Résumé
To evaluate the accuracy of spiral CT angiography in the diagnosis of central pulmonary embolism (PE).PURPOSETo evaluate the accuracy of spiral CT angiography in the diagnosis of central pulmonary embolism (PE).A retrospective study was undertaken in a population of 90 patients who were evaluated with spiral CT angiography (acquisition: 12 to 24 s with or without strict apnea; injection of 90 cm3 of 12, 20 or 30% contrast material at a rate of 4 to 7 cm3, selective pulmonary angiography of each lung (n = 55) and/or ventilation-perfusion (VP) scanning (n = 35).MATERIAL AND METHODSA retrospective study was undertaken in a population of 90 patients who were evaluated with spiral CT angiography (acquisition: 12 to 24 s with or without strict apnea; injection of 90 cm3 of 12, 20 or 30% contrast material at a rate of 4 to 7 cm3, selective pulmonary angiography of each lung (n = 55) and/or ventilation-perfusion (VP) scanning (n = 35).Among the 55 patients evaluated with both spiral CT and angiography, central pulmonary embolism was excluded in 19 patients (34%), assessed in 29 patients (53%) whereas CT examination was considered as inconclusive in 7 patients (13%) due to interpretive difficulties at the level of obliquely oriented arteries and/or presence of hilar lymph nodes (sensitivity: 90.5%; specificity: 82.6%). With spiral CT, the finding of 95 central emboli (3 main, 61 lobar and 31 segmental) corresponded exactly to the angiographic findings; spiral CT enabled direct visualization of intraluminal filling defects (n = 95) whereas the angiographic recognition of PE was based on direct (n = 57) and indirect (n = 38) signs. Spiral CT angiography was more sensitive and specific than VP scanning.RESULTSAmong the 55 patients evaluated with both spiral CT and angiography, central pulmonary embolism was excluded in 19 patients (34%), assessed in 29 patients (53%) whereas CT examination was considered as inconclusive in 7 patients (13%) due to interpretive difficulties at the level of obliquely oriented arteries and/or presence of hilar lymph nodes (sensitivity: 90.5%; specificity: 82.6%). With spiral CT, the finding of 95 central emboli (3 main, 61 lobar and 31 segmental) corresponded exactly to the angiographic findings; spiral CT enabled direct visualization of intraluminal filling defects (n = 95) whereas the angiographic recognition of PE was based on direct (n = 57) and indirect (n = 38) signs. Spiral CT angiography was more sensitive and specific than VP scanning.Spiral CT angiography appears as a sensitive and specific noninvasive method for the diagnosis of central PE.CONCLUSIONSpiral CT angiography appears as a sensitive and specific noninvasive method for the diagnosis of central PE.