The question had come from his boss. It was around the beginning of 2017, and the director of research for the Brady Urological Institute at Johns Hopkins asked Dr. Gonzalo Torga, a postdoctoral fellow, which of two genomic tests they should offer to their prostate cancer patients as a way of determining eligibility for certain drugs. Torga wasn’t sure, so his boss suggested they test the tests: Order a liquid biopsy from both companies for a number of patients, and then compare the results.
Their findings, the Hopkins scientists say, were disconcerting: In many cases, the two tests gave conflicting results for the same patient. They published their comparison in JAMA Oncology in December, arguing that they needed to warn other clinicians that these tests don’t give standard results. But the companies in question retort that this warning is based on sloppy study design and could mean that patients don’t get offered potentially helpful tests.
Others say that the whole episode shows a need for better clinician training — and regulation — of a kind of test that has become a hot, billion-dollar topic in oncology even as critics ask whether sales are outpacing the science. They point out that it’s hard enough to create a test that works well for a specific kind of cancer, let alone one that is being marketed, as these are, for a whole whack of different tumors.
About the Author Reprints
I’ve worked in a hospital lab doing analysis.
Motto: “Never trust just one analysis result”.