Journal of Clinical Oncology, 2005 ASCO Annual Meeting Proceedings.
Vol 23, No 16S (June 1 Supplement), 2005: 3010
© 2005 American Society of Clinical Oncology

This Article Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Download to citation manager Rights & Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Clark, E. Articles by Huang, F. Search for Related Content PubMed Articles by Clark, E. Articles by Huang, F.

Abstract

Identification of pharmacogenomic markers for predicting sensitivity to BMS-354825, a SRC/ABL kinase inhibitor

Bristol-Myers Squibb Co, Princeton, NJ

3010

Background: BMS-354825 is a SRC/ABL kinase inhibitor currently in clinical trials for both CML and solid tumors patients. Molecular markers predictive of response to BMS-354825 could assist in clinical development by identifying patients most likely to derive clinical benefit. Methods: Gene expression profiling were performed in 23 breast and 23 lung cancer cell lines for which sensitivity (IC₅₀) to BMS-354825 was pre-determined. Using the breast cell lines as a training set, 3 statistical analyses including: 1) KNN with a permutation test: 2) Pearson correlation between gene expression level and IC₅₀ values; and 3) a t-test were performed to identify 137 genes whose expression levels are highly correlated with drug sensitivity. Then, a predictive model, consisting of the top 6 genes correlated with drug sensitivity, was built using a weighted-voting algorithm (Whitehead GeneCluster program) to predict sensitivity to BMS-354825 in an independent test set (23 lung cancer cell lines). Biological studies were also conducted to validate these markers. Results: From the 137 genes whose expression levels are highly correlated with drug sensitivity (p<0.05), we have identified 64 genes that are modulated by BMS-354825 treatment. Using the top 6 genes, we correctly predicted the sensitivity of 19 (83%) out of 23 testing samples across tissue types. Co-regulated expression patterns of these 137 genes were observed in primary breast tumors indicating their potential utility as response predictors. Subsequent biological studies to inhibit SRC-family kinase activity by treatment with BMS-354825 and siRNA have suggested that many of these genes are involved in SRC-family signaling pathways. Immunohistochemical analysis of several top candidate markers demonstrated that the protein level of the markers is also correlated with the sensitivity of cell lines to BMS-354825, indicating both RNA and proteins are candidate markers for predicting response to BMS-354825. Conclusions: This work has identified pharmacogenomic biomarkers with utility in predicting response to BMS-354825 in cell lines. These markers have clinical impact and will be used in planned clinical studies.

No significant financial relationships to disclose.

Abstract presentation from the 2005 ASCO Annual Meeting