blind, treatment period with either ricolinostat 120 mg/d or matching placebo (1:1 randomization) followed by a 12-week open-label safety extension phase during which all participants received ricolinostat.

The protocol-specified primary outcome measure was mean change from baseline in average daily pain on the NRS self-reported in a daily diary. Secondary outcome measures included the change in UENS, the pain interference subscale of the Brief Pain Inventory Questionnaire (BPRS), the Patient Global Impression of Change (PGIC), the Neuropathy Total Symptom Score-6 (NTSS-6), and the Norfolk Diabetic Quality of Life-Diabetic Neuropathy (Norfolk Diabetic QOL-DN).

The planned sample of 274 patients was expected to provide 80% power to detect a group difference of 0.8 points on the NRS assuming an SD of 2.35. Patient treatment assignments were randomly generated by computer at Medpace, Inc (Medpace, Inc, Cincinnati, OH). Efficacy analyses included all randomized patients who received at least one dose of study drug with at least one postrandomization assessment, using mixed-model repeated-measures analysis for the NRS and ANCOVA for other measures assessed at baseline and endpoint only. Missing values for subjects who stopped treatment early were imputed using reference-based imputation based on the placebo group. Safety analyses included all subjects who received at least one dose of study drug.

The protocol was reviewed by a central or local institutional review board for each site in accordance with the Declaration of the World Medical Association. All subjects provided written informed consent to participate, and the study was listed on ClinicalTrials.gov (NCT03176472). Safety was reviewed during the study by an independent data monitoring committee.

Subjects were enrolled at 32 investigative outpatient sites in the United States between December 2020 and January 2023. Of 1369 individuals screened, 569 met initial inclusion/exclusion criteria and entered the lead-in pain observation phase, of whom 282 met pain severity criteria and were randomized to either ricolinostat or placebo. The most common reasons for failure during the pain observation period were mild pain severity (N = 147), excessive pain variability (N = 48), and poor diary compliance (N = 37). Subject flow through the study is shown in the consort chart (Fig. 1). Subject demographics and characteristics were generally similar across the randomized groups (Table 1), as were routine laboratory measures. Medications for diabetes control were also similar across groups, with 62/140 (44%) of placebo and 50/142 (35%) of ricolinostat subjects continuing a permitted pain medication during the study. Baseline severity is shown in Table 2.

Consort chart.

At 12 weeks, no significant differences between groups in change in pain severity or secondary measures were observed (Table 3). Additional change during the 12-week open-label extension was modest and similar in both groups (Table 3). Efficacy outcomes were not affected by baseline pain severity, concomitant pain medications, or other covariates. Safety and tolerability were generally good. Few discontinuations occurred as a result of adverse events. Treatment-emergent adverse events were experienced during the 12-week double-blind period by 52/140 (37%) of those on placebo and 53/142 (37%) on ricolinostat. Only 3 individual events were reported by more than 2% of the ricolinostat group: urinary tract infection (placebo 2/140, 1.4%; ricolinostat 5/142, 3.5%), nausea (placebo 5/140, 3.6%; ricolinostat 3/142, 2.1%), and diarrhea (placebo 8/140, 5.7%; ricolinostat 3/142, 2.1%). Among laboratory examinations during the 12-week period, 27/140 (19%) subjects on placebo and 38/142 (27%) on ricolinostat experienced reversible decreases from baseline in neutrophil counts >500 × 103/mm3 on 2 consecutive visits. The only subject with a decrease in neutrophil count below 1.25 × 103/mm3 was in the placebo group.