The search for better diagnostic tests for prostate cancer
Posted Mar 06 2010 12:00am
Three papers presented on Friday at the Genitourinary Cancers Symposium in San Francisco addressed the near-term development of tests that may be able to improve our ability to accurately diagnose clinically significant prostate cancer.
In the first of these papers, Aubin et al. presented data on a urine-based assay of messenger RNA (mRNA) permitting measurement of levels of TMPRSS2:ERG, which is a highly specific gene-fusion biomarker. (This particular test is being developed by Gen-Probe, Inc.) Urine was collected from 669 patients immediately following a digital rectal examination, and 635/669 urine samples gave sufficient mRNA for analysis. The authors were able to show that TMPRSS2:ERG was potentially superior to two current methods of predicting prostate biopsy outcome:
Serum PSA testing
The Prostate Cancer Prevent Trial-based risk calculator
In addition, they showed that higher TMPRSS2:ERG scores were associated with clinically significant as compared to clinically insignificant prostate cancers as defined using the Epstein criteria. And finally, in 101/635 patients who went on to have radical prostatectomy, Aubin et al. demonstrated that higher TMPRSS2:ERG scores were associated with:
Pathological upgrading from Gleason scores of 6 on biopsy to Gleason 7 based on post-surgical pathology
Higher post-surgical Gleason grade (> 6 vs. ≤ 6)
Higher tumor volume
In a second paper, also addressing the development of a TMPRSS2:ERG-based diagnostic test (under development by Dianon Systems), Day et al. described the enrollment of 471 patients, all of whom had been referred for prostate biopsy at one of seven community-based urology practices. In this study, post-DRE urine was collected from all 471 patients, all patients went on to have a prostate biopsy, and 463/471 patients were found to have enough mRNA in their urine sample for assessment of their TMPRSS2:ERG level. Most of the patients (453/471) were also tested for their PCA3, normal serum PSA, and free PSA levels, and 199/453 patients (44 percent) had a positive biopsy.
Day et al. were able to show that:
Their TMPRSS2:ERG prototype test was highly specific for finding prostate cancer on biopsy (89 percent specificity), but it was not very sensitive (only 39% sensitivity).
Combining the results of their TMPRSS2:ERG test with other information (including PCA3, serum PSA, percent free PSA, age, prostate volume, family history, and race) significantly improved predictive accuracy for a positive biopsy
Finally, Groskopf et al. presented data demonstrating the ability of the PCA3 test to predict outcomes of repeat biopsies based on patients in the placebo arm of the REDUCE trial. In this study, the authors collected urine samples from 1,140 patients who had had a negative baseline biopsy, who received a placebo (as opposed to active dutasteride) during the clinical trial, and who were scheduled for additional biopsies at year 2 and/or year 4 of the trial.
Groskopf et al. showed that 1,072/1,170 patients provided enough RNA for PCA3 assay, and that 18 percent of those 1,072 patients had a positive biopsy. They further showed that:
PCA3 levels were highly associated with a positive biopsy rate (in patients with PCA3 < 5, only 6 percent had a positive biopsy, whereas in patients with PCA3 > 100, 57 percent had a positive biopsy).
PCA3 level was highly correlated with biopsy Gleason score (i.e., Gleason score < 7 vs. ≥ 7)
PCA3 levels determined at year 2 of the study were significant predictors of biopsy outcome at year 4 of the study.
What can we make of all of this?
Well it seems pretty clear that the addition of PCA3 and TMPRSS2:ERG data to other information (PSA levels, race, ethnicity, etc.) can improve the predictive accuracy of biopsy outcomes. However, as yet, we do not know whether these tests are predictive for actually clinically significant disease. In other words, we still do not have a test that can tell a doctor and his patient that, “If we do not treat you now, you will have progressive prostate cancer within X years time.” While The “New” Prostate Cancer InfoLink does not in any way deny the value of tests that can help us to determine whether a biopsy is or is not a good idea (because such tests are certainly helpful), there is still a long way to go before we have the test that we really need.