PCA3, PSA, and free PSA in prediction of prostate cancer risk
Posted Dec 01 2010 12:00am
A new study has suggested that, although the PCA3 test is slightly better than the total PSA test in prediction of risk for a positive biopsy finding, it is not necessarily always better than the free PSA test.
Aubin et al. conducted an “add on” study among patients in the placebo arm of the REDUCE trial i.e., among the men who did not receive dutasteride. Their study was designed to validate the diagnostic accuracy of the PCA3 test as an indicator of current and future biopsy results and to compare that diagnostic accuracy to the accuracy of total and free PSA data.
REDUCE included 4,000+ men in the placebo arm of the trial. Of these 4,000+ men, 1,140 men provided PCA3 samples prior to their biopsies in year 2 and/or year 4 of the trial. PCA3 samples were not, however, collected at baseline or at the time of “for-cause” biopsies.
Here are the results reported by Aubin et al.:
PCA3 samples from 1,072/1,140 men were adequate for analysis.
PCA3 scores ranged from 0.40 to 287.5, and exceeded the cutoff value of 35 in 26.3 percent of men.
Median PCA3 scores were 15.9 at year 2 and 18.0 at year 4 in subjects tested at both time points.
Prostate cancer was detected (using 10-core biopsies) in 190/1,072 men (17.7 percent) during the 4-year follow-up period.
Most tumors detected had Gleason scores of ≤ 7.
Men with a negative biopsy at the 2-year and 4-year time points had median PCA3 values of 15.3 and 16.9, respectively.
Median PCA3 scores were statistically higher at years 2 and 4 in men with a positive biopsy result at year 4.
PSA levels increased with increasing prostate volume; PCA3 scores did not.
Median PCA3 scores were 33.8 for men with a positive biopsy results and 16.7 for those whose biopsy results were all negative.
Using the cutoff value of 35, the sensitivity and specificity of the PCA3 test were 48.4 and 78.6 percent, respectively.
There was a 3.5 fold increased risk of biopsy-detectable prostate cancer among men with PCA3 scores above the cutoff of 35.
PCA3 at year 2 was a significant predictor of biopsy outcome at year 4, whereas total serum PSA and percent free PSA were not predictive.
The differences between the predictive ability of the PCA3 test,the total PSA test, and the percent free PSA test were not consistently significant, although the PCA3 test was consistently more accurate than the total serum PSA test.
Based on these and other data, the authors conclude that although PCA3 data represent a major contribution to the ability to predict the likelihood of a positive biopsy, serum PSA data and other factors “provide additional information that refines risk assessment.”
It would seem to The “New” Prostate Cancer InfoLink that the actual value of the PCA3 test has become increasingly limited as we have learned more about this test. Rather like the percent free PSA test, it can help to distinguish between risk for benign and cancerous prostate conditions, but its ability to do this is limited, and its sensitivity and specificity are probably not sufficient to allow this test to replace the PSA test as first-line test for prostate cancer risk (or not at the current cost of the test). It’s role is therefore likely to evolve primarily as a second-line test for men with an elevated PSA that continues to risk after an initial negative biopsy.
In theory it might be possible to develop an accurate predictive model for risk of a positive biopsy based on a patient’s age, total PSA level, percent free PSA level, PCA3 result, and other factors that could more accurately predict the likelihood of a positive biopsy than the current prostate cancer calculator. However, the clinical value of such a model is open to question at this time. The biggest problem with such a model is that is a model designed to predict the risk of finding prostate cancer cells in a man’s prostate which we now know is not necessarily at all the same as the risk of finding clinically significant prostate cancer.