SAVER - Sodium Valproate for Epigenetic Reprogramming in the Management of High Risk Oral Epithelial Dysplasia

Proof of Concept: A recent large epidemiological study showed that long-term use of the antiepileptic agent sodium valproate (SV) reduces the incidence of head and neck cancer in high-risk patients. SV is known to inhibit histone deacetylase (HDAC) activity, which may explain the putative chemopreventative effect. HDAC inhibitors are licenced for the treatment of cutaneous T-cell lymphoma, and importantly another epigenetic modulator, azacytidine is used to prevent malignant transformation of high-risk myelodysplastic syndromes. Animal models also support a role for HDACi in cancer prevention activity. Finally, epigenetic changes have a proven central role in oral epithelial dysplasia (OED) and its progression to carcinoma. Hypothesis: SV will reduce the rate of malignant transformation in patients with high risk OED through epigenetic modifications. Aims: 1) To establish clinical activity of SV as chemopreventive therapy in high-risk OED. 2) To explore the hypothesis that SV reduces progression through epigenetic reprogramming. 3) To investigate the feasibility and acceptability of a phase III randomised controlled trial of SV chemoprevention. Design: Randomised (2 SV: 1 Placebo), double-blind, placebo controlled phase II clinical trial. Population: Patients with OED at high risk of progression to carcinoma (malignant transformation) who have a standard of care therapeutic plan for either close surveillance or surgical resection (patients on either pathway may be recruited). Interventions: Treatment arm: Sodium Valproate 1000mg/day Control arm: Placebo Outcomes & Assessments: 1) Clinical, histological and molecular activity measured using a previously established surrogate composite endpoint which combines changes in: a. Clinical appearance (dimension) b. Histological grade of dysplasia c. Allelic imbalance (3p14, 9p21, 17p13) This primary endpoint is measured before and after a 4-month of SV or placebo treatment. 2) Mechanistic endpoints will explore the hypothesised epigenetic mechanism of action: histone acetylation in circulating white cells, and corresponding tissue specific changes in pre- and post- treatment biopsies of the oral lesions. Additionally, expression of gene targets in a tissue- & disease-specific panel will be assessed by mRNA transcriptional profiling (Nanostring). DNA promoter methylation most strongly linked with OED malignant progression (P16, DCC, EDNRB) will also be studied, while downstream effects on proliferation, apoptosis and senescence will be assessed by IHC. 3) Feasibility: rates of recruitment, screening:randomisation, drop-out, completion of treatment, toxicity. An embedded qualitative study with a purposively selected sample of patients including topic-guided, face-to-face semi-structured interviews will focus on patients’ accounts. Sample Size and Analysis: Randomised phase II design based on exact binomial probabilities, allowing for unequal allocation ratios: 110 patients randomised 2:1 SV: Placebo, assuming 10% dropout. The primary endpoint is response rate using a composite end-point (clinical/pathology/molecular). Analyses shall be carried out assuming a binary response rate with further ordinal modelling carried out should the data allow. Feasibility shall be assessed based on the ability of the study to meet recruitment targets. Economic benefit: reduction in malignant transformation of OED will outweigh the modest costs of SV.