The treatment landscape of metastatic castration-sensitive prostate cancer (mCSPC) has expanded over the last decade, and rational biomarkers based on biological evidence are urgently required to enable optimal treatment for each individual patient and improve treatment outcomes. We aimed to develop a comprehensive genomic profiling assay using cell-free DNA (cfDNA) as a minimally invasive biomarker of mCSPC.

Approximately 10ml blood samples were collected from 10 patients with mCSPC before systemic therapies, and the plasma and buffy coat were separated. cfDNA and leukocyte DNA were extracted from the plasma and buffy coat, respectively, and targeted DNA sequencing was performed using a capture panel of the exon regions of 88 prostate cancer-related genes. We used our established pipeline, the eVIDENCE algorithm, to detect somatic mutations in cfDNA, and the CNV kit was used for copy number analysis. The proportion of tumor-derived cfDNA, the circulating tumor DNA (ctDNA) fraction, was estimated using two approaches: the variant allele frequency of somatic mutations and deviation in the allele fraction of heterozygous germline single-nucleotide variations. We also compared the somatic mutation profiles between cfDNA and patient-matched tumor tissues.

Nine of the 10 (90%) cfDNA samples had a quantifiable ctDNA fraction (median = 29.07%, range = 5.64%–64.4%). The most common mutated genes were FOXA1 (44.4% of ctDNA-detected samples), TP53 (22.2%), SPOP (22.2%), and AR (11.1%). Of the 22 somatic mutations detected in tumor tissues, 10 (45.5%) were shared in patient-matched cfDNA. Fifteen somatic mutations were detected only in cfDNA, including mutations in AKT1 and BRAF.

This study showed that a ctDNA assay is sufficient to identify driver DNA alterations present in matched metastatic tissue and supports the development of DNA biomarkers to guide patient management based on ctDNA.