Accuracy of high-resolution CT in identifying chronic pulmonary thromboembolic disease.

OBJECTIVE

The aims of this study were to determine the reliability of the high-resolution CT (HRCT) appearance of the lung parenchyma in distinguishing patients with chronic pulmonary thromboembolism (CPTE) from those with other pulmonary diseases and to compare HRCT with radionuclide scanning.

SUBJECTS AND METHODS

Sixty-seven patients for whom HRCT scans were available for review were included in the study. Twenty-eight had proven pulmonary arterial hypertension (PAH), 17 cases of which were caused by CPTE, and 39 had other pulmonary abnormalities. Diagnosis based on the HRCT appearance was attempted by two radiologists, who independently evaluated pulmonary parenchyma for a mosaic pattern of variable attenuation, for a measurable disparity in the size of pulmonary vessels, and for the presence of peripheral scars. HRCT findings were compared with radionuclide scan findings and pulmonary angiography findings.

RESULTS

For both readers (readers 1 and 2), sensitivity (94% and 100%, respectively) and specificity (96% and 98%, respectively) were high for distinguishing patients with CPTE from patients with other pulmonary abnormalities, including those with nonthromboembolic PAH. The average ratios of segmental vessel size were 2.2 for patients with CPTE and 1.1 for those with nonthromboembolic diseases. Mosaic attenuation was identified in all patients with CPTE but was also seen in 22% (reader 1) and 14% (reader 2) of patients with no evidence of CPTE. Radionuclide scans revealed a high probability for pulmonary emboli for all but one patient with CPTE but also revealed a high probability for three patients who had no emboli.

CONCLUSION

HRCT findings of disparity in the size of segmental vessels and a mosaic pattern of variable attenuation reliably distinguished patients with CPTE from those with nonthromboembolic PAH and from those with other pulmonary abnormalities. In addition, HRCT was more specific than radionuclide scanning in identifying patients with CPTE.