Locally advanced rectal cancer: MR imaging for restaging after neoadjuvant radiation therapy with concomitant chemotherapy. Part II. What are the criteria to predict involved lymph nodes?

Purpose: To prospectively determine diagnostic performance of predictive criteria for nodal restaging after radiation therapy with concomitant chemotherapy by using ultrasmall superparamagnetic iron oxide (USPIO)-enhanced magnetic resonance (MR) imaging in patients with rectal cancer. Materials and Methods: After institutional review board approval and informed consent were obtained, 39 patients (24 men, 15 women; mean age, 64 years) with rectal cancer underwent USPIO-enhanced two-dimensional (2D) T2-weighted fast spin-echo, three-dimensional (3D) T1-weighted gradient-echo, and 3D T2*-weighted MR for restaging. Two observers evaluated nodes for border irregularity, short- and long-axis diameters, and estimated percentage of white region (<30%, 30%-50%, or >50%) within the node (3D T2*-weighted images). Ratio of the measured surface area of the white region within the black node to the measured surface area of the total node (Ratio(A)) was calculated. Signal intensity (SI) in gluteus muscle (SI(GM)) and in total node (SI(TN)) were used to calculate SI(TN)/SI(GM) ratio. Histopathologic findings were reference standard. Receiver operating characteristic (ROC) curves were compared and interobserver agreement was determined. Results: Lesion-by-lesion analysis was feasible in 201 lymph nodes. Area under the ROC curve (AUC) of border and short- and long-axis diameters for observer 1 were 0.85, 0.87, and 0.88 and for observer 2 were 0.70, 0.89, and 0.87, respectively. AUC for estimated percentage of white region within the node, Ratio(A), and SI(TN)/SI(GM) ratio for observer 1 were 0.98, 0.99, and 0.62 and for observer 2 were 0.97, 0.98, and 0.65, respectively. AUC for USPIO-enhanced MR criteria was significantly better than AUC for conventional MR criteria (P < .01). All criteria except border irregularity and SI(TN)/SI(GM) ratio showed high interobserver agreement (kappa > 0.79). Conclusion: The most reliable predictors for identifying benign nodes after radiation therapy with concomitant chemotherapy by using USPIO-enhanced MR imaging for restaging in patients with rectal cancer were estimated percentage of white region within the node and Ratio(A). Measurements on standard 2D T2-weighted fast spin-echo images versus primary staging results offer reasonably good accuracy to identify benign lymph nodes after therapy. (c) RSNA, 2009.