An assessment of the carcinogenic potential of ezetimibe using nonclinical data in a weight-of-evidence approach.

Ezetimibe blocks intestinal absorption of sterols via interaction with the Neimann-Pick C1-Like 1 (NPC1L1) transporter and is approved for use in the treatment of primary hyperlipidemia (heterozygous familial and non-familial), homozygous familial hypercholesterolemia, and homozygous sitosterolemia. A recently completed randomized clinical trial [simvastatin and ezetimibe in aortic stenosis (SEAS)] testing the effectiveness of Vytorin (a combination of simvastatin and ezetimibe) in patients with aortic stenosis reported an unexpected safety finding: an increase in overall cancer incidence and cancer-associated mortality (all types) in the treated groups relative to the placebo control. A subsequent meta-analysis utilizing a much larger database from two ongoing clinical trials indicated that the observed findings in the SEAS trial were likely due to chance and not a true drug-induced effect. Nonetheless, it has been suggested by various commentators on the SEAS trial that ezetimibe may be carcinogenic. The extensive nonclinical database for ezetimibe was used to test the hypothesis that ezetimibe may be a direct or indirect carcinogen. Using two different in silico approaches, ezetimibe showed no structural alerts for genetic toxicity or carcinogenicity. Ezetimibe was not genotoxic in two reverse mutation assays, one in vitro clastogenicity assay, and two mouse micronucleus assays. No evidence of proliferative lesions was observed in three species in studies of 1-12 months in duration. Ezetimibe was not carcinogenic in standard 2-year bioassays in mice and rats. Additionally, in these 2-year bioassays, no drug-related non-neoplastic lesions were noted. The absence of drug-induced non-neoplastic or proliferative lesions in these studies indicates that ezetimibe treatment was not associated with findings characteristic of carcinogens (i.e., DNA reactivity or cell proliferation) Administration of pharmacologic doses of ezetimibe to mice did not alter hepatic expression patterns of genes associated with apoptosis, cell proliferation, or epithelial-mesenchymal transition. No evidence of drug-induced tumors was observed in mice in which the molecular target of ezetimibe (NPC 1L1) was knocked out over the life span of the animal. In conclusion, the nonclinical data do not support the proposed hypothesis based on the single observation from the SEAS trial and, rather, support the conclusion that ezetimibe does not represent a carcinogenic hazard to humans using this drug in a therapeutic setting.

Assessment of normal and mutant human presenilin function in Caenorhabditis elegans.

We provide evidence that normal human presenilins can substitute for Caenorhabditis elegans SEL-12 protein in functional assays in vivo. In addition, six familial Alzheimer disease-linked mutant human presenilins were tested and found to have reduced ability to rescue the sel-12 mutant phenotype, suggesting that they have lower than normal presenilin activity. A human presenilin 1 deletion variant that fails to be proteolytically processed and a mutant SEL-12 protein that lacks the C terminus display considerable activity in this assay, suggesting that neither presenilin proteolysis nor the C terminus is absolutely required for normal presenilin function. We also show that sel-12 is expressed in most neural and nonneural cell types in all developmental stages. The reduced activity of mutant presenilins and as yet unknown gain-of-function properties may be a contributing factor in the development of Alzheimer disease.