Magnetic Resonance Imaging of Pulmonary Embolism: Diagnostic Accuracy of Unenhanced MR and Influence in Mortality Rates.

OBJECTIVES

We evaluated the diagnostic value for pulmonary embolism (PE) of the True fast imaging with steady-state precession (TrueFISP) MRI, a method that allows the visualization of pulmonary vasculature without breath holding or intravenous contrast.

METHODS

This is a prospective investigation including 93 patients with suspected PE. All patients underwent TrueFISP MRI after undergoing CT pulmonary angiography (CTPA). Two independent readers evaluated each MR study, and consensus was obtained. CTPA results were analysed by a third independent reviewer and these results served as the reference standard. A fourth radiologist was responsible for evaluating if lesions found on MRI for both analysis were the same and if these were the correspondent lesions on the CTPA. Sensitivity, specificity, predictive values and accuracy were calculated. Evidence for death from PE within the 1-year follow-up was also assessed.

RESULTS

Two patients could not undergo the real-time MRI and were excluded from the study. PE prevalence was 22%. During the 1-year follow-up period, eight patients died, whereas PE was responsible for 12.5% of cases. Between patients who developed PE, only 5% died due to this condition. There were no differences between MR and CT embolism detection in these subjects. MR sequences had a sensitivity of 85%, specificity was 98.6% and accuracy was 95.6%. Agreement between readers was high (κ= 0.87).

CONCLUSIONS

Compared with contrast-enhanced CT, unenhanced MR sequences demonstrate good accuracy and no differences in the mortality rates in 1 year were detected.