Recurrence and colon cancer-specific death in patients with large bowel obstruction requiring urgent operation: a competing risks analysis.

Abstract

Clinical presentation with large bowel obstruction has been proposed as a predictor of poor long-term oncological outcomes after resection for colorectal cancer. This study examines the association between obstruction and recurrence and cancer-specific death after resection for colon cancer. Consecutive patients who underwent resection for colon cancer between 1995 and 2014 were drawn from a prospectively recorded hospital database with all surviving patients followed for at least 5years. The outcomes of tumour recurrence and colon cancer-specific death were assessed by competing risks multivariable techniques with adjustment for potential clinical and pathological confounding variables. Recurrence occurred in 271 of 1485 patients who had a potentially curative resection. In bivariate analysis, obstruction was significantly associated with recurrence [hazard ratio (HR) 2.23, CI 1.52-3.26, p<0.001] but this association became nonsignificant after adjustment for confounders (HR 1.53, CI 0.95-2.46, p=0.080). Colon cancer-specific death occurred in 238 of 295 patients who had a noncurative resection. Obstruction was not significantly associated with cancer-specific death (HR 1.02, CI 0.72-1.45, p=0.903). In patients who had a noncurative resection, the competing risks incidence of colon cancer-specific death was not significantly greater in obstructed than in unobstructed patients (HR 1.02, CI 0.72-1.45, p=0.903). Whilst the immediate clinical challenge of an individual patient presenting with large bowel obstruction must be addressed by the surgeon, the patient's long-term oncological outcomes are unrelated to obstruction per se.