CDK4/6 Inhibition and Radiation as a Treatment Strategy to Improve Local Disease Control in Breast Cancers With Poor Prognoses

Abstract

Locoregional control remains an issue in women with multi-node positive estrogen receptor-positive (ER+) breast cancer and many women with triple-negative breast cancer (TNBC). There is growing evidence that CDK4/6 inhibition (CDK4/6i) sensitizes breast cancer cells to ionizing radiation (RT) by suppressing the DNA damage response, but the role of RB is currently unclear. We sought to understand the role of RB and the implications of RB loss in ER+ and TNBC.Clonogenic survival assays were used to calculate radiosensitivity and calculate enhancement ratios (rER). Cellular viability was quantified 72 hours after drug addition to calculate half-maximal inhibitory concentrations (IC50s). Phospho- and total RB expression levels were assessed by immunoblotting. DNA repair was assessed with γH2AX and RAD51 immunofluorescence. GFP-based plasmid reporter systems were used to assess homologous recombination (HR) and non-homologous end joining (NHEJ) competency. Isogenic RB1 knockout cells were generated with CRISPR-Cas9, and siRNA was used for transient RB1 knockdown. G1 cell cycle arrest was quantified using propidium iodide-based flow cytometry. In vivo efficacy of CDK4/6 inhibition + RT was assessed using MDA-MB-231 xenografts.Four TNBC cell lines with intact RB expression were radiosensitized (rER 1.08 - 2.22) after CDK4/6i with palbociclib, ribociclib, or abemaciclib, but two RB null TNBC models were not radiosensitized with combination treatment (rER: 0.84 - 1.00). In ER+ and TNBC cell lines, response to CDK4/6i + RT was accurately predicted by the presence or absence of RB protein, and higher RB expression led to increased radiosensitization of breast cancer cell lines at lower doses. In addition, pRB expression decreased in wild type TNBC cell lines treated with CDK4/6i + RT. HR was suppressed with CDK4/6i in RB wild type - but not mutant - cell lines. NHEJ efficiency in RB wild type TNBC was unchanged with CDK4/6i. Genetic knockdown of RB1 led to the loss of radiosensitization in ER+ and TNBC cell lines (rER: 0.84 - 1.13) as well as a decrease in sensitivity to CDK4/6i monotherapy. Parental and Cas9 control TNBC cells arrested in G1 24 hours after CDK4/6i and remained arrested at 48 hours, but RB1 CRISPR TNBC cells did not arrest after drug treatment. In a TNBC xenograft model with wild type RB expression (MDA-MB-231), CDK4/6i + RT led to significant radiosensitization in vivo and delayed time to tumor doubling.RB loss mitigates CDK4/6 inhibitor-mediated radiosensitization of both ER+ and TNBC cells. Loss of RB expression prevents CDK4/6i-mediated radiosensitization, suggesting that RB expression may be a valuable clinical biomarker to predict efficacy of CDK4/6i + RT in future clinical trials. Our data also provide the preclinical rationale for CDK4/6i with RT not just in ER+ BC, but also in women with TNBC that express RB.