I have posted a number of previous notes about multiplexed biomarkers and also a note about the OvaSure test that was developed at Yale in collaboration with LabCorp (see: Encouraging Results from LabCorp/Yale Multiplex Lab Test for Ovarian Cancer). These multi-test panels that require an algorithm to interpret the results are referred to by the FDA as IVDMIAs. The OvaSure test consists of assays for leptin, prolactin, osteopontin, insulin-like growth factor II, macrophage inhibitory factor, CA-125, plus a calculated risk index. It has now available for ordering by physicians and a recent article in the New York Times (see: Cancer Test for Women Raises Hope, and Concern) discusses some of the controversy that has been swirling around it:
...many experts are saying [that OvaSure] might do more harm than good, leading women to unnecessary surgeries. The Society of Gynecologic Oncologists almost immediately issued a statement saying it did not believe the test had been validated enough for routine use...OvaSure did not go through review by the Food and Drug Administration because the agency generally has not regulated tests developed and performed by a single laboratory [so-called home-brew tests], as opposed to test kits that are sold to laboratories, hospitals and doctors. (All OvaSure blood samples are sent to LabCorp for analysis.) But the F.D.A. has now summoned LabCorp to discuss OvaSure, saying the data appear insufficient to back the company’s claims about the test. ....Dr. Myla Lai-Goldman, chief medical officer of LabCorp, said that OvaSure had been validated in several studies and that additional data were expected by the end of this year....In a study published in the journal Clinical Cancer Research in February, the test correctly classified 221 of 224 blood samples taken from women with ovarian cancer or from controls. It identified 95 percent of the cancers, and its false positive rate — detecting a cancer that was not there — was 0.6 percent....The biggest concern is not that the test will miss cancers but that it will say a cancer is there when it is not....Dr. Lai-Goldman at LabCorp said that OvaSure should be restricted to women at high risk of ovarian cancer and that the test should be repeated if the result is positive. Those measures would limit the number of false positives. LabCorp estimates that there are 10 million women at high risk. These include carriers of mutations in genes called BRCA1 or BRCA2, as well as women with histories of ovarian or breast cancer.
I believe that there is much to be learned about multiplexed tests from this early experience with OvaSure. I found one LabCorp document on the web (see: OvaSure For Women at High-Risk for Ovarian Cancer) emphasizing that the test should be restricted to high-risk women. Interestingly enough, this document does not define "high risk patient" with the same clarity as the NYT article above. Here is an excerpt from this same document relating to the sensitivity and specificity of the test (see: Defining Test Sensitivity and Specificity for the General Public):
In a clinical study across all disease stages, the six-marker panel ...demonstrated a sensitivity of 95.3% and a specificity of 99.4% in detecting disease. Greater than 99% sensitivity (119 of 120) was shown in late-stage disease (stage III and stage IV).1 In early stage disease (stage I and stage II), the assay demonstrated a sensitivity of 91.6%, providing a significant improvement over CA-125 alone (less than 60% of stage I and stage II combined) for ovarian cancer detection. All positive readings should be retested on a new sample drawn at least three weeks after the original sample was collected. Patients with positive results confirmed by retesting on a second sample should be followed by a women's health specialist who may order additional evaluations, such as sensitive imaging.
This OvaSure controversy has left me wondering about two things. First of all, to what extent have physicians been ordering this test inappropriately, either because of pressure from those patients anxious about ovarian cancer or because of a lack of understanding that the test should be restricted to high risk women? Secondly, I wonder to what extent LabCorp salespeople have been subtlety pushing the test beyond the officially recommended indications during physician office calls in order to increase sales. For some additional reading on OvaSure and the FDA, you can link to a note in Libby's Hope, a blog for ovarian cancer survivors (see: Lost In Translation? FDA Believes That LabCorp’s Ovarian Cancer Early Detection Test (OvaSure) Lacks Adequate Clinical Validation).
Comments