Electrocardiogram Findings in Long-term Pediatric Cancer Survivors Exposed to Anthracyclines

Background: Survival outcomes for pediatric cancers have improved in recent decades, now exceeding 80% for all cancers. Anthracycline-induced cardiomyopathy is an established late-effect of childhood cancer treatment, but the long-term effects of anthracyclines on cardiac conduction and rhythm properties have not been thoroughly investigated. The Children’s Oncology Group Long-Term Follow-Up Guidelines recommend obtaining a baseline electrocardiogram (ECG) at entry into survivorship for children exposed to anthracyclines, and repeating testing as clinically indicated. The objectives of this study are to describe the nature and prevalence of ECG changes throughout treatment and survivorship in pediatric long-term cancer survivors exposed to anthracyclines, and to analyze the association of ECG findings with cardiomyopathy.

Methods: A retrospective study was conducted using an institutional database of pediatric cancer survivors treated with anthracyclines who were seen in the Texas Children’s Long Term Follow Up Clinic between 1979 and 2019. All ECGs were interpreted by a pediatric electrophysiologist. ECG abnormalities were grouped into five categories: ectopy (premature atrial or ventricular contractions), conduction delay (bundle branch block; 1st degree, 2nd degree, or advanced atrioventricular block), hypertrophy (left, right, or biventricular), atrial enlargement, or prolonged corrected QT interval (QTc). Cardiomyopathy was defined as either being referred to cardiology for this indication or having at least two echocardiograms showing shortening fraction ≤28% and/or ejection fraction ≤50%. The association between cardiomyopathy diagnosis and specific ECG changes was analyzed using chi-squared or Fisher’s exact test as appropriate, with calculated odds ratios (OR) and significance defined as p <0.05.

Results: A total of 1246 patients with ECG data were identified. The mean age at cancer diagnosis was 6.14.9 years, and the mean total cumulative equivalent anthracycline exposure was 168108.3 mg/m2 (range 1-660). 6787 ECGs were obtained after starting chemotherapy, and 918 were abnormal (13.5%). There were 862 patients with all normal ECGs (69.2%) and 384 patients with at least one abnormal ECG (30.8%). Specifically, 50 patients had at least one ECG with ectopy (4%), 137 with conduction delay (11%), 75 with prolonged QTc (6%), 170 with hypertrophy (13.6%), and 69 with atrial enlargement (5.5%). There were 74 patients who met criteria for cardiomyopathy (6%). There was a statistically significant association between developing cardiomyopathy and having an abnormal ECG with conduction delay (OR 3.3; p<0.001), prolonged QTc (OR 7.5; p<0.001), hypertrophy (OR 3.4; p<0.001), or atrial enlargement (OR 7.1; p<0.001). There was no association between cardiomyopathy and ectopy (p=0.115). The negative predictive value of having all normal ECGs was 97%.

Conclusion: ECGs are recommended for long-term follow-up of cancer patients exposed to anthracyclines. Although most ECGs are normal in this population, abnormalities on ECG are associated with the development of cardiomyopathy. These results suggest ECG screening may be useful for risk-stratifying survivors of pediatric cancer for anthracycline-induced cardiomyopathy.