Abstract
BACKGROUND: Tumor budding (TB) is an autonomous risk factor indicative of the invasiveness and prognosis of colorectal cancer (CRC). Dual-layer spectral detector computed tomography (SDCT) is a novel energy-resolved imaging technology that enables multiparametric tissue characterization through quantitative spectral analysis. Nevertheless, the efficacy of quantitative measures derived from SDCT in differentiating the TB grade of CRC remains ambiguous. This research aims to evaluate the efficacy of SDCT in preoperatively predicting TB grade in CRC. METHODS: This retrospective study enrolled 74 consecutive CRC patients who underwent surgical resection with confirmed TB grade at the First Hospital of Lanzhou University between July 2023 and July 2024. All patients received whole-abdomen spectral computed tomography (CT) scans within one week preoperatively. Based on postoperative pathology, patients were stratified into low-grade TB (Group A, Bd1) and high-grade TB (Group B, Bd2 + Bd3) groups. Using the IntelliSpace Portal workstation, radiologists manually delineated regions of interest (ROIs) on axial images to measure spectral CT quantitative parameters in both arterial and venous phases, encompassing CT values at 40 and 120 keV (CT(40keV), CT(120keV)), iodine concentration (IC), effective atomic number (Zeff), the slope of the spectral Hounsfield unit (HU) curve (λHU), and to calculate normalized iodine concentration (NIC), normalized effective atomic number (NZeff), and arterial enhancement fraction (AEF). Statistically significant variables from univariate analysis were incorporated into multivariate logistic regression to construct a combined predictive model. Diagnostic performance was evaluated using receiver operating characteristic (ROC) curve analysis. RESULTS: The research included a total of 74 individuals with CRC (mean age: 61.16±13.08 years; male: 36, female: 38). Compared with the low-grade TB group, the high-grade TB group showed significantly higher proportions of both high-grade tumors (P=0.001) and elevated Ki-67 expression (P=0.009), along with a significantly increased likelihood of lymphovascular invasion (P=0.005). The CT(40keV), CT(120keV), IC, NIC, Zeff, NZeff, and λHU in the arterial phase and AEF of the low-grade TB group were elevated compared to the high-grade TB group (all P<0.05), but no significant differences in the spectral parameters were seen in the venous phase between the two groups (all P>0.05). Among the individual characteristics, NIC showed the highest area under the curve (AUC) value (0.723) in discriminating between low-grade and medium-high grade TB. The AUC of the combined model with each parameter in the arterial phase and AEF was 0.814, surpassing that of single spectral parameters, with a sensitivity of 84.62% and a specificity of 72.73%. CONCLUSIONS: Quantitative parameters of spectral CT may be associated with TB grade in CRC, and their potential value in preoperatively predicting different TB grades could be of great significance for guiding clinical treatment selection and prognostic evaluation.