Analysis of Circulating Cell-Free DNA Identifies Multiclonal Heterogeneity of BRCA2 Reversion Mutations Associated with Resistance to PARP Inhibitors

David Quigley, Joshi J Alumkal, Alexander W Wyatt, Vishal Kothari, Adam Foye, Paul Lloyd, Rahul Aggarwal, Won Kim, Eric Lu, Jacob Schwartzman, Kevin Beja, Matti Annala, Rajdeep Das, Morgan Diolaiti, Colin Pritchard, George Thomas, Scott Tomlins, Karen Knudsen, Christopher J Lord, Charles Ryan, Jack Youngren, Tomasz M Beer, Alan Ashworth, Eric J Small, Felix Y Feng, David Quigley, Joshi J Alumkal, Alexander W Wyatt, Vishal Kothari, Adam Foye, Paul Lloyd, Rahul Aggarwal, Won Kim, Eric Lu, Jacob Schwartzman, Kevin Beja, Matti Annala, Rajdeep Das, Morgan Diolaiti, Colin Pritchard, George Thomas, Scott Tomlins, Karen Knudsen, Christopher J Lord, Charles Ryan, Jack Youngren, Tomasz M Beer, Alan Ashworth, Eric J Small, Felix Y Feng

Abstract

Approximately 20% of metastatic prostate cancers harbor mutations in genes required for DNA repair by homologous recombination repair (HRR) such as BRCA2 HRR defects confer synthetic lethality to PARP inhibitors (PARPi) such as olaparib and talazoparib. In ovarian or breast cancers, olaparib resistance has been associated with HRR restoration, including by BRCA2 mutation reversion. Whether similar mechanisms operate in prostate cancer, and could be detected in liquid biopsies, is unclear. Here, we identify BRCA2 reversion mutations associated with olaparib and talazoparib resistance in patients with prostate cancer. Analysis of circulating cell-free DNA (cfDNA) reveals reversion mutation heterogeneity not discernable from a single solid-tumor biopsy and potentially allows monitoring for the emergence of PARPi resistance.Significance: The mechanisms of clinical resistance to PARPi in DNA repair-deficient prostate cancer have not been described. Here, we show BRCA2 reversion mutations in patients with prostate cancer with metastatic disease who developed resistance to talazoparib and olaparib. Furthermore, we show that PARPi resistance is highly multiclonal and that cfDNA allows monitoring for PARPi resistance. Cancer Discov; 7(9); 999-1005. ©2017 AACR.See related commentary by Domchek, p. 937See related article by Kondrashova et al., p. 984See related article by Goodall et al., p. 1006This article is highlighted in the In This Issue feature, p. 920.

Conflict of interest statement

DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST

AA and CJL are named inventors on patents describing the use of PARP inhibitors and as such stand to gain financially as part of the ICR “Rewards to Inventors” Scheme.

©2017 American Association for Cancer Research.

Figures

Figure 1. PARPi resistance by multiple large…
Figure 1. PARPi resistance by multiple large deletions
1a: Timeline of patient 1’s serum PSA level before and during treatment with talazoparib. 1b: H&E stains of the solid tumor biopsy pre- and post-talazoparib were consistent with approximately 60% tumor purity. 1c: DNA read coverage of patient 1’s germline, pre- and post-talazoparib solid tumor, and post-talazoparib liquid biopsy. Horizontal white lines indicate the span of deletions, numbered D1 through D7, with deletion frequency indicated at the right side. The location and frequency of the p.K1872X/wild type allele is noted in gold/blue. 1d: Schematic protein models for BRCA2 WT, p.K1872X, and reversion isoforms. 1e: PCR amplification of the region around deletions D1 and D2 in patient 1’s germline, pre-talazoparib, and post-talazoparib solid biopsy confirms the presence of deletions D1 and D2 in post-talazoparib DNA.
Figure 2. Olaparib resistance in Patient 2…
Figure 2. Olaparib resistance in Patient 2 was associated with reversion mutations
2a: Timeline of patient 2’s serum PSA level before and during treatment with olaparib, showing a drop in PSA from 355 to a nadir of 40, followed by increase associated with drug resistance. 2b: CT scans from patient 2 pre-olaparib, on treatment, and after resistance show a liver metastasis at 17 mm on 2/8/16, decreased to 6 mm on 6/16/16 after olaparib treatment began, and at 18 mm on 9/13/16 respectively. 2c: Bone scans of patient 2 at times matching 2b, showing widespread metastasis, treatment response, and recurrence of metastatic lesions. 2d: Counts of indels observed exclusively in resistant cfDNA, binned into 50 bp intervals across BRCA2. Exon bounds shown at top. The pathogenic stop gain in p.R259fs allele is at nucleotide 1050, noted with red arrow. 2e: Diagram of BRCA2 reversion mutation alleles observed in patient 2. Wild type BRCA2 nucleotide and protein sequence, p.R259fs nucleotide and protein sequence, and BRCA2 reversion alleles present in patient 2’s post-olaparib DNA in all of exon 9 and the first 48 nucleotides of exon 10. Black vs. grey bars indicates observed vs. inferred nucleotide sequence. Deleted nucleotides shown as a thin black line with tics for each deleted nucleotide, insertions highlighted in blue, mutations highlighted in orange, alternate splice acceptor highlighted in green. For clarity, nucleotides are draw on reversion alleles where insertions are present; otherwise nucleotide sequence corresponds to WT. Count of each allele is shown at right.

Source: PubMed

3
Předplatit