Abstract PS06-04: The prognostic and predictive impact of circulating tumour DNA (ctDNA) dynamics in patients with metastatic Triple Negative Breast Cancer (TNBC) on olaparib based therapy: Results from Cohort E of the PlasmaMATCH trial

Abstract Background: Early changes in ctDNA levels, ctDNA dynamics, may help identify which patients are responding to therapy earlier than imaging. Few studies have assessed ctDNA dynamics during PARP inhibitor therapy. Here we report paired baseline and early on treatment ctDNA analysis from cohort E of plasmaMATCH, that recruited patients with TNBC to treatment with olaparib (PARP inhibitor) plus ceralasertib (ATR inhibitor). Methods: The plasmaMATCH trial assessed the ability of ctDNA testing to allocate patients to mutation matched treatment cohorts (A-D). Patients with TNBC, and without mutations matching cohorts B-D, were enrolled on cohort E. Samples were collected for ctDNA analysis pre-treatment at cycle 1-day1 (C1D1) and cycle 2-day 1 (C2D1). A minimum of 14 days of treatment in the first cycle was required for inclusion in this analysis. Samples were sequenced using error-corrected gene targeted panels (Guardant360, or GuardantOMNI, Guardant Health). Circulating DNA ratio (CDR) was calculated as the ratio of C2D1 ctDNA level to C1D1, pre-specified using the weighted mean of variant allele fractions (AF) of clonal mutations at C1D1, excluding variants with AF < 0.3%, and variants in genes frequently mutated in clonal haematopoesis (GNAS, JAK2, IDH1, IDH2 and ATM). The optimal cut-point for predicting progression free survival (PFS) was assessed as the cut-point with the highest Harrell’s C-index. Results: Of the 75 patients that were enrolled into cohort E, 53 patients had samples sent for paired C1D1-C2D1 ctDNA sequencing, 2 failed sequencing, and all 51 (68%) patients had detectable ctDNA at C1D1. The ctDNA analysis set was representative of the overall enrolled population. The optimal ctDNA dynamics C-index cut-point for predicting PFS was 0 (undetectable ctDNA at C2D1). Median PFS with ctDNA CDR >0 (detectable ctDNA at C2D1) was 4.3 months (95% CI 2.4-5.8), and with undetectable ctDNA was 12 months (95% CI 8-NA) (HR 4.02, 95% CI 1.22-13.23, p=0.01). Splitting patients by median CDR was not predictive (HR=0.98; 95%CI: 0.52-1.82, p=0.94). Confirmed objective response rate was 85.7% (42.1-99.6) in patients with undetectable ctDNA at C2D1, and with detectable ctDNA was 11.4% (3.8-24.6) (OR 4.02, 95% CI 1.22-13.23, p=0.01). Of the 7 patients with undetectable ctDNA at C2D1, one had a BRCA2 germline mutation, and all other patients were wildtype for BRCA1/2 mutations in tumour and germline. All patients with undetectable ctDNA and BRCA1/2 wildtype had a confirmed response.In cohorts A-D (mutation targeted therapies, in predominantly ER positive cancer), the optimal ctDNA dynamics C-index CDR cut-point was 0.165 (HR 3.44, 95%CI: 2.06-5.75, p< 0.001), with median CDR cut-point also highly predictive (HR=2.14, 95%CI:1.36-3.36, p=0.001). Undetectable ctDNA was also strongly predictive (HR=4.41; 95%CI: 1.97-9.87, p< 0.001) in cohort A-D.In cohort E, a significant association was found between baseline ctDNA and PFS, with an optimal C-index cutpoint of 6.81% (HR=3.02, 95% CI: 1.39-6.56, p=0.001). Median PFS for baseline ctDNA ≤ 6.81% was 10.2 months (95% CI 3.7-17.2), and for baseline ctDNA >6.81% was 4.4 months (95% CI 2.2-5.5). Conclusions: ‘Clearance’ of ctDNA to become undetectable at C2D1 identified sporadic TNBC patients who benefited from olaparib and ceralasertib. Although ‘clearance’ of ctDNA was associated with good outcome on olaparib plus ceralasertib, median CDR was not predictive of treatment benefit. This contrasts the results of ctDNA dynamic assessment of cohort A-D, where median CDR was highly predictive of treatment benefit. ctDNA dynamic assessment may differ between mutation targeted therapies (cohorts A-D) that induce cell-cycle arrest, and PARP inhibitors (cohort E) that inhibit DNA repair mechanisms. Implementing ctDNA dynamics into clinical trials and care may require distinct analysis for different therapies. Citation Format: Iseult Browne, Javier Pascual, Rosalind Cutts, Belinda Kingston, Sarah Hrebien, Lucy Kilburn, Alex Pearson, Laura Moretti, Andrew Wardley, Iain Macpherson, Richard Baird, Rebecca Roylance, Iris Faull, Kimberly C Banks, Isaac Garcia-Murillas, Judith Bliss, Alistair Ring, Nicholas Turner. The prognostic and predictive impact of circulating tumour DNA (ctDNA) dynamics in patients with metastatic Triple Negative Breast Cancer (TNBC) on olaparib based therapy: Results from Cohort E of the PlasmaMATCH trial [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PS06-04.