Predicting risk of early relapse with circulating tumour DNA for advanced ER+ breast cancer patients in the PALOMA-3 trial
Year: 2019
Session type: Poster / e-Poster / Silent Theatre session
Abstract
Background
CDK4/6 inhibitors in combination with endocrine therapy have become standard of care in advanced oestrogen receptor positive (ER+) breast cancer. Many patients respond well to these treatments, but a subset relapse early. We analysed pre-treatment circulating tumour DNA (ctDNA) samples from the PALOMA-3 trial to identify prognostic and predictive genomic markers.
Method
The phase III PALOMA-3 trial randomised 521 patients with advanced ER+/HER2- breast cancer 2:1 to receive fulvestrant plus either palbociclib (P+F) or placebo (F). Mutations in plasma DNA were assessed with a 17 gene sequencing panel. Circulating tumor fraction (TF) and copy number gain in 11 genes was assessed with a second panel, TF being inferred from 827 SNPs. Identified genomic markers were associated with clinical characteristics using chi-squared and Cochran-Armitage testing, and progression free survival (PFS) with a Cox multivariable model.
Results
Mutations, TF and copy number data were available for 310 patients (203 P+F, 107 F). In a multivariable analysis including treatment with palbociclib as a variable, the markers remaining significantly associated with poorer PFS included TF (HR 1.20, 95% CI 1.09 – 1.32, p = 0.0001, HR per 10% increase in purity), TP53 mutation (HR 1.84, 95%CI 1.27 – 2.65, p = 0.0011) and FGFR1 gain (HR 2.91, 95%CI 1.61 – 5.25, p = 0.0004). There was no significant interaction with palbociclib treatment. One or more of these factors was identified in 131/310, 42.3% of patients. Mutations in TP53 were significantly associated with number of disease sites (q = 0.0086), soft tissue/LN metastases (q = 0.042) and visceral metastases (q = 0.046).
Conclusion
Circulating tumour fraction, FGFR1 gain and TP53 mutation were independently predictive of early relapse in PALOMA-3. With independent validation these factors could be used to identify high-risk groups for augmentation of treatment in clinical trials.