CHICAGO – Researchers are beginning to unravel the genomics of squamous cell lung cancer, revealing a disease characterized by complex genomes with frequent and unique rearrangements.
Exome and RNA sequence analyses of 178 patients identified 48,690 nonsilent mutations in total, Dr. Ramaswamy Govindan reported at the annual meeting of the American Society of Clinical Oncology.
Patrice Wendling/IMNG Medical Media
Dr. Ramaswamy Govindan
"This is not a disease like CML [chronic myeloid leukemia] with one mutation," he said. "This is a tobacco-related lung cancer with an average of 228 nonsilent mutations per tumor."
Squamous cell lung cancer was found to have 8.3 somatic mutations/megabase, far surpassing, for example, the 0.5 mutations/megabase found in acute myeloid leukemia. The average number of mutations was 360/tumor.
"It’s quite significant the amount of mutational burden," Dr. Govindan remarked. "Granted many of them are passenger mutations, but still it’s a fairly disordered genome."
Dr. Govindan and his fellow researchers with the Cancer Genome Atlas (TCGA) Lung Cancer Project are attempting to characterize the poorly understood genomic and epigenomic landscape of lung squamous cell carcinoma (LSCC), and to identify potential therapeutic targets. No molecularly targeted therapy has been approved for use in LSCC, which accounts for roughly 30% of lung cancer deaths or 45,000 deaths/year in the United States.
The researchers hope to sequence about 1,000 lung cancers in the next year, with data presented on 178 LSCC patients, most of whom smoked (96%), were male (74%), and had early stage I/II disease (76%). Their median age was 68 years.
The tumor protein 53 (TP53) gene was almost universally altered in the cohort, along with frequent loss of cyclin-dependent kinase inhibitor 2A (CDKN2A) function, said Dr. Govindan, an oncologist/hematologist and professor of medicine with Washington University School of Medicine in St. Louis. Other significantly mutated genes were phosphatase and tensin homlog (PTEN), Kelch-like ECH-associated protein 1 (KEAP1), nuclear factor-erythroid 2 related factor 2 (NFE2L2), human leukocyte antigen-A (HLA-A) and phosphoinositide-3-kinase catalytic alpha (PIK3CA).
Therapeutic targets were identified in 127 patients or roughly three-fourths of patients with LSCC. "So it’s really rich in targets," he said.
Most of the samples had distinct genes that are significant in terms of therapy and that are altered in a mutually exclusive fashion. Targets include the fibroblast growth factor receptors (FGFR), phosphoinositide-3 (p13) kinase pathway (47%), epidermal growth factor receptor (EGFR)/erythroblastic leukemia viral oncogene homolog 2 (ErbB-2), and the cyclin-cyclin dependant kinase complexes.
Dr. Govindan recommended that existing drugs should be studied in patients with squamous cell lung cancer, but cautioned that, "This doesn’t necessarily mean a drug that is currently inhibiting a p13 kinase will work in squamous cell lung cancer. That requires well-coordinated clinical trials."
Discussant Dr. Tetsuya Mitsudomi, chief of thoracic surgery at Aichi Cancer Center Hospital in Nagoya, Japan, said the research has identified several existing drug targets that should be tested in clinical trials. He pointed to drugs such as the experimental FGFR inhibitor BGJ398 for patients with FGFR1 amplification, p13K inhibitors for PIK3CA mutations, sunitinib (Sutent) for platelet-derived growth factor receptor–alpha amplification or mutation and dasatinib (Sprycel) to target discoidin domain receptor 2 (DDR2) mutations. A response to dasatinib and erlotinib was recently reported in a patient with squamous cell lung cancer harboring a DDR2 mutation (Cancer Discov. 2011;1:78-89), with a phase II trial planned in patients with advanced tumors harboring this mutation (NCT01514864).
The researchers also conducted whole genome sequencing on 19 tumors, detecting an average of 165 rearrangements/tumor. This is far more than has been seen in the TCGA database for breast or colon cancer, Dr. Govindan said.
The analysis also identified frequent loss of CDKN2A function through multiple mechanisms: homozygous deletion in 30%, methylation in 21%, and mutation in 17%. More important, the mechanisms are complimentary, and there is somatic rearrangement resulting in loss of the CDKN2A locus, Dr. Govindan said. The findings are significant because, although CDKN2A loss may result in the same type of phenotype, tumor response to a specific therapy may differ based on the inactivation mechanism.
"Alterations in CDKN2A are significant and we should really look at this in terms of therapy," he said later in the presentation.
mRNA expression profiling confirmed a recent report that lung squamous cell carcinoma is composed of four biologically distinct mRNA expression subtypes, suggesting the need for different therapies (Clin. Cancer Res. 2010;16:4864-75).
In the current analysis, the classical subtype was present in 36% of samples and enriched with p13K alterations, hypermethylation and the highest rate of tobacco use. In contrast, the basal subtype (25%) was characterized by neurofibromin 1 loss, the secretory (24%) by platelet-derived growth factor (PDGF) receptor-alpha alterations and the primitive subtype (15%) with PTEN and retinoblastoma 1 mutations.