Off-Target Effects of Cancer Therapy on Development of Therapy-Induced Arrhythmia: A Review

Background: Advances in cancer therapeutics have improved overall survival and prognosis in this patient population; however, this has come at the expense of cardiotoxicity including arrhythmia. Summary: Cancer and its therapies are associated with cardiotoxicity via several mechanisms including inflammation, cardiomyopathy, and off-target effects. Among cancer therapies, anthracyclines and tyrosine kinase inhibitors (TKIs) are particularly known for their pro-arrhythmia effects. In addition to cardiomyopathy, anthracyclines may be pro-arrhythmogenic via reactive oxygen species (ROS) generation and altered calcium handling. TKIs may mediate their cardiotoxicity via inhibition of off-target tyrosine kinases. Ibrutinib-mediated inhibition of CSK may be responsible for the increased prevalence of atrial fibrillation. Further investigation is warranted to further elucidate the mechanisms behind arrhythmias in cancer therapies. Key Messages: Arrhythmias are a common cardiotoxicity of cancer therapies. Cancer therapies may induce arrhythmias via off-target effects. Understanding the mechanisms underlying arrhythmogenesis associated with cancer therapies may help design cancer therapies that can avoid these toxicities.


Introduction
Cancer and cardiovascular disease are leading causes of morbidity and mortality in the developed world [1]. Though developments of novel cancer treatment have led to improved overall survival in this patient population, this has come at the cost of cardiotoxicity associated with these therapies [2][3][4]. Given the risk of cardiotoxicity and arrhythmias associated with cancer therapy, the European Society of Cardiology (ESC) guidelines on cardio-oncology recommend that an electrocardiogram (ECG) be obtained in all patients starting cancer therapy and routine ECG monitoring in patients at risk of cardiac arrhythmias due to cancer therapy [5]. Cardiac arrhythmias have been associated with a wide range of cancer therapies, including conventional chemotherapy, immune checkpoint inhibition (ICI) and targeted therapies, such as monoclonal antibodies and tyrosine kinase inhibitors (TKIs) [3,6]. The underlying pathophysiology of arrhythmias in patients with cancer is further compounded by the type of cancer therapy, which may lead to either the development of cardiomyopathy, increased inflammation, QT prolongation, and/or myocardial ischemia. Conventional chemotherapy drugs work by interfering with cancer cell growth and replication; however, these same effects may affect other tissues including cardiomyocytes through the karger@karger.com www.karger.com/crd same anticancer mechanisms or other unintended effects (reactive oxygen species [ROS] creation, abnormal calcium handling, etc.) [2]. Targeted therapies, including TKIs, have ushered in an era of more precise therapies for cancer. However, TKIs are often not specific and can inhibit other tyrosine kinases not involved in oncogenesis. Recent data suggest that the pro-arrhythmic effects of some cancer therapeutics, including anthracyclines and TKIs, may be mediated by these off-target effects. Understanding the role of off-target effects in arrhythmogenesis in cancer therapy may help guide future antineoplastic design that eliminates the pro-arrhythmic side effects.

Epidemiology of Arrhythmias in Cancer
Patients with cancer are at an increased risk of atrial and ventricular arrhythmias, which may hold prognostic significance. The prevalence and incidence of atrial fibrillation (AF) are elevated among patients with cancer compared with the general population [4,7]. A nationwide cohort study in Denmark found patients with cancer had an AF incidence of 17.4 per 1,000 personyears compared with 3.7 per 1,000 person-years in patients without cancer, with the risk of new-onset AF highest in the first 90 days after cancer diagnosis [7]. In another population-based case-control study, patients with a colorectal cancer diagnosis were almost 12 times more likely to also have been diagnosed with AF/flutter compared to age, sex, and county-matched controls [8]. Patients diagnosed with breast cancer were more than twice as likely to develop AF as matched controls, and after adjusting for demographics, breast cancer-specific features, and cardiovascular risk factors, new-onset AF was associated with greater all-cause mortality [9].
In one retrospective cohort of patients that had an implantable cardioverter-defibrillator (ICD) placed, the incidence of ventricular tachycardia or fibrillation (VT/ VF) was greater in patients with cancer than in patients without cancer [10]. Further, the frequency of VT/VF increased from 0.12 ± 0.21 episodes per month to 1.19 ± 0.32 episodes per month in patients that received new diagnoses of cancer after ICD implantation [10]. In two prospective studies, patients with cancer were significantly more likely to experience non-sustained ventricular tachycardia (NSVT) on 24-h ECG than age-and sexmatched controls without cancer [11,12]. Additionally, the presence of NSVT and greater premature ventricular contraction (PVC) burden were associated with increased mortality among patients with cancer independent of relevant clinical characteristics [11,12].

Risk Factors of Arrhythmias in Patients with Cancer
The association between cancer and atrial and ventricular arrhythmias may be attributed to overlapping cardiovascular risk factors in cancer patients such as increased age, hypertension, obesity, and smoking, along with adverse effects of cancer treatments [13][14][15][16][17][18][19][20][21]. Further, arrhythmia risk varies by cancer type and staging. Lung cancer, esophageal cancer, and multiple myeloma have been associated with an increased risk of AF as compared to other cancers, with advanced stage cancers associated with both AF and VT/VF [7,9,10,22]. Chronic inflammation may represent a pathophysiologic link between cancer and arrhythmia. The presence of chronic inflammation increases cancer risk, and elevated inflammatory markers including C-reactive protein have been found to be associated with increased risk of AF and thrombosis (an important risk factor for subsequent ventricular arrhythmia) [23][24][25][26][27][28]. Additionally, increased NLRP3 inflammasome signaling, an important regulator of innate immunity, has been implicated in the pathogenesis of AF [29]. Oncologic surgery and radiation further increase the risk of arrhythmias. In a large cohort study, 12.6% of patients who underwent pulmonary lobectomy or pneumectomy for lung cancer experienced postoperative atrial arrhythmias, and in a separate study, 4.4% of patients undergoing elective surgery for colorectal cancer developed postoperative AF [30,31]. Thoracic radiation therapy leads to myocardial fibrosis, accelerated atherosclerosis, and cardiac autonomic dysfunction [32][33][34][35]. Myocardial fibrosis creates substrate heterogeneity, which increases propensity for arrhythmia.
Cardiotoxicity, specifically the risk of clinically significant arrhythmias, remains a concern with antineoplastic medications (Table 1) [2, 3, 36]. Anthracyclines are used to treat a wide range of solid and hematologic malignancies and exhibit significant cardiotoxicity, which may manifest as arrhythmias and/or systolic dysfunction. In those with anthracycline-associated cardiomyopathy who received an ICD, 30.4% experienced VT/VF, similar to a control group that had alternative etiologies for cardiomyopathy, yet 56.6% experienced AF, a higher incidence than the control group [37]. Antimetabolites such as 5fluorouracil (5-FU), cladribine, cytarabine, fludarabine, and gemcitabine are associated with cardiotoxicity. Fluoropyrimidines such as 5-FU's manifestations of cardiotoxicity include chest pain and acute coronary syndromes due to coronary vasospasm, in addition to AF and other arrhythmias [2, 36,38]. The most common mechanisms of 5-FU related cardiotoxocity are ischemia and drugrelated myocardial toxicity, which may be related to the administration of 5-FU as a continuous infusion, and attenuated by bolus-based regimens [38]. Further, arrhythmias associated with 5-FU include AF and sinus bradycardia, whereas supraventricular tachycardia (SVT) has been associated with fludarabine and gemcitabine [38][39][40][41]. Alkylating agents including melphalan and cyclophosphamide may cause cardiotoxicity due to oxidative stress leading to arrhythmias such as AF and SVT, with incidences reported at 8% and 3%, respectively [2, [42][43][44]. Platinum-based alkylating agents such as cisplatin are associated AF in 8% of patients treated systemically and up to 32% of patients who receive intrapericardial or intrapleural treatment [36,45]. Among the targeted agents, TKI, mabs, and immune checkpoint inhibitors (ICI) have been shown to induce a variety of arrhythmias. Ibrutinib, a TKI, is frequently used to treat chronic lymphocytic leukemia and is associated with a 3.5-16% incidence of AF, with a relative risk of developing AF of 4.69 [46][47][48]. Vemurafenib, a BRAF Kinase inhibitor, and sunitinib, a multitargeted kinase inhibitor, both frequently cause QTc prolongation, and the anaplastic lymphoma kinase (ALK) inhibitors are associated with bradycardia [49][50][51][52][53]. Rarely, ICI may cause myocarditis, with an incidence of 1%, but is often fatal, with a 50% mortality risk in those that develop myocarditis [54]. Cardiotoxicity associated with ICI included AF, ventricular arrhythmias, and conduction disorders in 30%, 27%, and 17% of patients, respectively [55]. The vascular endothelial growth factor (VEGF) inhibitor bevacizumab has also been associated with cardiotoxicity leading to bradycardia and cardiomyopathy [56][57][58]. Further, the cardiotoxic effects of trastuzumab, commonly used to treat HER2-positive breast cancer, have been well described, and include decreased left ventricular ejection fraction and new-onset heart  failure, which are frequently reversible after discontinuation [59,60]. Lastly, carfilzomib, a second-generation proteasome inhibitor used to treat multiple myeloma can lead to adverse cardiovascular effects in a dosedependent manner, with arrhythmias reported in 2.4% of all cases [61,62].

Off-Target Effects Mediating Arrhythmias in Cancer Therapy
Arrhythmias are relatively common among patients with cancer undergoing treatment with traditional cytotoxic chemotherapy as well as targeted agents and immunotherapy [3]. While AF is the most common arrhythmia, other arrhythmias including ventricular arrhythmias, bradyarrhythmias, and repolarization and conduction abnormalities have been known to occur. Cancer therapies are thought to induce arrhythmias via a variety of pathophysiological mechanisms including direct cardiomyocyte toxicity, electrolyte disturbances, and offtarget effects (Fig. 1). Understanding the molecular mechanisms underlying arrhythmogenesis of cancer therapies may offer insight into potential therapies to mitigate arrhythmia risk in this patient population.

Anthracyclines
Anthracyclines (doxorubicin, daunorubicin, idarubicin, epirubicin, and mitoxantrone) are cytotoxic chemotherapies commonly used in the treatment of cancers, including breast, gynecologic, sarcoma, leukemia, lymphomas, and pediatric solid malignancies. Anthracyclines are known to cause irreversible cardiotoxicity, manifesting as cardiomyopathy and arrhythmias. Cumulative dose and risk factors, including age and cardiovascular comorbidities, may increase the incidence of Doxorubicin may induce arrhythmia via ROS production, activation of CaMKII, and downregulation of SERCA leading to cytosolic calcium overload. Increased ROS production activates NLPR3 inflammasome leading to increased inflammation and arrhythmia. Ibrutinib inhibits CSK, which is an inhibitor of c-SRC, leading to inflammation, fibrosis, and arrhythmia. cardiotoxicity [37]. Anthracyclines lead to an increase in toxic free radicals, fibrosis, and lipid peroxidation of cardiac membranes, which may lead to acute or lateonset cardiotoxicity. Acute cardiotoxicity generally presents as arrhythmias that are typically reversible after discontinuation. Late toxicities may lead to valvular damage, cardiomyopathy, and more significant atrial and ventricular arrhythmias [63]. In addition, anthracyclines are known to cause QT prolongation and are associated with the development of PVCs, ventricular tachycardia (VT), and AF [37,[64][65][66].
Anthracyclines may mediate their cardiotoxic and arrhythmogenic effects at least in part via activation of toll-like receptor (TLR) 4 leading to innate immune system activation and a pro-inflammatory response [67][68][69]. Further, mitochondrial ROS production may be increased by anthracyclines, providing a mechanism for long-term remodeling of the arrhythmogenic substrate [70]. Indeed increased ROS production due to cancer therapy (not limited to anthracyclines) and ROS-mediated activation of the NLRP3 inflammasome may contribute to increased risk of arrhythmias in patients with cancer [71]. An increased incidence of PVCs and SVT has been reported soon after receiving anthracycline infusion (1-24 h post-infusion), suggesting direct effect of anthracycline on arrhythmogenesis [66,72]. One study of a rat model of doxorubicin-induced cardiomyopathy suggested that the pro-arrhythmic phenotype may be in part due to abnormal intracellular calcium handling caused by downregulation of sarcoplasmic/endoplasmic reticulum calcium-ATPase (SERCA), resulting in delayed afterdepolarizations [73]. Changes in gene regulation by doxorubicin may be related to its inhibition of topoisomerase II α [74]. Additionally, anthracycline-mediated ROS production may also be implicated in abnormal calcium handling [70,75]. Another study using rats showed that infusion with doxorubicin led to downregulation of SERCA and no change in expression of the sodium/calcium exchanger 1 channel (NCX1) [76]. In other studies, late inward sodium current has been observed via direct activation of Na v 1.5 channels by ROS or indirectly due to calcium/calmodulin-dependent protein kinase II (CaMKII), both of which are implicated in anthracycline-mediated cardiotoxicity [77][78][79]. Additionally, persistent late inward sodium current and sodium overload have been shown to increase ROS production and thus leading to a vicious cycle [80]. Interestingly, ranolazine, an inhibitor of I Na , may attenuate abnormal calcium handling induced by doxorubicin by restoring SERCA expression, reducing ROS production, and reducing expression of Na v 1.5 channel and NCX1 [76]. Whether ranolazine or other medications that inhibit I Na are cardioprotective and antiarrhythmogenic in patients treated with anthracyclines is yet to be seen and merits further study.
TKIs and Off-Target Effects Various tyrosine kinases, including receptor tyrosine kinases (RTKs) and non-receptor tyrosine kinases (NRTKs), have been discovered [81]. Tyrosine kinases regulate growth, differentiation, adhesion, among other functions [81]. Abnormal activation of tyrosine kinases (due to either mutations, amplifications, or upregulation) has been implicated in the development of cancer as well as progression and metastatic potential [82]. The identification of tyrosine kinases involved in oncogenesis and metastasis has led to the development of small-molecule TKIs that have radically improved the prognosis of malignancies including, but not limited to, chronic lymphocytic leukemia (CLL), chronic myeloid leukemia (CML), lung cancer with mutations or rearrangements in certain genes, melanoma, and colorectal cancer [82]. However, TKIs are often not specific to the targeted tyrosine kinase and thus unintended off-target effects can occur. Some of these off-target effects can lead to cardiotoxicity, particularly arrhythmias.
Ibrutinib is a TKI approved for the treatment of various B-cell lymphomas including mantle cell lymphoma and CLL [83,84]. In CLL, treatment with ibrutinib has revolutionized survival with low-risk patients having a 93% 3-year overall survival and high-risk patients having 63% 3-year overall survival [85]. The antineoplastic effect of ibrutinib on B-cell neoplasms is mediated via inhibition of Bruton tyrosine kinase (BTK), an early signaling molecule within the B-cell antigen receptor (BCR) signaling pathway [86]. However, ibrutinib has been associated with a significant risk of AF with a 3-to 10-fold increase risk of AF [46,47,[87][88][89][90]. One study suggested ibrutinib-induced AF occurred in the setting of on and off-target BTK inhibition. This study showed cardiac messenger RNA (mRNA) expression of BTK and related kinases, including tec protein tyrosine kinase (TEC), was increased in AF compared with normal sinus rhythm in human atrial tissue [91]. This study also demonstrated a decrease in phosphoinositide 3-kinase (PI3K)-Akt pathway activation via ibrutinib administration in neonatal rat ventricular myocytes [91]. The PI3K-Akt pathway is regulated by both TEC and BTK and has been suggested to be important in cardiac stress response [92,93]. In mouse models of AF, reduced PI3K-Akt activity was shown to increases susceptibility to AF [92]. Another study identified C-terminal Src kinase (CSK) as a potential target of ibrutinib which may be implicated in the development of AF [94]. In this study, mice lacking BTK were treated with ibrutinib for 4 weeks and had increased AF compared with mice not treated with ibrutinib suggesting an off-target effect of ibrutinib being responsible for AF development. The mice treated with ibrutinib also developed increased left atrial fibrosis and enlargement compared with mice treated with vehicle. CSK, a non-receptor tyrosine kinase expressed at higher levels in atrial tissue than in ventricular tissue, was identified as a potential target of ibrutinib that may be responsible for the development of AF after treatment with ibrutinib. Interestingly, CSK-knockout mice had increased left atrial fibrosis and inflammation as well as increased atrial arrhythmias compared with control mice further supporting the hypothesis that off-target inhibition of CSK is important in ibrutinib-associated AF. The same study also investigated the risk of development of AF of other TKIs that inhibit CSK using a pharmacovigilance database and demonstrated increased odds (odds ratio of 2.4, 95% confidence interval 2.1-2.7) of reporting AF in TKIs that inhibit CSK (nilotinib, bosutinib, dasatinib, and vemurafenib) compared with those that do not [94].
CSK inhibits Src family tyrosine kinases (SFKs) via C-terminal phosphorylation thus inhibition of CSK leads to increased SFK activity [95]. SFKs are non-receptor tyrosine kinases involved in diverse cellular process including growth, division, inflammation, survival, and programmed death [95]. The loss of CSK has been shown to induce acute inflammatory response dysfunction and cause an exaggerated inflammatory response in mice [96]. Increased inflammatory infiltration of macrophages and elevated pro-inflammatory cytokines including interleukin-6 and tumor necrosis factor-α was also seen in atrial tissue of CSK-knocked out mice [94]. Another mechanism by which CSK inhibition may lead to increased arrhythmogenesis is via increased SRC kinase-mediated apoptosis of cardiomyocytes [97]. Indeed, c-SRC (which is inhibited by CSK) has been implicated in arrhythmias post-myocardial infarction and in adverse cardiac remodeling [98][99][100]. Therefore, CSK appears to be important in inhibiting pro-arrhythmogenic processes and off-target inhibition via TKIs may be responsible for their proarrhythmia properties (Fig. 2).
ALK rearrangements are found in approximately 5% of patients with non-small cell lung cancer (NSCLC). Patients with ALK-rearranged NSCLC are relatively younger at the time of diagnosis compared with non-ALK-rearranged NSCLC, tend to have never or light smoking history, and are at higher risk of venous thromboembolism and arterial thrombosis [101,102]. TKIs targeting ALK (crizotinib, alectinib, ceritinib, brigatinib, and lorlatinib) have been developed to treat patients with ALK-rearranged NSCLC and have improved longterm outcomes in these patients [103][104][105]. Sinus bradycardia is a common cardiotoxicity associated with ALK inhibitors [53,106]. In a meta-analysis of 1,737 patients treated with ALK inhibitors, the pooled incidence of bradycardia after a median follow-up of 1.26 years was 8% [53]. In another study of 1,053 patients treated with crizotinib, 41.9% of patients had at least one episode of sinus bradycardia (less than 60 beats per minute) and the mean decrease in heart rate from baseline being 25 beats per minute [50]. Interestingly, presence of sinus bradycardia may correlate with clinical tumor response to treatment with ALK inhibitor [107]. While most cases of ALK inhibitor-induced bradycardia are asymptomatic, symptomatic bradycardia (including dizziness) has been described with ALK inhibitor use [53]. In mice, crizotinib has been shown to have a dose-dependent effect on heart rate [108]. Additionally, crizotinib was found to inhibit funny current (I f ) in sinoatrial (SA) node cells isolated from mice via inhibition of hyperpolarization-activated cyclic nucleotide-gated channel (HCN) 4 [108]. Given that HCN4 is the predominant isoform expressed in SA node cells, inhibition of HCN4 by ALK inhibitors may offer a mechanistic explaining of ALK inhibitor-induced sinus bradycardia [109].

Conclusion
Cancer therapies have advanced and improved overall outcomes in patients with malignancy at the cost of cardiotoxicity including arrhythmias [6]. Cardiooncology is a nascent but rapidly growing field within cardiology that bridges the gap between cardiovascular disease and oncology. Cancer therapies may accelerate and promote the development of cardiac arrhythmias via several mechanisms including off-target effects, ROS production, and increased fibrosis. As more is discovered about the underlying mechanisms of arrhythmias associated with cancer therapy, potential therapies for the prevention and treatment of therapy-associated arrhythmias may be developed. Further studies are needed to gain more understanding of the underlying pathophysiology of cancer and therapy-associated arrhythmias in order to improve cardiovascular outcomes of patients with cancer.