Spectral CT in the evaluation of perineural invasion status in rectal cancer.

Abstract

To investigate whether preoperative spectral CT quantitative parameters can assess perineural invasion (PNI) status in rectal cancer. Sixty-two patients diagnosed with rectal cancer who underwent preoperative spectral CT were retrospectively enrolled and divided into positive and negative PNI groups according to histopathologic results. The CT attenuation value (HU) of virtual monochromatic images (40-70keV), spectral curve slope (K(HU)), effective atomic number (Zeff), and iodine concentration (IC) from spectral CT were compared between these two groups using t test or rank sum test. A nomogram was established by incorporating the independent predictors to assess the overall diagnostic efficacy. The area under the ROC curves (AUCs) were compared using the DeLong test. The preoperative spectral CT parameters (40-70keV attenuation, K(HU), Zeff, and IC) were significantly higher in the PNI-positive group compared to the PNI-negative group (all p < 0.05). The highest predictive efficiency of PNI was observed at 40keV attenuation, with an area under the curve (AUC), sensitivity, specificity, and accuracy of 0.847, 81.8%, 72.5%, and 75.8%, respectively. Binary logistic regression demonstrated that the clinical feature (cN stage) and 40keV attenuation were independent predictors of PNI status. The nomogram incorporating these two predictors (cN stage and 40keV attenuation) exhibited the best evaluation efficacy, with an AUC, sensitivity, specificity, and accuracy of 0.885, 86.4%, 77.5%, and 80.6%. Spectral CT quantitative parameters proved valuable in the preoperative assessment of PNI status in rectal cancer patients. The combination of spectral CT parameters and clinical features could further enhance the diagnostic efficiency.