Patterns and Timing of Recurrence following CRS and HIPEC in Colorectal Cancer Peritoneal Metastasis.

INTRODUCTION

Cytoreductive Surgery and Hyperthermic Intraperitoneal Chemotherapy (CRS/HIPEC) is an established treatment of Colorectal Peritoneal Metastases (CRPM). This study aims to determine the timing and patterns of recurrent disease on imaging following complete CRS/HIPEC.

METHODS

Retrospective analysis of a national peritoneal tumour service database identified CRPM patients with complete CRS/HIPEC(CC0) from 2005 to-2018. Patients with<2 years follow-up or and those where post-operative histology from the CRS/HIPEC procedure did not confirm CRPM from their original colorectal cancer were excluded. Time to recurrence was measured from surgery to first radiologically illustrated recurrence. CT was the primary modality used, supplemented by PET-CT or MRI if required. Outcomes of interest were survival data (including overall survival (OS), disease-free survival (DFS) and peritoneal-recurrence free survival (PRFS)), timing and patterns of recurrent disease.

RESULTS

146 of the 176 patients identified were eligible for inclusion. Median OS for all study patients was 45.2 months (95% CI 38-53 months), median DFS was 11.7 months (95% CI 9-14 months), and median PRFS was 25.2 months (95% CI 14.7-30 months). Recurrent disease was seen in 112 cases (77%), radiologically classified as intraperitoneal in 50 patients (44%), single site systemic in 21 patients (19%) and multi-site in 41 patients (37%). CT detection rate for disease recurrence was 88%. Subgroup analyses showed that PCI ≥12, positive nodal primary disease and synchronous peritoneal disease were associated with worse outcomes.

CONCLUSION

Patients selected for CRS/HIPEC for CRPM have an OS > 45 months, with the majority recurring systemically within a year. Peritoneal recurrence is a later event after several years. Surveillance programs in this group should be most intensive in the first 2 years after surgery, using CT with oral and intravenous contrast.