NEW YORK, Feb. 8 - Researchers have used proteomic analysis to develop a promising blood-based test for ovarian cancer, according to a recent study.
The noninvasive test appears to detect the cancer even in its earliest stages, at the point when treatment provides the most benefit but when conventional screens often fail to spot the disease.
In a report to be published in the Feb. 16 issue of The Lancet investigators from the US Food and Drug Administration, the National Cancer Institute, MD Anderson Cancer Center, the Simone Protective Cancer Institute, Northwestern University, and Bethesda, Md.-based Correlogic Systems, used mass spectroscopy and bioinformatics analysis to characterize the blood-borne proteins of patients with known ovarian cancer.
In order to create a proteome profile characteristic of the disease, the researchers used SELDI-TOF machines from Ciphergen Biosystems to analyze protein content in 100 blood samples, generating more than 15,000 data points per patient. They then used Correlogic's Proteome Quest bioinformatics tool to isolate protein abundance patterns that would be able to conclusively identify cancer patients.
Researchers then used the system to screen 116 serum samples: 50 were from women previously diagnosed with ovarian cancer and 66 were taken from healthy women or women with non-malignant disease. The algorithm accurately identified all 50 cancer cases and correctly classified 63 or the 66 non-malignant samples. Of the 50 cancer patients, 18 had hard-to-detect stage I disease.
Ovarian cancer, the fifth-leading cause of cancer death among women in the US, accounted for roughly 14,500 deaths in 1995. Stage I ovarian cancer has a 5-year survival rate of more than 90 percent, and many patients can be cured through surgery alone.
But the cancer antigen 125 test now most commonly used to screen for the disease only picks up about 50 percent to 60 percent of early-stage cancers. More than 80 percent do not discover their disease until its later stages, when it has a 35 percent 5-year survival rate.
The team intends to test the proteomic analysis system in larger trials, said lead author Emanuel Petricoin, who is co-director of the joint FDA/NCI program in clinical proteomics. "The concept that patterns of proteins instead of single biomarkers can be used as a potential diagnostic is a brand new paradigm," he said in a statement.
The authors point out that while the test appears to have excellent sensitivity--each of the 50 cancers were accurately identified--its current false-positive rate means that it would not yet be suitable to use alone as a screening technique in the general population.