A Comparison of Emergency First Presentations of Colorectal Cancer in Under-50 and Over-50 Year-Old Patients.

INTRODUCTION: Colorectal cancer (CRC) is the second commonest malignancy related death in Western Europe with incidence increasing in young adults. 31% of UK patients with CRC present as emergencies. We compare the incidence, characteristics, management and outcomes in two cohorts presenting as CRC emergencies; under-50 and over-50 years old.

MATERIALS AND METHODS: Retrospective analysis was performed on 322 patients with emergency presentations of CRC over a 9-year period (January 2005-December 2013, West Suffolk Hospital, UK). Data were analyzed for demographics, symptoms, investigations, stage, grade, genetics, tumor location, management, and mortality.

RESULTS: 300 patients over 50 years old presented with CRC emergencies; 153 women (51%):147 men (49%); median age 77 years (interquartile range: 67-84). 22 patients under 50-years-old; 12 women (55%):10 men (45%); median age 43 years ([Interquartile Range (IQR)]: 35-46 years). Bowel obstruction was less common in under-50s (18.2% vs. 40.7%; p = 0.04). No over-50s had a positive family history for CRC; 7 under-50s did. A higher proportion of under-50s presented with Dukes A carcinomas (14.3% vs. 0.4%; p = 0.002), but no difference in other Dukes stages. Surgery was performed in a higher proportion of under-50s (95.5% vs. 77.0%; p = 0.04) and a higher proportion had same day surgery (71.4% vs. 28.1%; p = 0.01). Overall mortality was lower in under-50s (36.4% vs. 64.0%; p = 0.02). No significant differences occurred in in-hospital mortality (4.7% vs. 8.0%; p = 0.55), overall one-year survival (31.8% vs. 41.7%; p = 0.36), or median survival to death or study conclusion (27.1 vs. 19.6 months; p = 0.13).

CONCLUSION: Emergency CRC had comparable outcomes between young and old cohorts, during the study time period. Younger patients were more likely to undergo operative interventions but overall survival was comparable. Our study was limited by the reporting biases intrinsic to retrospective analyses and by a small under-50 sample size. Further large-scale studies are warranted to support observations.