Food and Drug Administration
NCI Press Office
Using a test that can be completed in 30 minutes using blood that can be obtained from a finger stick, researchers were able to differentiate between serum samples taken from patients with ovarian cancer and those from unaffected individuals.
The research, a joint effort between the FDA/NCI Clinical Proteomics Program and Correlogic Systems Inc., unites two exciting disciplines: proteomics the study of the proteins inside cells and artificial intelligence computer programs.
The diagnostic test relied on software that is able to detect patterns of key proteins in the blood. Using a sophisticated artificial intelligence computer program developed by Correlogic, scientists were able to "train" the computer to distinguish between patterns of small proteins found in the blood of cancer patients vs. control samples. The artificial intelligence program identified a pattern consisting of only a handful of proteins, among thousands, that could be used to distinguish between women with ovarian cancer and women with non-cancerous conditions.
"The idea that rather than a single biomarker, an entire pattern of proteins contains important diagnostic information, is an exciting new paradigm," said J. Carl Barrett, Ph.D., director of the NCI's Center for Cancer Research, which oversees the proteomics program.
The scientists first used serum samples from known cancer patients and unaffected individuals to establish proteomic patterns which were present at different levels in the two groups. Once these patterns were identified, the researchers compared them with the patterns of the same proteins in serum samples from other patients with and without cancer. The researchers correctly identified 50 out of 50 cancers and 63 of 66 non-cancer samples.
The researchers analyzed the serum proteins with mass spectroscopy, a technique used to sort proteins and other molecules based on their weight and electrical charge. The identity of the key proteins and the role they may play in cancer is unknown, but being investigated.
An important finding was the ability to correctly identify, in a small sample of patients, all stage I ovarian cancer cases. Currently, more than 80 percent of ovarian cancer patients are diagnosed at a late clinical stage and have a 20 percent or less chance of survival at five years. In contrast, the 20 percent of women diagnosed with early-stage disease have an excellent prognosis, with over 95 percent alive at five years after diagnosis. The results of this study indicate that proteomic technology may help clinicians diagnose the disease much earlier than current methods.
The authors of the paper caution that further study is needed to confirm the sensitivity and accuracy of this technique as a diagnostic tool. They hope that by combining the proteomic approach with other methods of ovarian cancer diagnosis, such as ultrasound, its accuracy can be further improved.
"Simple, accurate, and non-invasive methods for early detection of epithelial ovarian cancer may improve quality of life and survival and reduce unnecessary suffering for patients," said Kathryn Zoon, director of the Center for Biologics Evaluation and Research (CBER) at the FDA. The majority of the individuals included in the study had an increased risk of ovarian cancer, due to a family history of the disease, or mutations in BRCA1 or BRCA2 genes, which increase risk for both breast and ovarian cancers. Researchers on the study from Northwestern University Medical School, Chicago, considered it important to test the method in this population, as these are the women most in need of effective screening options. "The most important next goal is validating the promise of these results in large, multi-institutional trials. Early detection means we can treat the cancer before it has spread," said Lance Liotta, M.D., Ph.D., the senior investigator on the study from the NCI's Center for Cancer Research. Such trials are under way at the NCI, evaluating proteomics both alone and in combination with current screening methods for ovarian cancer.
"We're particularly excited about the potential of this technique to diagnose additional types of diseases. It may also be able to provide an early warning of impending toxicity," said the first author of the study, Emanuel Petricoin, Ph.D., of the FDA's CBER.