| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | COVER ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
||||||||
|
||||||||||
NEWS & VIEWS |
|
Applying the new test to serum samples from 50 women with ovarian cancer, including 18 Stage I cases, and 66 controls with benign reproductive system disorders or without reproductive system pathology yielded a sensitivity of 100 percent, specificity of 95 percent, and a positive predictive value of 94 percent.
The overall five-year survival rate for all stages of ovarian cancer combined is 52 percent. Although the five-year survival rate for women with localized ovarian cancer is 95 percent, only 26 percent of cases are diagnosed at that stage. No screening tests are currently recommended for women at average risk of ovarian cancer because none have sufficient specificity to avoid the large number of procedures that would follow false-positive results.
Tests such as serum CA125 levels and transvaginal ultrasonography are often recommended for women at increased risk because the predictive value of these tests improves as the prevalence of disease in the screened population increases.
"We applied this technology to ovarian cancer because of the tremendous clinical need," said first author Petricoin, a senior investigator and co-director of the joint FDA/NCI Clinical Proteomics Program.
 |
A New Paradigm in Screening |
|---|
 TOP A New Paradigm in... |
|---|
In contrast, the new test recognizes changes in the expression pattern of over 15,000 peptides and small proteins. "It’s the pattern—not the proteins—that reveals the disease," noted Lance A. Liotta, MD, co-director of the FDA/NCI Clinical Proteomics Program.
To develop the test, the investigators fed mass spectroscopy data from a "teaching set" of serum protein samples into a computer. The samples were from 50 women with ovarian cancer and 50 without the disease. The computer analysis of mass spectroscopy data recognized thousands of peaks representing variable amounts of certain kinds of proteins, each with a characteristic position based on its mass to charge ratio—a kind of "signature line" for each sample.
An artificial intelligence program then "trained" the researchers’ software to recognize signature patterns present only in those with ovarian cancer. After this training, the software was then used to analyze mass spectroscopy data from the second set of samples obtained from women with or without ovarian cancer.
The test is the first of many the group hopes to develop using this strategy. Other tests will target cancers of the breast, lung, prostate, and colon, the researchers said.
"This is a very exciting new development—the first quantum leap in testing for this disease in 20 years," said Carolyn Runowicz, MD, Professor at Albert Einstein College of Medicine in New York City and Vice-Chair of Obstetrics and Gynecology at St. Luke’s-Roosevelt Hospital.
Runowicz, a member of the American Cancer Society Gynecologic Cancer Advisory Board, said more work still needs to be done to help lower—or clarify—the rate of false-positives, because a false-positive causes unnecessary anxiety for the patient, and in some cases may lead to more invasive procedures.
Although the 5% false-positive rate sounds low, said Runowicz, that percentage among the millions at average ovarian cancer risk who might potentially be screened could affect large numbers of women. And because the study was a small one done among women at high risk of developing ovarian cancer, studies involving more women at average risk are needed to learn if it is as useful for them, she said.
"But this is a test with real potential to make a major difference in the fight against ovarian cancer," she said.
Petricoin and colleagues mention in their article that, in fact, studies of the test in low- and high-risk women are currently in progress. They are also studying how to use other screening procedures together with the proteomic test to develop a screening and work-up algorithm that avoids too many false-positive results. Another optimistic note: the authors suggested that the new method is amenable to automation and cost-effective, high-throughput testing.
| ||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | COVER ARCHIVE | SEARCH | TABLE OF CONTENTS |
