Entrectinib results in ventricular tachycardia and Brugada phenocopy through inhibition of sodium current.

Background
Entrectinib is a tropomyosin receptor kinase (TrK) inhibitor currently approved for the treatment of ROS1-positive non-small cell lung cancer (NSCLC) and neurotrophic tyrosine receptor kinase (NTRK) gene fusion-positive solid tumors [1]. ROS1 fusion is found in 1–2% of all NSCLCs, and 30% of patients with ROS1 fusion will develop central nervous system (CNS) involvement [2]. Entrectinib has shown promising results in patients with and without CNS involvement, making it an attractive first-line option for the treatment of patients with molecularly diagnosed malignancies [3]. However, the cardiovascular adverse event profile of entrectinib has not yet been fully elucidated. Here, we present a case of ventricular arrhythmia and Brugada phenocopy that occurred three days after the initiation of entrectinib. Next, we aim to investigate the underlying molecular mechanisms by which entrectinib induces a Brugada phenocopy, utilizing human-induced pluripotent stem cell-derived cardiomyocytes as an experimental model system.

Case report
A 40-year-old man with metastatic ROS1-positive adenocarcinoma of the lung developed acute-onset dyspnea three days after initiating entrectinib 600 mg daily. Upon evaluation by emergency medical services (EMS), a wide-complex rhythm at 215 beats per minute was recorded (Fig. 1A). The rhythm exhibited left bundle branch morphology with a left inferior axis and a precordial transition at V4, consistent with ventricular tachycardia arising in the anterior outflow tract region. He received intravenous lidocaine (100 mg) and synchronized cardioversion, which restored sinus rhythm.

The post-conversion electrocardiogram (ECG) showed diffusely low voltage, coved ST segment elevation in the anterior precordial leads, with horizontal ST elevation in II, III, and aVF, and a corrected QT interval of 445 milliseconds (Fig. 1B). The cardiac troponin I level was slightly elevated to 0.04 ng/mL (normal ≤ 0.03 ng/mL). Computed tomography of the chest was negative for pulmonary embolism, and emergency coronary angiography revealed mild stenosis of the distal left anterior descending artery and otherwise normal arteries. Echocardiography showed normal left ventricular systolic function with an ejection fraction of 64%, and a moderate-sized pericardial effusion with no evidence of tamponade.
An ECG performed one hour after initial evaluation showed evolution and partial resolution of the ST changes. The troponin I level peaked four hours after arrival at 0.05 ng/mL. The patient was admitted and started on metoprolol tartrate 12.5 mg twice daily, and entrectinib was discontinued. Cardiac magnetic resonance imaging (CMR; Siemens 1.5T) performed on hospital day three demonstrated no myocardial or pericardial late gadolinium enhancement (LGE). Myocardial tissue characterization with parametric maps, including T1-time, T2-time, and extracellular volume (ECV), was within normal limits. On hospital day four, the patient underwent diagnostic pericardiocentesis with drainage of 360 mL of straw-colored fluid. Cytology revealed pleomorphic malignant cells consistent with a malignant pericardial effusion.
An ECG performed prior to hospital discharge on hospital day seven showed complete resolution of the patient’s ST segment changes and normalization of voltage (Fig. 1B). The patient was discharged with a wearable external cardiac defibrillator (ZOLL LifeVest). His discharge medications included metoprolol succinate 50 mg daily. Comprehensive cardiac genetic testing using NextGen sequencing on an Illumina sequencing system was negative for known pathogenic variants and variants of uncertain significance (VUS) across 175 genes known to be associated with hereditary cardiac conditions, including SCN5A, the gene encoding the cardiac sodium channel NaV1.5. At the time of a two-month follow-up visit, following the availability of genetic evaluation, no further ventricular arrhythmias had occurred, and the patient’s wearable defibrillator was discontinued. At the time of manuscript preparation, the patient was receiving treatment with a different small molecular TKI targeted to ROS1, larotrectinib, with an excellent response from an oncologic standpoint and without recurrence of clinical arrhythmia or electrocardiographic changes.
The ECG changes we observed are consistent with a type I Brugada ECG pattern, and a detailed evaluation of this patient’s presentation supports a possible diagnosis of drug-induced Brugada syndrome (BrS), also known as Brugada phenocopy, related to treatment with entrectinib. However, we could not exclude the possibility that the observed ST elevations resulted from pericarditis and/or myocarditis due to the known metastatic disease or from acute changes post-cardioversion. Therefore, we aimed to test the hypothesis that entrectinib leads to ventricular tachycardia and Brugada phenocopy through inhibition of sodium currents in NaV1.5.

Materials and methods

Ethical approval
This study was reviewed and approved by the Institutional Review Board at Vanderbilt University Medical Center, Nashville, Tennessee, USA. (IRB # 9047).

Statistical analysis
Statistical analyses were performed with GraphPad Prism (GraphPad, San Diego, CA). Results are presented as the mean ± standard error mean (SEM). Two-tailed Student’s t-tests compared two groups, and multiple group comparisons were analyzed by either one- or two-way ANOVA followed by Tukey post-hoc analysis. A p-value of < 0.05 was considered statistically significant.

Generation of Human-Induced Pluripotent Stem Cells (hiPSCs)
Peripheral blood mononuclear cells (PBMCs) were isolated from the peripheral blood cells of the consenting participants as previously described [4]. PBMCs were reprogrammed into iPSCs using electroporation of non-integrating episomal vectors (Epi5™ Episomal iPSC Reprogramming Kit, ThermoFisher; Neon Transfection System, ThermoFisher).

Differentiation of Human-Induced Pluripotent Stem Cell-Derived Cardiomyocytes (hiPSC-CMs)
Ventricular-specific iPSC-derived cardiomyocytes were generated as previously described [5, 6]. iPSCs were grown to approximately 80% confluency before initiating cardiac differentiation. At day 0, cells were treated with “differentiation media” composed of RPMI 1640 supplemented with GlutaMAX and HEPES (Life Technologies), further supplemented with 0.2 mg/mL L-ascorbic acid 2-phosphate and 0.5 mg/mL human recombinant albumin (Sigma). Cells were also treated with 6 µM CHIR99021 (LC Laboratories) from day 0 to day 2. On day 3, cells were treated with differentiation media that also contained 5 µM IWR-1 (Merck Millipore). On day 5, the media was changed, and cells were fed with differentiation media. On days 7–15, cells were fed every other day with “culture media” composed of RPMI 1640 supplemented with GlutaMAX and HEPES (Life Technologies), further supplemented with 1x B27 plus insulin (ThermoFisher). On day 15, cells were digested with 10x TripLE Select (ThermoFisher) and passaged for metabolic selection. Cells were allowed to recover from passaging for 7 days prior to metabolic selection with “selection media” containing RPMI 1640 without glucose and HEPES (Life Technologies) supplemented with 0.2 mg/mL L-ascorbic acid 2-phosphate and 0.5 mg/mL human recombinant albumin (Sigma), as well as a final concentration of 4 mM lactate/HEPES. Metabolic selection occurred for five days with media changes every 2–3 days. Upon completing metabolic selection, cells were treated with culture media changes occurring every 2 days. Cells were cultured for a total of 45 days, from the initiation of the differentiation protocol, before undergoing functional studies.

Western blotting
hiPSC-aCMs were homogenized using sonication in homogenization buffer and 1x Complete EDTA-free protease inhibitor cocktail tablet (Roche). Protein concentrations were quantified by Pierce BCA protein assay (BioRad). Equal quantities of protein extract were denatured in 2x Laemmli buffer (Bio-Rad) with 10% beta-mercaptoethanol at 60 °C for 10 min. Samples were separated on a 4–20% gradient SDS-PAGE gel and transferred to a PVDF membrane for immunostaining. Primary antibody concentrations were as follows: anti-total NaV1.5 1:2,500 (Abcam rabbit pAb, catalog #ab56240) and anti-GAPDH 1:10,000 (Thermo-Fisher Scientific mouse mAb, catalog #AM 4300). Secondary antibody concentrations were anti-mouse IgG-HRP conjugate 1:10,000 and anti-rabbit IgG-HRP conjugate 1:10,000 (Promega). Clarity Western ECL substrate (BioRad) was used to develop immunoblots, which were imaged on an iBright 1500 imaging system (Invitrogen).

Whole cell current-clamp action potential recordings
In whole cell rupture patch current-clamp mode, action potentials (APs) from isolated human iPSC-CMs on day 35–45 of differentiation were elicited by injection of a brief stimulus current (1–2 nA, 2–6 ms and a 0.5 Hz of pulse stimulations) at room temperature (22 ~ 23 ˚C). The bath (extracellular) solution contained (in mM): NaCl 145, KCl 4.0, CaCl2 1.8, and MgCl2 1.0, HEPES 5.0, glucose 10, with a pH of 7.4 (adjusted by NaOH). The pipette (intracellular) solution contained (in mM): KCL 130, ATP-K2 5.0, MgCl2 1.0, CaCl2 1.0, BAPTA 0.1, and HEPES 5.0, with a pH of 7.3 (adjusted by KOH). Glass microelectrodes with tip resistances of 1 ~ 2 mΩ were used. Susceptibility to afterdepolarizations was evaluated using a 0.1 Hz pacing rate while all other studies were paced at 0.5 Hz. Ten successive traces were averaged for analysis of action potential durations at 90% repolarization (APD90).

Cell survival assays
Cell-Titer-Glo luminescence cell viability assays (Promega) were utilized to determine cell survival following treatment with entrectinib (Selleckchem). hiPSC-aCMs were plated on 96-well plates and allowed to mature for seven days with daily media exchanges. Increasing concentrations of entrectinib were added at day 7, and cells were allowed to incubate overnight prior to treatment with the luminescence viability assay reagents, which are used to quantify cell viability as a measure of ATP production. Absorbance measurements were obtained on a GloMax plate reader (Promega).

Results
To determine the effect of entrectinib treatment on sodium current in human cardiomyocytes, we utilized human pluripotent stem cell-derived ventricular cardiomyocytes (hiPSC-vCMs) as a model system. Population-control derived hiPSC-vCMs were generated as previously described and were considered mature for experimental analysis after 45 days [4–6]. To determine whether entrectinib treatment results in sodium-channel current blockade through either fast or slow kinetics, hiPSC-vCMs were treated with entrectinib (1 µM) for either a brief (15 min) or prolonged exposure (48 h). Post-treatment, whole-cell patch clamp recordings were performed. Treatment of hiPSC-vCMs with entrectinib (1 µM) for 48 h resulted in a significant decrease in sodium currents and a positive shift in the voltage-dependence of inactivation (Fig. 2A, n = 12 cells/condition, p < 0.01). This effect was not observed with brief (15-minute) treatment, where we noted no change in peak sodium current (at -30 mV, 5960 ± 244 vs. 5883 ± 229 pA pre- vs. post-treatment, n = 4 cells/condition, p > 0.05). To rule out cellular toxicity as the mechanism by which entrectinib treatment led to decreased sodium currents, we determined the median lethal dose (LD50) of entrectinib on hiPSC-aCMs using a cell survival assay. Following 24 h of treatment with entrectinib, the LD50 for entrectinib was calculated to be 330 µM (Fig. 2B), well above the concentration (1 µM) utilized in our studies. Western Blot analysis was performed to exclude changes in NaV1.5 expression following entrectinib treatment. Prolonged (48-hour) treatment with entrectinib did not alter NaV1.5 protein expression (Fig. 2C).

Discussion
Here we report the case of a 40-year-old man with no personal or familial history of cardiovascular disease who developed monomorphic ventricular tachycardia and coved ST-segment elevations in the anterior precordial and inferior leads three days after starting entrectinib for the treatment of ROS1-positive lung adenocarcinoma. These ECG changes are consistent with a type I Brugada ECG pattern, and a detailed evaluation of this patient supports a likely diagnosis of drug-induced Brugada syndrome (BrS), also known as Brugada phenocopy, related to treatment with entrectinib.
The diagnosis of drug-induced Brugada syndrome is a diagnosis of exclusion, and care was taken in this case to exclude other potential causes of the patient’s ECG changes and ventricular tachycardia. Stress CMR excluded underlying structural heart disease, myocardial ischemia, stress cardiomyopathy, and myocarditis. Electrical cardioversion may transiently cause ST-segment elevations, however, these effects usually resolve within minutes. Malignant pericardial effusions rarely cause arrhythmias, and there was no evidence of pericarditis on cardiac MRI, nor any classic symptoms of pericarditis to explain the patient’s ECG findings. A negative result on cardiac genetic testing, while not definitive, makes a genetic cause of arrhythmia less likely, and no arrhythmia episodes were observed after discontinuation of the drug. An unusual aspect of this case is presentation with monomorphic ventricular tachycardia; while this has been reported in BrS, it is rare, with patients more commonly manifesting ventricular fibrillation [7]. To our knowledge, this is the third reported case of a Brugada ECG pattern related to entrectinib, and the first to include the results of genetic testing to evaluate for predisposing genetic conditions [8, 9].
Safety data from small clinical trials of entrectinib suggest that the drug carries a lower arrhythmic risk compared to Bruton’s Tyrosine Kinase (BTK) inhibitors, a class of TKI with known pro-arrhythmogenic properties [10]. However, in these small, short-duration studies, electrocardiographic abnormalities and cardiovascular toxicities were not unusual. Among 161 clinical trial patients included in the entrectinib Food and Drug Administration (FDA) safety analysis, 3.1% developed QT interval prolongation of > 60 milliseconds, and 3.4% developed heart failure [3]. The modest incidence of arrhythmias seen in these early studies may underestimate the true arrhythmic potential of the drug, as early clinical trials often include patients who are less likely to experience adverse side effects, and phase II and III trials often lack the long-term follow-up needed to identify the true scope of adverse events. Large-scale studies of the drug are unlikely to occur given the low incidence of ROS1-positive tumors; however, case reports and translational research studies can provide valuable insight into a drug’s risk profile when other forms of evidence are lacking.
Cardiotoxicity, including inflammatory cardiomyopathies, was observed in clinical trials of entrectinib, and case reports have described both lymphocytic and eosinophilic myocarditis as well as acute myopericarditis. In three clinical trials (ALKA-372-001, STARTRK-1, and STARTRK-2), a total of 344 patients received entrectinib, and congestive heart failure (CHF) and myocarditis were detected in 3.4% and 0.3% of patients, respectively [11]. Three case reports of myocarditis following initiation of entrectinib have been documented. Two cases involved lymphocytic myocarditis, with one case experiencing a fulminant course, and eosinophilic myocarditis was reported in one case [11–13]. New-onset ventricular tachycardia was noted in each case. Acute myopericarditis resulting in a large pericardial effusion with tamponade physiology has also been reported [14]. In our case report, cardiac MRI did not show evidence of myocarditis, and markers of myocardial necrosis were only mildly elevated (troponin I level peaked at 0.05 ng/mL with the upper limit of normal < 0.04 ng/mL); however, an endomyocardial biopsy was not performed, and we thus cannot definitively exclude a component of myocarditis.
Emerging evidence suggests that myocardial inflammation may play a crucial role in the development of Brugada syndrome [15]. Brugada pattern and ventricular arrhythmias have long been recognized as sequelae of fever [16]. However, conflicting clinical and biochemical evidence exists regarding inflammatory responses following the inhibition of ROS1 and NRTK with small-molecule kinase inhibitors, including entrectinib. Crizotinib is a small-molecule TKI targeting ALK and ROS1, and crizotinib-induced pneumonitis is a rare but potentially fatal form of interstitial lung disease associated with ROS1 inhibition, suggesting a pro-inflammatory effect of ROS1 inhibition [17]. Studies of eosinophilic esophagitis, which results from allergic inflammation, have shown that IL-13 stimulates NTRK1 expression, suggesting an anti-inflammatory effect of NTRK1 inhibition [18]. Activation of the NLRP3 inflammasome has been implicated in the pathogenesis of inflammatory and autoimmune disease [19]. Entrectinib has been shown to potently inhibit the assembly and activation of the NLRP3 inflammasome and has been proposed as a potential therapy for the treatment of inflammasome-related diseases [20]. Due to the interplay between biochemical feedback pathways in highly regulated kinase networks and the common off-target effects of small-molecule kinase inhibitors, elucidating the inflammatory state that occurs following the inhibition of NTRK and ROS1 by entrectinib is highly complex, and additional basic and translational research is needed [21].
Our findings indicate that brief treatment with entrectinib did not affect sodium currents. In contrast, prolonged treatment decreased the peak sodium current and shifted the voltage dependence of inactivation, which may lead to Brugada phenocopy and predispose individuals to ventricular arrhythmia. Prolonged treatment did not alter NaV1.5 abundance, indicating that the changes in sodium current do not arise from direct channel inhibition but rather through other pathways yet to be defined. Elucidating the downstream signaling pathways by which entrectinib treatment results in ventricular tachycardia and Brugada phenocopy may provide novel insights into the molecular mechanisms of arrhythmogenesis. As such, further research into entrectinib-mediated regulation of sodium currents is warranted and necessary.