General Principles for the Validation of Proarrhythmia Risk Prediction Models: An Extension of the CiPA In Silico Strategy

Figures

• 

  Figure 1. A generic flowchart of proarrhythmia risk prediction model development process. Different principles are applied to different steps along the model development process. In the first step, a defined end point (type of proarrhythmia risk) consistent with the context of use (Principle 1) is chosen. This is enabled by testing a list of drugs based on a defined clinical risk categorization or quantification system. The drug list is split into a training set and a validation set. The training set undergoes some defined pharmacological experimental protocols to generate experimental data as model input (Principle 3). The experimental variability in the pharmacological assessment needs to be captured and propagated through the model via uncertainty quantification (Principle 6). In the meantime, a platform to mimic the response of physiological system is chosen. Such platforms can be, for example, induced pluripotent stem cell‐derived cardiomyocytes (iPS‐CMs), in vivo animals, ex vivo tissues, or in silico cardiomyocyte models. Such platforms may be subjected to a development process, for instance to adjust the structure and parameters of an in silico model to better replicate cardiac…