Pooled Analysis of Gene Expression and Associated Treatment Response in the MONALEESA-2, -3, and -7 Trials

Improvements in OS have been seen in some trials with CDK4/6 inhibitors, but not all. With the differences in CDK4 versus CDK6 inhibition among ribociclib and other CDK4/6 inhibitors, the association between cell cycle–related genes and outcomes (based on pooled analysis of gene expression using tumor samples from the MONALEESA trials) was evaluated.

Data for gene expression were generated using pretreatment archived tumor samples (primary, 73%; metastatic, 27%) with a custom NanString nCounter panel of 781 genes, including cycle cycle–related genes, along with those in other signaling pathways and those relevant to breast cancer biology. Samples were pooled from 1139 pre- and postmenopausal patients with HR-positive/HER2-negative advanced breast cancer across the 3 MONALEESA trials including patients on first- and second-line therapy, with data categorized into training (80%) and test (20%) data sets. The training data set was used to analyze each gene individually (with continuous modeling) for association with progression-free survival (PFS). Genes with a gene × treatment interaction P value of less than .10 were evaluated in the test data set, and genes or gene signatures were classified based on expression level (low, medium, or high). For each subset, median PFS was calculated using the Kaplan-Meier method, and hazard ratios (HRs) of treatment benefit were estimated. A Cox proportional hazards model was used to adjust for clinical covariates, and a machine learning approach, which used available gene expression data and select clinical factors and their interactions with treatment arms, was applied to predict PFS.

Gene expression levels of CDKN2B and the expression ratio of CCND1/CDKN2A revealed a predictive relationship with benefit from ribociclib in both the training and test data sets. PFS benefit with ribociclib was found to be consistent regardless of CDK4/6 expression ratio or level of expression of CCNE1, CDK2, RB1, combined cell cycle–related genes, E2F gene signatures, RB gene signature, combined DNA-replication genes, or combined proliferation-related genes. A clinico-genomic signature found to be prognostic for PFS benefit with ribociclib was revealed. Patients with a low signature score had a longer median PFS versus those with a high signature score in both the ribociclib (HR, 0.37; 95% confidence interval [CI], 0.22-0.62) and placebo (HR, 0.30; 95% CI, 0.15-0.59) arms.

This largest pooled analysis to date of the association of gene expression data with CDK4/6 inhibitor treatment response in patients with HR-positive/HER2-negative advanced breast cancer revealed PFS benefit with ribociclib plus ET versus placebo plus ET to be consistent regardless of expression levels of most cell cycle–related genes. Variation in the magnitude of benefit with ribociclib was observed depending on CDKN2B expression levels, CCND1/CDKN2A expression ratio, and machine learning–derived signature scores; however, it should be noted that the CDK4/6 inhibitor signature requires validation in additional data sets.

Source:

Bardia A, Su F, Solovieff N, et al. Pooled gene expression analysis and association with treatment response in patients with HR+/HER2- advanced breast cancer in the MONALEESA-2, -3, and -7 trials. San Antonio Breast Cancer Symposium 2022. Abstract PD17-08.