This discovery phase formed the basis for the next round of tests, in which 59 prostate cancer samples and 70 control samples were tested against the 186 autoantibodies. In this phase, the researcher identified a panel of 22 compounds that best distinguished the prostate cancer blood samples from the controls. Using these 22 markers, only two of 70 controls incorrectly tested positive for prostate cancer, and seven of 59 prostate cancer samples were falsely negative.
Next, the researchers validated their findings using the remaining 128 blood serum samples. They found eight of 68 controls and 11 of 60 prostate cancer samples were misclassified. This means 88 percent of the time, samples that were not cancerous were correctly identified and 81.6 percent of the time, samples that were cancerous tested positive.
"These 22 biomarkers appear to be the right number. If you used too many or too few, the accuracy went down a bit. Our findings held up when we tested the model on an independent set of blood serum samples," Chinnaiyan says.
The results proved to be more reliable at predicting cancer than prostate specific antigen, which is a single biomarker. PSA testing results in a false positive around 80 percent of the time, leading to unnecessary prostate biopsies. The normal range for the PSA test is less than 4.0 nanograms per milliliter (ng/mL) in most men. For men over 40 years old with a family history of prostate disease or for African-American men over 40 years old, some doctors suggest that a level higher than 2.5 ng/mL should be checked with more tests, because these two groups of men have an increased risk of prostate cancer.
The 22-biomarker test was reliable at identifying prostate cancer even in the PSA ranges of 4-10 ng/ml or 2.5-10 ng/ml, intermediate PSA scores that do not always suggest cancer. The study authors suggest the 22 biomarkers could be used f
Source:University of Michigan Health System