The sum of gains and losses of genes encoding the protein tyrosine kinase targets predicts response to multi-kinase inhibitor treatment: Characterization, validation, and prognostic value.

Validated predictive biomarkers for multi-tyrosine kinase inhibitors (MTKI) efficacy are lacking. We hypothesized that interindividual response variability is partially dependent on somatic DNA copy number alterations (SCNAs), particularly those of genes encoding the protein tyrosines targeted by MTKI (called target genes). Genomic alterations were investigated in MTKI responsive and non responsive patients with different histological subtypes included in the ProfiLER protocol (NCT 01774409). From March 2013 to August 2014, 58 patients with advanced cancer treated with one of 7 MTKIs were included in the ProfiLER trial and split into one discovery cohort (n = 13), and 2 validation cohorts (n = 12 and 33). An analysis of the copy number alterations of kinase-coding genes for each of 7 MTKIs was conducted. A prediction algorithm (SUMSCAN) based on the presence of specific gene gains (Tumor Target Charge, TTC) and losses (Tumor Target Losses, TTL) was conceived and validated in 2 independent validation cohorts. MTKI sensitive tumors present a characteristic SCNA profile including a global gain profile, and specific gains for target genes while MTKI resistant tumors present the opposite. SUMSCAN favorable patients achieved longer progression-free and overall survival. This work shows that the copy number sum of kinase-coding genes enables the prediction of response of cancer patients to MTKI, opening a novel paradigm for the treatment selection of these patients.