Definition and Prediction of Early Recurrence and Mortality Following Cytoreductive Surgery and Hyperthermic Intraperitoneal Chemotherapy for Colorectal Peritoneal Metastases: Towards Predicting Oncologic Futility Preoperatively

Abstract

Early recurrence (ER) is a significant challenge for patients with colorectal peritoneal metastases (CRPM) following cytoreductive surgery with hyperthermic intraperitoneal chemotherapy (CRS HIPEC). Preoperative risk stratification for ER would improve preoperative decision making. We conducted a retrospective study examining patients who underwent CRS HIPEC for CRPM from 2000 to 2018. Optimal definition of ER was determined via minimum p-value approach based on differentiation of post-recurrence survival. Risk factors for ER were assessed in a derivation cohort by uni- and multivariate logistic regression. A predictive score for ER was generated using preoperative variables and validated in an independent cohort. 384 patients were analyzed, 316 (82%) had documented recurrence. Optimal length of post-operative RFS to distinguish ER (n = 144, 46%) vs. late recurrence (LR) (n = 172, 63%) was 8 mos (p<0.01). ER patients had shorter median OS post-CRS-HIPEC (13.6 vs. 39.4 mos, p<0.01). Preoperative BMI (OR 1.88), liver lesions (OR 1.89), progression on chemotherapy (OR 2.14), positive lymph nodes (OR 2.47) and PCI score (16-20: OR 1.7; >20: OR 4.37) were significant predictors of ER (all p<0.05). Using this model, patients were assigned risk scores from 0 to 9. Intermediate (scores 4-6) and high-risk patients (score 7-9) had observed rates of ER of 56% and 79% and overall 2-year survival rates of 27% and 0% respectively. The model showed fair discrimination (AUC 0.72) and good calibration (Hosmer-Lemeshow GOF p = 0.68). ER predicts markedly worse OS following surgery. Preoperative factors can accurately stratify risk for ER and identify patients in whom CRS-HIPEC for CPRM is futile.