The clinical behavior and course of prostate cancer is highly variable, reflecting the heterogeneity of disease. Although most men currently diagnosed with prostate cancer have low-risk, favorable disease characteristics, some may harbor aggressive features and progress despite definitive treatment. Thus, initial risk stratification is important but has been traditionally based on the conventional serum biomarker PSA as well as biopsy Gleason score and clinical stage.
The desire for personalized and precision medicine combined with the elucidation of the pathogenesis of prostate cancer have led to the development of a new generation of genomic biomarkers and tissue-based gene expression tests (table) (Curr Oncol Rep 2016; 18:30; Eur Urol 2015; 68:1033-44). These novel approaches, as explained in this article, may provide additional information to assist and improve clinical decision-making at several junctures for men with localized prostate cancer.
It is important to distinguish between prognostic and predictive markers (J Clin Oncol 2005; 23:9067-72). Prognostic markers are those that are related to a prespecified clinical outcome, usually a time-to-event outcome such as overall or recurrence-free survival. They can be used for clinical management by informing on disease aggressiveness and potential candidacy for disparate strategies like active surveillance for indolent disease and multimodal therapy for more virulent disease.
Predictive markers are those that provide information on the likely benefit from a specific treatment, and can be used to help determine which modality may be best for that individual patient. Thus, prognostic markers reflect the effects of tumor and/or patient characteristic on outcome, while predictive markers reflect the effects of treatment on outcome.
This article focuses on contemporary, commercially available molecular tests for clinically localized prostate cancer after diagnosis (figure). Several other tests are available to aid in the diagnosis of prostate cancer, including the PCA3 assay (non-coding RNA [Progensa]) and assessment of epigenetic changes (methylation [ConfirmMDx]), but are beyond the scope of our discussion. Despite the development, study, and availability of these tools, clinical decisions are often still made based only on stage, grade, and serum PSA, possibly enhanced with the incorporation of more sophisticated nomograms and models. Nevertheless, there remains a need for more refined prognostic and predictive biomarkers in prostate cancer.
After being diagnosed with prostate cancer, patients and their families most often want to know: How aggressive is the cancer? Do I need treatment? Will this cancer kill me? The Prolaris test may answer some of these questions regarding the natural history of untreated prostate cancer.
This RT-PCR assay, using prostate biopsy tissue, measures the gene expression of 46 genes (set of 31 cell-cycle progression [CCP] and 15 housekeeping genes) to determine a proliferation score. Cuzick et al found that the CCP expression panel was the strongest predictor of disease-specific survival in men with prostate cancer managed conservatively when compared with clinical parameters (Lancet Oncol 2011; 12:245-55). Based on the CCP score and clinicopathologic variables, the test report provides an estimate of 10-year prostate cancer-specific mortality risk with conservative management (Lancet Oncol 2011; 12:245-55; J Clin Oncol 2013; 31:1428-34).
This can help inform a man’s decision regarding active surveillance versus definitive treatment at the time of initial diagnosis. It should be noted that the CCP score may have more utility in those with Gleason score >7 (Lancet Oncol 2011; 12:245-55).
The Oncotype DX Genomic Prostate Score (GPS) can also be used to help answer this question. It is a multi-gene RT-PCR expression assay that is performed on fixed-paraffin embedded prostate needle biopsy tissue, measuring 12 genes across four pathways associated with prostate cancer aggressiveness (Curr Oncol Rep 2016; 18:30). The data are algorithmically combined to calculate the GPS score (0-100). The GPS predicts the aggressiveness of cancer as reflected by pathologic stage at the time of prostatectomy.
In a validation study of nearly 400 men with low- and intermediate-risk disease who were candidates for active surveillance, the GPS predicted high grade and stage on surgical pathology (Eur Urol 2014; 66:550-60). Other reports in men with very low- to intermediate-risk prostate cancer found an association between GPS and adverse pathology, biochemical recurrence, and metastatic recurrence (Eur Urol 2015; 68:123-31). Thus, the GPS and reporting of likelihood of favorable pathology (both Gleason ≤3+4 and pT2) may aid in the choice between surveillance and treatment.
ProMark is a proteomic test based on the evaluation of eight selected biomarkers via quantitative immunofluorescence in diagnostic biopsy tissue. The ProMark score is reported on a scale from 0 to 100, and has been validated to predict adverse pathology on radical prostatectomy specimens defined by Gleason score >3+4 and/or ≥pT3a (Clin Cancer Res 2015; 21:2591-600). Of note, the study population as well as the approval for Medicare coverage pertains to patients with biopsy Gleason scores of 3+3 and 3+4.
Testing of PTEN and TMPRSS2-ERG is also available through Bostwick Laboratories (ProstaVysion) and Metamark. There is ample evidence that PTEN, a tumor suppressor gene on chromosome 10q, is important in prostate carcinogenesis and PTEN loss is evaluable by either FISH or immunohistochemistry (Prostate Cancer Prostatic Dis 2016; 19:1-6). PTEN as assessed in both biopsy and prostatectomy specimens predicts prostate cancer death in men with Gleason <7 managed conservatively as well as upgrading on prostatectomy and disease progression after surgery. TMPRSS2-ERG gene fusion is common in prostate cancer and has been associated with poorer prognosis (Eur Urol 2015; 68:1033-44).
It remains to be determined whether detection of PTEN deletion and ERG rearrangement/fusion provides additional clinically useful information, how it affects treatment decision, and whether the data complement other biomarkers.
When considering definitive treatment for localized prostate cancer, patients’ relevant concern is frequently about the likelihood their cancer will return. Using either prostate biopsy or radical prostatectomy specimens, the Prolaris score (reported as a range from –3 to +3) has been validated as an independent prognostic factor for biochemical recurrence and metastatic progression after surgery. In addition, Freedland et al demonstrated that the score significantly predicted outcome, including biochemical recurrence, in men treated with external beam radiotherapy(Int J Radiat Oncol Biol Phys 2013; 85:848-53).
Thus, the Prolaris score, both before and after prostatectomy, can inform whether a patient within a specific D’Amico risk group has prostate cancer that is more (positive score) or less (negative score) aggressive than average; the test is applicable to low-, intermediate-, and high-risk groups. This information can be used alone or in conjunction with other prognostic models such as the post-surgical CAPRA-S (J Clin Oncol 2013; 31:1428-34).
The optimal use and incorporation of this information into clinical practice remain to be determined, as does the true impact of these decisions on patient outcome. However, there is evidence that the Prolaris score provides information that alters initial management. In a cohort comprised of primarily low-risk cancers, Crawford et al reported that the score reduced the overall recommendation for surgery or radiotherapy by 49% and 30%, respectively (Curr Med Res Opin 2014; 30:1025-31). Conversely, a less favorable score increased the recommendation for active treatment by 23%.
The GPS also provides additional information with respect to whether an individual’s cancer is higher or lower risk than expected within a National Comprehensive Cancer Network (NCCN) risk group, and may also alter the actual risk group stratification. It is important to remember that the outcome measure being considered is likelihood of favorable pathology.
The Decipher test, described in more detail below, may also provide information in this scenario. A study of diagnostic biopsy specimens in 57 men suggested that Decipher predicted the risk of metastasis at 10 years after radical prostatectomy (Urology 2016; 90:148-52). The Decipher biopsy report includes estimates of high-grade disease at radical prostatectomy as well as risk of metastasis at 5 years and 10-year prostate cancer-specific mortality, but further studies are necessary to validate these findings and to determine whether this information may help in selection of active surveillance.
After radical prostatectomy, the information obtained regarding pathologic stage, grade, and surgical margin status is critical in providing more accurate prognosis and guidance for additional treatments. Despite robust risk assessment tools like CAPRA-S, evidence from clinical trials, and joint guidance from the AUA and American Society for Radiation Oncology, management of the higher risk patient varies and application of postoperative radiotherapy haphazard. Can newer biomarkers help the situation? Can we identify the man who benefits from radiotherapy or the man who can be spared potential morbidity?
The Decipher test is a genomic classifier that uses an oligonucleotide microarray to profile RNA from formalin-fixed, paraffin-embedded radical prostatectomy specimens (J Urol 2013; 190:2047-53; PLoS ONE 2008; 3:e2318). It examines a 22-gene panel derived from transcriptome-wide sequencing of a prostate cancer cohort enriched for recurrence following surgical treatment and reports a score from 0 to 1. Karnes et al prospectively evaluated over 1,000 men at risk for recurrence after prostatectomy (J Urol 2013; 190:2047-53). In this cohort, Decipher stratified men into a low score group having 3 times lower cumulative incidence of metastases and a high score group having 4 times higher cumulative incidence of metastases. Decipher was an independent predictor of metastases after adjusting for postoperative therapy and other clinical variables.
In addition to metastatic progression, Decipher has been shown to predict biochemical recurrence as well as cancer-specific survival after prostatectomy. Klein et al investigated the value of adding Decipher to risk stratification tools such as CAPRA-S and the Stephenson nomogram in predicting metastatic disease (Eur Urol 2015; 67:778-86). In multivariable analysis, Decipher was the only variable associated with metastatic progression.
The clinical application of Decipher remains to be clarified, but it does appear to provide additional post-prostatectomy risk stratification. Retrospective data suggest that the test can be used to determine the need for adjuvant treatment after radical prostatectomy. In those men with low scores (<0.4), there was no difference in development of metastases after adjuvant or salvage radiotherapy. However, men with scores ≥0.4 treated with adjuvant radiation had significantly better outcomes when compared with salvage radiation-6% versus 23% cumulative incidence of metastases at 5 years (J Clin Oncol 2015; 3398:944-51).
Thus, the Decipher test not only appears to provide information regarding future clinical behavior, but also may help to select men most likely to benefit from early intervention after prostatectomy.
It recent years, progress has been made in the diagnosis of prostate cancer as well as judicious application of treatments. Nevertheless, molecular-based tools are necessary to improve choices and outcomes for individual patients. Evidence suggests that many of the tests provide prognostic information on both the natural history of untreated disease as well as after surgery or radiation. It remains to be determined whether any biomarker can help predict response to initial treatment modalities. After surgery, only the Decipher test currently helps select patients who will benefit from adjuvant therapy.
Although the current NCCN Prostate Cancer Guidelines (v2.2017) mention tissue-based tests and summarize the evidence, no clear recommendations regarding incorporation into routine practice are provided. This is in stark contrast to the breast cancer guidelines, which integrate molecular tests including HER-2 and Oncotype DX. The goal should be to combine clinicopathologic, molecular, and imaging data along with patient preferences into a unified tool to assisted decision-making along the entire spectrum of prostate cancer states.
Subscribe to Urology Times to get monthly news from the leading news source for urologists.