Cs. Ng et al., A CT sign of chronic pulmonary arterial hypertension: The ratio of main pulmonary artery to aortic diameter, J THOR IMAG, 14(4), 1999, pp. 270-278
The aim of this study was to determine whether the ratio of the diameters o
f the main pulmonary artery and of the ascending aorta (rPA), as assessed o
n computed tomography (CT), is predictive of pulmonary arterial hypertensio
n (PAH). We undertook a retrospective review of 50 patients with a wide ran
ge of pulmonary and cardiovascular diseases, who had undergone bath chest C
T and pulmonary arterial pressure measurements at right heart catheterizati
on. Two independent observers made measurements of the diameter of the main
pulmonary artery and of the ascending aorta on a single defined CT section
. Body surface area (BSA, n = 48), pulmonary arteriolar resistance (n = 39)
, total lung capacity (n = 40), and aortic pressures (n = 50) were also rec
orded. rPA and pulmonary arterial diameter (dPA) were positively related to
mean pulmonary artery pressure (R-s = 0.74, p < 0.0005 for both analyses).
For patients younger than 50 years of age, mean pulmonary artery pressure
correlated more strongly with rPA than dPA (R-s = 0.77, p < 0.00005, compar
ed with R-s = 0.59, p < 0.005); and vice versa for patients older than 50 y
ears of age (R-s = 0.63, p < 0.005, compared with R-s = 0.75, p < 0.00005).
Using a mean pulmonary artery pressure greater than 20 mm HE as indicative
of PAH and a value of rPA > I, the sensitivity, specificity, and positive
and negative predictive values for determining PAH were 70% (26/37), 92% (1
2/13), 96% (26/27), and 52% (12/23), respectively. On multivariate analysis
, rPA was positively related to mean pulmonary artery pressure (p < 0.0005)
, and negatively related to age (p < 0.0005), but was not related to BSA. B
y contrast, dPA showed some dependency on BSA (p < 0.0005), as well as on m
ean pulmonary arterial pressure. In patients younger than 50 years of age,
we have found a strong correlation between rPA and mean pulmonary artery pr
essure in a heterogeneous study population, and this relationship is indepe
ndent of BSA and sex. The presence of the sign "rPA > 1" is simple in pract
ical CT reading to determine; if this is identified, there is a very high p
robability of pulmonary arterial hypertension, and clinicians should be ale
rted to this possibility.