Electrocardiographic Abnormalities and Their Association with Outcomes in Randomized Clinical Trials of Pulmonary Arterial Hypertension

Rationale: Pulmonary arterial hypertension (PAH) is a progressive disease with manifestations including right atrial enlargement, right ventricular dysfunction, dilation, and hypertrophy. Electrocardiography (ECG) is a noninvasive, inexpensive test that is routinely performed in clinical settings. Prior studies have described separate abnormal findings in the electrocardiograms of patients with PAH. However, the role of composite ECG findings reflective of right heart disease (RHD) for risk stratification, clinical trial enrichment, and management of patients with PAH has not been explored. Objectives: To describe a pattern of RHD on ECG in patients with PAH and to investigate the association of this pattern with clinical measures of disease severity and outcomes. Methods: We harmonized individual participant data from 18 phase III randomized clinical trials of therapies for PAH (1998-2013) submitted to the U.S. Food and Drug Administration. RHD was defined as the presence of right ventricular hypertrophy, right axis deviation, right atrial enlargement, or right bundle branch block on ECG. Random effects linear regression, multilevel ordinal regression (cumulative link model), and Cox proportional hazards models were used to assess the association of RHD by ECG with 6-minute walk distance (6MWD), World Health Organization (WHO) functional class, and clinical worsening after a priori adjustment for age, sex, body mass index, and PAH etiology. Effect modification of treatment and ECG abnormalities was assessed by including an interaction term. Results: A total of 4,439 patients had baseline ECG, and 68% of patients had evidence of RHD. RHD on ECG was associated with higher pulmonary vascular resistance (P < 0.001) and higher mean pulmonary artery pressures (P < 0.001). Patients with RHD on ECG had 10 meters shorter 6MWD (P = 0.005) and worse WHO functional class (P < 0.001) at baseline. RHD on baseline ECG was associated with increased risk of clinical worsening (hazard ratio, 1.42; 95% confidence interval; 1.21, 1.67; P < 0.001). Patients with RHD had greater treatment effect in terms of 6MWD, WHO functional class, and time to clinical worsening than those without (P for interaction = 0.03, 0.001, and 0.03, respectively). Conclusions: RHD by ECG may be associated with worse outcomes and potentially greater treatment effect. Electrocardiograms could be an inexpensive, widely available noninvasive method to enrich clinical trial populations in PAH.