An article just published on the Medscape web site addresses data on the potential of GenomeDx Bioscence’s new “Decipher” test which “could” (according to one urologic oncologist) help prevent over-treatment of at least some men with prostate cancer. One might ask whether we can — as yet — decipher the value of the Decipher test.
The article in Medscape Oncology is based on and discusses data from two more poster presentations at the recent Genitourinary Cancers Symposium in Orlando — one by Cooperberg et al. and the other by Badani et al.
The new Decipher test is very similar to the OncotypeDx genomic test that has been used for a while to assess the aggressiveness of different types of breast cancer. Basically, both tests provide a method for risk stratification based on biologic material in the tumor tissue. To date, the Decipher test has only been available to select physicians and their patients as part of clinical trials, and these trials are continuing at the present time. However, GenomeDx Bioscience hopes to be able to make their new test more widely available later this year.
The test assesses the levels of expression of 22 markers associated with aggressive prostate cancer, and — so far — uses samples of paraffin-embedded tumor tissue (removed during prostatectomy) on which to carry out the assay. The assay outcome is a percentage score that, according to the company, indicates whether a man is at high or low risk for metastatic disease, and therefore whether he is a candidate for further treatment after initial surgery, such as radiation and/or androgen-suppression therapy.
According to the presentation by Cooperberg and his colleagues in Orlando, in men who had previously been treated by radical prostatectomy, and who were known to be at high risk of recurrence based on standard clinical and pathologic variables, both the Decipher genomic classifier percentage score (DGC) and the older CAPRA-S score demonstrated value as significant predictors of prostate cancer-specific mortality. However, the DGC was capable of further “down-risking” > 50 percent of men stratified as high risk based on the CAPRA-S alone. In other words, they said, the DGC can offer independent prognostic information, and — potentially — a prognostic model that integrates data from the DGC and the CAPRA-S scores may be capable of improving our abilities and accuracy in the prediction of metastatic and lethal forms of prostate cancer.
The poster by Badani et al. reported that patients’ DGC scores appear to influence physician’s treatment recommendations and their confidence in their decisions regarding the management of high-risk prostatectomy patients. In other words, their data suggest that clinical use of data from the Decipher test may be able to impact treatment recommendations.
Now we do want to draw your attention to that little word “may” in reports of the conclusions of both of the above studies. What both studies did was to look at data based on men for whom decisions had already been made and then try to see whether data from the Decipher test correlated with later outcomes. In other words, if men had not had adjuvant radiation and/or ADT and still did well, did the Decipher test also predict this. Conversely, could the Decipher test also be used to project (with accuracy) which men would not do well without adjuvant radiation and/or ADT?
Based on these studies, it does appear that the Decipher test has potential. However, at present that potential appears to be limited to the management of men who started out as high-risk patients (based solely on “traditional” methods such as the D’Amico risk profile or their CAPRA-S score), went on to have first-line radical prostatectomy, and then need to be assessed (post-surgery) for whether they also need adjuvant radiation therapy (with or without ADT). However, it would be nice to see the currently available data being corroborated by a true, randomized, double-blind, prospective study in which men are given or not given this test and then assigned to either further treatment or to simple monitoring so that we could compare the outcomes of their management on a truly prospective basis in clinical practice.
Widespread application of tests like the Decipher test without such supportive data from truly prospective clinical trials can all too easily lead us down paths along which we see what we expect and want to see … sometimes even when what we think we are seeing simply isn’t there at all. This has been the problem with the PSA test, which works extremely well as a test to assess the actual progression of prostate cancer in men who have been treated, but much less well as a prognostic tool in helping to make decisions about which men should actually receive treatment in the first place.
Another question regarding the potential of the Decipher test is going to be this one. If it can, in fact, be used with accuracy to predict which men really do need second-line treatment after first-line surgery, then can it also be used with accuracy to increase our ability to determine which patients can be managed with active surveillance and which men really do need first-line treatment of some type (based on tissues from biopsy cores)? If it could, then that would make it a test with considerable value