Preoperative Peritoneal MRI: Usefulness to Highlight Potential Hidden Lesions for Complete Cytoreductive Surgery in Patients with Colorectal Cancer with Surgical History.

BACKGROUND

In patients with colorectal cancer and peritoneal metastases (CRC-PM), the completeness of cytoreductive surgery (CRS) is crucial. However, a history of moderate (Prior Surgical Score, PSS-2) or extensive (PSS-3) abdominal surgery may compromise the exploration, increasing the risk of undetected CRC-PM. This retrospective monocentric study investigated the value of preoperative peritoneal magnetic resonance imaging (MRI) in identifying potentially occult lesions in patients with PSS-2/3 CRC-PM scheduled for CRS.

PATIENTS AND METHODS

Consecutive patients with pathologically confirmed CRC-PM and PSS-2/3, selected for radical treatment, were included. All underwent preoperative peritoneal MRI ≤ 7 days before CRS, between January 2015 and December 2020. MRI, surgical, and pathological reports were reviewed focusing on seven anatomical sites of interest (perihepatic, pelvic, retroperitoneum, abdominal wall, anastomosis, inguinal canal, and cardiophrenic space).

RESULTS

Overall, 248 patients were included; 242 (97.6%) underwent complete CRS (CC-0). Among them, 212 (85.5%) were PSS-2 and 36 (14.5%) PSS-3. The sensitivity, specificity, and accuracy of MRI in detecting lesions were, respectively, 65%, 91%, and 82% (perihepatic region); 53%, 81%, and 63% (pelvis); 41%, 91%, and 69% (retroperitoneum); 46%, 91%, and 79% (abdominal wall); and 44%, 98%, and 74% (anastomotic sites). In the inguinal canal and cardiophrenic space, preoperative MRI led to ten resections in ten patients, with neoplastic cells detected in eight cases (80%).

CONCLUSIONS

Preoperative peritoneal MRI demonstrated good specificity and a promising negative predictive value (NPV) but modest sensitivity in detecting lesions across seven anatomically challenging regions. Further studies are warranted to better define its added value over standard preoperative imaging protocols.