|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Originally published as JCO Early Release 10.1200/JCO.2008.20.9783 on May 26 2009 © 2009 American Society of Clinical Oncology.
Epigenetic Events Highlight the Challenge of Validating Prognostic Biomarkers During the Clinical and Biologic Evolution of Prostate CancerInstitute for Genome Sciences and Policy, Duke University Medical Center, Durham, NC The association of molecular events with prostate cancer behavior presents dual opportunities for increased molecular insight and improved care. The promise of molecular oncology relies on the belief that optimal clinical care will result from a comprehensive understanding of the molecular pathogenesis of cancer and the specific genetic events present within an individual's tumor. In this issue of Journal of Clinical Oncology, Richiardi et al1 have associated methylation of APC with poor prognosis in two independent cohorts of prostate cancer patients. The strong association between epigenetic silencing of APC and poor prognosis additionally supports a role for WNT/β-catenin signaling in aggressive prostate cancer and introduces a novel prognostic biomarker with which clinicians may be able to better anticipate prostate cancer behavior. Epigenetic silencing of targeted genes is one of the earliest molecular alterations commonly found in prostate neoplasia. Methylation of the CpG-enriched promoter region of GSTP1 has been recognized for more than a decade as a frequent, early event with potential as a diagnostic biomarker. More comprehensive screens of promoter methylation patterns demonstrate consistent changes associated with the development and progression of prostate cancer2,3 and, in the first publication from the Cancer Genome Atlas Project supported by the National Human Genome Research Institute of the National Institutes of Health, epigenetic silencing was complementary to genetic loss in activating critical signaling pathways.4 Thus, epigenetic control of gene expression is fundamental to cancer biology; as a result, assays for epigenetic alterations have the potential to become accurate diagnostic, prognostic, and predictive biomarkers.
Richiardi et al1 report on a retrospective analysis of promoter methylation of three genes, GSTP1, RUNX3, and APC, in two cohorts of patients with prostate cancer from Turin, Italy. Men were diagnosed with prostate cancer after biopsy of the prostate, transurethral resection of the prostate, or radical prostatectomy, and each had available paraffin-embedded tumor tissue and long-term follow-up. The first cohort was comprised of specimens collected in the 1980s, before widespread use of prostate-specific antigen (PSA) screening (n = 216; 1982 to 1988), and the second cohort included specimens collected in the 1990s (n = 243; 1993 to 1996) after the adoption of PSA screening. Promoter methylation of the APC gene was associated with increased risk of prostate cancer–specific mortality in both cohorts (1980s hazard ratio [HR], 1.42; 95% CI, 0.98 to 2.07; 1990s cohort HR, 1.57; 95% CI, 0.95 to 2.62; combined HR, 1.49; 95% CI, 1.11 to 2.00). Promoter methylation of GSTP1 had no association with disease-specific mortality in either cohort, and promoter methylation of GATA2 was associated with increased prostate cancer–specific mortality in the earlier cohort (odds ratio [OR], 2.17; 95% CI, 1.48 to 3.18) but not in the more contemporary cohort. When evaluated as single events, the degree of risk associated with promoter methylation for APC or GATA2 was less than that for high-grade tumors (Gleason sum A substantial body of research focuses on the role for GSTP1 as a tissue-, blood-, urine-, and semen-based biomarker. Although GSTP1 promoter methylation occurs with greater frequency in malignant epithelial cells, it is commonly observed in premalignant prostate epithelial cells and has relatively limited utility as a tissue-based diagnostic biomarker for prostate cancer. However, GSTP1 promoter methylation in cell-free serum DNA5,6 and in urine7 may be more specific for prostate cancer and eventually may be helpful for diagnosis. As a prognostic biomarker, conflicting reports focus on GSTP1 promoter methylation; some find a positive association with risk of recurrence after radical prostatectomy,8 and others find no association.9,10 In the current report, Richiardi et al1 found no association between GSTP1 promoter methylation and the robust prognostic outcome of prostate cancer–specific mortality. Overall, GSTP1 promoter methylation may have a future role as a biomarker, but its diagnostic or prognostic potential when used as a single molecular event appears quite limited. A critical and as yet unresolved question raised by studies that focused on epigenetic changes in prostate cancer is whether a general hypermethylation phenotype is more important than the decreased expression of the specific genes targeted. Although GSTP1 is commonly targeted for promoter methylation, GSTP1/P2 knockout mice do not develop cancers at increased rates, either as an individual event11 or when crossed with a p53-null background,12 although they are more susceptible to drug-induced skin cancer.11 Alternatively, Richiardi et al1 found that APC promoter methylation was associated with increased prostate cancer–specific mortality and that prostate specific–loss of APC (APC –/–) in mice activates β-catenin and results in reliable malignant transformation.13 These transgenic mice did not develop metastasis, but they had to be killed relatively early because of the growth of their primary cancers. The lack of progression to metastatic disease suggests the requirement of additional genetic events that could also be accomplished through epigenetic silencing. In the study in this issue of JCO, Richiardi et al,1 as well as in several previous papers,14,15 report a progressively worse prognosis for prostate cancer as the number of interrogated genes are found to have promoter methylation. This suggests that increasing efficiency of a general hypermethylation phenotype may be the critical underlying biology that results in silencing of a sufficient number of genes for transformation and progression of prostate cancer. Thus, to fully exploit the potential of epigenetic markers, it is likely that approaches that assay multiple sites are going to prove more informative than any single event and should include genes with a proven functional role in prostate carcinogenesis. The manuscript by Richiardi et al1 also highlights how the advent of PSA screening has impacted the biology of prostate cancer. The reported demographic differences between the 1980s and the 1990s cohorts reflect the well-documented clinical evolution of prostate cancer during the past two decades, driven primarily by the widespread adoption of PSA screening.16 However, the impact of PSA screening on prostate cancer biology remains less clear. In the study by Richiardi et al,1 each gene had significantly different frequencies of promoter methylation between the two cohorts, such as decreased methylation of GSTP1 (76.1% to 62.8%; P = .002) and RUNX3 (84.6% to 48.1%; P < .001) and increased methylation of APC (35.9% to 45.1%; P = .047). This observation supports a biologic evolution of prostate cancer that is occurring along with the clinical changes that result from PSA screening. Although the differences between the cohorts had a relatively modest impact on the association between any specific event and prognosis, the changing underlying biology has implications for biomarkers that are developed and validated on archived specimens. Prostate cancer challenges biomarker development because of clinical and biologic disease heterogeneity and because of its prolonged natural history. Although a plethora of molecular biomarkers have been associated with prognosis, no molecular biomarker has been found to contribute sufficiently to the clinically derived prognostic nomograms to warrant broad clinical utilization. The changes in frequency of promoter methylation reported by Richiardi et al1 are likely emblematic of other genetic events in prostate cancer and raise a catch-22 for biomarker development: a biomarker in localized prostate cancer needs to be validated on samples with at least 10 years of follow-up to detect robust survival data but, at the same time, the older specimens with the best clinical annotation may be biologically distinct from the tumors to which the biomarker will be clinically applied. Although the changes that result from the advent of the PSA era are obvious and profound, it is likely that more incremental changes in clinical practice are also likely to influence both the clinical presentation and biology of prostate cancer. Thus, although retrospective validation in independent data sets remains an important step in the development of biomarkers for prostate cancer, prospective validation in patient populations that are likely to clinically and biologically represent the patients for whom the test is eventually intended remains critical. However, even these studies cannot fully ensure that additional evolution of prostate cancer will impact the eventual utility of the biomarker being studied. Although there are no easy solutions for this clinical conundrum, approaches that include robust clinical characterization, utilize clinically based nomograms, and incorporate benchmark molecular events known to be important in prostate cancer pathogenesis will probably meet with the best success. Detailed clinical characterization of retrospective or prospective participants is critical to ensure future relevance of a study. One limitation of the study by Richiardi et al1 is the difference between the two cohorts with respect to pathologic grading. As the Gleason sum remains the characteristic of prostate cancer with the strongest prognostic strength, use of different systems to capture grade detracts from the strength of the comparisons between the two cohorts. Although the distribution of individual clinical characteristics for prostate cancer will likely evolve over time, the use of nomograms, such as the Kattan nomogram or Partin tables, will help to mitigate the impact of these individual changes and will help to more consistently assess individual patient risk. Finally, benchmark molecular events, such as the recently discovered translocations that involve ETS family members or GSTP1 promoter methylation, will put novel biomarkers into a biologic context that will continue to be interpretable, regardless of changes in the frequency with which biologically similar prostate cancers are observed in the future. Future improvements in our ability to anticipate prostate cancer prognosis are likely more dependent on the incorporation of molecular events than on additional refinement of clinical variables. Although the challenges remain substantial, the need continues to increase. The number of men who know they have prostate cancer is increasing and, although active surveillance is probably appropriate for a large percentage of men with localized disease, our current knowledge of prostate cancer biology and our abilities to anticipate clinical progression often are insufficient to avoid a desire for definitive treatment. The manuscript by Richiardi et al1 reports on a potential prognostic role of promoter methylation of the APC gene and, at the same time, highlights the evolving nature of prostate cancer. Only by understanding the challenges introduced by disease heterogeneity, a prolonged natural history, and an evolving biology and by adopting practices that take these characteristics of prostate cancer into account will we be able to develop robust molecular biomarkers and significantly impact the care of men diagnosed with prostate cancer. AUTHOR'S DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST The author(s) indicated no potential conflicts of interest. Acknowledgment P.G.F. is a Damon Runyon Lily Clinical Investigator. Partial support for the preparation of this manuscript was provided by Department of Defense Grants PC051083 and PC0616664, and by National Institutes of Health Grant CA123175. REFERENCES
1. Richiardi L, Fiano V, Vizzini L, et al: Promoter methylation in APC, RUNX3, and GSTP1 and mortality in prostate cancer patients. J Clin Oncol 27:3161–3168, 2009. 2. Yegnasubramanian S, Kowalski J, Gonzalgo ML, et al: Hypermethylation of CpG islands in primary and metastatic human prostate cancer. Cancer Res 64:1975–1986, 2004. 3. Yegnasubramanian S, Haffner MC, Zhang Y, et al: DNA hypomethylation arises later in prostate cancer progression than CpG island hypermethylation and contributes to metastatic tumor heterogeneity. Cancer Res 68:8954–8967, 2008. 4. McLendon R, Friedman A, Bigner D, et al: Comprehensive genomic characterization defines human glioblastoma genes and core pathways. Nature 455:1061–1068, 2008.[CrossRef][Medline] 5. Ellinger J, Haan K, Heukamp LC, et al: CpG island hypermethylation in cell-free serum DNA identifies patients with localized prostate cancer. Prostate 68:42–49, 2008.[CrossRef][Medline] 6. Bastian PJ, Palapattu GS, Lin X, et al: Preoperative serum DNA GSTP1 CpG island hypermethylation and the risk of early prostate-specific antigen recurrence following radical prostatectomy. Clin Cancer Res 11:4037–4043, 2005. 7. Gonzalgo ML, Pavlovich CP, Lee SM, et al: Prostate cancer detection by GSTP1 methylation analysis of postbiopsy urine specimens. Clin Cancer Res 9:2673–2677, 2003. 8. Rosenbaum E, Hoque MO, Cohen Y, et al: Promoter hypermethylation as an independent prognostic factor for relapse in patients with prostate cancer following radical prostatectomy. Clin Cancer Res 11:8321–8325, 2005. 9. Bastian PJ, Ellinger J, Heukamp LC, et al: Prognostic value of CpG island hypermethylation at PTGS2, RAR-beta, EDNRB, and other gene loci in patients undergoing radical prostatectomy. Eur Urol 51:665–674, 2007 discussion 674.[CrossRef][Medline] 10. Woodson K, O'Reilly KJ, Ward DE, et al: CD44 and PTGS2 methylation are independent prognostic markers for biochemical recurrence among prostate cancer patients with clinically localized disease. Epigenetics 1:183–186, 2006.[Medline] 11. Henderson CJ, Smith AG, Ure J, et al: Increased skin tumorigenesis in mice lacking pi class glutathione S-transferases. Proc Natl Acad Sci U S A 95:5275–5280, 1998. 12. Gate L, Majumdar RS, Lunk A, et al: Influence of glutathione S-transferase pi and p53 expression on tumor frequency and spectrum in mice. Int J Cancer 113:29–35, 2005.[CrossRef][Medline] 13. Bruxvoort KJ, Charbonneau HM, Giambernardi TA, et al: Inactivation of Apc in the mouse prostate causes prostate carcinoma. Cancer Res 67:2490–2496, 2007. 14. Rosenbaum E, Carducci MA: Pharmacotherapy of hormone refractory prostate cancer: New developments and challenges. Expert Opin Pharmacother 4:875–887, 2003.[CrossRef][Medline] 15. Ellinger J, Bastian PJ, Jurgan T, et al: CpG island hypermethylation at multiple gene sites in diagnosis and prognosis of prostate cancer. Urology 71:161–167, 2008.[CrossRef][Medline] 16. Ung JO, Richie JP, Chen MH, et al: Evolution of the presentation and pathologic and biochemical outcomes after radical prostatectomy for patients with clinically localized prostate cancer diagnosed during the PSA era. Urology 60:458–463, 2002.[CrossRef][Medline]
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2009 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
8; 1980s cohort OR, 2.17; 95% CI, 1.48 to 3.18; 1990s cohort OR, 3.27; 95% CI, 2.00 to 5.37). However, there was a progressive increase in prostate cancer mortality risk as the number of genes with methylated promoters increased (no methylated promoters HR, 0.80; 95% CI, 0.31 to 2.08; one methylated promoter HR, 1.00 [reference]; two methylated promoters HR, 1.53; 95% CI, 1.24 to 3.15; three methylated promoters HR, 1.97; 95% CI 1.24 to 3.15; P for trend =.002). This study underscores the potential role for epigenetic biomarkers and highlights the challenges in validating molecular biomarkers while the natural history of prostate cancer is evolving. 
