Right to Life - Archive

Algorithm Blood Test to Detect Ovarian Cancer

The analysis of proteins in blood samples may be able to help doctors detect early-stage ovarian cancer [study publ. 1Dec05 online early edition of the Proceedings of the National Academy of Sciences; Long Island Newsday].

Although nearly 25,000 U.S. women are diagnosed with ovarian cancer each year and 14,000 women die each year of the cancer, ovarian cancer is one of the most difficult cancers to detect at an early stage.

Dr. Lance Liotta [National Cancer Institute] and Dr. Emanuel Petricoin [FDA] have been working to find a highly accurate detection method for ovarian cancer.

Along with researchers from Bethesda,Md.-based Correlogic Systems — which developed the test — the team last year published a preliminary study in the Lancet that found a new computer-based diagnostic test can accurately distinguish between blood samples from women with ovarian cancer and samples from women without the disease.

The researchers, who had identified a pattern of five proteins unique to ovarian cancer patients’ blood, used the computer program to look for that pattern among the study participants who had confirmed ovarian cancer and patients who were cancer-free.

The scientists evaluated 50 women with ovarian cancer and 66 women without the disease and found that the test correctly identified all 50 samples from the cancer patients, including 18 samples from women with “Stage I” cancer, or the early form of the disease.

The test also recognized 63 of the 66 samples that were non-cancerous (Kaiser Daily Reproductive Health Report, 2/8/02).

Stony Brook University biostatistician Wei Zhu then refined the math of the algorithm; clinical trials are being organized.

Zhu estimates that the blood test could be available within a year

[Ricks, Long Island Newsday, 12/2; 3Dec03 Kaiser Daily Reproductive Health Report (not prolife) kaisernetwork.org]