Despite different available methods for colorectal cancer (CRC) screening and their proven benefits, morbidity, and mortality of this malignancy are still high, partly due to lowcompliance with screening. Minimally invasive tests based on the analysis of blood specimens may overcome this problem. The purpose of this reviewwas to give an overviewof published studies on blood markers aimed at the early detection of CRC and to summarize their performance characteristics. The PUBMED database was searched for relevant studies published until June 2006. Only studies with more than 20 cases and more than 20 controls were included. Information on the markers under study, on the underlying study populations, and on performance characteristics was extracted. Special attention was given to performance characteristics by tumor stage. Overall, 93 studies evaluating 70 different markers were included. Most studies were done on protein markers, but DNA markers and RNA markers were also investigated. Performance characteristics varied widely between different markers, but also between different studies using the same marker. Promising results were reported for some novel assays, e.g., assays based on SELDI-TOF MS or MALDI-TOF MS, for some proteins (e.g., soluble CD26 and bone sialoprotein) and also for some genetic assays (e.g., L6 mRNA), but evidence thus far is restricted to single studies with limited sample size and without further external validation. Larger prospective studies using study populations representing a screening population are needed to verify promising results. In addition, future studies should pay increased attention to the potential of detecting precursor lesions.
Two recently developed monoclonal antibody (MAbkbased anti-mucin assays, CA M26 and CA M29, were studied in 250 cancer patients and compared to 3 well-established marker tests, viz., CA 125, CA 15.3 and SCC, in order to assess their clinical usefulness as serum tumor markers. Pre-treatment sera were obtained from patients with predominantly lowstage epithelial malignancies comprising 200 adenocarcinomas (of the ovary, endometrium, breast and large intestine) and 50 squamous-cell carcinomas (of the uterine cervix). Pretreatment sera of 50 patients with benign ovarian tumors were included to evaluate levels in benign disease. CA M26 and CA M29 cut-off levels were established in 89 healthy controls. In patients with adenocarcinomas, overall positivity for CA M29 was 2476, ranging from 10% in breast cancer to 60% in ovarian cancer. Overall positivity was highest for CA 125 (30%) and lowest for CA M26 (I 8%) with CA M29 (24%) being similar to CA 15.3 (25%). In adenocarcinomas the combined CA M26-CA M29 assays equalled results obtained with the CA 125-CA 15.3 combination (33% vs. 36%). Elevation of 2 or more markers was highly indicative of advanced disease (p < 0.025). A majority of positive patients showed either CA M26 or CA M29 elevations, indicating that both antibodies detect distinct epitopes. After adjustment for tumor site and stage, the profile of CA M26 as a single marker differed significantly from the profiles of CA 125 and of CA M29. CA M26 was frequently (32%) elevated in patients with squamous-cell carcinoma of the cervix and CA M26 levels were often independently elevated. CA M26 seems to be valuable as an additional marker in breast cancer and perhaps as a new marker in cervical cancer. CA M29 may be useful in ovarian cancer in addition to CA 125.