Abstract 2810: Computationally predicted sensitivity of clinical cohorts identifies drug relationships and biomarkers associated with response to PCM-075, a PLK1 selective inhibitor

Polo-like kinase 1 (PLK1) is a serine-threonine kinase which regulates various cellular processes, including mitosis, DNA replication, and the stress response, and is over-expressed in many malignancies. PCM-075, a PLK1 selective inhibitor, is currently in a phase1b/2 clinical trial in AML in combination therapy with low dose cytarabine (LDAC) (NCT03303339). To identify potential drug relationships and biomarkers associated with response to PCM-075, a two-step computational approach was conducted. Using a modification of the ‘imputed drug-wide association study9 (IDWAS) method(1), we first performed ridge regression model training on IC50 values and gene expression signatures for PCM-075 and 138 other drugs from the Genomics of Drug Sensitivity in Cancer (GDSC) database . In the second step, we applied these models to clinical patient datasets (The Cancer Genome Atlas (TCGA) populations containing 32 Cancer types and 9,968 total patients), generating a predicted sensitivity value to all drugs for each patient. Hierarchical clustering of imputed sensitivities across all patients showed that the PCM-075 activity profile is most closely related to cytarabine, etoposide and cisplatin, all of which are compounds that impact DNA replication and lead to apoptosis. Correlation of predicted response values was significant both within and between cancer types, with the acute myeloid leukemia (AML) cohort showing highest increased sensitivity to PCM-075, cytarabine and etoposide relative to the other 31 cancer types examined. Interestingly, other PLK1 inhibitors present in the GDSC (BI-2536 and GW843682X) do not closely cluster with PCM-075. Association of imputed TCGA sensitivity values with somatic variant data shows a high degree of concordance between predicted biomarkers, including TP53 mutation status, further suggesting a functional relationship. Given that this computational analysis suggests that similar pathways are affected by both cytarabine and PCM-075, we tested this experimentally by assessing the level of interaction between these two compounds within an AML cell line (HL-60). The combination resulted in statistically significant synergy (BRAID, k~2.7(1.31-4.25)) for anti-proliferative activity. The mechanistic cause of synergy may be related to DNA damage caused by cytarabine which is known to inhibit and degrade the PLK1 protein. In conclusion, we have identified a pathway signature that clusters PCM-075 with a subset of current oncology therapeutics. Synergy between these compounds in combination suggests the potential for use in treating AML and potentially other cancer types. 1.Genome Res. 27: 1743-1751 (2017). Citation Format: Penn Whitley, Peter J. Croucher, Barbara Valsasina, Dario Ballinari, Italo Beria, Jeffrey N. Miner, Mark G. Erlander. Computationally predicted sensitivity of clinical cohorts identifies drug relationships and biomarkers associated with response to PCM-075, a PLK1 selective inhibitor [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 2810.