AbstractPurpose: To identify potential predictors of response and resistance mechanisms in patients with hormone receptor–positive (HR+), HER2-negative (HER2−) advanced breast cancer (ABC) treated with the cyclin-dependent kinase 4 and 6 (CDK4/6) inhibitor abemaciclib ± endocrine therapy (ET), baseline and acquired genomic alterations in circulating tumor DNA (ctDNA) were analyzed and associated with clinical outcomes. Experimental Design: MONARCH 3: postmenopausal women with HR+, HER2− ABC and no prior systemic therapy in the advanced setting were randomly assigned to abemaciclib or placebo plus nonsteroidal aromatase inhibitor (NSAI). nextMONARCH: women with HR+, HER2− metastatic breast cancer that progressed on/after prior ET and chemotherapy were randomly assigned to abemaciclib alone (two doses) or plus tamoxifen. Baseline and end-of-treatment plasma samples from patients in MONARCH 3 and nextMONARCH (monotherapy arms) were analyzed to identify somatic genomic alterations. Association between genomic alterations and median progression-free survival (mPFS) was assessed. Results: Most patients had ≥1 genomic alteration detected in baseline ctDNA. In MONARCH 3, abemaciclib+NSAI was associated with improved mPFS versus placebo+NSAI, regardless of baseline alterations. ESR1 alterations were less frequently acquired in the abemaciclib+NSAI arm than placebo+NSAI. Acquired alterations potentially associated with resistance to abemaciclib ± NSAI included RB1 and MYC. Conclusions: In MONARCH 3, certain baseline ctDNA genomic alterations were prognostic for ET but not predictive of abemaciclib response. Further studies are warranted to assess whether ctDNA alterations acquired during abemaciclib treatment differ from other CDK4/6 inhibitors. Findings are hypothesis generating; further exploration is warranted into mechanisms of resistance to abemaciclib and ET. See related commentary by Wander and Bardia, p. 2008

Supplementary Figure S5. PLK1 targeting decreases the expression of ER signaling genes.

Supplementary Figure S1. RNAi screen to identify kinases that regulate both ER transcriptional activity and viability of MCF7/LTED cells.

Supplementary Figure S2. Validation of PLK1, RPS6KA2 and GSG2 knockdown.

Supplementary Table and Figure Legends. Legends for Supplementary Tables S1-S2 and Supplementary Figures S1-S6.

Supplementary Figure S4. Treatment with volasertib abrogates cell growth and induces apoptosis.

Supplementary Table S1. Off-target effect of ER siRNA on TK-Renilla activity.

Distribution of clinically actionable mutations according to PEPI score, PCNA proliferation signature and breast cancer subtypes

Supplementary Figure 9: Changes in sTILs after NAC do not correspond to an observed change in T cell clonality.

Ki67 expression in patient tumor specimens.

A. Subgroup analyses of percent change in Ki67 from baseline to week 2 were performed on KE population based on progesterone receptor status, tumor size, baseline disease stage, tumor grade, baseline lymph node involvement, and baseline Ki67 (index 20%). B. Ki67 changes at baseline, 2 weeks, and end of treatment. C. Geometric mean percentage suppression of Ki67 from baseline to end of treatment by treatment arm. D. Mean percentage rate of response as determined by complete cell cycle arrest (responder: Ki76<2.7) at end of treatment, by treatment arm. E. Effect of study treatment discontinuation on cell proliferation. F. Tissue expression of Ki67 from representative patients who were on treatment or off treatment for more than 4 days.

Median Ki67 expression by tumor grade.

Variant filtering algorithm and mutation allele frequency distribution a) Variant filter algorithm. b) Mutation Allele Frequency (MAF) histogram of 11861 non-synonymous variants. c) MAF of 330 non-synonymous variants after filtering out potential germline variants.

A. Analysis of the RNAseq samples. B. Top downregulated gene set enrichment analysis (GSEA) gene sets across treatment arms. C. Top upregulated GSEA gene sets in the combination treatment arm.

Dataset 1: Metadata and clonotype data for TCRβ sequencing (Adaptive) in 15 pairs of NAC-treated patients (tumor data).

E2F4 gene signature modulation and association with overall survival in ER+/HER2- breast cancer treated with endocrine therapy. a) E2F4 gene signature expression at baseline (pre) and after 2-weeks of aromatase inhibitor treatment (post), data from ACOSOGZ1031b study. Tumors were divided according to the E2F4 score at baseline in high, medium or low tertiles. After two-weeks of aromatase inhibitor treatment, there was a decrease in the E2F4 score in all groups. However, the proportion of tumors still expressing a high post-treatment E2F4 score was greater in patients with a high E2F4 score at baseline (60%) than the other groups. b) Overall survival in patients with ER+ breast cancer treated with adjuvant endocrine therapy in the METABRIC database (n=1498) according to E2F4 signature score tertiles. c) Overall survival in patients with ER+ tumors from METABRIC according to E2F4 score tertiles in Luminal A and d) Luminal B tumors.

Supplementary Figure S1. RNAi screen to identify kinases that sensitize cells to CDK4/6 inhibition. Supplementary Figure S2. Calculation of the sensitivity index (SI) value. Supplementary Figure S3. PDK1 and CDK4/6 inhibitors are synergistic against a panel of cancer cell lines. Supplementary Figure S4. Inhibition of the PI3K/PDK1 pathway results in durable responses and is well-tolerated in vivo. Supplementary Figure S5. Ribociclib-resistant cells are cross-resistant to other CDK4/6 inhibitors and upregulate P-PDK1. Supplementary Figure S6. Inhibition of PI3K/PDK1/AKT/mTOR signaling pathway resensitizes ribociclib-resistant cells to CDK4/6 inhibition.

Supplementary Figure S3. PLK1 knockdown decreases hormone-independent ER transcriptional activity.

Supplementary Table S1. Details of antibodies used in this study. Supplementary Table S2. Q-PCR primers. Supplementary Table S3. HALLMARK gene-set enrichment analysis (GSEA) of baseline tumor samples from the NeoPalANA trial. Supplementary Table S4. Palbociclib IC50 values in MCF7 P, MCF7 EDR, T47D P-LM, and T47D EDR palbociclib-resistant (PalboR)-LM cell lines. Supplementary Table S5. Top-20 Gene Ontology (GO) terms significantly enriched in PalboR cell lines. Supplementary Table S6. HALLMARK gene sets significantly up-regulated in cell lines with acquired resistance to palbociclib. Supplementary Table S7. RPPA of MCF7 P, MCF7 EDR, and their PalboR derivatives.