A Mixed Treatment Comparison Meta-Analysis of Pharmacotherapeutic Monotherapy and Placebo for Pulmonary Artery Hypertension

    loading  Checking for direct PDF access through Ovid

Abstract

Pulmonary artery hypertension (PAH) is rare, progressive, and difficult to treat. Earlier trials have found that active drug treatments improve hemodynamics, surrogate outcomes, and mortality compared with placebo. Few direct trials of active treatments exist and a mixed treatment comparison network meta-analysis allows comparisons of direct and indirect treatment effects. Randomized controlled trials of patients aged 10 to 80 years with idiopathic or secondary PAH treated with epoprostenol, treprostinil, iloprost, beraprost, bosentan, ambrisentan, macitentan, sildenafil, tadalafil, vardenafil, or riociguat monotherapy were reviewed using the Oxford scoring. Reviewers abstracted data on 6-minute walk distance, change in cardiac index, change in pulmonary artery pressure, change in pulmonary vascular resistance, clinical worsening, and death along with safety endpoints. A Bayesian mixed treatment comparison network meta-analysis using Markov chain Monte Carlo analysis was used to construct network geometry and evaluate treatments. Twenty trials met the criteria and were included, creating a star-shaped evidence network, with some diversity and limited co-occurrence. There were 4328 patients with a mean age of 51 years and idiopathic hypertension in World Health Organization functional class II (34%) or III (60%) PAH. Significant disease progression was observed in every treatment arm, and no individual treatment was consistently better than placebo (direct comparisons) or any other active treatment (indirect comparisons) for 6-minute walk distance, cardiac index, pulmonary artery pressure, pulmonary vascular resistance, clinical worsening, death, or any safety endpoint. No treatment was clearly preferred directly against placebo or indirectly for any endpoint. Providers should select PAH treatments on the basis of individualized treatment responses, preferences, and tolerability.

Related Topics

    loading  Loading Related Articles