RESUMO
PURPOSE: The purpose was to compare first-pass and delayed-phase thoracic computed tomography (CT) venography for the evaluation of suspected central thoracic venous pathology. MATERIAL AND METHODS: CT images and medical records of all patients who underwent thoracic CT venography over a 5-year period were retrospectively reviewed. Both first-pass (18s) and delayed-phase (60s) venous images were obtained in all patients. The images were reviewed in random order by three readers using a semiquantitative visual grading scheme for image quality, including artifacts and the uniformity of venous enhancement. In addition, enhancement was quantitatively evaluated. The presence and type of venous pathology, and overall diagnostic confidence were recorded and compared. Reference verification was performed when available. RESULTS: Eighteen patients formed the study group, mean age 49.5years and 28% male. Dual-arm injection was successful in 72% of exams. All readers reported more streak artifacts on first-pass imaging than delayed imaging (72%-94% vs. 27%-44%, respectively; P<.05). First-pass imaging had significantly higher measured enhancement across all central venous segments than delayed imaging (mean HU range 212-906 HU vs. 173-414 HU; P<.05) but also had significantly more heterogeneous enhancement (mean S.D. range 75-1058 HU vs. 67-378 HU; P<.05). For overall diagnosis, reader agreement, accuracy, and confidence levels were higher for delayed-phase images (P<.05). CONCLUSION: Indirect thoracic CT venography using delayed-phase imaging alone may be sufficient for evaluating clinically suspected central venous abnormality.