Diagnostic performance of the main pulmonary artery-to-ascending aorta diameter ratio for pulmonary hypertension defined by mean pulmonary arterial pressure in children with left-to-right shunt congenital heart disease.

BACKGROUND: The computed tomography (CT)-derived main pulmonary artery-to-ascending aorta diameter ratio (MPA/AA) may identify pulmonary hypertension (PH). In left-to-right (L-R) shunt congenital heart disease (CHD), pulmonary overcirculation may enlarge the pulmonary artery without elevated mean pulmonary arterial pressure (mPAP), complicating threshold interpretation. We evaluated MPA/AA for mPAP-defined PH using right heart catheterization (RHC) and derived upper reference limits in RHC-negative children.

METHODS: Among 231 eligible children who underwent CT and RHC within 3 months, the primary diagnostic analysis was performed in 164 children with L-R shunt CHD. A heterogeneous non-L-R cardiovascular cohort (n=67) was used only to assess model transportability. Disease-context upper reference limits were estimated in 109 children without mPAP-defined PH on RHC.

RESULTS: In the primary cohort, median MPA/AA was higher in children with mPAP-defined PH than in those without it (1.56 vs. 1.22; P<0.001) and correlated with mPAP (Spearman r=0.69; P<0.001). The MPA/AA model had an area under the curve of 0.94 [95% confidence interval (CI): 0.90-0.97]. At a cutoff of 1.40, the model presented with 84.5% sensitivity and 92.6% specificity, whereas at a cutoff of 1.44, it presented with 96.3% specificity and 73.6% sensitivity. Bootstrap validation showed little overfitting. The maximum spline-derived 95th-percentile upper reference limit in RHC-negative children was 1.39. Discrimination was retained in the non-L-R cohort, although calibration was weaker.

CONCLUSIONS: In children with L-R shunt CHD, CT-derived MPA/AA adds useful information for identifying mPAP-defined PH on clinically indicated CT, but it does not replace RHC or justify CT screening.