Blood-Based Tumor DNA Testing Reveals How Prostate Cancer Rewires Itself Under Treatment

Prostate cancer is a moving target. Under the pressure of modern therapies - hormone-blocking drugs, PARP inhibitors, taxane chemotherapy - the cancer adapts. It rewires its own molecular circuitry to maintain the survival signals that treatment is designed to cut off. The challenge for oncologists has always been detecting those adaptations before the disease has fully outrun the current treatment.

A new multi-center study, published February 26 in Clinical Cancer Research, suggests that the answer may already be circulating in patients' blood. Using serial liquid biopsies - blood tests that detect fragments of DNA shed by tumor cells - researchers tracked how metastatic prostate cancer evolves across more than 1,700 patients as they moved through multiple lines of therapy. The central finding: alterations in the androgen receptor (AR) consistently emerged after treatment and consistently predicted worse outcomes, regardless of which drug class patients had received.

The Case for Real-Time Molecular Monitoring

Genomic testing has become standard in prostate cancer management, but most tests rely on tumor tissue collected at diagnosis or at the time of one specific biopsy. That approach has a fundamental problem: prostate cancer, particularly at advanced stages, is heterogeneous and dynamic. A tissue sample taken two years ago from one metastatic site reflects neither the current state of the disease nor the molecular changes driven by subsequent treatment.

Liquid biopsy offers a different approach. Circulating tumor DNA (ctDNA) - small fragments released by dying cancer cells into the bloodstream - reflects the overall mutational landscape of the disease as it exists at the time of the blood draw. Serial testing, taken at multiple time points as the disease progresses, provides what the researchers call a moving picture rather than a snapshot.

The study, led by Chinmay T. Jani at Sylvester Comprehensive Cancer Center at the University of Miami Miller School of Medicine, analyzed paired ctDNA samples taken before treatment initiation and again after treatment discontinuation - a window that typically corresponds to disease progression. The project was initiated and developed at UC San Diego under the mentorship of Rana McKay at the Moores Cancer Center, and involved collaborators from UC San Francisco, Scripps Research Institute, and Guardant Health.

AR Alterations Emerge Across All Treatment Classes

The study examined patients with metastatic castration-resistant prostate cancer (mCRPC) - disease that no longer responds to hormone-deprivation therapy. This population receives a range of second- and third-line treatments, including androgen receptor pathway inhibitors (ARPIs) like enzalutamide and abiraterone, PARP inhibitors targeting DNA repair defects, and taxane chemotherapy.

Across all three treatment classes, the same pattern appeared. After therapy, patients were significantly more likely to carry AR amplifications or AR mutations - particularly in regions that allow the receptor to remain active despite treatment. These alterations function as a master switch: they keep the cancer's growth signals turned on even when the therapy is designed to silence them.

The presence of AR alterations was not merely common - it was consequential. Across all treatment classes, patients with AR-altered tumors experienced shorter overall survival, faster treatment discontinuation, and earlier need for the next line of therapy. The effect persisted regardless of whether the patient had received an ARPI, a PARP inhibitor, or a taxane.

"AR alterations emerged as a consistent marker of more aggressive disease," Jani said. "They mattered regardless of which therapy patients received."

Resistance Mechanisms in PARP Inhibitor-Treated Patients

Among patients treated with PARP inhibitors - drugs that exploit defects in DNA repair pathways, particularly BRCA mutations - the study identified specific resistance mechanisms. Some tumors acquired BRCA reversion mutations, which restore the DNA repair function that PARP inhibitors target, effectively rendering the drug ineffective. Others accumulated alterations in genes including TP53, EGFR, and PIK3CA, which are markers of genomic instability and broad treatment resistance.

This type of resistance characterization - knowing not just that a treatment stopped working but identifying the specific molecular escape route - points toward potential rational combination strategies. Targeting AR while inhibiting the PARP pathway, or designing next-generation PARP inhibitors that remain active against reversion mutations, are directions supported by this kind of data.

Scale and Study Design

With more than 1,700 patients, this is one of the largest linked clinical genomic analyses of mCRPC performed to date, drawing on a database that links ctDNA results to clinical outcomes across multiple institutions. The real-world design - as opposed to a controlled clinical trial - reflects how liquid biopsy is actually used in practice and captures the full diversity of patient characteristics, treatment sequences, and disease trajectories.

The observational nature of the study means it cannot establish causation - it cannot prove that AR alterations directly cause worse outcomes rather than simply co-occurring with more aggressive disease. Nor can it determine whether acting on ctDNA results to change treatment earlier would improve survival; that question requires a prospective interventional trial. The study provides biological rationale and outcome correlation, not a treatment algorithm.

What It Means for Precision Oncology

The findings argue that single-timepoint genomic testing is insufficient for managing advanced prostate cancer. Tumors that appear to have one genomic profile at diagnosis may look entirely different after six months of ARPI therapy. Decisions about subsequent treatment - when to switch, what to switch to, whether to escalate - could be better informed by knowing what the disease has become, not what it was.

"Serial ctDNA testing gives us a moving picture, not a snapshot," Jani said. "That has real implications for precision oncology."

"This work highlights the power of real-time molecular monitoring to inform precision medicine in advanced prostate cancer," said McKay. "As we develop the next generation of therapies, including novel AR-targeted agents and rational combinations, understanding how tumors evolve under treatment pressure will be critical to delivering the right drug to the right patient at the right time."