Prediction of pathological risk factors in rectal cancer using combined extracellular volume fraction from T1 mapping...

Accurate preoperative prediction of pathological risk factors in rectal cancer is critical for guiding treatment decisions and improving patient outcomes. While the apparent diffusion coefficient (ADC) and extracellular volume fraction (ECV) each provide insights into tumor biology, their combined predictive value remains underexplored.

To assess the predictive performance of ECV, derived from T1 mapping, and ADC, from diffusion-weighted imaging (DWI), both individually and in combination, for evaluating pathological features in rectal cancer.

This retrospective study included 51 patients with histologically confirmed rectal adenocarcinoma, who underwent 3.0 T MRI between October 2023 and October 2024. Quantitative ECV and ADC values were extracted from T1 mapping and DWI, respectively. Logistic regression models, incorporating Ridge and Elastic Net regularization, were used to predict T stage, vascular invasion, and nerve invasion. Five-fold cross-validation was applied, and model performance was evaluated using AUC, sensitivity, specificity, and F1 score. DeLong's test was used to compare AUCs between models.

ECV and ADC values were significantly associated with pathological features. ECV was higher and ADC was lower in advanced T stage (T3-4), vascular invasion-positive, and nerve invasion-positive groups (P < 0.05). The combined ECV + ADC model achieved the highest AUCs: 0.906 for T staging, 0.811 for vascular invasion, and 0.861 for nerve invasion, outperforming single-parameter models. However, differences were not statistically significant (P > 0.05).

The combination of T1 mapping-derived ECV and DWI-derived ADC improves the noninvasive prediction of pathological risk factors in rectal cancer. This dual-biomarker approach may enhance preoperative assessment and support personalized treatment strategies.