Prostate cancer patient selection for cell-free DNA testing

From the Pritchard and Nelson labs and Prostate Cancer Program, University of Washington and Human Biology Division

Solid tumor heterogeneity, not only between patients but also within a unique tumor, is a critical barrier to understanding the diseases and identifying adapted therapeutic options. Precision medicine is tailored to overcome these issues and bring innovative and efficient therapies to the patients. However, assessing tumor heterogeneity is challenging. Surgical access to the tumor is not always granted, and possible biopsies are invasive and limit the analysis to the one core collected. To address these limitations, two blood-based strategies are being developed: the isolation of either circulating tumor cells or cell-free DNA. Cell-free DNA (cfDNA) can be found in the plasma and previous studies have demonstrated the relevance of sequencing this DNA to determine patient mutational landscape correlating with tissue-based analysis. However, Dr. Colin Pritchard, co-director of the genetics and solid tumors lab at University of Washington and director of precision diagnostics for the Brotman Baty Institute for Precision Medicine, Dr. Michael Schweizer (Clinical Research Division), and Dr. Peter Nelson (Human Biology Division) saw limitations to this technique in the context of prostate cancer (PC). Dr. Pritchard explains: “We noticed that ultra-sensitive approaches to cell-free DNA testing were picking up very low-level mutations in the blood that aren’t relevant to treating the patient’s prostate cancer.  We wondered if we could improve performance of cell-free DNA testing by identifying clinical features that predict when a plasma cell-free DNA study most accurately reflects what is going in the patient’s tumors, and therefore is best suited to guide treatment”. This question was addressed in a recent issue of the journal Prostate.


Heatmap showing that successful cell-free tumor DNA (cfDNA) sequencing in men with advanced prostate cancer is correlated with disease burden.
Successful cell-free tumor DNA (cfDNA) sequencing in men with advanced prostate cancer is correlated with disease burden. Illustration provided by Dr. Pritchard

The authors obtained blood from advanced PC patients, as well as tumor tissue and germline samples. Next generation sequencing (NGS) was performed on cfDNA from plasma and DNA from other samples for each patient using an optimized version of the University of Washington OncoPlex (UW-OncoPlex), a technique Dr. Pritchard helped develop.  OncoPlex is a targeted DNA sequencing method that identifies mutations in over 250 known cancer-related genes.  Of the 93 patients who consented, 72 patients with prostate adenocarcinoma were included in the analysis. For each patient, Dr. Schweizer, first author of the study, and colleagues compared the sequencing results between the different samples; successful cfDNA sequencing was defined as the identification of at least one somatic mutation present in the tumor tissue but not in the germline sample. Interestingly, the level of prostate-specific antigen (PSA), a marker used as a screening method for PC, was significantly elevated in patients in which cfDNA analysis was successful. In addition, 81% of “successful” patients had high volume disease compared to 38% within unsuccessful cases. Even more striking, 100% of successful and 59% of unsuccessful patients presented with metastases. Finally, prostate cancer patients whose tumors no longer responded to hormonal therapies, the so-called castration-resistant cancers, were more likely to have somatic mutations detected in the cfDNA. Altogether, these observations suggest that ability to detect somatic mutations in cfDNA correlates with disease aggressiveness, and implies that stratification of patients based on metastatic burden, disease volume, hormonal dependency and PSA levels would help discriminate between patients who would benefit from blood based genetic testing and those who should preferentially be oriented towards a tissue-based analysis. Dr. Pritchard explains: “Our study highlights the importance of selecting the right patients for cell-free circulating tumor DNA testing. Just as with a tissue biopsy for cancer diagnosis, it is important to biopsy the right sample at the right time, or else the study is not informative. In the case of men with advanced prostate cancer, we define clinical features that help determine when cell-free DNA testing will be most informative and useful.”

These findings have the potential to challenge the scientific community to establish the most effective way to apply blood-based genetic testing in prostate cancer. “Our study raises questions about whether clinical features such as disease burden should be part of the testing criteria for cell-free DNA testing in prostate cancer”, says Dr. Pritchard.  “There is not yet a standard-of-care  around this important question for prostate cancer, or for any cancer where cell-free DNA testing is beginning to be used.”


This work was supported by the National Institute of Health, the UW/FHCRC Institute for Prostate Cancer Research, the Fred Hutch Solid Tumor Translational Research Program and the Pacific Northwest Prostate Cancer SPORE.

Fred Hutch/UW Cancer Consortium member Drs. Schweizer, Cheng, Yu, Montgomery, Nelson and Pritchard contributed to this research.

Schweizer MT, Gulati R, Beightol M, Konnick EQ, Cheng HH, Klemfuss N, De Sarkar N, Yu EY, Montgomery RB, Nelson PS, Pritchard CC. Clinical determinants for successful circulating tumor DNA analysis in prostate cancer. Prostate 2019;79:701-708.  doi: 10.1002/pros.23778.