BACKGROUND AND PURPOSERisk of cardiac conduction slowing (QRS/PR prolongations) is assessed prior to clinical trials using in vitro and in vivo studies. Understanding the quantitative translation of these studies to the clinical situation enables improved risk assessment in the nonclinical phase.
EXPERIMENTAL APPROACHFour compounds that prolong QRS and/or PR (AZD1305, flecainide, quinidine and verapamil) were characterized using in vitro (sodium/calcium channels), in vivo (guinea pigs/dogs) and clinical data. Concentration-matched translational relationships were developed based on in vitro and in vivo modelling, and the in vitro to clinical translation of AZD1305 was quantified using an in vitro model.
KEY RESULTSMeaningful (10%) human QRS/PR effects correlated with low levels of in vitro Na v 1.5 block (3-7%) and Ca v 1.2 binding (13-21%) for all compounds. The in vitro model developed using AZD1305 successfully predicted QRS/PR effects for the remaining drugs. Meaningful QRS/PR changes in humans correlated with small effects in guinea pigs and dogs (QRS 2.3-4.6% and PR 2.3-10%), suggesting that worst-case human effects can be predicted by assuming four times greater effects at the same concentration from dog/guinea pig data.
CONCLUSION AND IMPLICATIONSSmall changes in vitro and in vivo consistently translated to meaningful PR/QRS changes in humans across compounds. Assuming broad applicability of these approaches to assess cardiovascular safety risk for non-arrhythmic drugs, this study provides a means of predicting human QRS/PR effects of new drugs from effects observed in nonclinical studies.
AbbreviationsFTIM, first time in man; hCav1.2, human cardiac calcium channel; hNav1.5, human cardiac sodium channel; rCav1.2, rat cardiac calcium channel; PD, pharmacodynamic; PK, pharmacokinetic; PPB, plasma protein binding; QTc, heart ratecorrected QT