Background
There is an unmet need for non-hormonal treatments for endometriosis. Peritoneal mesothelial cells of women with endometriosis exhibit detrimental metabolic reprogramming that favours formation and survival of endometriosis lesions. Our preclinical data show that correcting this metabolic phenotype using dichloroacetate is possible, but this approach has not been evaluated in humans. We aimed to determine the feasibility of a clinical trial using dichloroacetate to treat endometriosis-associated pain.
Methods
We performed a single-arm, open-label, single-site study in 30 pre-menopausal women with chronic pelvic pain and surgically confirmed American Society of Reproductive Medicine Stage I–II endometriosis. Participants were recruited from a tertiary-referral endometriosis centre based within the Royal Infirmary of Edinburgh. Potentially eligible patients were approached by their attending clinician for their possible interest in participating and then were followed-up by the study team. Participants were prescribed oral dichloroacetate for 12 weeks (ie, 6·25 mg/kg total bodyweight twice daily, optional increase to 12·5 mg/kg total bodyweight twice daily after 6 weeks). Genotyping was performed on a subset of participants to explore the effect of the dichloroacetate-metabolising enzyme (GSTZ1) haplotype on side-effects. The primary outcome was recruitment and retention, based on multiple timepoints, with rates of 50% or more for recruitment and 80% or more for retention deemed acceptable for progression to a subsequent trial. Retention was assessed based on the proportion of recruited participants who at visits 3 and 5 submitted their average numerical rating scale scores, completed the questionnaires, and attended per-protocol blood testing. Adverse events were assessed at each study visit by the study team in addition to ad hoc reporting by participants. This trial was registered with ClinicalTrials.gov (NCT04046081) and is completed.
Findings
Between Nov 19, 2019, and Aug 13, 2021, 93 women were screened for the study, of whom 77 were approached. Of these, 54 (70%) were eligible and 30 (56%) were recruited. Therefore, the prespecified recruitment rate was met. Five of six retention metrics met the prespecified threshold. Overall retention was 19 (66%) of 29 participants who met all retention metrics, which did meet the prespecified criterion for overall retention. 18 (60%) participants completed all 12 weeks of treatment, and all completing participants reported taking all, or almost all (75–99%), of the prescribed dichloroacetate. 19 (86%) of 22 participants described the trial as a positive experience. Side-effects were common but typically mild as jointly judged by participants and the study team. Six participants developed a transient peripheral neuropathy, of whom two consequently stopped treatment. Both these individuals had a GSTZ1 haplotype associated with slow metabolism of dichloroacetate.
Interpretation
Recruitment to clinical trials of dichloroacetate for endometriosis appears feasible. Genotyping to inform dosage could have the potential to mitigate side-effects. Dichloroacetate is a promising treatment for endometriosis-associated pain but requires assessment in placebo-controlled trials. This study has supported funding for a double-blind, placebo-controlled trial (EPiC2).
