Value of dual-energy computed tomography quantitative parameters in differentiating neoplastic from bland portal vein thrombosis.

BACKGROUND

Portal vein thrombosis (PVT) is a common clinical pathological state involving distinct pathophysiological processes. Accurate discrimination of PVT nature is of utmost importance for guiding treatment strategies, but histopathology has limitations and imaging lacks quantitative indices. This study aimed to evaluate the feasibility and diagnostic value of quantitative parameters from dual-energy computed tomography (DECT), an advanced technique that allows for such quantitative evaluation, in distinguishing neoplastic from bland PVT.

METHODS

Computed tomography (CT) images of 173 patients with PVT (bland group, n=74; neoplastic group, n=99) were retrospectively analyzed. Portal venous phase iodine-based decomposition images were reconstructed to contrast iodine indices among groups, including thrombus iodine concentration (IC), normalized iodine concentration in the aorta (NIC-A), normalized iodine concentration in the portal vein (NIC-V), electron density (Rho), effective atomic number (Z), dual-energy index (DEI), and spectral slope (K). Diagnostic performance was evaluated using the area under the receiver operating characteristic curve (AUC), sensitivity, specificity, and accuracy. Univariate and multivariate analyses selected DECT parameters and created a nomogram for prediction.

RESULTS

(I) The values of IC, NIC-A, NIC-V, Rho, Z, DEI, and K were significantly higher in the neoplastic group compared to the bland group (P<0.001). (II) The AUC using IC, NIC-A, and NIC-V to differentiate between neoplastic and bland PVT were 0.963, 0.970, and 0.969, respectively; using Rho, Z, DEI, and K, they were 0.732, 0.952, 0.949, and 0.933, respectively. All the quantitative parameters achieved high sensitivity and high specificity in distinguishing neoplastic from bland PVT. (III) A nomogram was developed to predict neoplastic PVT probability; its AUC, sensitivity, and specificity reached remarkable levels, with values of 0.994, 98.59%, and 97.96% in the training cohort and 0.940, 100%, and 92.00% in the test cohort, respectively.

CONCLUSIONS

The DECT quantitative parameters demonstrate significant potential as non-invasive markers for distinguishing between neoplastic and bland PVT.