Gene Expression Signatures Predict Erlotinib Sensitivity in NSCLC



ORLANDO – A set of epithelial-to-mesenchymal transition genes and a novel five-gene expression signature appear to predict disease control with erlotinib in refractory non–small cell lung cancer patients whether they have endothelial growth factor receptor mutations or not.

These candidate biomarkers have potentially broad impact, as they could help identify erlotinib (Tarceva) sensitivity in the 88% of patients with wild-type endothelial growth factor receptor (EGFR), Dr. John V. Heymach said at the annual meeting of the American Association for Cancer Research.

Dr. Heymach described the two gene profiles in an update of the phase II Biomarker-Integrated Approaches of Targeted Therapy for Lung Cancer Elimination (BATTLE) 1 trial. They were identified retrospectively from gene expression profiling of pretreatment core needle biopsies in 101 BATTLE trial patients and by studying 54 non–small cell lung cancer lines to find genes involved in the epithelial-to-mesenchymal transition (EMT).

Currently, the treatment benefit of erlotinib – a tyrosine kinase inhibitor that acts on EGFR and is approved for the treatment of non–small cell lung cancer (NSCLC) and pancreatic cancer - can be predicted in only about 12% of patients who have certain mutations and amplifications of EGFR, said Dr. Heymach of the University of Texas M.D. Anderson Cancer Center, Houston.

Previously reported results from the BATTLE 1 trial, which were presented at the 2010 AACR annual meeting, focused on prespecified markers as predictors of response to EGFR inhibition; these updated findings focused on novel gene markers that were not prespecified, but were discovered retrospectively from the biopsies taken as part of the earlier BATTLE trial work.

The investigators found a five-gene expression signature, including Lipocalin-2 (LCN2), NPR3, OGG1, TRIM72, and a gene of unknown function called C5orf23, which was predictive of disease control in patients treated with erlotinib who lacked EGFR mutations. Disease control occurred by 8 weeks in 83% of those with the signature, compared with 0% of patients who lacked the signature (P less than .001).

They also found that LCN2 was involved in the EGFR pathway, and was associated with epithelial-type tumor cells. The findings suggest it is a promising potential target for therapy.

The EMT genes that were found to predict disease control in this study did so by a different measure: Disease control by 8 weeks occurred in 64% of those with cells that were still epithelial type, while disease control occurred in only 10% of those with mesenchymal type (P = .02). A gene called Axl, which is a tyrosine kinase receptor, was found to be associated with mesenchymal-type cells, and could also be a potential therapeutic target, Dr. Heymach said.

The predictive value of both the five-gene expression signature and the EMT signature will be tested prospectively in the upcoming BATTLE II trial, which will also test markers from the P13K-AKT pathway, EGFR signatures, and KRAS mutations. The trial will have four treatment arms, including erlotinib, sorafenib (Nexavar), erlotinib plus an AKT inhibitor, and the AKT inhibitor with an MEK inhibitor.

Dr. Thomas J. Lynch Jr., the discussant following Dr. Heymach’s presentation of his findings during a late-breaking abstract session at the AACR meeting, praised the "mining of data" from the landmark BATTLE 1 trial, which resulted in these findings.

While the finding are important, perhaps their greatest value is in "elucidating new targets in non–small cell lung cancer more than necessarily determining who benefits from marginally active therapy," said Dr. Lynch, director of the Yale Cancer Center and physician-in-chief at Smilow Cancer Hospital at Yale-New Haven (Conn.).

The BATTLE 1 trial, conducted by a team of researchers at M.D. Anderson, is the first completed prospective, adaptively randomized study in heavily pretreated non–small cell lung cancer patients that mandated tumor profiling with real-time core needle biopsies. The results of the trial demonstrate the feasibility of this approach, and create a new paradigm for translational research, and they represent a substantial step toward realizing personalized lung cancer therapy, according to the investigators. The findings are published in the inaugural issue of Cancer Discovery, which debuted at the 2011 AACR conference (Cancer Discovery 2011;1:OF42-9).

This study was funded by the U.S. Department of Defense, the M.D. Anderson and University of Texas Southwestern Lung Cancer Specialized Program in Research Excellence, and M.D. Anderson’s Cancer Center Support Grant from the National Cancer Institute. Dr. Heymach disclosed that he has received research support from AstraZeneca and Bayer, and has served on advisory boards for AstraZeneca, Genentech, and Bayer.

Next Article: