EGFR gene copy number alterations are not a useful screening tool for predicting EGFR mutation status in lung adenocarcinoma

We investigated if gene copy number (GCN) alterations of the epidermal growth factor receptor (EGFR), as detected by silver enhanced in situhybridisation (SISH), could be used to select patients for EGFRmutation testing. Resected lung adenocarcinoma specimens with adequate tumour were identified. EGFRSISH was performed using the Ventana Benchmark Ultra platform. EGFRGCN was classified according to the Colorado Classification System. EGFRmutations were scanned by high resolution melting and confirmed by Sanger sequencing. Thirty-four of 96 tumours were EGFRSISH positive (35%), and 31 of 96 tumours harboured one or more EGFRmutations (32%). Of 31 EGFR-mutant tumours, 18 were EGFRSISH positive (58%). There was a statistically significant relationship between the presence of an EGFRmutation and EGFRGCN (p=0.003). Thirteen of 31 EGFR-mutant tumours were EGFRSISH negative (42%), and 16 of 65 EGFR-wild type tumours were EGFRSISH positive (24%). The sensitivity, specificity, positive predictive value and negative predictive value were 58%, 75%, 52.9% and 79%, respectively. Despite a significant relationship between EGFRGCN alterations and EGFRmutations, our results indicate that EGFR GCN as detected by SISH is not a suitable way to select patients for EGFRmutation testing.