Discrimination of metastatic lymph nodes in patients with gastric carcinoma using diffusion-weighted imaging

Purpose: To determine the accuracy of diffusion-weighted imaging (DWI) in discrimination of metastatic lymph nodes (LNs) in gastric carcinoma with rigorous histopathological correlation. Materials and Methods: With Institutional Review Board (IRB) approval, 28 patients with gastric carcinoma underwent magnetic resonance imaging (MRI) and DWI before surgery. LNs were resected at surgery and thereafter submitted for histopathological analyses. All histopathologically identified LNs (≥5 mm) that exactly matched the location and size of nodes on MRI/DWI were submitted to lesion-by-lesion analyses. Short-axis diameter, border irregularity, enhanced patterns, and apparent diffusion coefficient (ADC) values of each LN were recorded. Each measurement was compared between metastatic and benign LNs, confirmed by nodal histopathology. A receiver operating characteristic (ROC) curve was generated to evaluate the capability of morphological and ADC measurements in distinguishing metastatic lymph nodes. Results: The median ADC value of metastatic nodes was significantly lower (1.28 × 10^-3 mm²/sec; interquartile range, 1.20-1.31) than that of benign (1.55; 1.47-1.73) nodes (P < 0.001). DWI showed greater diagnostic accuracy in determining metastatic nodes (AUC = 0.857) than the combined morphological measurements of short-axis, border irregularity, and enhanced patterns (AUC = 0.746, P = 0.03). Adding ADC values to the combined morphologic criteria demonstrated the greatest predictive power (AUC = 0.889). Conclusion: DWI may provide great potential in effective discrimination of metastatic LNs in gastric carcinoma.