Novel Reference Equations for Pulmonary Artery Size and Pulsatility Using Echocardiography and Their Diagnostic Value in Pulmonary Hypertension

Abstract

BackgroundAccording to the most recent pulmonary hypertension (PH) guidelines, a main pulmonary artery (MPA) diameter>25 mm on transthoracic echocardiography (TTE) supports the diagnosis of PH. However, the size of the pulmonary artery(PA) may vary according to body size, age, and cardiac phases. Research questions1)What are the reference limits for PA size on TTE, considering differences in body size, sex, and age? 2)What is the diagnostic value of PA size for classifying pulmonary hypertension? 3)How does the selection of different reference groups (healthy volunteers versus patients referred for right heart catheterization (RHC)) influence the diagnostic odds ratio (DOR)? Study Design and MethodsThe study included a reference cohort of 248 healthy individuals as controls, 693 PH patients proven by RHC, and 156 non-PH patients proven by RHC. In the PH cohort, 300 had group-1 PH, 207 had group-2 PH, and 186 with group-3 PH. MPA and right PA(RPA) diameters and areas were measured in the upper sternal short-axis and the suprasternal notch views. Reference limits (5th-95th percentile) were based on absolute values and height-indexed measures. Quantile regression analysis was used to derive median and 95th quantile reference equations for the PA measures. DORs and probability diagnostic plots for PH were then determined using healthy controls and non-PH cohorts. ResultsThe 95th percentile for indexed MPA diameter was 15mm/m in diastole and 19mm/m in systole in both sexes. Quantile regression analysis revealed a weak age effect (pseudo R2 of 0.08 to 0.10 for MPA diameters). Among measures, the MPA size in diastole had the highest DOR, 156.2(68.3-357.5), for detection of group-1 PH. Similarly, the DORs were also high for group-2 and 3 PH when compared to controls but significantly lower compared to non-PH cohort. InterpretationThe study presents novel reference limits for MPA based on height indexing and quantile regression.