This meta-analysis focused on the association between anthracyclines and the risk of arrhythmia. Our results identified a consistent association between anthracycline use and arrhythmia, with a statistically significant 90% increase in the risk of arrhythmia and 114% increase in the risk of supraventricular arrhythmia. According to network meta-analyses, epirubicin treatment was associated with the highest probability of being associated with the risk of arrhythmia compared to other anthracycline and non-anthracycline treatments. Overall, these findings support a link between arrhythmia and anthracyclines and provide a relatively rational explanation regarding why anthracyclines significantly increase the risks of cardiac events and mortality in addition to heart failure.
Anthracyclines are widely used to treat adult and pediatric cancers. Despite their therapeutic efficacy, anthracyclines are associated with both acute-onset and late-onset cardiac toxicities. Meta-analyses have reported an overt cardiotoxicity incidence of 6.3% and a subclinical cardiotoxicity incidence of 17.9%. Among these AICs, the induction of arrhythmia may not be as serious as cardiomyopathy or heart failure. However, anthracyclines are associated with electrocardiographic alterations. In 1981, a patient with non small-cell lung cancer developed acute-onset AF with a rapid ventricular response during the first administration of Adriamycin; thereafter, physicians became aware of possible life-threatening arrhythmias during or soon after Adriamycin administration. However, anthracycline-induced arrhythmia did not gain much attention until Bender first described the prolongation of the QT interval with anthracycline therapy in 1984. Two years later, epirubicin was reported to cause transient cardiac arrhythmias and electrocardiogram alterations. The impact of anticancer drugs on cardiac repolarization (i.e., QT prolongation) and the potential risk of life-threatening arrhythmia torsades de pointes, which are common challenges in the field, have been gradually noticed and reduced.
In our study, supraventricular arrhythmia was the most common adverse effect of anthracyclines. Previous studies have reported that AF, one of the representatives of supraventricular arrhythmia, was detected using Holter monitoring for 10.3% of patients treated with doxorubicin. Studies have revealed that AF onset always occurred before the development of heart failure in patients, and that the incidence of AF increases with anthracycline doses; therefore, new-onset AF might be a sensitive indicator of AICs. An experimental study of sheep also reported that atrial remodeling favoring AF development was observed in doxorubicin-induced heart failure models. Experiments involving rats also confirmed that anthracycline exposure resulted in an 86% incidence of ventricular tachycardia or ventricular fibrillation.
Although QT prolongation is not a perfect marker of the arrhythmia risk, it has become a primary safety metric for oncologists. The variability in the QT interval, a probable marker of an arrhythmogenic substrate, reflects an increase in regional differences in ventricular repolarization and has been associated with life-threatening ventricular arrhythmias and sudden death. Our results also suggest that QT prolongation is common in anthracycline-induced arrhythmia, similar to the results of previous reports.
Left ventricular dysfunction caused by AIC has been believed to be irreversible; however, early initiation of standard medical treatment for heart failure may lead to left ventricular functional recovery with AIC. In this meta-analysis, we found that anthracycline use may increase arrhythmia-related risks. Close monitoring of ECG changes would clarify the occurrence of acute-onset and delayed-onset cardiac toxicity caused by anthracyclines, which may have some significance in preventing irreversible AIC and the occurrence of serious adverse cardiac events. Further studies of life-long cardiac follow-up for patients undergoing anthracycline-based chemotherapy are necessary.
Our study had several strengths. First, the results are generalizable because of the large number of participants involved in the arrhythmia analyses and the use of raw data. Subgroup, sensitive, and regression analyses were performed to support these findings. Furthermore, to our knowledge, this is the first study to conduct indirect comparisons using a network meta-analysis to assess differences in the arrythmia risks associated with individual anthracyclines.
Our study had several potential limitations as well. First, our study was not designed to explore the interaction between comorbidity and the arrhythmic adverse effects of anthracyclines; therefore, the observed risk does not reflect the risk for special populations such as patients with conduction disorders and patients with prior cardiovascular events. Second, many of these studies did not provide information regarding important lifestyle factors that influence the arrhythmic risk, patient compliance, and chemotherapy regimen. Third, a misclassification bias may have occurred because arrhythmia screening is not sufficiently intensive to detect several paroxysmal arrhythmias such as paroxysmal AF and infrequent premature contractions. Fourth, the meta-analysis was highly heterogeneous among single-arm studies. The high heterogeneity can be attributed to different indications for the types of carcinomas and anthracyclines, different dosages of anthracyclines, different treatment and follow-up durations, different coexisting conditions, different countries, different ages, and different sexes. We used a random-effects model for our calculations to account for the possibility of study-dependent variations in effects and conducted a sensitivity analysis. Finally, this analysis was not sufficiently strong to provide the basis for any changes in practice. These findings need to be formally assessed by larger, prospective, real-world studies and clinical trials with strict ECG monitoring.