Although biomarkers have been shown to correlate with survival in prostate cancer, the use of biomarkers in the management of these patients may be dependent upon the clinical state and the class of drug evaluated.
Daniel P. Petrylak, MD
Although a variety of biomarkers have been shown to correlate with survival in men with prostate cancer, the use of biomarkers in the management of these patients may be dependent upon the clinical state and the class of drug evaluated. Speaking at the 6th Annual Interdisciplinary Prostate Cancer Congress, held in New York City on March 16, Daniel P. Petrylak, MD, director of the Genitourinary Oncology Program at Yale Cancer Center in New Haven, Connecticut, reviewed the evidence for biomarkers to date. “There is nothing at this point that has fallen out as absolutely prognostic or predictive in an individual situation,” he said.
One area of interest is the prognostic value of promoter methylation of genes associated with various parameters of prostate cancer progression.1,2Henrique et al evaluated methylation ofAPC,CCND2,GSTP1,RARB2, andRASSF1Aas biomarkers for prognosis.2High-level methylation ofAPCwas the only epigenetic event significantly associated with a worse disease-free or disease-specific survival. There was a trend toward worse outcome in patients treated with androgen- deprivation therapy (ADT) who hadAPChypermethylation (P=.079).APChypermethylation was associated with a shorter time to biochemical relapse (disease-free survival) in patients treated with prostatectomy (P=.043), but not in patients treated with external-beam radiation therapy or ADT.
Okegawa et al assessed whether circulating tumor cells (CTCs) with tumorrelated methylated DNA could be used to predict survival in patients with hormone- refractory prostate cancer.3The researchers analyzed blood samples from 76 patients; CTC evaluations occurred 3.5 to 78.5 months after the initial diagnosis. Serum DNA methylation status was assessed forAPC,GSTP1,RASSF1A,MDR1, andPTGS2.
A total of 47 patients (62%) had 5 or more CTCs and a median overall survival (OS) of 12 months compared with 26 months for patients with fewer than 5 CTCs (P<.001). The presence of 5 or more CTCs was significantly correlated with the detection of 1 or more methylated loci (P <.001). The median OS for patients with methylated DNA was 12 months versus 48 months or more for patients without methylated DNA (P<.001).
Posttreatment CTC levels were shown to predict OS in a study of 231 patients with progressive CRPC4; data from the study led to FDA approval of the Cell- Search assay. However, in his talk, Petrylak said that although CTCs are useful for predicting treatment response or survival with cytotoxic and hormonal therapies, about half of patients do not have detectable CTC levels by current detection methods. More-sensitive CTC detection techniques are needed, he said.
TheTMPRSS2-ERGgene fusion, which is present in approximately half of newly diagnosed prostate cancer cases, has also been investigated as a biomarker, though the clinicopathologic impact of the fusion is not clear, Petrylak said. Different subtypes ofTMPRSS2-ERGhave been detected, and the Edel subtype may represent a more lethal genotype, susceptible to higher recurrence, more likely to evolve to CRPC and to progress to metastasis.5A role has been postulated forTMPRSS2-ERGas a predictive biomarker for sensitivity to abiraterone acetate; however, a phase II trial found that the fusion has a limited role in this area.6
In summary, Petrylak said, although a variety of biomarkers correlate with survival, none are “ready for prime time.” However, he said, research on biomarkers should be an element in all subsequent phase III trials.