MRI-Determined Tumor Contact Area as a Predictor of Pathological Extraprostatic Extension in Clinical T2 Prostate Cancer.

Objectives: To assess the validity of magnetic resonance imaging-determined tumor contact area (MRI-TCA) as a predictive factor for pathological extraprostatic extension (EPE) in cT2N0M0 prostate cancer patients. Methods: We retrospectively analyzed 72 cT2N0M0 prostate cancer patients who underwent multiparametric MRI (mpMRI) followed by robot-assisted laparoscopic prostatectomy (RARP) between February 2014 and April 2021. Patients whose MRI-based index lesion did not match the pathological specimens were excluded. MRI-TCA was approximated using an elliptical shape and calculated by two different methods: MRI-TCA1: Calculated using the tumor contact length (TCL) in the axial plane and the longer TCL in either the sagittal or coronal plane, capturing tumor dimensions across two planes. MRI-TCA2: Calculated using the TCL in the axial plane and tumor thickness derived from MRI slice data, reflecting the tumor's contact area within the MRI volume. We compared postoperative prostate-specific antigen (PSA) recurrence-free survival by stratifying patients based on the optimal thresholds of MRI-TCL, MRI-TCA1, MRI-TCA2, pathological-TCL, and pathological-TCA. Results: Sixteen patients (22.2%) were pathologically positive for EPE. MRI-TCL, MRI-TCA1, and MRI-TCA2 were significantly greater in patients with EPE-positive (EPE+) tumors than in those with EPE-negative (EPE-) tumors (p < 0.0001, p < 0.0001, and p = 0.0026, respectively). No statistically significant differences were found between MRI-TCL and MRI-TCA1 (p = 0.914) or between MRI-TCL and MRI-TCA2 (p = 0.112) in predicting pathological EPE. A significant difference in postoperative PSA recurrence rate was observed in the stratified analysis based on pathological-TCA (p = 0.022). Conclusion: Both MRI-TCA1 and MRI-TCA2 are clinically accessible and effective parameters for predicting pathological EPE in cT2N0M0 prostate cancer patients. However, neither method demonstrated clear superiority over MRI-TCL. Pathological-TCA was shown to be a significant predictor of both pathological EPE and postoperative PSA recurrence.