Comparative safety, pharmacokinetic and pharmacodynamic evaluation of three oral selenium compounds in cancer patients

Abstract

Selenium (Se) is an essential trace element involved in many biological processes in humans. Se compounds have demonstrated growth inhibition of malignant cells and tumours in many experimental models. Furthermore, combining Se with cancer therapies has yielded promising results in both preclinical studies and human trials, where reduced toxicity and enhanced anti-tumour efficacy is reported. However the optimal form and dose of Se in combination with anticancer therapies in terms of safety and efficacy has yet to be established, with particular concern about possible second malignancies due to DNA-damaging properties of both high-dose Se compounds and cancer therapies. A key component of the toxicological study of new compounds is the assessment of their genotoxic potential. To achieve this a quantitative polymerase chain reaction-based method was developed that measured DNA damage within regions of nuclear and mitochondrial DNA relative to untreated controls, this assay was then validated in cancer cells and peripheral blood mononuclear cells (PBMCs). Subsequently, the method was optimised on a newly-acquired qPCR platform (Magnetic induction cycler) and used for measuring DNA damage in the clinical trial. The primary research objective of determining the optimal form and dose of Se to be used alongside anticancer therapies was addressed through a randomised, double-blinded clinical trial. The safety, tolerability and pharmacokinetic (PK) and pharmacodynamic (PD) profiles of sodium selenite (SS), Se-methylselenocysteine (MSC) and seleno-l-methionine (SLM) were compared in patients with chronic lymphocytic leukaemia (CLL) and a cohort of patients with metastatic solid cancers. Patients received 400g of elemental Se orally, as SS, MSC or SLM, for 8 weeks. Safety evaluations and bloods for PK and PD analyses were taken twice prior to treatment and at Days 2, 28 (week 4), 56 (week 8) and 84 (week 12) from commencing treatment. Each Se compound group (n=8) included four patients with CLL and four with solid cancers. Twenty-four patients were treated and all Se compounds were well-tolerated, with no grade 3-4 toxicities attributed to the study drugs. The baseline-corrected area under the concentration-time curve for plasma Se was markedly higher with SLM than MSC or SS, consistent with previous reports. Assessment of DNA damage in normal and malignant PBMCs revealed negligible genotoxicity. In CLL patients, no fall in blood lymphocyte count was seen in those treated with SS, whereas the majority of those treated with SLM or MSC showed a reduction, with one patient achieving a partial response. Persistent reductions in plasma VEGF were observed in 7/12 patients with CLL and 4/12 with solid cancers. Minimal changes were observed in intracellular glutathione or markers of the endoplasmic reticulum stress response. In conclusion, Se compounds are well-tolerated and non-genotoxic at this dose. A reduction in plasma VEGF occurred at lower doses than predicted whereas changes in ER stress markers and intracellular glutathione are consistent with the dose-response seen in tumour xenografts and cell lines. Evaluation of these compounds at the higher doses of 1600g and 6400g per day, where PD effects are predicted to be greater, is planned for a future study.