Purpose: To retrospectively quantify right ventricular dysfunction (RVD) and the pulmonary artery obstruction index at helical computed tomography (CT) on the basis of various criteria proposed in the literature and to assess the predictive value of these CT parameters for mortality within 3 months after the initial diagnosis of pulmonary embolism (PE).
Materials and methods: Institutional review board approval was obtained, and informed consent was not required for retrospective study. In 120 consecutive patients (55 men, 65 women; mean age +/- standard deviation, 59 years +/- 18) with proved PE, two readers assessed the extent of RVD by quantifying the ratio of the right ventricle to left ventricle short-axis diameters (RV/LV) and the pulmonary artery to ascending aorta diameters, the shape of the interventricular septum, and the extent of obstruction to the pulmonary artery circulation on helical CT images, which were blinded for clinical outcome in consensus reading. Regression analysis was used to correlate these parameters with patient outcome.
Results: CT signs of RVD (RV/LV ratio, >1.0) were seen in 69 patients (57.5%). During follow-up, seven patients died of PE. Both the RV/LV ratio and the obstruction index were shown to be significant risk factors for mortality within 3 months (P = .04 and .01, respectively). No such relationship was found for the ratio of the pulmonary artery to ascending aorta diameters (P = .66) or for the shape of the interventricular septum (P = .20). The positive predictive value for PE-related mortality with an RV/LV ratio greater than 1.0 was 10.1% (95% confidence interval [CI]: 2.9%, 17.4%). The negative predictive value for an uneventful outcome with an RV/LV ratio of 1.0 or less was 100% (95% CI: 94.3%, 100%). There was a 11.2-fold increased risk of dying of PE for patients with an obstruction index of 40% or higher (95% CI: 1.3, 93.6).
Conclusion: Markers of RVD and pulmonary vascular obstruction, assessed with helical CT at baseline, help predict mortality during follow-up.
Copyright RSNA, 2005.