ORLANDO - Two sets of gene expression profiles predict response to a common lung cancer drug for patients that have no guiding indicators for their treatment now, scientists from The University of Texas MD Anderson Cancer Center report at the AACR 102nd Annual Meeting.

"The only way we have to predict erlotinib's treatment benefit for patients now is by checking for certain mutations and amplifications of the epidermal growth factor receptor (EGFR), and that only covers about 12 percent of patients," said John Heymach, M.D., Ph.D., associate professor in MD Anderson's Department of Thoracic/Head and Neck Oncology.

Erlotinib, known commercially as Tarceeva, inhibits EGFR.

"We know there are other groups of patients who benefit from erlotinib, but right now we can't identify them in advance," Heymach said. "These two biomarker sets have potentially broad impact by covering the 88 percent of patients who lack EGFR mutations."

Using a five-gene signature and a group of genes involved in a certain type of cell transition, Heymach and colleagues developed biomarkers that indicate treatment benefit for patients with normal EGFR. They also identified one gene in each set a potential targets for new drugs, Heymach said.

John Heymach, M.D., Ph.D., is an associate professor in MD Anderson's Department of Thoracic/Head and Neck Oncology.

(Photo Credit: MD Anderson)

The biomarkers were developed by analyzing data from the Biomarker-integrated Approaches of Targeted Therapy for Lung Cancer Elimination (BATTLE) clinical trial conducted at MD Anderson. BATTLE analyzed the use of pre-specified biomarkers to guide treatment among four options.

"Last year, we presented results on pre-specified markers to determine how effective they would be at predicting response to EGFR inhibition," Heymach said. "This year, we looked for gene responses that had not been chosen in advance for BATTLE."

The researchers analyzed core needle biopsies of 101 BATTLE clinical trial patients and also looked at 54 non-small cell lung cancer lines to find genes involved in epithelial-to-mesenchymal transition (EMT). Cancer cells that change from stationary cells (epithelial) that line organs into mobile mesenchymal cells become more invasive, resistant to treatment and are associated with metastasis.

The endpoint for BATTLE was disease control after 8 weeks. The team found:

• A predictive five-gene expression signature for patients treated with erlotinib who lacked EGFR mutations. Of those with the signature, 83 percent had disease control at eight weeks compared with zero patients who lacked the signature. The five genes are Lipocalin-2 (LCN2), NPR3, OGG1, TRIM72, and a gene of unknown function called C5orf23.
• Lipocalin-2 is involved in the EGFR pathway, Heymach said, so it's a promising potential target for therapy. Lipocalin-2 was associated with epithelial-type tumor cells.
• A set of EMT genes that predicted disease control by a different measure. Of those whose cells were still epithelial type, 64 percent had disease control at eight weeks. Only 10 percent of patients with mesenchymal-type cells reached that milestone.
• A gene called Axl was associated with mesenchymal-type cells and also is a potential therapeutic target.

Both sets of signatures will be tested prospectively in the upcoming BATTLE II clinical trial, along with markers from the PI3K-AKT pathway, EGFR signatures and KRAS mutations. The four treatment arms of the trial are erlotinib, sorafenib (Nexavar), erlotinib plus an AKT inhibitor and the AKT inhibitor with a MEK inhibitor.