Case Study 3: Human induced pluripotent stem cell derived cardiomyocyte (hiPSC-CM) multi electrode array (MEA) assay

This case study compares the use of induced pluripotent stem cell cardiomyocytes (iPSC CMs) to study cardiac safety with current standard in vitro assays used in pre-clinical electrophysiological cardiac safety screening, from GlaxoSmithKline. The In Vitro model replaces studies of cardiac safety in telemetered dogs. For more information on this case study please see the reference or contact the NC3Rs.

Background: Cardiovascular adverse drug reactions (ADRs) have led to many marketed drug withdrawals and termination of promising pre-clinical and clinical drug development candidates. Drug-induced cardiac electrophysiological alterations, most notably QT prolongation and torsadogenic incidence/risk has caused widespread concern and action from both regulators and pharmaceutical companies. The current strategy that many pharmaceutical companies use to detect and quantify drug-induced cardiac electrophysiological alterations range from ion channel screening, in vitro tissue studies and in vivo evaluation in conscious telemetered animals. hiPSC-CM with MEA technology could be used for predictive electrophysiological cardiac safety screening. For wider acceptance by drug companies and regulators, hiPSC-CMs must be validated against existing preclinical cardiac safety assays to determine assay predictivity in relation to current standards. Validating a particular cell type with reference compounds is key to setting up a robust preclinical assay for interspecies comparison of cardiac safety endpoints. 

Aim: To evaluate the utility of the hiPSC-CM MEA assay for use in pre-clinical cardiovascular risk assessment hiPSC-CMs studied on MEAs were pharmacologically validated using a set of ten compounds with a mixture of hERG, Na+ and Ca2+ channel blocking properties were studied and a comparison of MEA acquired parameters was made to the QT and QRS intervals on ex vivo RVW, in vivo rodent/non-rodent and clinical (where data was available) electrocardiograms to determine translation of pharmacological effects at a given concentration.

Results: Ten compounds with a mixture of ion channel blocking properties were selected for investigation. E-4031 and Cisapride were selected for their pure hERG channel blockade, Flecainide, Quinidine, Terfenadine and Verapamil for their mixed ion channel blocking properties, Nifedipine for its pure calcium (Ca2+) channel blocking properties, Mexiletine for sodium (Na+) channel block; while two proprietary GSK compounds (Anonymised; and denoted as GSK A and GSK B) were picked due to the abundant pre-clinical data available from their development. Both GSK A and GSK B had issues with electrophysiological findings during the developmental phase as outlined and were added to examine the utility of the assay to pick up non-reference compounds. All ion channel blockers produced functional effects on repolarization and/or depolarization around their IC25 and IC50 values, and excessive blockade of hERG and/or blockade of sodium current precipitated arrhythmias. There was a high degree of correlation between data obtained in the hiPSC-CM MEA assay compared to those obtained in the classical in vitro electrophysiological assay system the RVW (see Table), and, to other preclinical assays and clinical data.

Compound

Concentration for p<0.05 in the RVW

Concentration for p<0.05 in the MEA

MEA/RVW Fold Difference (Log units)

Nifedipine Repolarization

0.03

0.03

0

Cisapride Repolarization

0.1

0.1

0

Terfenadine Repolarization

1

0.1

-1§

Terfenadine Conduction

10

1

-1§

Verapamil Repolarization

1

0.03

-1.52§

Flecainide Conduction

3

1

-0.47§

Flecainide Repolarization

3

1

-0.47§

Quinidine Repolarization

0.1

0.3

0.47*

GSK A Repolarization

10

1

-1§

GSK B Repolarization

0.6

n/a due to beat rate changes

n/a

Table 1: The table shows the concentration at which a statistically significant change in QT/QRS was seen in the RVW and compares this to the concentration at which a statistically significant change is seen in FPDc /spike amplitude in the hiPSC-CM MEA assay. The logarithmic scale difference is shown in the last column. In most cases, MEA fell within half a logarithmic unit of the wedge data and in certain cases was better at predicting an effect.

Impact: The initial successful recapitulation of in vitro data from the RVW with 10 compounds with known electrophysiological effects in the hiPSC-CM MEA has led us to broaden our studies to include a larger validation set including more compounds of unknown action and compounds known to have no pro-arrhythmic action in vivo. Preliminary data (not yet reported) continue to support a strong translation between data from the hiPSC-CM MEA assay and the RVW. We are currently evaluating the potential to use the MEA model in higher throughput format as an early screen for drug-induced cardiac electrophysiological alterations in pre-clinical safety evaluation. Based on this data, hiPSC-CMs could provide an alternative in vitro model to screen for drug-induced cardiovascular electrophysiological risk and reduce the number of animals currently used in in vitro/ex vivo studies.

Reference: Harris, K et al. (2013) Comparison of electrophysiological data from human-induced pluripotent stem cell-derived cardiomyocytes to functional preclinical safety assays Toxicol. Sci.; 134(2):412-426