Population level predictions of QT prolonging effects for hydroxychloroquine and hypokalaemia
Our hiPSC-CM data suggests differences in drug effects between individual patient lines as well as a more pronounced prolongation of repolarisation in the context of hypokalamia. To interrogate this further, we performed population level simulations of the cardiac action potential to assess the variability in drug response between individual subjects and to illustrate the consequences of the electrolyte-dependent changes in hERG potency. In addition to the kalaemic-dependence of potency of hydroxychloroquine against hERG measured in our assays, these simulations also incorporate block of the inward rectifier potassium current (IK1), the cardiac sodium current (INa), and the cardiac L-type calcium current (ICaL) from the literature (22). However, it should be noted that the contribution of this multichannel pharmacology to the overall effect on repolarisation is relatively small, since the degree of block at the highest dose of hydroxychloroquine simulated (3000 ng/ml) were 4.1 %, 20 % and 7.3 % for ICaL, IK1 and INa respectively, compared to 70 % block of hERG at the same concentration (see Supplementary table 2). Populations of action potentials representing baseline (no drug, normal electrolytes), 1000 ng/ml hydroxychloroquine and hydroxychloroquine + mild hypokalaemia are shown in Figure 4A. Frequency distribution histograms of repolarisation duration illustrating the effect of hypokalaemia on population response to 1000 ng/ml and 3000 ng/ml hydroxychloroquine (an estimate of mean plasma Cmax and maximum plasma Cmax respectively, see Discussion) are shown in Figure 4B-D and summarised in Figure 4E. In normokalaemia, the time to 90 % repolarisation (APD90) was increased to 355 ± 49 ms (mean ± SD) and 4444 ± 60 ms for 1000 ng/ml and 3000 ng/ml hydroxychloroquine respectively compared to baseline (274 ± 40 ms). Furthermore, within these distributions, a range in the severity of prolongation of repolarisation was observed, even for those individuals with the same baseline APD90s (Supplementary Movie 1). Mild hypokalaemia further increased the duration of repolarisation to 383 ± 53 ms and 481 ± 63 ms for 1000 ng/ml and 3000 ng/ml hydroxychloroquine. The percentage of the population with APD90 >500 ms in response to 3000 ng/ml hydroxychloroquine increased from 17.4 % in normokalaemia to 34.8 % in hypokalaemia, reflecting a significant increase in proarrhythmic risk associated with mild hypokalaemia.