Predicting pulmonary hypertension with standard computed tomography pulmonary angiography

Abstract

The most common feature of pulmonary hypertension (PH) on computed tomography pulmonary angiography (CTPA) is an increased diameter-ratio of the pulmonary artery to the ascending aorta (PA/AAAX). The aim of this study was to investigate whether combining PA/AAAX measurements with ventricular measurements improves the predictive value of CTPA for precapillary PH. Three predicting models were analysed using baseline CTPA scans of 51 treatment naïve precapillary PH patients and 25 non-PH controls: model 1: PA/AAAX only; model 2: PA/AAAX combined with the ratio of the right ventricular and left ventricular diameter measured on the axial view (RV/LVAX); model 3: PA/AAAX combined with the RV/LV-ratio measured on a four chamber view (RV/LV4CH). Prediction models were compared using multivariable binary logistic regression, ROC analyses and decision curve analyses (DCA). Multivariable binary logistic regression showed an improvement of the predictive value of model 2 (−2LL = 26.48) and 3 (−2LL = 21.03) compared to model 1 (−2LL = 21.03). ROC analyses showed significantly higher AUCs of model 2 and 3 compared to model 1 (p = 0.011 and p = 0.007, respectively). DCA showed an increased clinical benefit of model 2 and 3 compared to model 1. The predictive value of model 2 and 3 were almost equal. We found an optimal cut-off value for the RV/LV-ratio for predicting precapillary PH of RV/LV ≥ 1.20. The predictive value of CTPA for precapillary PH improves when ventricular and pulmonary artery measurements are combined. A PA/AAAX ≥ 1 or a RV/LVAX ≥ 1.20 needs further diagnostic evaluation to rule out or confirm the diagnosis.