Diagnosis of Pulmonary Arterial Hypertension in Children by Using Cardiac Computed Tomography

Abstract

Objective: To establish diagnostic criteria for pulmonary arterial hypertension (PAH) in children by using parameters obtained through noninvasive cardiac computed tomography (CCT).

Materials and methods: We retrospectively measured parameters from CCT images of children from a single institution in a multiple stepwise process. A total of 208 children with mean age of 10.5 years (range: 4 days-18.9 years) were assessed. The variables were classified into three groups: the great arteries; the ventricular walls; and the bilateral ventricular cavities. The relationship between the parameters obtained from the CCT images and mean pulmonary arterial pressure (mPAP) was tested and adjusted by the children's body size. Reference curves for the pulmonary trunk diameter (PTD) and ratio of diameter of pulmonary trunk to ascending aorta (rPTAo) of children with CCT images of normal hearts, adjusted for height, were plotted. Threshold lines were established on the reference curves.

Results: PTD and rPTAo on the CCT images were significantly positively correlated with mPAP (r > 0.85, p < 0.01). Height was the body size parameter most correlated with PTD (r = 0.91, p < 0.01) and rPTAo (r = -0.69, p < 0.01). On the basis of the threshold lines on the reference curves, PTD and rPTAo both showed 88.9% sensitivity for PAH diagnosis, with negative predictive values of 93.3% and 92.9%, respectively.

Conclusion: PTD and rPTAo measured from CCT images were significantly correlated with mPAP in children. Reference curves and the formula of PTD and rPTAo adjusted for height could be practical for diagnosing PAH in children.

Keywords: Children; Computed tomography; Pulmonary arterial hypertension.

Fig. 1. Parameters measured on end-diastole phase of contrast-enhanced CCT images to diagnose pulmonary arterial hypertension.

A, B, D. Images were obtained in transverse planes. C. Image was obtained in tilted oblique axial plane to reveal whole pulmonary trunk from pulmonary annulus to bifurcation. E, F. Images were at middle ventricular level with image (E) on cardiac short-axis plane and image (F) on cardiac four-chamber plane. Twenty measurements were performed and annotated 1–20, and all of these measurements could also be easily performed on picture archiving and communication system. 1 = maximal diameter of left pulmonary artery before branching, 2 = maximal diameter of right pulmonary artery before branching, 3 = maximal diameter of middle pulmonary trunk before bifurcation, 4 and 5 = perpendicular diameters of ascending aorta on same image that measured pulmonary trunk, 6 and 7 = perpendicular diameters of descending aorta at level through diaphragm, 8 and 9 = perpendicular diameters of inferior vena cava on same image that measured descending aorta, 10–12 = middle ventricular myocardial thickness of RV, septum, and LV respectively, 13 and 14, 15 and 16 = perpendicular height and width intersects at midpoints of RV and LV, respectively, 17 and 18, 19 and 20 = perpendicular length and width intersects at midpoints of RV and LV respectively, 21 and 23, 22 and 24 = area of LV and RV respectively. CCT = cardiac computed tomography, LV = left ventricle, RV = right ventricle

Fig. 2. Relationship of PTD (A) and rPTAo (B) to height in children determined using CCT images of normal hearts.

Dashes represent best-fitting linear regression of mean of study subjects from our study step II. Solid lines indicate thresholds of 25-mm Hg mean pulmonary arterial pressure calculated from subjects from our study step III. Subjects from steps II and III were not same. PTD = pulmonary trunk diameter, rPTAo = PTD-to-ascending aorta diameter ratio