Prediction the Existence and Severity of Pulmonary Hypertension Based on CT Pulmonary Angiographic Data

Purpose: To use automated volumetric analysis tool based on data from computerized tomographic pulmonary angiography (CTPA) as a screening tool for the existence and severity of pulmonary hypertension (PH).

Methods: Patients who underwent computed tomography pulmonary angiography (CTPA) and echocardiography within 24 hours with evaluation of systolic pulmonary artery pressure (sPAP) between 2009 and November 2015 were included. Using an automated software, the volumes of the four cardiac chambers indexed to body surface index (BSA) were calculated. The diameters of the pulmonary artery and the severity of reflux of contrast material were also assessed.
A prediction model for PH was built by using backward logistic regression.

Results: The final study group included 961 patients, of whom 221 with pulmonary embolism. Patients were randomly assigned into two groups: learning (70% of patients) and validation (30% of patients).
Age, reflux severity grade, all heart chambers volume, ventricular volume ratio, and pulmonary artery diameter were included in the first step of prediction PH (sPAP above 36mmHg) in patients without pulmonary embolism. The logistic model showed good discrimination ability (area under the curve = 0.775, discrimination slop 0.37). After validation the model showed 87% sensitivity, 50% specificity and a negative predictive value of 81%.
Gender, Right atrium and ventricle volume, ventricular volume ratio, and pulmonary artery diameter were included in the second step of predicting severe PH (sPAP above 60mmHg). The logistic model showed good discrimination ability (area under the curve = 0.770). After validation the model showed 91% sensitivity, 53% specificity, and a negative predictive value of 97%.
A different logistic model was built for the prediction of PH in patient with PE.

Conclusion: This study shows the ability to screen the presence of pulmonary hypertension and to estimate its severity based solely on automated volumetric analysis from CTPA data