Can Liquid Biopsies and Tumor Biomarkers Personalize Prostate Cancer Treatment?

Experimental, minimally invasive “liquid biopsy” blood tests might soon help to personalize prostate cancer treatment by predicting androgen resistance and survival benefits from particular treatments.

[[{"type":"media","view_mode":"media_crop","fid":"49438","attributes":{"alt":"Image © harmpeti /","class":"media-image","id":"media_crop_9508680305752","media_crop_h":"0","media_crop_image_style":"-1","media_crop_instance":"5965","media_crop_rotate":"0","media_crop_scale_h":"0","media_crop_scale_w":"0","media_crop_w":"0","media_crop_x":"0","media_crop_y":"0","title":"Image © harmpeti /","typeof":"foaf:Image"}}]]Experimental, minimally invasive “liquid biopsy” blood tests might soon help to personalize prostate cancer treatment by predicting androgen resistance and survival benefits from particular treatments, researchers announced at the 2016 American Society of Clinical Oncology (ASCO) Annual Meeting, held June 3–7 in Chicago.  

Liquid biopsies detect circulating tumor cells (CTCs) or bits of tumor DNA (ctDNA). Not all tumors shed cells or DNA into a patient’s bloodstreams, but most do. And when they do, they can reveal a lot about themselves-including molecular signatures that can be targeted with specific treatments.

Recent years have seen an explosion of candidate biomarkers for prostate cancer and other malignancies, including both liquid biopsies and tumor-sample gene panels. Most candidate biomarkers have been prognostic gene-mutation signatures that can estimate patient survival regardless of what treatment strategies are attempted. These prognostic tests can be useful for risk-stratifying patients who are participating in clinical studies, or in communicating prognosis to a patient and his loved ones.

But other proposed biomarkers are predictive: their results correlate with a given patient population’s outcomes following treatment with a particular treatment strategy. Of particular interest among prostate cancer experts at this year’s ASCO meeting was the development of biomarkers that predict tumor androgen-receptor (AR) resistance.

One suspected factor in AR resistance is nuclear androgen-receptor splice variant 7 (AR-V7), which lacks a normal AR ligand-binding domain. Detecting AR-V7 protein in prostate CTCs can predict the failure of abiraterone and enzalutamide (two AR signaling inhibitors), according to a cross-sectional cohort study reported at the ASCO meeting (abstract 5013). The study was also published online in JAMA Oncology.[1]

Among patients whose CTCs harbor AR-V7, taxane-based chemotherapy regimens are associated with significantly longer progression-free and overall survival (OS) as well (OS: 8.9 vs 4.6 months; multivariate hazard ratio for risk of death on taxane vs abiraterone or enzalutamide therapy, 0.24 [95% CI, 0.10–0.57]; P = .035).

A total of 161 evaluable men diagnosed with metastatic castration-resistant prostate cancer (mCRPC) were included in the study; pretreatment blood samples were taken and tested using immunofluorescence to identify nucleated cells with the AR-V7 protein. Response to AR therapy was defined as a decline in patients’ PSA levels of greater than 50%.[1]

None of the patients with AR-V7–positive CTCs responded to AR therapy.[1]

The findings imply that AR-V7 testing can help ensure that patients receive appropriate treatment earlier in disease progression, when therapy might be more effective.

“We found that a novel liquid biopsy for AR-V7 was able to identify, with specificity, patients who will not benefit from these therapies and should instead start chemotherapy independent of the line of therapy being administered,” said lead study author Howard I. Scher, MD, chief of genitourinary oncology services at Memorial Sloan Kettering Cancer Center in New York.

The findings bolstered earlier findings from a study at the Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins, which had employed messenger RNA (mRNA) rather than protein immunofluorescence to test CTCs for AR-V7.[2] An expanded analysis of the Johns Hopkins cohort was also presented at the ASCO meeting this year (abstract 5012); lead study author Emmanuel S. Antonarakis, MD, and colleagues reported that their original findings remained statistically significant in multivariate analyses, for both first-line and second-line abiraterone or enzalutamide treatment.

Together, the Sloan Kettering and Johns Hopkins studies “provide compelling evidence that AR-V7–positive CTCs are a predictive biomarker for failure to respond to AR-directed therapy,” concluded R. Bruce Montgomery, MD, and Stephen R. Plymate, MD, of the University of Washington in Seattle, in an invited commentary published alongside the study from Scher’s team in JAMA Oncology.[3]

The findings must yet be validated in a prospective study and important details still need to be worked out. The relative sensitivity of different AR-V7 tests remains somewhat unclear, for example; a lower proportion of patients with AR-V7–positive CTCs were detected using protein-immunofluorescence testing, than had previously been detected using mRNA tests.[3]

If AR splice variants are “truly driving tumor biology” in prostate cancer, then AR-V7 tests might be promising targets for drug development, as well as useful tests for discerning which existing therapies are likely best for a given patient, Drs. Montgomery and Plymate noted.[3]  

AR-V7 testing among men with mCRPC could offer the US healthcare system $150 million in net savings each year, according to researchers at the Johns Hopkins University School of Medicine in Baltimore (abstract 283).

Liquid biopsies are not the only promising area of predictive biomarker development. Researchers from the Mayo Clinic in Rochester, Minnesota, reported on their development and validation of a primary tumor biopsy-based gene panel that predicts tumor resistance to androgen-deprivation therapy (abstract 5018). Seeking a genomic signature that will predict which patients with prostate cancer are more likely to fail adjuvant androgen-deprivation therapy and experience metastatic disease, the authors obtained tumor expression profiles for 1,632 candidate genes from 1,212 patients from the Decipher GRID database. They identified a 48-gene androgen-resistance signature that predicts patients who are at high risk of failing hormone therapy, and at risk of metastasis, suggesting that the signature could be used to identify patients “who could be spared of the toxicities surrounding androgen-deprivation therapy.”

As with AR-V7 biomarker tests, however, the proposed androgen-resistance signature has not been tested in randomized studies.

“This study is a good first step but further validation is needed,” said Philip K. Kantoff, MD, of the Memorial Sloan Kettering Cancer Center at an ASCO meeting discussion panel on biomarkers for androgen-receptor therapies.


1. Scher HI, Lu D, Schreiber NA, et al. Association of AR-V7 on circulating tumor cells as a treatment-specific biomarker with outcomes and survival in castration-resistant prostate cancer. JAMA Oncology. Published online June 4, 2016.

2. Antonarakis ES, Lu C, Luber B, et al. Androgen receptor splice variant 7 and efficacy of taxane chemotherapy in patients with metastatic castration-resistant prostate cancer. JAMA Oncology. 2015;1(5):582-591.

3. Montgomery RB, Plymate SR. AR-V7 protein in circulating tumor cells-the decider for therapy? JAMA Oncology. Published online June 4, 2016.