The comparative sensitivity of three in vitro safety pharmacology models for the detection of lidocaine-induced cardiac effects

Abstract

IntroductionIn the current ICH S7B guideline, in vitro evaluation of proarrhythmic liability is limited to the risk of QT interval prolongation, whilst the effect of new chemical entities on cardiac conductivity is often overlooked. The aim of this work was to compare the effects of the sodium channel blocker, lidocaine in three in vitro safety pharmacology models: hNav1.5 channel test, atrial action potential (AP) and Purkinje fiber AP and to identify the most sensitive model for detecting cardiac conduction slowing. MethodsWhole-cell patch‐clamp methods were used to record the sodium current (INa) encoded by hNav1.5 in stably transfected HEK293 cells at ambient temperature. Transmembrane APs were recorded in rabbit Purkinje fibers and rabbit and guinea-pig left stimulated atria at physiological temperature. Parameters involved in depolarization or repolarization were reported. ResultsLidocaine (from 10 to 1000μM) decreased the amplitude of INa (IC50: 256±37μM) in a concentration-dependent manner. In the Purkinje fiber assay, lidocaine (10, 30 and 100μM) had no effects on maximal upstroke velocity (Vmax), but shortened AP duration at 90% repolarization (APD90). At 30 and 100μM, lidocaine also increased AP triangulation. In guinea-pig atria, lidocaine decreased Vmax starting from 30μM and conduction velocity (CV) at 100μM, but had no effects on other parameters. In rabbit atria, lidocaine decreased Vmax and CV at 100μM without affecting APD90. The effects of 100μM lidocaine on Vmax and CV were more marked in rabbit than in guinea-pig atria. ConclusionRabbit atria are more sensitive than rabbit Purkinje fibers or guinea-pig atria for detecting lidocaine-induced cardiac conduction slowing. These data suggest that isolated rabbit atria in addition to the hNav1.5 channel assay could be relevant models to predict drug-induced conduction slowing.