Tailored-surgery for locally advanced colon cancer based on 3D mathematical reconstruction surgical planner: Observational comparative non-randomized study.

Background: This study investigates the effectiveness of a three-dimensional reconstruction mathematical model (3D-IPR) for preoperative planning in locally advanced colon cancer (LACC) with threatened surgical margins. The objective was to evaluate the utility of a 3D-IPR surgical planner tool in cases of LACC with threatened surgical margins. Additionally, the study aims to compare the diagnostic accuracy of the 3D-IPR model against conventional CT scans in determining the infiltration of neighboring structures.

Methods: This Single-center, prospective, observational, comparative, non-randomized study. INCLUSIóN CRITERIA: Patients over 18 years old undergoing surgery for LACC as indicated by a radiologist's analysis of CT scans. Preoperative confirmation of neoplasm by colonoscopy.

Exclusion Criteria: patients who had received neoadjuvant chemotherapy, suspected carcinomatosis on preoperative CT and patients with unresectable tumors. All patients were selected consecutively.

Procedures: Intervention involved using a 3D-IPR model as a surgical planning tool for patients with LACC. The 3D-IPR provided detailed metrics about the tumor and surrounding structures to assist in surgical planning. Surgical procedures were guided by the radiological assessments from CT scans and intraoperative findings, with surgeries categorized based on surgical margins as R0, R1, or R2. The primary endpoint was the diagnostic accuracy of the 3D-IPR model in determining tumor infiltration of neighboring structures compared to conventional CT scans. The measure used to assess this outcome was the definitive pathological report of tumor infiltration, which served as the gold standard for comparison. Demographic, intraoperative, morbidity, mortality, and pathological data were analyzed.

Results: 21 patients were assessed, 1 excluded with a final sample of 20 patients. 3D-IPR model demonstrated higher diagnostic accuracy for tumor infiltration of neighboring structures compared to conventional CT scans, with sensitivity, specificity, Positive Predictive Value, and Negative Predictive Value of 70 %, 90 %, 87.5 %, and 75 %, respectively. Surgeries were predominantly minimally invasive (70 %), with no major complications or mortality within 30 days, and a 0 % conversion rate to open surgery.

Conclusions: The 3D-IPR model significantly enhances preoperative planning accuracy, reducing the risk of incomplete resections and improving surgical outcomes. This technology offers a reliable basis for surgical decisions, potentially improving patient prognosis and survival rates.