61P - Deconvoluting the intra-tumour heterogeneity and subclonal evolution of CDK4/6 inhibitor resistance in ER+ breast cancer

Background

Subclonal heterogeneity and evolution are characteristics of breast cancer (BC), playing a key role in tumour development, progression, and resistance to current therapies. Single-cell (s.c) sequencing studies suggest that subclonal heterogeneity drives therapy response in BC. Here, we sought to identify whether pre-existing transcriptomically defined cell subpopulations underpin CDK4/6 inhibitors (CDK4/6i) resistance in ER+ BC, and to identify the differential resistance mechanisms between different CDK4/6i.

Methods

ER+ BC cell lines were infected with a lentiviral barcode library and were long-term exposed to DMSO, abemaciclib or palbociclib until resistance was achieved. Barcodes were detected at DNA and s.c RNA level coupled with s.c transcriptomic and whole exome sequencing, as well as protein analysis of the G1/S cell cycle checkpoint.

Results

S.c lineage tracing identified stochastic CDK4/6i resistance in MCF7 cells, while resistance in T47D cells was conserved. In all cell models, abemaciclib exerted a stronger selective pressure than palbociclib. Detection of barcodes from dead cells highlighted differential selective pressures of the drugs. S.c transcriptomic sequencing identified shared and differential mechanisms of resistance between cell lines and specific CDK4/6i, suggestive of different subclonal, non-genetic mechanisms driving resistance. Some of these subclones were pre-defined, whereas others were demarked by acquisition of pathway up-regulation. For instance, RB1 copy number loss was identified in palbociclib-resistant T47D cells, indicating RB1 as a genetic driver of resistance. Interestingly, only a subclonal cluster was identified with loss of RB1 expression in palbociclib-resistant MCF7 cells. Abemaciclib-resistant MCF7 cells lost RB1 protein levels, suggesting differential non-genetic mechanisms of resistance to CDK4/6i in MCF7 cells.

Conclusions

We identified that resistance to CDK4/6i can be stochastic or conserved as well as pre-existing or acquired according to the cell model, suggesting transcriptomic and epigenomic mechanisms of resistance, with drug-tolerant or drug-induced persistent properties.

Editorial acknowledgement

Clinical trial identification

Legal entity responsible for the study

The authors.

Funding

Pfizer.

Disclosure

All authors have declared no conflicts of interest.