Plasma samples were culled from 3334 men with advanced prostate cancer, including 1674 subjects from the TRITON2/3 studies of rucaparib and 1660 non-trial clinical samples. The observed GA landscape was compared to 2006 metastatic biopsies, with concordance assessed in 837 patients. In keeping with previous reports of ctDNA burden, 94% (3127) of subjects had detectable ctDNA with 8.8% (295) with mutations in BRCA1/2. Concordance with tissue evidence of BRCA1/2 mutations was observed in 93% of evaluable subjects (67/72) and 20 subjects had evidence of such mutations only in ctDNA. Notably, subclonal reversion mutations in BRCA1/2 were observed in 10 of 1660 routine clinical specimens, suggesting a mechanism for PARPi resistance, at least in a subpopulation evaluated.
Alterations in AR, the gene encoding the androgen receptor, were detected in 42% (940/2213) samples, including amplifications and hotspot mutations. Among the mutations detected are specific alterations which confer resistance to commonly used highly-potent ARSIs, such as abiraterone acetate and enzalutamide. The authors also describe a subset of samples with rare compound “double” mutations and novel potentially activating mutations in AR. Additional GAs were detected in relevant signaling pathways including PI3K/AKT/mTOR (14%), WNT/beta-catenin (17%), and RAS/RAF (5%). Microsatellite instability was rare (1.4% of 2213 patients).
These data lend further support to the relative reliability (as compared to tissue assays) of using plasma for evaluating relevant tumor genomic alterations in the advanced metastatic setting, reflecting genomics data demonstrating that dominant metastatic clones found at autopsy can be found in the circulating compartment1. This is particularly powerful as detection of resistant subclones that may not be in a tissue-based sample, either because these cells reflect occult or unsampled metastatic samples, could impact therapeutic decisions. It should be considered that use of subjects from the TRITON studies, which comprised approximately half of the cohort may result in higher rates of observed GAs in BRCA1/2 than in daily practice, given the enrichment in such genomic alterations as “ground truth” in this group. As noted by the authors, the limitations of the assay in these studies includes an inability to detect deletions in BRCA1/BRCA2, as well as other clinically-relevant commonly-deleted prostate cancer genes (e.g. PTEN). Further evaluation using orthogonal assays, such as RNAseq, would add additional detail, particularly along the AR signaling axis, to these promising results. Finally, the authors astutely recommend that a degree of caution must be taken when interpreting liquid biopsy results, given the influence of alterations representing clonal hematopoiesis.
Publication of full length publication can be found in the February 8th issue of Clinical Cancer Research.
Presented by: Hanna Tukachinsky, PhD, Foundation Medicine Inc., Cambridge, MA
Written by: Jones Nauseef MD, PhD. Fellow, Division of Hematology and Oncology, Weill Cornell Medicine/New York Presbyterian Hospital. Twitter: @DrJonesNauseef during the 2021 ASCO Genitourinary Cancers Symposium (ASCO GU), February 11th to 13th, 2021
1. Woodcock DJ, Riabchenko E, Taavitsainen S, et al., Prostate cancer evolution from multilineage primary to single lineage metastases with implications for liquid biopsy. Nature Comm. 11:5070 (2020). DOI:10.1038/s41467-020-18843-5.