Case Study 4: Assessment of cardiomyocyte contraction in human-induced pluripotent stem cell-derived cardiomyocytes

This case study describes the use of cardiac induced pluripotent stem cells in a high-throughput calcium imaging-based system to assess the effects of the drug on cardiac contractility, from AstraZeneca. This model replaces studies which are usually carried out in rodents or dogs. For more information on this case study please see the reference or contact the NC3Rs.

Background: Cardiotoxicity is a major cause of drug withdrawal from the market and attrition during drug development.  Drug-induced cardiotoxicity results from both functional effects and structural changes to the heart.  Currently, in vitro screens to identify subtypes of functional cardiotoxicity in new drug candidates rely on over-expressing cell lines and primary cells, the latter of which is not amenable to high throughput screening. In particular current approaches to assess cardiomyocyte contractility in vitro are limited to low and medium throughput assays ranging from single organ/cell assays (low throughput) and 24 well plate assays (medium throughput).  Examples include the isolated whole heart model utilizing preclinical species (Guo et al, 2009), isolated ventricular cardiomyocytes from preclinical species (Harmer et al, 2012) or recent advances in “heart on a chip” technology (Grosberg et al, 2012, Grosberg et al, 2011).  These studies allow detailed physiological assessment of cardiac contractility by quantifying contraction and relaxation kinetics; however, they are limited by assay throughput (e.g. 1 compound per day for the whole heart assay, 2 compounds per day for the isolated cardiomyocytes or 24 data points/day for the “heart on a chip” technology).  This limitation in throughput does not allow such models to be applied early in the drug discovery process where chemical design can be influenced. 

Aim: Validate the use of hiPS-CMs for the prediction of changes in cardiac contraction in a format applicable to early drug discovery

Results: Our data demonstrated that hiPS-CMs provide a suitable model for the detection of changes in cardiomyocyte contraction.  When combined with measurement of Ca2+ flux (frequency and amplitude) using a live cell fast kinetic fluorescent assay (FLIPR), changes in cardiac contraction can be detected in a screen with a throughput (384 well based) applicable to toxicity testing in early drug discovery, allowing front-loading and screening cascades to be developed.  Moreover, a number of the inotropic (31 compounds) and non-inotropic compounds (20 compounds) used to validate this approach have been tested in a primary, adult, dog ventricular cardiomyocyte assay that measures the velocity and force of contraction (Harmer et al, 2012).  Overall, the sensitivity and specificity compares well between both assays (81% and 70% specificity in hiPS-CM Ca2+ transients and dog cardiomyocytes, respectively, 87% and 75% sensitivity in hiPS-CM Ca2+ transients and dog cardiomyocytes, respectively).  These results indicate that despite the reported limitations of hiPS-CMs, they do have utility as an early predictor of changes in cardiac contraction in a format applicable to drug discovery.  If used in this way, this approach can contribute towards an improved balance between efficacy and safety of novel compounds. 

Impact: Implementation of this approach into the drug discovery process has allowed replacement of the previous low throughput in vitro dog contractility assay.  This has had a 3Rs impact of saving ~25 dogs/year, meaning no animals are used for in vitro contractility assessments, increased throughput (10 times higher, compared to dog contractility assay) has also allowed improved compound quality going into subsequent in vivo studies (Gp or rat and dog telemetry) and allowed SAR cycles to be conducted.  The consequence of this is improved internal decision making allowing early identification/removal of the risk of changes in cardiac contraction reducing cardiac related drug attrition.

Pointon A et al (2015) Assessment of cardiomyocyte contraction in human-induced pluripotent stem cell-derived cardiomyocytes.Toxicol Sci.144(2):227-37. doi: 10.1093/toxsci/kfu312