Toward Ovarian Cancer Biomarkers

By the time doctors detect ovarian cancer in a patient, it's usually too late for successful treatment. More timely diagnosis could be one eventual payoff of efforts to identify early-stage biomarkers. Toward that end, a team of Swedish researchers led by Peter James at Lund University reports the results of a large-scale proteomics analysis to find biomarkers of human ovarian cancer (J. Proteome Res., DOI: 10.1021/pr060593y). Using 64 tissue samples from patients at different stages of ovarian cancer, they identified proteins whose expression levels change significantly among normal, benign, borderline, and malignant tissues. Many of these proteins were not previously linked with ovarian cancer, but some of them have been associated with other kinds of cancer. The researchers are using immunohistochemical methods to confirm that the proteins are differentially expressed in ovarian tumors. After raising antibodies against the proteins, the scientists validated five of the proteins; they are currently raising antibodies against 30 more. Because of the variation in expression levels from patient to patient, the researchers suggest that many biomarkers—possibly as many as 100—will be needed to diagnose ovarian cancer reliably.