Role of circulating tumour cells in predicting recurrence after excision of primary colorectal carcinoma

Background: This study assessed the potential for reverse transcriptase-polymerase chain reaction (RT-PCR)-based circulating turnout cell identification to predict colorectal cancer recurrence. Methods: mRNA for carcinoembryonic antigen and cytokeratin 20 was identified by RT-PCR in blood from patients with colorectal cancer, before and after primary turnout resection. Cancer recurrence was assessed at follow-up, and the accuracy of RT-PCR and primary turnout lymph node positivity in predicting recurrence was estimated. Results: One hundred and ninety-six patients with colorectal cancer were studied over a median follow-up of 1393 days from surgery. Regression analysis selected 24-h post-resection RT-PCR positivity (hazard ratio for a positive test in predicting recurrence 8.66 (95 per cent confidence interval (c.i.) 3.08 to 24.33)) before lymph node involvement (hazard ratio 7.92 (95 per cent c.i.. 3.26 to 19.20)). When 24-h post-resection RT-PCR was combined with lymph node positivity, the hazard ratio increased to 18.54 (95 per cent c.i. 4.01 to 85.11), attributing a 3 per cent recurrence risk to 52 per cent, and a 50 per cent recurrence risk to 48 per cent, of patients with colorectal cancer resected with curative intent. Conclusion: RT-PCR positivity within 24 h of primary colorectal cancer resection is a strong predictor of colorectal cancer recurrence, and may be useful clinically.