Identification of a predictive gene signature to recMAGE A3 antigen-specific cancer immunotherapy in metastatic melanoma and non-small-cell lung cancer

Purpose: To evaluate the presence of a gene expression signature present before treatment as predictive of response to treatment with MAGE‑A3 immunotherapeutic in metastatic melanoma patients and to validate its predictivity in adjuvant therapy of early-stage lung cancer. Patients were participants in two Phase II studies of the recombinant MAGE‑A3 antigen combined with immunological adjuvants. mRNA from tumor samples (biopsies) collected before MAGE-A3 immunotherapy was analyzed by microarray hybridization and by quantitative polymerase chain reaction (qRT-PCR). The melanoma microarray dataset was used to discover and crossvalidate a gene expression signature and classifier discriminative of Responders (R) versus Non-Responders (NR) patients; the gene signature and classifier were then applied to an adjuvant lung cancer study. Patients that were not included for analysis are denoted as NE (Non-evaluable). GSK Biologicals