Abstract PS4-03-02: P-climb: pet-ct and liquid biopsy for monitoring breast cancer bone metastasis

Abstract Background: Bone metastases (mets) develop in 50-70% of patients with metastatic breast cancer (MBC) and are associated with significant morbidity due to skeletal-related events (SREs), associated with significant decline in quality of life. Of the 180,000 patients living with MBC in the United States, 30-45% have bone mets as their only site of disease (‘bone-only’/BO-MBC). Patients with BO-MBC are frequently excluded from clinical trials because bone mets are not ‘measurable.’ In a large nationwide cohort, using the FoundationOne®Liquid CDx circulating tumor DNA (ctDNA) assay we previously demonstrated that ctDNA was just as likely to be detected and ctDNA tumor fraction (TF) remained prognostic in patients with BO-MBC. We hypothesize that ctDNA TF in combination with PET-CT will provide superior monitoring of bone mets. Methods: In this contemporaneously enrolled, retrospectively analyzed study, patients with any hormone receptor subtype receiving any line of systemic therapy undergoing standard-of-care PET-CT were identified for blood collection within 37 days of each PET-CT for ctDNA analyses. Patients were categorized as {BO or bone-lymph node (BO/L-MBC)} versus {bone mets + visceral mets (BV-MBC)} based on disease status. Plasma was banked from 20mL of blood collected at the time of clinical blood draws and underwent analysis by the laboratory developed test FoundationOneMonitor ctDNA assay for research assessment of ctDNA TF (accounting for aneuploidy, variant allele frequency, fragment length, clonal hematopoiesis [CH] and genomic alterations for 324 genes at each timepoint). PET-CT scans were deidentified and reviewed by a study radiologist on VISAGE PACS to determine SUVmax (PET Maximum Standardized Uptake Value) plus number of skeletal/non-skeletal lesions. Outcome measures included progression-free survival (PFS) and time to next therapy (TTNT). Results: A total of 42 patients (21 BO/L-MBC, 21 BV-MBC) completed a total 160 PET-CT & ctDNA pairs(2-7 per patient). ctDNA TF was detected at first available blood draw for 38% (8/21) with BO/L- and 57% (12/21) BV-MBC respectively (Fisher exact p=0.27). Overall, ctDNA TF value demonstrated higher correlation to SUVmax in both BO/L-MBC and BV-MBC cohorts at T2 (Spearman rho range 0.55-0.60, both p≤0.01) than T1 (range 0.43-0.47, p=0.03 and 0.05), with no significant correlation between TF and total number of lesions by PET-CT. Primary analyses focused on pairs of baseline (T1) and next restaging (T2) scan. T1:T2 TF change was not significantly different between the two cohorts (Wilcoxon rank sum p=0.90). For the total study population, TF change and SUVmax change from T1:T2 were each significantly associated with clinician-interpreted imaging response (TF change: Kruskal-Wallis p=0.006; SUVmax change: ANOVA p=0.002). When evaluating TF change from T1:T2, patients whose TF remained increased (relative to stable or declined) had significantly shorter PFS (hazard ratio [HR] 3.7, 95% confidence interval [CI] 1.7-7.8, log-rank p