Abstract

Aims In colorectal cancer (CRC), the presence of lymph node (LN) metastases is an important prognostic factor. Approximately 20% of patients diagnosed as having node-negative (pN0) CRC will relapse. Pathological nodal stage misclassification due to sampling error resulting from the small volume of tissue tested has been proposed to explain this recurrence rate in pN0 patients. The authors compared the assessment of node positivity by histopathology (HP) with a molecular method which can accommodate larger tissue volumes.

Methods Detection rate of guanylyl cyclase C (GCC) mRNA was determined in 1495 LNs from 99 CRC patients. Using a subset of 647 LNs, multiple levels of HP analysis were compared with GCC mRNA molecular detection. Finally, clinicopathological factors were correlated with the molecular detection of GCC and clinical outcome in 123 patients with pN0 colon cancer.

Results GCC mRNA was detected in 8.0% of the 560 nodes initially identified as HP-negative, whereas two repeat HP examinations detected 3.0% of these cases. In HP-positive LNs, the GCC mRNA detection rate was 90% (78/87) when half-LN were tested. Testing the entire LN remaining after HP by GCC increased the detection rate of HP-positive LNs to 95% (p=0.027). In comparison, 75% (65/87) and 92% (80/87) of the LN positive by clinical HP remained positive when one or two subsequent sections were examined by HP. Finally, patients with pN0 disease who were GCC-positive exhibited an earlier time of recurrence (hazard ratio, 3.54; 95% CI 1.40 to 8.98; p=0.0077).

Conclusions Molecular detection of tumour cells in LNs may have prognostic value in identifying patients diagnosed as having pN0 colon cancer who will relapse following surgery.