5578 - Detection of circulating tumor DNA in de novo metastatic castrate sensitive prostate cancer

Background

De novo metastatic castrate sensitive prostate cancer (mPC) represents approximately 10% of prostate cancer diagnoses but almost 50% of mPC related deaths. Biomarkers are required to guide therapy intensification at time of diagnosis, but scant tumor material is available since most patients do not undergo prostatectomy. Plasma circulating tumor DNA (ctDNA) is a promising minimally-invasive biomarker in castration-resistant disease but remains untested in castrate-sensitive disease.

Methods

We collected plasma cell-free DNA (cfDNA) at or near time of diagnosis from 51 de novo mPC patients enrolled at two academic centres. CfDNA and matched diagnostic needle biopsies were subjected to deep targeted sequencing across all exons of 73 prostate cancer relevant genes and analyzed independently for somatic alterations.

Results

22 of 31 (71%) ADT-naive patients had detectable ctDNA (fraction range 0.5-70%). A further 20 patients received between 1 and 49 days of ADT prior to cfDNA collection (median 23) and had significantly lower ctDNA fractions than ADT-naive patients (mean 6.0% vs 22.7%; p = 0.009). Although there was no relationship between Gleason score, serum PSA or age at diagnosis and ctDNA fraction, 11 of 13 patients (86%) with lung and/or liver metastases had detectable ctDNA. Excluding one case with hypermutation and mismatch repair deficiency detected only in ctDNA, 83% of non-silent mutations were concurrently identified in both tissue and ctDNA while 9.3% and 7.8% were unique to tissue or ctDNA respectively. 9 patients had truncating mutations and loss of heterozygosity across DNA repair genes BRCA2, ATM, CDK12 or MSH2. No AR gene alterations were detected.

Conclusions

Plasma ctDNA is detected in the majority of patients with de novo mPCa and somatic mutations identified in ctDNA are highly concordant with the matched diagnostic prostate biopsy. Exposure to ADT prior to plasma collection significantly reduces ctDNA detection rates and ctDNA fraction. cfDNA analysis can detect important driver alterations and is complementary to tissue-based analyses.

Clinical trial identification

Legal entity responsible for the study

Vancouver Prostate Centre, Department of Urologic Sciences, University of British Columbia, British Columbia, Canada.

Funding

Has not received any funding.

Editorial Acknowledgement

Disclosure

All authors have declared no conflicts of interest.