Can in silico models predict drug-induced cardiac risk in vulnerable populations?

Area of Science:

• Computational biology
• Cardiovascular pharmacology
• Drug safety assessment

Background:

• Traditional drug-induced QT prolongation assessments use limited healthy cohorts.
• Vulnerable patient populations are often excluded, limiting predictive accuracy.
• Computational modeling offers a novel approach to assess drug effects across diverse cardiac conditions.

Purpose of the Study:

• To evaluate virtual cardiac populations for preclinical assessment of drug-induced QT interval prolongation and arrhythmic risk.
• To investigate drug response variability across different heart pathologies and sexes.
• To establish a predictive computational framework for safer drug development.

Main Methods:

• Generation of a virtual cohort of 512 subjects with healthy and diseased hearts (heart failure, cardiomyopathies, ischaemia, infarction).
• Simulation of drug administration (moxifloxacin, quinidine, bepridil, flecainide) to assess QT prolongation and arrhythmia risk.
• Analysis of drug response variability based on cardiac pathology and sex.

Main Results:

• Patients with heart failure, hypertrophic and dilated cardiomyopathy showed greater QT prolongation to moxifloxacin.
• Females consistently exhibited higher QT prolongation than males across tested drugs.
• Contraindicated drugs significantly increased arrhythmia risk in diseased populations, particularly in females, leading to lethal arrhythmias.

Conclusions:

• Computational models effectively capture inter-individual variability in drug response across pathologies and sexes.
• Virtual cardiac populations provide a valuable predictive framework for preclinical drug safety evaluations.
• This approach supports the development of safer, personalized drug therapies for cardiovascular conditions.