The Wall Street Journal is carrying a story about the potential inaccuracy of breast cancer tests, specifically tests for drugs like Herceptin. This is an issue that has always troubled me. It gets even further amplified given the number of “home brews” around. When a new test is released, is it compared to a baseline/gold standard? Should a test be offered if it is only a few percentage points better than existing tests (but often significantly more expensive). It is the reason the FDA is wrestling with the issue of diagnostic approval and what needs to be tested and how, especially classifier. These are all questions we need to answer before companion diagnostics can become more common.
That said, I did find the tone of the article somewhat troublesome.
recent studies that turned up problems in testing point to a potential snag for such drugs: They depend on accurate lab results.
Duh!!! That’s the entire point. All tests will have a false positive and false negative rate, and no test should be done in a vacuum, not yet anyway. What we need to do is make sure that people taking the tests are properly trained and appropriate quality control is followed.
Further Reading
Steve Murphy
Technorati Tags: Personalized Medicine, Diagnostics, Breast Cancer
One Trackback
[...] Deepak Singh at Business|Bytes|Gene|Molecules writes really deep-minded posts these days, the one I chose now is Your Personal Health: A little warning. [...]