Fecal Assay for COX-2 mRNA Helps Diagnose Colorectal Cancer


Laurie Barclay, MD


Aug. 12, 2004 — A fecal assay for cyclooxygenase-2 (COX-2) mRNA is
highly sensitive and specific for diagnosing colorectal cancer (CRC),
according to the results of a study published in the August issue of
Gastroenterology.


"COX-2 is overexpressed frequently in aerodigestive tumors,
especially in colorectal tumors. Therefore, it may be a suitable
biomarker for [CRC] screening," write Shigeru Kanaoka and colleagues,
from the Hamamatsu University School of Medicine in Japan. "RNA-based
stool assays have been reported in several preliminary studies, one
of which showed that CD44 variant expression in human feces could be
detected in 68% of CRC patients by a combination of reverse-
transcription polymerase chain reaction (RT-PCR) and Southern
hybridization."

In this study, 29 patients with CRC and 22 control patients without
neoplastic disease of the colon or rectum had RNA isolated from
routinely collected stool samples, as well as nested RT-PCR testing
for expression levels of carcinoembryonic antigen (CEA) and COX-2.

For the fecal COX-2 assay, sensitivity for CRC was 90% (95%
confidence interval [CI], 73% - 98%), and specificity was 100% (95%
CI, 85% - 100%). For the fecal CEA assay for CRC, sensitivity was
100% (95% CI, 88% - 100%), and specificity was 5% (95% CI, 2% - 23%).
COX-2 mRNA was detected in three of four patients with Dukes' stage
A, 13 of 14 patients with Dukes' stage B, and 10 of 11 patients with
Dukes' stage C or D. COX-2 mRNA was detected in five of seven
patients with proximal cancer and in 21 of 22 patients with distal
cancer.

"The COX-2 assay was superior to the CEA assay for detecting CRC in
terms of specificity, although both assays had high sensitivity," the
authors write. "It is possible to detect COX-2 mRNA in feces from
patients irrespective of the clinical stage of CRC."

Study limitations include small sample size and inability to
determine whether colonoscopy together with forceps biopsy influenced
COX-2 expression. The authors therefore recommend further studies
with larger numbers of patients and control subjects to determine the
sensitivity for a much broader spectrum of tumors, to assess the
specificity for a much broader spectrum of control subjects, and to
determine whether the fecal COX-2 assay is as sensitive and specific
as the fecal occult blood test in average-risk persons.

"Our method has certain advantages compared with DNA-based stool
assays; approximately 1 g of fecal pellet is sufficient for the assay
and more sensitive in terms of analyzing a single molecule," the
authors conclude. "It currently remains unclear, however, that the
fecal COX-2 assay is a useful screening test for CRC in the clinical
setting."

Gastroenterology. 2004;127:422-427

Reviewed by Gary D. Vogin, MD