Cardiotoxicity Evaluation of Tyrosine Kinase Inhibitors using Human Induced Pluripotent Stem Cell-derived Healthy and Long QT Syndrome Cardiomyocytes

Background Cardiotoxicity associated with tyrosine kinase inhibitors (TKIs) has been reported in several clinical trials and preclinical studies, whereas the toxicity in populations with congenital electrophysiological dysfunction remains unclarified. We studied cardiotoxicities of four US Food and Drug Administration (FDA)-approved TKIs with different targets by measuring changes in cardiomyocyte (CM) contractility. Contractions of human induced pluripotent stem cell-derived (hiPSC)-CMs from healthy donors and long QT syndrome (LQT) patients were studied with an impedance-based bioanalytical method (xCELLigence® real-time cell analysis [RTCA] cardio system) to quantify TKI toxicity. Both healthy donor (wild type, WT)- and LQT patient-derived hiPSC-CMs exhibited a functional CM phenotype. The four TKIs inhibited the growth and contractility of CMs with different potencies. Crizotinib and dasatinib decreased the beating of both WT- and LQT-hiPSC-CMs with approximately equal potencies. Sunitinib weakened spontaneous pulsations of both kinds of cells, but the LQT-hiPSC-CMs showed higher sensitivity. In contrast, lapatinib exerted a milder effect on LQT- hiPSC-CMs. and LQT1 patients both patient-specific donors improve and incubated CO every The WT and LQT cells were plated at densities of 3 × 10 4 and 6 × 10 4 cells/well, respectively. extreme caution with sunitinib. Following exposure to the tested drugs, the and exhibited different outcomes involving their mechanical beating properties and varying sensitivities. These findings could contribute to providing a platform for precision medicine and drug screening in the future.

However, certain adverse reactions are unexpected and cannot be accurately predicted from previous standard preclinical studies. To date, several preclinical approaches have been used to test the cardiotoxicity of compounds, such as using hearts-on-chips, engineered heart tissue models, and the most widely used, human induced pluripotent stem cell-derived cardiac myocytes (hiPSC-CMs). Particularly, hiPSC-CMs have provided a new platform for building in vitro models that have been widely used in activities such as disease modeling, drug development and screening, and evaluation of cardiotoxicity. Compared with animal models, hiPSC-CMs have no species and extrapolation limitations. In addition, in contrast to non-cardiac human cell lines, hiPSC-CMs express the native cardiac proteins required to reconstitute the complex cardiac structure and phenotype (e.g., sarcomere organization, calcium handling, metabolism, and electrophysiology) [4].
Based on these characteristics, patient-derived cardiomyocytes (CMs) contribute to the understanding of mechanisms underlying genetic diseases and enable the study of differences between a drug's activity in healthy and particular patient groups [5,6]. Differential patient-specific reactions to a particular drug are driven by interactions between genetic, epigenetic, and environmental factors, and drug pharmacodynamics could also be influenced by inherited polymorphisms in target enzymes, transporters, ion channels, and receptors [7].
Multiple patient-specific hiPSC-CMs have been established successfully and used in studies of pathogenic pharmacodynamic mechanisms and compound screening [8,9]. For instance, patient-specific (long QT syndrome 1 [LQT1] and LQT2) CMs have been used to explore cardioactive drug effects, and the results appear to indicate that these cells simulate clinical observations (4). These studies provide further evidence to support the notion that these cells can be used to evaluate the arrhythmia-inducing propensity of drugs [10].
LQT is an inherited arrhythmia syndrome, which manifests as palpitations, syncope, sudden cardiac death, and clinical susceptibility to malignant ventricular arrhythmias. LQT is always associated with sudden cardiac death caused by torsade de pointes (TdP). The electrocardiogram (ECG) indicators of LQT are prolonged QT intervals and increased QT dispersion. Inherited forms of LQTs are caused by mutations in cardiac ion channel coding genes. The potassium voltage-gated channel subfamily Q member 1 (KCNQ1, LQT1), potassium voltage-gated channel subfamily H member 2 (KCNH2, LQT2), and sodium voltage-gated channel alpha subunit 5 (SCN5A, LQT3) are the most common LQT pathogenic genes, and LQT1 is the most common inherited LQT.
Mutation of the KCNQ1 gene reduces the main component of the cardiac repolarization outward current, the slow delayed rectifier current (I Ks ) channel, which delays ventricular repolarization and QT interval extension. The I Ks channel has two major components: KCNQ1 (also called K v 7.1 or K v LQT1, the major ion-conducting, voltagesensing component) and potassium voltage-gated channel subfamily E regulatory subunit 1 (KCNE1, also known as minK or I sK , a small regulatory subunit). In the human heart, I Ks is mainly responsible for shortening ventricular action potentials under stressful conditions such as with high β-adrenergic tone.
Therefore, adverse cardiac events in LQT1 patients occur mostly during exercise or under emotional stress [11]. To date, over 250 KCNQ1 mutations have been found linked to LQT1 and new mutations continue to be identified. Mutations in the transmembrane, linker, and pore regions of KCNQ1 are usually defined as highprobability, disease-causing mutations that tend to cause severe cardiac events in patients at a younger age than mutations in the C-terminal region [12].
The numerous adverse cardiac events associated with TKIs have raised the question whether patient groups with existing ion channel dysfunction have similar susceptibility to TKI cardiotoxicity. Drug-induced cardiotoxicity usually manifests clinically as contractile dysfunction. The xCELLigence® real-time cell analysis (RTCA) cardio system, which utilizes impedance technology to quantify CM-beating properties, has been previously reported as an emerging method to quantify cardiac contractility [13]. In the present study, to the best of our knowledge this is the first time, we compared the responses of healthy and LQT1 CMs to TKIs with different targets using healthy donor (wild type, WT)-and LQT syndrome patient-derived hiPSC-CMs, with the aim of providing a reference for the appropriate clinical drug treatment of particular patient groups.

Cell culture
WT-hiPSC-CMs and LQT-hiPSC-CMs were purchased from Cellapy (CA2101106, Beijing, China) and HELP (HELP4120, Nanjing, China), respectively. Both cell types were obtained as frozen 1 mL aliquots containing approximately 1 and 2 million cells, respectively. The cells were thawed and plated in 96-well E-plates (ACEA Biosciences, Hangzhou, China) pre-coated with 0.1% gelatin. The cells were incubated in maintenance medium at 37 °C in an atmosphere of 5% CO 2 , and the culture medium was refreshed every 2 days. The WT and LQT cells were plated at densities of 3 × 10 4 and 6 × 10 4 cells/well, respectively.

Clinical data
The somatic cells for generation of the WT-hiPSC-CMs were obtained from a 39-year-old healthy female volunteer in Beijing by Cellapy, while those for the LQT1-hiPSC-CMs were from a 53-year-old Uyghur female patient with LQTS1 by HELP. The patient's LQTS1 was due to an identified missense mutation (G to A substitution at nucleotide 656), resulting in an amino acid substitution at position 219 of glycine to glutamic acid (G219E) in the KCNQ1 channel.

Real-time impedance-based bioanalyses of hiPSC-derived CMs
Spontaneous CM contraction and cell health were monitored in real-time by impedance using the xCELLigence® RTCA cardio system (ACEA Biosciences, Hangzhou, China). Impedance signals were monitored and recorded for 20 s per sweep. The cell index (CI), which is related to cell adherence and growth, was used as a measure of cell vitality.

Data and statistical analyses
For RTCA CM monitoring, the CI, beating rate, beating amplitude, and half-maximal inhibitory concentration (IC 50 ) for each well were calculated off-line using RTCA cardio software 1.0 and normalized to the corresponding baseline values measured prior to treatment with the compounds. A threshold level of 12 was used to suppress noise for improved peak recognition and the raw data of the beat curves were displayed using a threshold level of 0. The data are presented as means ± standard error of the mean (SEM) and the statistical significance of differences was estimated using a one-way analysis of variance (ANOVA) or Student's t test. A value of p < 0.05 was considered significant.

WT-and LQT-hiPSC-CMs exhibit functional CM phenotypes in culture
Analysis of the hiPSC-CMs based on marker protein expression and electrophysiological measurements revealed that the expected phenotypes were replicated in these cell models, namely that the WT-hiPSC-CMs had a normal phenotype and the LQT-hiPSC-CMs showed QT-interval prolongation. The relevant certificate of analysis is provided in the Supplementary data. Spontaneously contracting CMs were observed in both WT-and LQT-hiPSC-CMs as assessed by light microscopy.
The CI values increased and gradually stabilized with the growth and adherence of both WT-and LQT-hiPSC-CMs ( Fig. 1a and 1c). Within 168 h after plating, the WT-hiPSC-CMs achieved a stable state in both beating rate and amplitude earlier than the LQT-hiPSC-CMs did ( Fig. 1b and 1d, Fig. 2a and 2b). The spontaneous beating rate of both WT-and LQT-hiPSC-CMs was stable at approximately 40 bpm, and the amplitude steadily increased over time and stabilized at approximately 0.09, with a rhythmic irregularity < 40% (Fig. 2c).

Effects of crizotinib
Crizotinib reduced the CI of both WT-and LQT-hiPSC-CMs in a concentration-dependent manner, indicating its inhibitory effect on the growth and adherence of cells at concentrations > 3 µmol/L ( Fig. 3a and 3b). In addition, 3 µmol/L and 10 µmol/L crizotinib abolished the cellular automaticity of both WT-and LQT-hiPSC-CMs, but the latter CMs showed signs of recovery ( Fig. 3c and 3d). Further analysis of the beating rate indicated that crizotinib inhibited WT-and LQT-hiPSC-CMs in a concentration-dependent manner. The beating rate was reduced at concentrations > 0.1 µmol/L, and the spontaneous pulsations gradually recovered over time ( Fig. 3e and 3 g). In contrast, a detailed trace analysis revealed that 0.3 µmol/L and 1 µmol/L increased the amplitude of both WTand LQT-hiPSC-CMs ( Fig. 3f and 3 h). While spontaneous beating ceased early after administration of crizotinib, the amplitude significantly increased after beating was restored. Analysis of the concentration-related effects of crizotinib on WT-and LQT-hiPSC-CMs demonstrated IC 50 values of 0.89 µmol/L and 0.76 µmol/L, respectively (Table 1).

Effects of sunitinib
Investigation of the effects of sunitinib demonstrated that it decreased the CI of WT-hiPSC-CMs in a concentration-dependent manner (Fig. 4a). Sunitinib at a concentration of 3 µmol/L affected pulse signals, while 10 µmol/L entirely abolished spontaneous beating (Fig. 4c). Thirty minutes after administration, the spontaneous beating rhythms of WT-hiPSC-CMs were markedly decreased at 1 µmol/L and 3 µmol/L in a concentrationdependent manner without self-recovery within the observation period (Fig. 4e). Similarly, sunitinib at 1 µmol/L and 3 µmol/L increased the beating amplitude 0.5 h after administration, but a gradual recovery was observed over time (Fig. 4f). In the LQT-hiPSC-CMs, sunitinib reduced the CI values in a concentration-dependent manner. However, compared with the WT-hiPSC-CMs at the highest concentration (10 µmol/L), sunitinib reduced CI values to 0 within approximately 0.5 h, which indicates that this concentration completely inhibited cell growth (Fig. 4b). Likewise, for pulse patterns, the LQT-hiPSC-CMs showed a higher sensitivity to sunitinib than the WT-hiPSC-CMs did, where sunitinib weakened pulse signals at concentrations ≥ 0.3 µmol/L in the former (Fig. 4d). Further analysis demonstrated that sunitinib decreased the beating rate and amplitude of the LQT-hiPSC-CMs in a concentration-dependent manner, and a gradual recovery was observed up to 6 h after exposure ( Fig. 4g and 4 h). Interestingly, the LQT-hiPSC-CMs showed weak beating at a lower concentration (0.3 µmol/L) of sunitinib than the WT-hiPSC-CMs did. Evaluation of the concentration-dependent effects of sunitinib on WT-and LQT-hiPSC-CMs demonstrated IC 50 values of 4.71 µmol/L and 0.52 µmol/L, respectively ( Table 1).

Effects of dasatinib
Dasatinib reduced the CI values ( Fig. 5a and 5b) and affected the pulse patterns ( Fig. 5c and 5d) of both WT-and LQT-hiPSC-CMs in a concentration-dependent manner. In the WT-hiPSC-CMs, 30 µmol/L dasatinib affected pulse patterns, which was observed as inhomogeneity in the pulse rhythm and amplitude (Fig. 5c). Analysis of the beating rate and amplitude within 24 h after exposure indicated that dasatinib concentration-dependently inhibited the beating rate at ≥ 3 µmol/L 0.5 h after exposure (Fig. 5e). In contrast, the effect of dasatinib was relatively milder on the beating amplitude than it was on the beating rate. Furthermore, at the highest concentration (30 µmol/L), dasatinib increased the cell amplitude ( Fig. 5f). In the LQT-hiPSC-CMs, dasatinib affected pulse patterns at 10 µmol/L and 30 µmol/L, and 30 µmol/L dasatinib temporarily interrupted spontaneous beating 0.5 h after exposure (Fig. 5d). Lower concentrations (0.3 µmol/L and 3 µmol/L) of dasatinib temporary increased the beating rate of LQT-hiPSC-CMs 0.5 h after exposure, but the cells soon recovered (Fig. 5g). However, 10 µmol/L and 30 µmol/L dasatinib exerted remarkable inhibitory effects on the beating rate without self-recovery within 24 h after exposure. Consistently, in the LQT-hiPSC-CMs, dasatinib affected the beating amplitude less than it did the beating rate, and with the highest concentration (30 µmol/L), the beating amplitude increased after recovery of spontaneous beating (Fig. 5h). Evaluation of the concentration-dependent effects of dasatinib on WT-and LQT-hiPSC-CMs demonstrated IC 50 values of 27.34 µmol/L and 23.30 µmol/L, respectively ( Table 1).

Effects of lapatinib
Investigations of the effects of lapatinib on the WT-hiPSC-CMs demonstrated that it reduced CI values in a concentration-dependent manner (Fig. 6a), and interrupted pulse signals at concentrations ≥ 10 µmol/L (Fig. 6c).
In this study, lapatinib exerted a mild inhibitory effect on the beating rate and amplitude of WT-hiPSC-CMs at 1 µmol/L and 3 µmol/L, and self-recovery was observed within the study period. However, at 10 µmol/L and 30 µmol/L, lapatinib totally abolished the beating of WT-hiPSC-CMs without recovery ( Fig. 6e and 6f). Furthermore, a concentration-dependent inhibition of CI values for the LQT-hiPSC-CMs was observed at ≥ 10 µmol/L (Fig. 6b), and beating of the cells in some wells was arrested from 1 h to 12 h at 10 µmol/L. In addition, the cells in all wells treated with 30 µmol/L ceased beating within 24 h after exposure (Fig. 6d). Further analysis showed that lapatinib at lower concentration (0.3 µmol/L to 3 µmol/L) increased the beating rate of LQT-hiPSC-CMs minimally, while 10 µmol/L markedcly decreased the rate and even stopped the beating early after exposure (Fig. 6g). In addition, lapatinib exerted a concentration-dependent inhibitory effect on the beating amplitude of the LQT-hiPSC-CMs (Fig. 6h). Evaluation of the concentration-dependent effects of lapatinib on WTand LQT-hiPSC-CMs demonstrated IC 50 values of 6.87 µmol/L and 10.89 µmol/L, respectively (Table 1).

Discussion
TKI-induced cardiotoxicity remains a major consideration in the treatment of cancer. In this study, using a stem cell-based strategy, we compared the beating parameters of CMs derived from a healthy donor and LQT1 patient before and after exposure to TKIs. Our results demonstrated that hiPSC-CMs simulate many physiological characteristics of human CMs, which is consistent with the increasing use of hiPSC-CMs as potential in vitro models for identifying drug targets as well as determining drug responses and toxicities.
In the current study, before exposure to different TKIs, the spontaneous beating rate of both WT-and LQT-hiPSC-CMs was stable at approximately 40 bpm. Paci et al. [14] demonstrated that the rate of spontaneous beating of ventricular-and atrial-like hiPSC-CMs was 35.3 ± 2.2 and 50 ± 10, respectively, by developing two computational models of hiPSC-CM action potential. This previously published result supports our finding and indicates that the CMs we studied likely exhibited a ventricular-like phenotype. Another in silico model based on hiPSC-CMs was able to reproduce the LQT syndrome with large-scale simulations, and this LQT1 model showed a beating rate of 44 ± 14 [15], similar to the results of our study. After stabilizing the cells in our study, they were treated with different TKIs to determine the sensitivity of the two types of CM models to these agents.
Crizotinib, a receptor TKI with several targets including c-ros oncogene1 (ROS1) and MET protooncogene (MET), is mainly used to treat anaplastic lymphoma kinase (ALK)-positive advanced non-small cell lung cancer (NSCLC). The most frequent crizotinib-related cardiotoxicities reported in clinical trials were QT interval prolongation and bradycardia [16,17]. Consistent with these clinical findings, slowing of the beating rate was observed in our study, and both WT-and LQT-hiPSC-CMs lost all synchronous contractions 0.5 h after treatment with 3 µmol/L crizotinib.
Similarly, Doherty et al. [18] evaluated the cardiotoxicity of crizotinib and discovered that the cessation of beating induced by 10 µmol/L crizotinib lasted for up to 24 h. Our results indicated that WT-and LQT-hiPSC-CMs exhibited a similar sensitivity to crizotinib. Previous studies considered that the pathogenesis of crizotinibinduced bradycardia is probably related to the drug's antagonism of L-type calcium channels, chronotropic effects on the sinoatrial node, anti-mesenchymal-epithelial transition effects, or anti-MET effects [19,20]. Moreover, crizotinib could affect other ion channels, and it was shown to block human ether-a-go-go-related gene (hERG) and Na v 1.5 ion channels with IC 50 values of 1.7 µmol/L and 3.5 µmol/L, respectively.
In addition to blockade of ion channels, reactive oxygen species (ROS) accumulation, activation of the apoptotic cascade, and increased cholesterol synthesis contribute to the cardiotoxicity of these drugs [18]. Therefore, the cardiotoxicity of crizotinib does not seem to involve the I Ks , but is likely related to inhibition of Ca v ion channels and metabolism. The multi-ion channel inhibition suggests that LQT1 patients do not necessarily have higher sensitivity to cardiotoxicities than non-LQT1 patients.
Sunitinib targets vascular epidermal growth factor receptor 1 (VEGFR1)-3, platelet-derived growth factor receptor (PDGFR)-α and -β, and colony-stimulating factor 1 receptor (CSF1R) [19]. This agent is known for its therapeutic effect on renal cell carcinoma, chronic myeloid leukemia, imatinib-resistant gastrointestinal stromal tumor, pancreatic cancer, and neuroendocrine tumors [2]. The cardiotoxicity of sunitinib has been reported in several clinical trials and it causes hypertension and myocardial ischemia frequently because it inhibits blood circulation.
In the cardiovascular system, the tyrosine kinase inhibition induced by sunitinib impairs cellular signal transduction, cell cycle regulation and metabolism, and transcription [21]. In our study, 10 µmol/L sunitinib potently affected the growth of both WT-hiPSC-CMs and LQT-hiPSC-CMs. Previous studies have illustrated the major pathways mediating the toxicity of sunitinib, including inhibition of VEGF and PDGRF signaling and coronary microvascular dysfunction [20]. Activation of the endothelin-1 system, inhibition of stem cell growth factor receptor, and cellular energy compromise due to inhibition of AMP-activated protein kinase (AMPK) also contribute to progression of the toxicity of sunitinib [21].
Therefore, the negative effect of sunitinib on the overall survival rate of CMs is mainly mediated by activation of apoptotic pathways and disruption of energy metabolism. Moreover, LQT-hiPSC-CMs exhibited a more severe dysfunction of spontaneous beating. A previous study indicated that sunitinib blocked the hERG channel but had less effect on Na v 1.5 and Ca v ion channels, which possibly explains the aggravated burden of repolarization in LQT patients [18]. These results suggest that LQT1 patients should be more closely monitored under such treatments.
Dasatinib was approved by the US Food and Drug Administration (FDA) in 2006 as a small-molecule BCR-ABL1targeted TKI. Thus, this agent was first used to treat patients with Philadelphia chromosome positive (Ph + ) chronic myeloid leukemia (CML) in the chronic accelerated and acute phases, who were intolerant or resistant to imatinib. Cardiovascular toxicities including pleural effusions, pulmonary arterial hypertension, and QT interval prolongation can occur during treatment with dasatinib.
Compared to other TKIs, dasatinib is thought to have relatively few cardiac side effects because the mean QT interval change is only 3-13 ms [22]. In agreement with this notion, in this study, the CMs were less affected at the highest concentration (10 µmol/L) of dasatinib than with the other tested TKIs. However, the risk of reduction in heart rate cannot be ignored, particularly with LQT1 patients. The spontaneous beating rate was inhibited in a concentration-dependent manner to the same degree in both CM models, and the amplitude showed a compensatory increase. Although temporary cessation of beating occurred in the LQT-hiPSC-CMs 0.5 h after exposure, the amplitude still showed an increasing trend.
Recently, a study indicated that dasatinib decreased heart rate and cardiac output in a dose-dependent manner, and cardiac troponin I levels were also increased in vivo in treated dogs [23,24]. These results indicate that minor myocardial damage caused by dasatinib and mitochondrial dysfunction might be the most plausible explanations for the observed results. Moreover, dasatinib had a stronger effect on the contraction rate than it did on the amplitude, which could explain the negative inotropic effect observed in vivo.
Lapatinib targets human epidermal growth factor receptor 2 (HER2, HER2/neu, or ErbB2) and epidermal growth factor receptor (EGFR)/HER1 and is used in combination with capecitabine for the treatment of ErbB2 overexpression in advanced or metastatic breast cancers [25]. The most common clinical manifestation of the cardiotoxicity of lapatinib is decreased left ventricular ejection fraction (LVEF), although the underlying mechanism is still not completely understood [26]. Likewise, in our study, lapatinib decreased the spontaneous beating at a notably lower concentration than the other TKIs and caused beating cessation at ≥ 10 µmol/L. However, in the LQT-hiPSC-CMs, we observed an increased rate at lower concentrations. The amplitude from the LQT-hiPSC-CMs was decreased in a lapatinib concentration-dependent manner. Recently, it was reported that lapatinib suppressed I Ks , I Kr , and I K1 , but not I Ca , and suppressed the amplitude of peak I Na , with IC 50 values of 1.84 µmol/L and 0.8 µmol/L, against I Ks and I Kr , respectively. Furthermore, the QTc (corrected QT) prolongation mediated by lapatinib was reversed by isoproterenol (an activator of I Ks ) in mice [27,28], which may explain the more significant decrease in the rate observed with WT-hiPSC-CMs than with LQT-hiPSC-CMs, to a certain degree.
The stability of ion channels, ion carriers, and ion flow in CMs is essential for their efficient operation. Mutations in KCNQ1 that cause LQT1 can reduce I Ks and prolong action potential duration (APD) by two mechanisms: 1) impaired trafficking of KCNQ1 proteins to the plasma membrane and 2) dysfunction of membrane channel proteins [29]. The significantly decreased CI indicates that cytotoxicity caused by the "off-target effect" played an important role in the cardiotoxicity of TKIs. TKIs mainly modulate CM functions via various signaling pathways and mechanisms such as mitochondrial toxicity, inhibition of AMPK and PDGFR, and effects on the vasculature [2,[30][31][32].
The effect on electrical activity of CMs is the secondary cause of TKI-induced cardiotoxicity, and inhibition of certain ion channels might have severe detrimental effects in patients with congenital ion channel diseases. The APD of ventricular-like hiPSC-CMs derived from an LQT1 patient treated with ML277 was shortened significantly, and LQT1-specific CMs were more sensitive to the effect of I Kr blockade than non-LQT1-specific CMs, which prolonged the QTc more significantly in the former cells [9,33,34]. This confirms the compensatory enhancement of I Kr in LQT1 patients due to the lack of I Ks , which indicates that drugs that block I Kr have a higher risk of inducing arrhythmia in LQT1 patients than drugs that do not block this channel, similar to sunitinib.
In contrast, multichannel blockers probably do not exert a more potent effect on LQT1 patients than on non-LQT1 patients because of comprehensive physiological regulatory processes. In particular, similar to non-LQT1 patients, those with LQT1 might not develop arrhythmias caused by agents that block I Ks , such as lapatinib.
Moreover, the effect on sodium channels may indirectly affect intracellular calcium concentrations via Na + -Ca 2+ exchange, leading to regulation of the sarcoplasmic reticulum Ca 2+ load and effects on contractility.
In summary, crizotinib, sunitinib, and lapatinib were considered to have a high risk for affecting cardiac myocardial contraction, consistent with the findings of previous studies, whereas dasatinib was found to be relatively safe [35]. Moreover, LQT1 patients should be monitored when using sunitinib. The current study provides valuable information on the mechanisms by which TKIs affect cardiac myocardial contraction of both WT-and LQT-hiPSC-CMs. Because of the high throughput and accuracy of this current platform, it is expected to contribute to saving costs during drug screening and facilitating drug discovery pipelines. Patient-derived hiPSC-CMs simulate the characteristics of genetic cardiac diseases and, therefore, could facilitate clarification of the mechanisms involved in the development and progression of diseases caused by genetic factors. Furthermore, patient-derived hiPSC-CMs could be used to determine whether a patient's genetic composition would affect their likelihood of developing drug-induced cardiotoxicity. The results of this study also provide a reference for appropriate drug treatment with TKIs and their management.
The present study has some obvious limitations that are worth mentioning. The G219E mutation in the voltagesensing domain of KCNQ1 and the resulting pathogenic mechanisms have not been reported yet. The comparison of CMs derived from a healthy control donor and the patient did not account for potential gene polymorphisms. Finally, whether CMs carrying the G219E mutation in KCNQ1 could reflect the responses of all LQT1 patients to the investigated TKIs still requires comprehensive verification.

Conclusion
In conclusion, we found that WT-hiPSC-CMs and LQT-hiPSC-CMs from different sources exhibited similar responses to several TKIs. To the best of our knowledge, this is the first such report. Of the four tested drugs, crizotinib, sunitinib, and lapatinib induced a higher degree of cardiotoxicity than dasatinib, which we considered to be a relatively safe and effective drug, as previously reported. Moreover, LQT1 patients should exercise extreme caution with sunitinib. Following exposure to the tested drugs, the healthy hiPSC-and LQT-hiPSC-CMs exhibited different outcomes involving their mechanical beating properties and varying sensitivities. These findings could contribute to providing a platform for precision medicine and drug screening in the future.