Prognostic Factors for Surgical Failure in Malignant Bowel Obstruction and Peritoneal Carcinomatosis

Abstract

Introduction: Patients with peritoneal metastasis frequently develop malignant bowel obstruction (MBO). Medical palliative management is preferred but often fails. Conversely, the role of palliative surgery remains unclear and debated. This study aims to identify patients who could benefit from invasive surgical interventions and factors associated with successful surgical palliation. Materials and Methods: In this retrospective study, 98 consecutive patients who underwent palliative surgery for MBO over 5 years were reviewed. We evaluate as the primary outcome surgical failure to select patients who could benefit from palliative surgery, avoiding unnecessary surgery. A prognostic score was developed based on a logistic regression model to identify patients at risk of surgical failure. The score was evaluated for overall accuracy by receiver operating characteristic curve analysis. Results: Palliative surgery was achieved in 76 (77.5%) patients. The variables that were found to be significant factors for surgical failure are recurrent disease (P = 0.015), absence of bowel obstruction (P < 0.001), absence of bowel distension (P < 0.001), and mesenteric involvement (P = 0.001) and retraction (P < 0.001). The absence of bowel distension (P = 0.046) and bowel obstruction (P = 0.012) emerged as independent predictors of surgical failure. Carcinomatosis level assessment for peritoneum score, based on these factors, was built to evaluate the risk of surgical failure. Conclusion: Our proposed scoring system might help select patients most likely to benefit from palliative surgery.

Keywords: ileostomy; malignant bowel obstruction (MBO); palliative outcomes; peritoneal carcinomatosis (PC); surgical palliation.

Figure 1

Receiver operating characteristic (ROC) analysis of carcinosis level assessment for peritoneum (CLAP) score. Area under ROC curve is 0.866 (0.782–0.926) p value < 0.001. According to Youden index J the best discriminating value was >4. Sensitivity, specificity, negative predictive value and positive predictive value for each score level are reported in the table.