Multimodal noninvasive prediction of pulmonary hypertension in IPF

Introduction and Objective: Pulmonary hypertension (PH) complicating idiopathic pulmonary fibrosis (IPF) is challenging to diagnose given inaccuracy of transthoracic echocardiogram (TTE) measurements. However, it has significant prognostic implications and is therefore important to accurately identify. Methods: We conducted a cross-sectional study of patients with IPF who underwent RHC as part of their evaluation. A variety of commonly available noninvasive variables were evaluated for their ability to predict pulmonary arterial pressure in a linear regression model, including the traditionally used right ventricular systolic pressure (RVSP) estimated from TTE. Results: There were 105 eligible patients identified from January 2006 to July 2016. The average age was 62.7 ± 7.7 years, 35 had RHC proven PH and 43% ultimately underwent lung transplantation. A linear model including three terms: RVSP (ANOVA P <.01), the ratio of FVC/DLCO from PFTs (P =.05) and pulmonary artery to aorta diameter ratio from CT (P <.01) was found to predict the mean pulmonary artery pressure more reliably than RVSP alone (R².39 vs.29, P <.05), with a lower rate of incorrect classification of PH status in these individuals (27.6 vs 35.2%, P =.05) and high negative predictive value (87.2%). Conclusion: If used in conjunction with RVSP from TTE, parameters from PFTs and the CT scan more accurately predict the presence or absence of PH than any of the variables in isolation. Using these in concert may allow greater discrimination in deciding which patients to subject to diagnostic right heart catheterization.