OBJECTIVES: To investigate whether the pulmonary artery (PA)-to-ascending aorta (Ao) ratio is associated with outcome in unselected patients referred for cardiac magnetic resonance (MR) imaging.
METHODS: This study prospectively enrolled 650 consecutive patients (47.2% women; mean age, 56.1 years ± 17.7 [standard deviation]). Diameters of PA and Ao were measured in axial black blood images. On the basis of previous results, a PA-to-Ao ratio of 1.0 or greater was chosen as the cutoff for further analysis. Univariable and multivariable Cox regression models were used to investigate the primary end point, which was defined as a composite of cardiovascular hospitalization and death.
RESULTS: A PA-to-Ao ratio of 1.0 or greater was present in 131 (20.2%) patients. Patients with a PA-to-Ao ratio of 1.0 or greater were predominantly women (P = .010); more frequently presented with atrial fibrillation (P < .001), diabetes (P < .001), and impaired renal function (P < .001); and had higher N-terminal pro-B-type natriuretic peptide (NT-proBNP) levels (P < .001), larger left (P = .023) and right ventricles (RV; P = .002), and worse RV function (P < .001). Patients were followed for 17.8 months ± 12.9, during which 110 patients (16.9%) reached the primary end point. By Kaplan-Meier analysis, event-free survival was significantly worse in patients with a PA-to-Ao ratio of 1.0 or greater (log-rank test, P < .001). A PA-to-Ao ratio of 1.0 or greater was independently associated with outcome by multivariable Cox regression analysis, in addition to age, NT-proBNP serum levels, and RV size.
CONCLUSIONS: A PA-to-Ao ratio of 1.0 or greater identified patients at risk, most likely because of elevated PA pressures. On the basis of these results, the PA-to-Ao ratio should routinely be reported at cardiac MR imaging.