Abstract 3987: Predicting MCL1 inhibitor sensitivity in large cell line panels using a gene expression signature

• Published in: Cancer research (Chicago, Ill.) Vol. 79; no. 13_Supplement; p. 3987
• Main Authors: Wernitznig, Andreas, Rudolph, Dorothea, Samwer, Matthias, Schweifer, Norbert, Trapani, Francesca, Wunberg, Tobias, Arnhof, Heribert, Lee, Teakyu, Sensintaffar, John L., Olejniczak, Edward T., Benes, Cyril H., Fesik, Stephen W., Kraut, Norbert
• Format: Journal Article
• Language: English
• Published: 01.07.2019
• ISSN: 0008-5472; 1538-7445
• DOI: 10.1158/1538-7445.AM2019-3987

Abstract MCL1, an anti-apoptotic member of the BCL-2 family of proteins, is a key regulator of cancer cell survival and a known resistance factor to anti-cancer drugs, making it a highly desirable target for therapeutic intervention. Recently several MCL1 inhibitors have entered Phase I clinical development. Data derived from large cancer cell line panels suggest, that cell lines of hematopoietic origin are more broadly sensitive to MCL1 inhibition, than cell lines derived from solid tumor types. In particular, for the therapy of solid tumor patients with an MCL1 inhibitor, a patient selection biomarker would be highly desirable. Published in vitro data using various MCL1 inhibitors, siRNA or CRISPR/Cas9 technology show that tumor cell lines with low BCL-XL gene expression are mostly sensitive to MCL1 inhibition, down-regulation or inactivation. Here we show that by adding the gene expression data of additional six genes (including the MCL1 binding partner BAK1) to BCL-XL, a supervised learning predictor was applied and could reach a performance of almost 80% correctly classified solid tumor cell lines. This new predictor has been applied to either tumor samples, adjacent normal tissues or normal tissue samples from TCGA and GTEx. Briefly, most normal tissue samples are categorized as being sensitive. Moreover, solid tumor samples in contrast to solid tumor cell lines are predicted to be broadly sensitive to MCL1 inhibition with a predicted anti-tumor effect rate of 60% to 95%. In summary, our work describes the translational challenges using cell line-derived predictors on ex vivo tumor samples. Citation Format: Andreas Wernitznig, Dorothea Rudolph, Matthias Samwer, Norbert Schweifer, Francesca Trapani, Tobias Wunberg, Heribert Arnhof, Teakyu Lee, John L. Sensintaffar, Edward T. Olejniczak, Cyril H. Benes, Stephen W. Fesik, Norbert Kraut. Predicting MCL1 inhibitor sensitivity in large cell line panels using a gene expression signature [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 3987.