Genomic biomarkers for cardiotoxicity in rats as a sensitive tool in preclinical studies
Article first published online: 4 APR 2013
Copyright © 2013 John Wiley & Sons, Ltd.
Journal of Applied Toxicology
Volume 33, Issue 10, pages 1120–1130, October 2013
How to Cite
Nishimura, Y., Morikawa, Y., Kondo, C., Tonomura, Y., Fukushima, R., Torii, M. and Uehara, T. (2013), Genomic biomarkers for cardiotoxicity in rats as a sensitive tool in preclinical studies. J. Appl. Toxicol., 33: 1120–1130. doi: 10.1002/jat.2867
- Issue published online: 26 JUL 2013
- Article first published online: 4 APR 2013
- Manuscript Accepted: 29 JAN 2013
- Manuscript Revised: 25 JAN 2013
- Manuscript Received: 21 DEC 2012
- genomic biomarker;
- cardiac troponin;
- drug development
The development of safer drugs is a high priority for pharmaceutical companies. Among the various toxicities caused by drugs, cardiotoxicity is an important issue because of its lethality. In addition, cardiovascular toxicity leads to the attrition of many drug candidates in both preclinical and clinical phases. Although histopathological and blood chemistry examinations are the current gold standards for detecting cardiotoxicity in preclinical studies, the large number of withdrawals from clinical studies owing to safety problems indicate that a more sensitive tool is required. We recently identified 32 genes that were candidate genomic biomarkers for cardiotoxicity in rats. Based on their functions, the present study focused on 8 of these 32 genes (Spp1, Fhl1, Timp1, Serpine1, Bcat1, Lmcd1, Rnd1 and Tgfb2). Diagnostic accuracy for the genes was determined by a receiver-operating characteristic (ROC) analysis using more cardiotoxic and non-cardiotoxic compounds. In addition, an optimized support vector machine (SVM) model that was composed of Spp1 and Timp1 was newly constructed. This new multi-gene model exhibited a much higher diagnostic accuracy than that observed for plasma cardiac troponin I (cTnI), which is one of the most useful plasma biomarkers for cardiotoxicity detection. Furthermore, we determined that this multi-gene model could predict potential cardiotoxicity in rats in the absence of any cardiac histopathological lesions or elevations of plasma cTnI. Overall, this multi-gene model exhibited advantages over classic tools commonly used for cardiotoxicity evaluations in rats. Our current results suggest that application of the model could potentially lead to the production of safer drugs. Copyright © 2013 John Wiley & Sons, Ltd.