Breast Cancer in the News: How Good Are the Tests?
The drug Herceptin has been remarkably effective at treating Her2 positive breast cancer, reducing the risk of death by up to 40 percent. But the tests to detect Her2, as well as those used to find estrogen receptor and progesterone receptor, are, in a word, unreliable, offering false positive or negative rates of up to 20 percent. That is partly because there is no uniform test applied and the FDA has not been able to regulate the testing adequately. The College of American Pathology (CAP) and the American Society of Clinical Oncology (ASCO) issued Her2 testing guidelines and have recently come out with estrogen receptor testing guidelines as well.
How big a problem is this and what is being done about it? We asked some of the Alliance Medical Advisory board members and here is what they had to say.
K.M. Steve Lo, MD, Oncologist, Bennett Cancer Center
There is a definite error rate in Her2, ER, and PR testing. Oncologists rely very heavily on these results to recommend the best therapy. Fortunately, most major hospitals in this area, including Stamford Hospital, follow the CAP guidelines, so, in the past five years, I have not seen any cases where the ER, PR, and Her2 testing done in Stamford Hospital differed from independent testing from other major institutions or conflicted with reverse PCR results used in Oncotype DX testing. However, if a patient is coming to me for a second opinion from another community hospital, I often have our pathology lab independently confirm the ER, PR, Her2 testing to ensure that it was done properly.
David L. Rimm, MD/PhD, Yale University School of Medicine
In my opinion, this is THE ELEPHANT in the room. Due to inconsistent standards in pathology labs and variable quality tissue submitted for testing, there may be false negative rates as high as 20 percent for both Her2 and ER/PR testing. That means as many as 15 to 20 percent of women may be undertreated! The topic of laboratory standardization of tests is a major issue in the pathology societies especially as new tests come online. For example, if we have these problems with simple single analyte tests, how are we to monitor and quality control tests that have 21 (Oncotype Dx) or 70 (Mammoprint) analytes? This topic is likely to get a lot more attention before it goes away. Hopefully, it will “go away” over the next few years, driven by CAP-based guidelines, proficiency testing and standardization.
Barbara A. Ward, MD, Medical Director, Breast Center, Greenwich Hospital
As a surgeon, I am asked to do my part in seeing that tissue samples are placed in formalin as quickly as possible for proper fixation. The results may be examined via the Hercept test, followed by a FISH analysis for confirmation, that may be reanalyzed via the Oncotype or Mammoprint test. Oncologists are typically looking for correlation in results. For example, it would be unusual for a low grade cancer with favorable features to be Her2 positive. In that case, we would want a Hercept result confirmed by a FISH analysis. If an outside lab had a negative Hercept test on a core biopsy for a high grade cancer, we would typically repeat FISH testing on the surgical specimen once the cancer has been removed.
To read more, see Cancer Fight: Unclear Tests for New Drug, by Gina Kolata, New York Times,
April 19, 2010
