Low Attenuation on High Resolution Computed Tomography in Pulmonary Embolism: An Experimental Study in Pigs

PURPOSE: To evaluate the incidence and type of low attenuation seen on high resolution computed tomography (HRCT) performed after artificially induced pulmonary embolism. MATERIALS AND METHODS: Using permanent embolic materials, pulmonary embolism was induced in ten Yorkshire pigs. Pre- and postembolic pulmonary angiography was performed, and HRCT was performed immediately and 1, 3, and 6 weeks after embolization. The incidence and type of low attenuation of all segments, as seen on HRCT, was evaluated. Low attenuation was classified as mottled, lobular, segmental, or peripheral. The pigs were sacrified after 6 weeks and contact radiographs were obtained. RESULTS: Low attenuation developed in eight of ten pigs. Pulmonary angiography revealed arterial occlusion in 15 large and 19 small segmental arteries (34 of 45 segments). In the remaining 11 segments, follow-up HRCT demonstrated areas of low attenuation. This was present in 25 of 35 segments (71%) as seen on HRCT images obtained immediately; in 16 of 41 segments (39 %) on images obtained 1 week after embolization; in 17 of 41 segments (41 %) on those acquired at 3 weeks; and in 25 of 45 segments (56 %) on those acquired at 6 weeks. The overall incidence of low attenuation was 83/166 (50 %). The types of low attenuation were mottled in 32/83 cases, lobular in 13/83, segmental in 13/83, and peripheral in 25/83. In large segmental arterial occlusion, the incidence of low attenuation on HRCT was 100% immediately, 57% at 1 week, 60% at 3 weeks, and 80 % at 6 weeks. In small segmental arterial occlusion, the incidence was 47%, 25 %, 11 %, and 21 % respectively. The overall incidence of low attenuation was 40/55 (73 %) in large segmental arterial occlusion and 18/71 ( 25%) in small segmental arterial occlusion. CONCLUSION: Low attenuation on HRCT is a finding of pulmonary embolism and is more common on HRCT performed immediately after embolization (71%) and in large segmental arterial occlusion (73%). Low attenuation on HRCT is an ancillary finding and may be useful in the diagnosis of pulmonary embolism.

An Experimental Study of the Radiologic-Pathologic Findings of Pulmonary Embolism

PURPOSE: To evaluate the low attenuation of mosaic pattern in pulmonary embolism, as observed on HRCT, and to correlate the findings with the pathologic features of resected lung. MATERIALS AND METHODS: Using permanent embolic materials, pulmonary embolism was induced in eight Yorkshire pigs. Pre-and post-embolic pulmonary angiography was performed and after 6 weeks, the incidence and pattern of parenchymal change in low attenuation (mosaic pattern), as seen on HRCT, was evaluated. The animals were then sacrified and contact radiography of the lung was performed. Thirty-eight segments of pathology were taken from the area in which the presence of embolism had been suggested. Pathologic and HRCT findings were then correlated. RESULTS: On HRCT, low attenuation was seen in 23 of 36 segments (64%) and showed variable patterns : crescent peripheral hyperlucency (61%, n=14), heterogeneous mottled hyperlucency (17%, n=4), lobular hyperlucency (13%, n=3), and homogeneous segmental hyperlucency (9%, n=2). Parenchymal low attenuation was seen on HRCT in 10 of 11 segments (91%) in which large segmental arterial occlusion occurred, and in 3 of 16 segments (19%) in which there was small segmental arterial occlusion. Abnormal pathologic findings were pulmonary congestion, dilatation of pulmonary arteries, interlobular septal thickening, and thrombus formation. Among the 38 pathologic segmental specimens, 29 were from the area in which HRCT findings were positive, and in which pulmonary embolism subsequently occurred. In only four of nine segments (44%) in the area in which HRCT fingings were negative was pulmonary embolism subsequently. CONCLUSION: HRCT findings of pulmonary embolism at six weeks after embolization showed variable patterns of low attenuation, diminished diameter of pulmonary arteries, and normal diameter of bronchi. In cases with large segmental arterial occlusion, the finding of low attenuation was more common ; this may be due to reduced blood flow to the embolic area, in combination with bronchiolar spasm. For the early diagnosis of pulmonary embolism, these findings may be useful.