Baseline Electrocardiographic Abnormalities in Pre- Treatment Cancer Compared With Non-Cancer Patients: A Propensity Score Analysis

Background: Most studies have compared post-treatment electrocardiogram (ECG) abnormalities in cancer patients to the general population. To assess baseline cardiovascular (CV) risk, we compared pre-treatment ECG abnormalities in cancer patients with a non-cancer surgical population. Methods: We conducted a combined prospective (n = 30) and retrospective (n = 229) cohort study of patients aged 18 80 years with diagnosis of hematologic or solid malignancy, compared with 267 pre-surgical, non-cancer, age-and sex-matched controls. Computerized ECG interpretations were obtained, and one-third of the ECGs underwent blinded interpretation by a board-certified cardiologist (agreement r = 0.94). We performed contingency table analyses using likelihood ratio Chi-square statistics, with calculated odds ratios. Data were analyzed after propensity score matching. Results: The mean age of cases was 60.97 +/- 13.86; and 59.44 +/- 11.83 years for controls. Pre-treatment cancer patients had higher likelihood of abnormal ECG (odds ratio (OR): 1.55; 95% confidence interval (CI): 1.05 to 2.30), and more ECG abnormalities (chi 2 = 4.0502; P = 0.04) compared with non-cancer patients. ECG abnormalities were higher in black compared to non-black patients (P = 0.001). In addition, baseline ECGs among cancer patients prior to cancer therapy demonstrated less QT prolongation and intra-ventricular conduction defect (P = 0.04); but showed more arrhythmias (P < 0.01) and atrial fibrillation (AF) (P =0.01) compared with the general patient population. Conclusions: Based on these findings, we recommend that all cancer patients receive an ECG, a low-cost and widely available tool, as part of their CV baseline screening, prior to cancer treatment.