Background: Cyclin-dependent kinases 4 and 6 inhibitors (CDK4/6i) are now standard of care (SOC) for pts with HR+/HER2- MBC. However, aside from HR positivity, there are no biomarkers for pt selection. Although mutations (muts) in genes such as RB1, FAT1, or FGFR, are associated with CDK4/6i resistance, they are rare events in HR+/HER2- MBC. The goal of this study is to examine the ctDNA mut profile, and the impact of commonly mutated genes (ESR1, TP53 and PIK3CA) on progression free survival (PFS) in pts received pal as SOC in a real world experience. Methods: Chart review was performed for pts with HR+/HER2- MBC who received pal and ctDNA Guardant360® (G360) testing between 01/2015 and 06/2020 at Siteman Cancer Center. The Kaplan-Meier method was used to generate time-to-event curves with calculated mPFS. A stratified log-rank test was used for all comparisons with P-values reported. Hazard ratios (HRs) and associated 95% confidence intervals (CI) were calculated with a stratified Cox proportional-hazards model. Results: Among the 73 pts identified, median age was 61.5 years (range 33.7-83.3). 15 (20.5%) pts received pal as 1st-line, 22 (30.1%) pts as 2nd-line, and 36 (49.3%) pts as 3rd-line and beyond. All 73 pts had muts by G360. Top 3 mutated genes are PIK3CA (n=25, 34.2%), TP53 (n=20, 27.4%) and ESR1 (n=17, 23.3%), followed by GATA3 (n=12, 16.4%) and MYC (n=6, 8.2%). For association analysis with PFS, 17 pts were excluded from further analysis due to having had multiple interim therapies between G360 testing and pal therapy, leaving 56 pts who had G360 immediately prior to pal (n=11), during pal (n=13), at progressive disease (PD) on pal (n=20), or had 1 additional therapy after PD on pal (n=12). Among the 56 pts, the median age on pal was 62.8 years (36.6-78.4). Median duration on pal was 7.7 months (mo). Pal was received as 1st-line in 14 (25.0%), 2nd-line in 17 (30.4%) and 3rd-line plus in 25 (44.6%) pts. Muts in PIK3CA, TP53 and ESR1, were identified in 18 (32.1%), 16 (28.6%), and 13 (23.2%) pts, respectively, with rates similar to the entire 73-pt cohort. Among 13 ESR1 mutant pts, 7 had letrozole and 6 had fulvestrant as pal’s endocrine partner. Concurrent muts in 2 of 3 genes were observed in 11 (19.6%) pts (6 with PIK3CA and TP53, and 5 with ESR1 and TP53). The mPFS was 7.4 and 12.6 mo (p=0.03) in pts with or without PIK3CA mut, 7.9 and 12.8 mo (p=0.85) with or without ESR1 mut, and 16.0 and 13.3 mo (p=0.46), with or without TP53 mut, respectively. mPFS for pts with concurrent muts in 2 of the 3 genes was 7.4 mo, compared to 11.7 mo and 13.3 mo in pts with 1 or no mut in these 3 genes (p=0.46). In multivariate analysis that included age, visceral mets and line of therapy, the HRs for PFS were <1 for pts without mut in PIK3CA or ESR1 (Table 1). Pts with 1-5 and 6-10 mut were 26% and 36% less likely to have PD compared to pts with >10 muts, respectively. Conclusions: In this retrospective real world experience, ctDNA muts in PIK3CA and ESR1 as well as higher number of muts in ctDNA were associated with worse outcome in pts received pal. Although several studies have shown CDK4/6i benefit regardless of PIK3CA or ESR1 mut status, our data indicates that muts in these genes render worse prognosis on CDK4/6i and supports PIK3CA and ESR1 muts as driver events in HR+ MBC. PI3Ki or novel ER targeted agents are combined with CDK4/6i in clinical trials and results are eagerly awaited. In addition, further study for pts with higher ctDNA mut number is warranted.

Multivariate Cox Proportional Hazards Regression Model on PFS
VariablesPFS
HR (95% CI)P-value
ESR1 (non-mut vs. mut) 0.64 (0.26-1.56) 0.32 
TP53 (non-mut vs. mut) 1.70 (0.80-3.60) 0.16 
PIK3CA (non-mut vs. mut) 0.42 (0.17-1.02) 0.06 
Age (<65 vs. >=65 yo) 0.92 (0.45-1.87) 0.81 
No. Lines of pal (1st- vs. 3rd-line) 0.61 (0.25-1.48) 0.28 
No. Lines of pal (2nd- vs. 3rd-line) 0.46 (0.20-1.04) 0.06 
No. of muts (1-5 vs. >10) 0.74 (0.24-2.33) 0.61 
No. of muts (6-10 vs. >10) 0.64 (0.20-2.07) 0.46 
Visceral mets (no vs. yes) 1.25 (0.52-2.98) 0.61 
Multivariate Cox Proportional Hazards Regression Model on PFS
VariablesPFS
HR (95% CI)P-value
ESR1 (non-mut vs. mut) 0.64 (0.26-1.56) 0.32 
TP53 (non-mut vs. mut) 1.70 (0.80-3.60) 0.16 
PIK3CA (non-mut vs. mut) 0.42 (0.17-1.02) 0.06 
Age (<65 vs. >=65 yo) 0.92 (0.45-1.87) 0.81 
No. Lines of pal (1st- vs. 3rd-line) 0.61 (0.25-1.48) 0.28 
No. Lines of pal (2nd- vs. 3rd-line) 0.46 (0.20-1.04) 0.06 
No. of muts (1-5 vs. >10) 0.74 (0.24-2.33) 0.61 
No. of muts (6-10 vs. >10) 0.64 (0.20-2.07) 0.46 
Visceral mets (no vs. yes) 1.25 (0.52-2.98) 0.61 

Citation Format: Jing Xi, Nusayba Bagegni, Foluso Ademuyiwa, Ashley Frith, Rama Suresh, Anna Roshal, Caron Rigden, Leonel Hernandez-Aya, Lindsay Peterson, Mateusz Opyrchal, Katherine Clifton, Katherine Weilbaecher, Ron Bose, Cynthia Ma. Circulating tumor DNA (ctDNA) mutation (mut) profile in relation to paboclicib (pal) efficacy in hormone receptor positive (HR+) and HER2 negative (HER2-) metastatic breast cancer (MBC) [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS2-18.