Cardiotoxicities induced by trastuzumab have long been established. Although the FDA has already mentioned the cardiac adverse effects ranged from asymptomatic left ventricular dysfunction, arrhythmias, and hypertension to cardiac death, it did not specify the types or the intensity of arrhythmias [14]. A study published in 2022 revealed that women with BC had a higher incidence of atrial fibrillation than the general population (HR 1.98 (95% CI: 1.73–2.27)) [15]. However, the majority of evidence focused on heart dysfunction or heart failure. Our study evaluated HF and Af in patients with HER2 positive BC who received trastuzumab.
Our study found that CKD, rather than DM, HTN, or hyperlipidemia, were the risk factors for getting HF and Af, partially compatible with the results of previous studies. The combination use of anthracyclines, smoking, high body-mass index, and older age [14, 16, 17] have been reported as risk factors for trastuzumab-induced cardiotoxicity. On the contrary, previous radiation to the chest wall, cyclophosphamide, or paclitaxel were said to play minor roles in developing cardiac AEs [17]. The effect of baseline EF and HTN were inconclusive [8, 17, 18]. A more recent meta-analysis retrieved data from 17 articles revealed that HTN, DM, old age, and exposure to anthracycline were the risk factors for trastuzumab-induced cardiac dysfunction [19].
Considering the impacts on cardiac outcomes on concurrent cardiovascular agents, we analyzed how aspirin, NSAIDs, statins, metformin, beta-blockers, and RAAS inhibitors affect the heart AEs. Aspirin, beta-blockers, and RAAS inhibitors were associated with a higher incidence of heart failure and atrial fibrillations. It was impressive because beta-blockers and RAAS agents were standard treatments for CHF. One possible reason was that patients taking those drugs were prone to have AEs. Aspirin, NSAIDs, statins, metformin, beta-blockers, and RAAS drugs might represent the severity of the underlying diseases.
In the subgroup analysis, we found that metformin and statins therapy was associated with a lower incidence of Af. The results were partially consistent with previous studies that statin therapy [20] and metformin [21] were significantly associated with a reduced risk of incidence or recurrence of Af. Aspirin was not recommended to prevent stroke in patients with Af as monotherapy [22] and might result in the association of aspirin and lower Af rate.
Unexpectedly, the results revealed that the trastuzumab group had a lower risk of developing Af than the control arm. We believed the result was valid since a similar population was analyzed for HF. The result was consistent with the previous studies that trastuzumab was associated with increased heart failure incidence. How trastuzumab affects the heart is still under investigation. HER2 might be the key to preventing dilated cardiomyopathy [23]. Unlike anthracyclines, trastuzumab does not cause cardiomyocyte loss [24] but may lead to HER2 receptors decrease.. Guarneri et al. reported that cardiomyocytes in patients who received trastuzumab had type II cardiotoxicity, which was, only electronic microscopic changes were observed [25]. Grazette et al. showed mitochondrial dysfunction in rat myocytes after trastuzumab use [26]. Both heterodimerization and heterodimerization with other HERs activate the intracellular domain and promote the RAS-MAPK and phosphatidylinositol 3’-kinase-AKT-mammalian target of rapamycin (mTOR) pathways, which are essential in mitochondrial function, sarcoplasmic reticulum calcium regulation [27–29]. The interference of the above pathways results in less limited cell proliferation and cessation of cell death[30]. In cardiomyocytes, trastuzumab leads to a decline of pro-survival signaling, especially the NRG/ErbB pathway, and further slows down the cellular defensive and energy-generating systems, and in the end, prevents the cardiomyocytes from functioning properly [30].
Recently a meta-analysis from 15 studies involving 8,124 patients has reported an overall 1.2% incidence of Af in BC patients receiving trastuzumab. The incidence seemed not influenced by the formulation of trastuzumab, the additional use of neoplastic agents, anthracycline exposure status, or concurrent radiotherapy [31]. Unexpectedly, we found that trastuzumab treatment patients had a reduced risk of getting Af. Through a chick embryo model, Yoshiyama et al. demonstrated a significant decrease in heart rate and arrhythmias, mainly Af, after high doses of trastuzumab [32]. In 2020, Chang et al. demonstrated in a male rabbit model that trastuzumab affects the electroactivity of the sinoatrial node (SAN) and pulmonary veins (PVs) through the PI3K pathway. At higher concentrations (10 and 30 uM), trastuzumab produced PV conduction block. While in lower concentration, trastuzumab might lead to Af occurrence. The attenuation of the electroactivity further led to a slower heart rate and the PVs might be the trigger in drug-induced atrial fibrillation [33].
Also, trastuzumab was found to affect the transmembrane calcium-mediated currents through the sodium/calcium exchange, subcellular sarcoplasmic reticulum calcium fluxes, and other calcium channels [34, 35]. Another report showed that trastuzumab at a concentration of 10 mg/L did not trigger more arrhythmias in rate hearts [36]. The trastuzumab block of HER2-mediated anti-oxidative pathways might further accumulate reactive oxygen species (ROS) [12]. The effects of increased ROS and oxidative stress on the different ion channels of cardiac myocytes are heterogeneous and varied in cell depolarization and repolarization [37]. It might partially explain how trastuzumab affects heart rhythm and Af.
Unspecified arrhythmias were reported in the relatively little literature regarding arrhythmias in trastuzumab use in HER2 high expression BC patients. As early as 2001, Behr et al. tried to approach trastuzumab-induced cardiotoxicity by applying 111In-DTPA-trastuzumab to 20 patients. Only one developed arrhythmia among the seven patients showing isotope uptake in myocardium on image [38]. In one small cohort including 27 metastatic BC patients, trastuzumab was believed to cause arrhythmias in 6 patients, one of which developed sinus bradycardia and others were unspecified [39]. In a larger study including 68113 BC patients, the chemotherapy group was shown to have higher Af incidence, while in the subgroup analysis, there were no significant differences regarding trastuzumabuse [40]. Other abnormal electrocardiogram findings, including branch bundle blocks, QT prolonged, ventricular bigeminy, and non-sustained ventricular tachycardia in HER2 expressed BC patients, were shown as case reports [41–43]. A small cohort of 20 metastatic HER2 BC patients reported no arrhythmias after trastuzumab infusion [44]. The evidence regarding arrhythmias associated with trastuzumab in BC patients was still vague, controversial, and not well-established. According to the above pre-clinical and clinical evidence, we suggested that the AE of arrhythmias of trastuzumab might refer to ventricular tachycardia and ventricular fibrillation rather than Af.
BC is the most common female cancer in Taiwan[48]. Chang et al. conducted a small open-labeled, single-centered, observational study in Northern Taiwan, in which seventy-three female patients with HER2 overexpression BC were included. Through the 5 year study period, fourteen (19.2%) patients had marked left ventricular dysfunction, and 7 patients developed new mitral regurgitation [45]. Chien et al. conducted a nationwide cohort study last from January 2006 to December 2013 in Taiwanese female patients with HER2 positive BC, aimed to identify if trastuzumab increases the risk of HF and cardiomegaly. In their study, 23006 women with a median age of 51.0 years were included. The 1-year hazard ratio of trastuzumab was 1.86 (95% CI 1.08–3.19) [46]. The mean age of patients diagnosed with HER2 BC was 52.46 years, similar to the data adopted by Chien. It was also compatible with previous studies that Asian women with BC were 10-year younger than Western ones. Our research showed that trastuzumab was a risk factor for getting HF. However, compared with the above report, we extracted more recent years and longer duration from the NHIRD and investigated the effect of trastuzumab in Af.
To sum up, our study was unique in the following aspects: first, we evaluated the cardiac AE of specifically Af rather than HF only or unspecified arrhythmias. Second, the impacts of representative medication in treating the pre-mentioned comorbidities, which were seldom mentioned in previous reports, were analyzed; third, a national-wide health system support database research with large patient numbers and long studied periods was conducted.
There were some limitations to our study. First, the misclassification of false-positive diagnosis of administrative codes for HF and Af might occur as limitations of population-based, including administrative data. Second, more frequent cardiac screening in patients on trastuzumab led to detection bias. Third, education in patients receiving trastuzumab might induce health behavior bias. No information regarding the value of LVEF and the severity of the cardiac outcomes were available using the NHIRD. Also, this study only detected patients who used NHI-supported trastuzumab; those who paid for the regimen themselves were omitted. EF is not a routine examination in real-world clinical practice in non-HER2 BC patients. Better comparisons across exposure groups could be achieved if the evaluation of heart function was performed in non-HER2 patients.
Furthermore, although we attempted to match both groups with age, comorbidities including DM, HTN, hyperlipidemia, CKD, medication usage including aspirin, NSAID, statins, metformin, beta-blockers, residual confounding still existed. Our study reflects the real-world clinical practice of trastuzumab treatment in Taiwan under the NHI coverage. Our data showed that around 12% of all BC patients had received trastuzumab. HER2-positivity accounts for 20–25% of all BC patients. Patients with HER2-positive node-negative early BC may have never received trastuzumab according to the reimbursement criteria. Nevertheless, these patients would still receive cardiotoxic chemotherapy (such as anthracyclines) and thus may have risks of Af. Moreover, other well-characterized risk factors for Af, such as hypertension and diabetes, and concurrent medications may have masked the less prominent role of trastuzumab in developing Af.
Indeed, clinical reports have been focused on the more prominent cardiomyopathic role of trastuzumab; the incidence of Af (1.2%) is lower than that of HF (1–4%) [14, 15]. Nevertheless, given the possible arrhythmogenic effects of trastuzumab [12], further studies are necessary to characterize the potential electrophysiological side effects.