Abstract C075: The use of 4HF Cancer Data Miner platform for an in Silico pharmacogenomic study predicting tumor response to the CDK4/6 inhibitor Palbociclib

Abstract

4HF Biotec recently developed an in-silico platform dedicated to cancer data investigations. It connects cancer-related molecular and drug sensitivity information enabling target and biomarker discovery. We used the platform to perform an integrative pharmacogenomic study for the CDK4/6 inhibitor Palbociclib. CDK4/6 inhibitors are molecules blocking E2F1 mediated G1/S cell cycle progression. Palbociclib is already approved in the clinic for treatment of HR-positive, ERBB2 negative advanced and metastatic breast cancer. Clinical trials in other tumor types such as leukemia, head and neck, ovarian and non-small cell lung cancer are also ongoing. In this context, large preclinical studies may help for identification of companion diagnostics improving patient selection. In a data-driven approach, we searched for molecular events associated with cancer response and resistance to Palbociclib. A publicly available dataset was used that contains 843 tumor cell lines which were tested for response to Palbociclib in a 2D monolayer assay in vitro (GDSC). Palbociclib IC50s values were integrated to our platform and were evaluated for associations with the molecular features of the cell lines including mutated genes (whole exome sequencing), gene deletions or amplifications (SNP6.0 microarrays) and gene expression (Affymetrix HGU133 plus 2.0) or combinations. The resulting biomarker candidates were analyzed for predictivity in an independent dataset of 142 tumor cell lines obtained from 4HF Biotec (internal data). In vitro, the best efficacies were observed in cell lines from hematological cancers (leukemia and lymphoma) and solid tumors such as renal, neuroblastoma, head and neck cancers. Very sensitive cell lines were also seen in most of the other tumor types tested. Surprisingly, the panel of breast cancer cell lines was among those with the lowest proportion of sensitive cell lines (6 out of 43) with those derived from small cell lung and uterus cancers. Statistical analyses identified alterations in RB1 gene by far the most prominent biomarker associated with resistance to Palbociclib (adjusted p value 2.53.10-9) and that can occur by gene deletions, mutations or loss of expression. Other genomic alterations such as the loss of chromosome 9 containing CDKN2A and CDKN2B were associated with increased response rates. In contrast, cell lines with chromosome amplification in 17q containing ERBB2 or chromosome 1 containing MYCL were mainly resistant to the compound (p Citation Format: Vincent Vuaroqueaux, Hoor Al Hasani, Gerhard Kelter, Hans R Hendriks, Heinz-Herbert Fiebig. The use of 4HF Cancer Data Miner platform for an in Silico pharmacogenomic study predicting tumor response to the CDK4/6 inhibitor Palbociclib [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics; 2019 Oct 26-30; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2019;18(12 Suppl):Abstract nr C075. doi:10.1158/1535-7163.TARG-19-C075