The differentiation between tumor and bland thromboses is important as the management differs. Retrospectively, we aim to evaluate the utility of FDG PET in detecting and differentiating tumor from bland thromboses and if FDG PET provides additional value to contrast-enhanced CT for tumor thrombus detection.
Twenty-four sites of venous thromboembolism, detected on PET/CT, were retrospectively reviewed. Classification of type of thrombosis was based on histology and radiological follow-up. We evaluated the presence of contrast-enhanced CT findings that were suggestive of tumor thrombosis; sign of invasion, neovascularity, and enhancement. Metabolic activity by means of SUVmax was measured by drawing ROI at the site of thrombosis. Mann-Whitney U test was used to compare the mean SUVmax between thromboses and internal references. We used ROC analysis to identify the optimal cutoff value of SUVmax for detection of tumor thrombosis.
Twenty-four sites of venous thromboembolism were identified in 15 patients. All tumor thromboses demonstrated at least 1 positive sign on contrast-enhanced CT, whereas 33% of bland thromboses had the same finding. The difference between tumor and bland thrombus SUVmax was statistically significant (P < 0.005). On ROC analysis, a cutoff of SUVmax 2.25 (sensitivity, 78%; specificity, 100%) was suggested to differentiate tumor from bland thrombosis.
PET/CT is able to differentiate tumor from bland thrombosis, with an optimal cutoff value of SUVmax 2.25. The metabolic information increases the diagnostic accuracy of tumor thrombus and is a useful adjunct to the described features on contrast-enhanced CT.
From the Department of Diagnostic Radiology, Queen Mary Hospital, University of Hong Kong, Pokfulam, Hong Kong.
Received for publication March 21, 2012; revision accepted June 1, 2012.
Conflicts of interest and sources of funding: none declared.
Reprints: Elaine Yuen Phin Lee, BMedSci, BMBS, MRCP, FRCR, Department of Diagnostic Radiology, University of Hong Kong, Department of Diagnostic Radiology, Room 406, Block K, Queen Mary Hospital, University of Hong Kong, 102 Pokfulam Road, Hong Kong. E-mail: email@example.com.