Genetic classifiers for prostate cancer: A new era on the horizon?

Since the introduction of the prostate-specific antigen (PSA) test, there has been a clear shift towards diagnosis at an earlier stage of prostate cancer. However, patients who receive curative treatment in the contemporary era still face significant risks of recurrence and disease progression. In most cases of tumor recurrence, salvage treatment is performed. As can be seen from general clinical experience, not all forms of salvage therapy guarantee control of disease status or cure. Although debate continues on the benefit of adjuvant therapy in comparison with salvage therapy, it cannot be denied that there may well be a proportion of patients who would benefit from earlier adjuvant intervention rather than salvage therapy given the identification of disease recurrence or progression. Thus, a definite need exists for more accurate predictors of the prognosis and outcome of prostate cancer. It is now clear that PSA and other PSA-related markers are not enough to do the job. Even though the Gleason score is widely regarded as a powerful predictor of outcome of prostate cancer, the possibility of intra- and interobserver variability exists. Certainly, the limitations in prognostications of prostate cancer patients are the primary reason for over- and undertreatment.

To overcome such limitations, several genetic classifiers have recently been introduced. Two new molecular classifiers (Prolaris, Myriad Genetics Inc., UT, USA and Oncotype DX, Genomic Health, CA, USA) are now available in the United States for the selection of appropriate candidates for active surveillance [1, 2]. More recently, a 22-marker genomic classifier (Decipher, GenomeDx, CA, USA), which can be used to identify patients with more aggressive prostate cancer who could benefit from earlier adjuvant therapy after initial radical prostatectomy, has become available for clinical application as well [3]. So far, published studies on these new tools have demonstrated promising results, providing hope for eradication of over- and undertreatment. If situations regarding these new "gadgets" continue to evolve in a positive way, we may someday be able to mention that these new genetic tools have opened up a new era in the management of prostate cancer, just as we did with the PSA test.

Meanwhile, before making such a bold prediction, we must remember a few issues. First, prostate cancer is not generally considered a genetic disease. Although some patients have a family history of prostate cancer, it is still not clear whether these cases are any different from other nonfamiliar prostate cancers. Do we treat a prostate cancer patient any differently today because of his family history or race? I say no or not yet. For objective evaluation and further development of genetic tools, we need to first elucidate the actual role of genetics in the carcinogenesis and pathogenesis of prostate cancer. Second, studies on the usefulness of these new genetic molecular tools have mostly relied on historical data and pathologic specimens from prostatectomies performed in the past. We need to see the performances of new genetic tools in action via larger-scale, multi-institutional prospective trials. Third, as with many other new tools in development, the issue of cost should be considered. The aforementioned genetic classifiers cost about United States dollar 3,000 to 4,000 per case (higher than the surgeon's fee for performing radical prostatectomy in Korea), which would add significantly to the medical cost. Because the economic situation is not great in all regions of the world, it would be difficult or impossible for any such new tools to receive coverage from government-sponsored medical insurance system without demonstrating a "game-changing" impact. More evidence is needed for objective evaluation of new genetic classifiers. Also, the cost for new tools must come down for widespread clinical usage.

To spare many men from unnecessary treatment and to improve the chances of cure for others, a breakthrough in the stratification of prostate cancer is greatly warranted. For a personalized approach to help the cause, genetic classifiers sound almost perfect. On the other hand, today, we still don't have a clear picture of how much genetics contributes to the pathogenesis of prostate cancer. Furthermore, new genetic classifiers do not provide black-and-white results. They only give risk scores or probabilities, which can again cause dilemmas for patients and physicians. Although it may not seem fancy, evidence-based medicine usually does no harm. More information on the actual roles of various genomic and molecular markers and pathways in prostate cancer pathophysiology is needed to support the application of genetic tools in the management of prostate cancer. Simply speaking, more "beef" is needed to really start up a new era.