# Predictive Utility of Structured MRI Reporting for Rectal Cancer Outcomes

## TL;DR

This study shows that structured MRI reports help predict complications and recurrence in rectal cancer patients, improving surgical planning and risk assessment.

## Contribution

The study demonstrates how structured MRI reporting improves prediction of rectal cancer outcomes compared to traditional free-text reports.

## Key findings

- Sphincter involvement is strongly linked to increased postoperative complications.
- Extramural vascular invasion correlates with higher local recurrence rates.
- Advanced T staging predicts a higher risk of 30-day anastomotic leaks.

## Abstract

Background/Objectives: This retrospective study evaluates the predictive role of magnetic resonance imaging (MRI) in complications and recurrence in rectal cancer patients undergoing surgery and neoadjuvant therapy, highlighting the impact of structured reporting templates on MRI quality. Compared to traditional free-text reports, structured radiology reports offer a point-by-point evaluation, improving clarity and completeness by thoroughly addressing all relevant findings. MRI is critical in rectal cancer staging, guiding treatment based on tumor characteristics like T stage, sphincter involvement, vascular invasion, and lymph node status. Methods: A retrospective analysis of MRI and reports from 67 rectal cancer patients at the time of diagnosis, who were subsequently treated with neoadjuvant radiochemotherapy and surgery, was conducted. MRI report features, including tumor location, morphology, T stage, sphincter infiltration, mesorectal fascia involvement, lymph nodes, and extramural vascular invasion, were evaluated against European Society of Gastrointestinal and Abdominal Radiology (ESGAR) recommendations. Multivariate and univariate analyses were performed to correlate MRI findings with postoperative outcomes such as complications, local recurrence, bleeding, and 30-day anastomotic leaks. Results: Sphincter involvement showed a strong association with increased complications (multivariate β = 0.410, univariate r = 0.270). Extramural vascular invasion was linked to higher rates of local recurrence (multivariate β = 0.199, univariate r = 0.127). Lymph node involvement correlated with an elevated risk of postoperative bleeding (multivariate β = 0.133, univariate r = 0.293). Additionally, advanced T staging predicted a higher incidence of 30-day anastomotic leaks (multivariate β = 0.210, univariate r = 0.261). These findings may provide clinically relevant insights to support personalized surgical planning and improve preoperative risk stratification. Conclusions: Detailed MRI reporting, aligned with structured templates, significantly guides surgical and therapeutic strategies in rectal cancer management. However, the retrospective nature of the study and the limited sample size may affect the generalizability of the results.