Longitudinal circulating tumor DNA (ctDNA) whole-exome sequencing (WES) in the phase Ib/II trial of palbociclib and bazedoxifene reveals genomic dynamics and clonal evolution with the acquisition of treatment resistance in hormone receptor-positive, HER2-negative (HR+ HER2-), advanced breast cancer (ABC).

Abstract

1058 Background: Patients (pts) with HR+ HER2- ABC ultimately develop endocrine resistance. To gain insights into the genetic mechanisms of resistance we performed WES on serial plasma samples from endocrine resistant pts treated on a clinical trial (NCT02448771). Methods: Plasma samples were collected at baseline (n=36), day 1 of cycle 2 (n=33), and at the end of treatment (EOT, n=33). Samples were subjected to ultra-low passage (ULP, 0.19-0.57X) WGS to determine ctDNA tumor fraction (TF) for the selection of samples (TF>0.03) for subsequent WES (193X). Somatic single nucleotide variations, somatic copy number alteration (SCNA), phylogeny, tumor mutational burden, mutational signatures, and germline analyses were performed. Results: All 102 samples underwent successful ULP and 68 WES. Overall, most frequent pathogenic mutations were in ESR1 and PIK3CA. At baseline, 32% of pts had ESR1 mutation and 21% PIK3CA mutation. There was no association between ESR1 mutations and PFS. In contrast, baseline PIK3CA mutations were detected only in pts who did not have a clinical benefit, and were associated with worse PFS compared to pts with wild-type PIK3CA (1.8 vs. 3.9 months, respectively, HR=0.2, 95% CI 0.06-0.6, P=0.0019, log-rank test). Additionally, pts with a baseline truncating mutation, mostly in tumor suppressor genes ( TP53, MEN1, RB1, CDKN1B, NF1, TP53BP1, TP63, SMAD2/4, ARID1A, KMT2C), also had a significantly worse PFS (1.7 vs 3.8 months, HR=0.3, 95% CI 0.1-0.7, P=0.006, log-rank test). At EOT, 20% (4/20) of pts with matched baseline samples had newly acquired mutations that are suggestive of mechanisms of acquired resistance and offer potential therapeutic targets (e.g. ERBB2, PIK3CA). SCNA analysis showed that in all pts there were at least 2 SCNAs in cancer-related driver genes, most common in CCND1 and ELF3. Moreover, in all samples we identified at least 1 SCNA related to a potential mechanism of resistance. To better understand tumor heterogeneity and sub-clonal architecture we performed an evolutionary analysis (sufficient TF≥0.15, available in n=7). Phylogenetic analysis revealed sub-clonal dynamics that could explain the acquisition of resistance in at least three pts (3/7), and identified novel genes which might have role in endocrine resistance (e.g. DCAF13, ZFHX3). Conclusions: Our results demonstrate the feasibility and utility of serial WES in a clinical trial. Serial ctDNA WES and evolutionary studies enabled us to discover novel potential genomic mechanisms of tumor progression, and identified PIK3CA mutations as a candidate biomarker of resistance to the combination of palbociclib and bazedoxifene, which may apply to other next generation endocrine treatments. Clinical trial information: NCT02448771.