Bridging the Gap Between hiPSC-CMs Cardiotoxicity Assessment and Clinical LVEF Decline Risk: A Case Study of 21 Tyrosine Kinase Inhibitors

Abstract

Objectives: There is growing concern over tyrosine kinase inhibitor (TKI)-induced cardiotoxicity, particularly regarding left ventricular dysfunction and heart failure in clinical treatment. These adverse effects often lead to treatment discontinuation, severely impacting patient outcomes. Therefore, there is an urgent need for more precise risk assessment methods. This study aimed to assess the cardiotoxicity of TKIs, refine in vitro to in vivo extrapolation (IVIVE) methodologies to improve predictive accuracy, and identify critical in vitro parameters for assessment. Methods: By leveraging high-throughput cardiotoxicity screening with human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs), a mechanism-based toxicodynamic (TD) model for TKIs was constructed. A QSP-PK-TD model was developed by integrating pharmacokinetic (PK) and quantitative systems pharmacology (QSP) models. This model incorporates critical drug exposure factors, such as plasma protein binding, tissue-plasma partitioning, and drug distribution heterogeneity to enhance extrapolation accuracy. Results: The QSP-PK-TD model validated the reliability of IVIVE and identified the area under the curve of drug effects on mitochondrial membrane potential (AEMMP) and cardiomyocyte contractility (AEAAC) as key in vitro parameters for assessing TKI-induced cardiotoxicity. Incorporating critical drug exposure factors obviously improved qualitative and quantitative extrapolation accuracy. Conclusions: This study established a framework for predicting in vivo cardiotoxicity from in vitro parameters, enabling efficient translation of preclinical data into clinical risk assessment. These findings provide valuable insights for drug development and regulatory decision-making, offering a powerful tool for evaluating TKI-induced cardiotoxicity.

Keywords: QSP-PK-PD model; TKIs; cardiotoxicity; hiPSC-CMs; in vitro to in vivo extrapolation (IVIVE).

Figure 1

Individual plots of TKIs on survival fraction of hiPSC-CMs for TKIs with no cardiotoxicity (A) and TKIs with cardiotoxicity (B).

Figure 2

Individual plots of TKIs on relative ATP level of hiPSC-CMs for TKIs with no cardiotoxicity (A) and TKIs with cardiotoxicity (B).

Figure 3

Individual plots of TKIs on contractile force of hiPSC-CMs for TKIs with no cardiotoxicity (A) and TKIs with cardiotoxicity (B).

Figure 4

Quantitative assessments based on the QSP-PK-TD model. The different subfigures represent the correction of drug concentration using different factors. Cp, plasma drug concentration. Fu, plasma protein binding. Kp, tissue–plasma partition coefficient. H, heterogeneous distribution.

Figure 5

Sensitivity analysis of the QSP-PK-TD model. The systematic parameters in QSP model were from a previous study [21]. HR, heart rate. TPR, total peripheral resistance. SV, stroke volume. LVEDV, left ventricular end-diastolic volume. LVESV, left ventricular end-systolic volume. index_SV, LVEDV correction coefficient of SV. FB_LVESV, feedback constant of LVESV on dissipation of LVEDV. kin_HR, kin_TPR, kin_LVEDV, and kin_SV are the zero-order production rate constants, while kout_HR, kout_TPR, kout_LVEDV, and kout_SV are the first-order dissipation rate constants for HR, TPR, LVEDV, and SV, respectively.

Figure 6

Qualitative assessments based on the QSP-PK-TD model. The different subfigures represent the correction of drug concentration using different factors. Cp, plasma drug concentration. Fu, plasma protein binding. Kp, tissue–plasma partition coefficient. H, heterogeneous distribution.

Figure 7

The combined effects of PK-TD parameters on AEAAC.

Figure 8

The combined effects of PK-TD parameters on AEMMP.

Figure 9

Schematic diagram of TKI-induced cardiomyocyte injury. C_{in_TKI} is the effective concentration of intracellular TKIs, and Cth is the threshold concentration of TKIs. E_{max_inj} is the maximum effect coefficient; hill_{_inj} is the hill index. EC_50inj is TKI concentration for half the maximum cell injury. MMP, is the mitochondrial membrane potential; kin_{_ATP} and kin_{_MMP} are the zero-order production rate constants for ATP and MMP, respectively. kout_{_ATP} and kout_{_MMP} are the first-order elimination rate constants for ATP and MMP, respectively. EC_50MMP is the TKI concentration at which MMP production reaches half the maximum; EEC, TKI concentration at which average contractile force reaches half the maximum. E_inj, EMMP, and EAAC are the effect of drug exposure on cell survival, MMP, and average contractile force.