Performance of dual-layer spectral CT-based extracellular volume fraction to assess Ki-67 proliferation status in rectal cancer: comparison with diffusion kurtosis imaging

Abstract Objective To evaluate and compare MRI-based diffusion kurtosis imaging (DKI), diffusion-weighted imaging (DWI), and dual-layer spectral CT (DLSCT)-derived extracellular volume (ECV) fraction for predicting the Ki-67 proliferation index in rectal cancer (RC). Materials and methods This retrospective study included 117 patients with pathologically confirmed rectal adenocarcinoma who underwent preoperative MRI and DLSCT. Patients were divided into high (Ki-67 > 50%) and low (Ki-67 ≤ 50%) Ki-67 groups based on postoperative immunohistochemistry. Parameters from DKI (mean diffusivity (D mean ), mean kurtosis [K mean ]), DWI (apparent diffusion coefficient [ADC]), and ECV fraction were calculated and compared. Diagnostic performance was assessed using the area under the receiver operating characteristic curve (AUC) and decision curve analysis (DCA). Correlations with Ki-67 were evaluated using Spearman’s coefficient. Results The high Ki-67 group exhibited significantly higher K mean and ECV fraction, and lower D mean and ADC values (all p < 0.05). K mean and ECV showed positive correlations with Ki-67, while ADC showed a negative correlation. ROC analysis demonstrated that ECV fraction provided the optimal diagnostic efficacy (AUC = 0.815), which was significantly better than ADC (AUC = 0.665), D mean (AUC = 0.641), K mean (AUC = 0.721), and combined DKI parameters (AUC = 0.744). DCA confirmed a favorable net benefit for the ECV fraction. Additionally, ECV was associated with key pathological and staging features, like T-stage, N-stage, extramural venous invasion, and other aggressive pathological features. Conclusion DLSCT-derived ECV fraction demonstrated superior performance to diffusion-based MRI parameters for predicting the proliferation status of RC, providing a more comprehensive assessment of the tumor microenvironment. Key Points Question Noninvasive preoperative biomarkers are lacking to assess Ki-67 proliferation status and guide personalized treatment strategies for patients with rectal cancer. Findings Dual-layer spectral CT (DLSCT)-derived extracellular volume (ECV) fraction outperformed MRI diffusion parameters for Ki‑67 and correlated with T/N stage, differentiation, and extramural vascular invasion. Critical relevance statement This study validates Dual-layer spectral CT (DLSCT)-derived extracellular volume (ECV) fraction as a non‑invasive biomarker that outperforms MRI diffusion parameters for Ki‑67 assessment and correlates with tumor aggressiveness, advancing personalized treatment decision‑making in rectal cancer radiology.