Value of Diffusion-Weighted Imaging in Differentiating Benign from Malignant Portal Vein Thrombosis: A Cross-Sectional Prospective Study

Document Type : Original Article

Authors

Department of Diagnostic and Interventional Radiology, National Liver Institute, Menoufia University, Shebin El-Kom, Menoufia, Egypt.

Abstract

Background: It is still difficult to make noninvasive distinction between tumor invasion and bland clot in portal vein thrombosis [PVT]. The histopathologic examination is the gold standard to assess PVT. However, open laparotomy or percutaneous biopsy have been supplanted by imaging diagnostics in clinical practice to characterize PVT.  
Aim of the work: By measuring apparent diffusion coefficient [ADC] values, this study aimed to validate diffusion weighted magnetic resonance imaging [MRI] for distinguishing benign from malignant PVT.
Patients and methods: Diffusion weighted sequences and dynamic liver MRI were performed on 159 adult patients with imaging-confirmed PVT. To determine ADC values and signal intensity ratios, regions of interest were positioned in each thrombus and within the adjacent spinal cord.
Results: Malignant PVT [n=129] occurred in older patients than benign PVT [n = 30] [mean age 62.1 ± 7.3 vs 50 ± 13.3 years; p < 0.001]. Portal hypertension, lower limb swelling, and CHILD score C were prevalent in malignant than benign PVT [72.9%, 45.7%, 61.2% vs 33.3%, 16.7% and 0.0% respectively]. Mean thrombus ADC was significantly lower in malignant than benign PVTs [1.2 ± 0.14 × 10⁻³ mm²/s vs 1.4 ± 0.05 × 10⁻³ mm²/s respectively]. ROC analysis for ADC yielded an area under the curve [AUC] of 0.677 [p = 0.02] with a cutoff ≤ 1.2 × 10⁻³ mm²/s, achieving 73.2% sensitivity and 56.7% specificity. The ADC ratio [PVT/cord] was also lower in malignant cases [1.6 ± 0.3 vs 1.9 ± 0.36; p < 0.001] but demonstrated poor discriminatory performance [AUC = 0.60; p = 0.453].
Conclusion: When combined with mean ADC values, DW MRI is a valuable noninvasive imaging method that is highly effective at characterizing tissue. In addition, it can be used to distinguish between benign and malignant PVT. 

Keywords

Main Subjects

Files

Share

How to cite

Statistics