9342-8530 Quebec Inc DGCRF

RE: New study

[CUR2004]

A three-marker algorithm for projecting prostate cancer risk, need for biopsy
Posted on June 9, 2011 by Sitemaster

Rate This
A new study has suggested that a clinical algorithm that combines data on serum PSA with detection of TMPRSS2:ERG fusion and prostate cancer gene 3 (PCA3) in urine collected after a digital rectal exam (DRE) can predict prostate cancer on subsequent biopsy better than any one of the three potential markers alone.
Salami et al. collected urine from 48 consecutive patients at two medical centers after they had been given a DRE but before prostate biopsy at 2 centers. These urine samples were used to detect PCA3 levels and TMPRSS2:ERG fusion transcript expression. Serum PSA was measured by clinical assay.
Here are the results of the analysis:
The urine samples provided informative RNA specimens in 45/48 patients.
Prostate cancer was present on subsequent biopsy in 15 patients (whether this was 15/45 patients with good urine-derived data or 15 of the total 48 patients is not clear from the abstract).
The presence of TMPRSS2:ERG in post-DRE urine was strongly associated with prostate cancer (odds ratio [OR] = 12.02; P < 0.001).
PCA3 had the highest sensitivity in predicting prostate cancer diagnosis (93 percent).
TMPRSS2:ERG had the highest specificity in predicting prostate cancer diagnosis (87 percent).
TMPRSS2:ERG had the greatest discriminatory value in predicting prostate cancer.
Combining data from serum PSA, PCA3, and TMPRSS2:ERG in a multivariable algorithm optimized for clinical utility optimized the prediction of a positive biopsy compared to any one of the three markets alone.
The authors conclude that use of a clinical algorithm to recommend biopsy for all patients with PSA ≥ 10 ng/ml, while restricting biopsy among those with PSA <10 ng/ml only to men with detectable PCA3 or TMPRSS2:ERG in post-DRE urine, performed better than use of any one of the individual biomarkers alone in predicting the presence of prostate cancer.
The “New” Prostate Cancer InfoLink would note, however, that the proposed algorithm still tells us nothing about the ability to differentiate between clinically significant and clinically indolent forms of prostate cancer. The value of the proposed algorithm at this time is restricted to the elimination of unnecessary biopsies. This is a worthy goal in itself, but only addresses part of the problem. In addition, the proposed algorithm clearly needs to be tested prospectively as opposed to retrospectively before it could be applied to clinical decision making on an ongoing basis.