 |
| Image courtesy Johns Hopkins – Credit: Elizabeth Cook and Kaitlin Lindsay |
The team at Johns Hopkins University of School of Medicine published its findings in the Jan. 18, 2018 journal Science. The test, called CancerSEEK simultaneously evaluates the levels of eight cancer proteins and the presence of cancer gene mutations from circulating DNA in the blood. Five of the cancers currently have no screening test.
The investigators explored several hundred genes and 40 protein markers for cancers, and methodically narrowed the number of potential markers down to 16 genes and eight proteins that were strongly linked to cancers.
For the study the researchers evaluated these 16 markers in 1,005 patients with a variety of cancers including ovarian, liver, stomach, pancreatic, esophageal, colorectal, lung and breast cancers. The median sensitivity, meaning the ability of the test to correctly identify the cancer type was 70 percent, in other words the test correctly identified the cancer type 70 percent of the time. The test’s accuracy ranged from a high of 98 percent for ovarian cancer and a low of 33 percent for breast cancer.
The other measure of accuracy for such tests is specificity, meaning how often the test incorrectly identified a patient as having cancer when they did not. The CancerSEEK test had a specificity of 99 percent, meaning it only indicated cancer in 1 percent of the tested people who did not have cancer.
The study follows the findings of a similar test by Dutch researchers last November published in the journal Cancer Cell. That test identified cancers of the lung, breast, pancreas, brain, liver, colon and rectum 71 percent of the time.
Both blood tests need to be tested in populations for whom the cancer status is unknown to determine how reliable they are as a cancer test.
No comments:
Post a Comment