Suzanne Jenkins1, James Chih-Hsin Yang2, Pasi A Jänne3, Kenneth S Thress4, Karen Yu5, Rachel Hodge6, Susie Weston7, Simon Dearden8, Sabina Patel6, Mireille Cantarini7, Frances A Shepherd9. 1. AstraZeneca, Macclesfield, United Kingdom. Electronic address: Suzanne.Jenkins@astrazeneca.com. 2. National Taiwan University Hospital, National Taiwan University Cancer Center, Taipei City, Republic of China. 3. Dana-Farber Cancer Institute, Boston, Massachusetts. 4. AstraZeneca, Waltham, Massachusetts. 5. Roche Molecular Systems, Inc., Pleasanton, California. 6. AstraZeneca, Cambridge, United Kingdom. 7. AstraZeneca, Alderley Park, United Kingdom. 8. AstraZeneca, Macclesfield, United Kingdom. 9. Princess Margaret Cancer Centre, Toronto, Canada.
Abstract
INTRODUCTION: Osimertinib is an oral, central nervous system-active, EGFR tyrosine kinase inhibitor (TKI) for the treatment of EGFR T790M-positive advanced NSCLC. Here we have evaluated EGFR mutation frequencies in two phase II studies of osimertinib (AURA extension and AURA2). METHODS: After progression while receiving their latest line of therapy, patients with EGFR mutation-positive advanced NSCLC provided tumor samples for mandatory central T790M testing for the study selection criteria. Tumor tissue mutation analysis for patient selection was performed with the Roche cobas EGFR Mutation Test (European Conformity-in vitro diagnostic, labeled investigational use only) (Roche Molecular Systems, Pleasanton, CA). Patients should not have been prescreened for T790M mutation status. The cobas test results were compared with those of the MiSeq next-generation sequencing system (Illumina, San Diego, CA), which was used as a reference method. RESULTS: Samples from 324 and 373 patients screened for AURA extension and AURA2, respectively, produced valid cobas test results. The T790M detection rates were similar between AURA extension and AURA2 (64% and 63%, respectively). The pooled T790M rate was 63%, with no difference by ethnicity (63% for Asian and non-Asian patients alike) or immediately prior treatment with an EGFR TKI (afatinib, 69%; erlotinib, 69%; and gefitinib, 63%). A higher proportion of patients had T790M detected against a background of exon 19 deletions versus L858R mutation (73% versus 58% [p = 0.0002]). In both trials the cobas test demonstrated high sensitivity (positive percent agreement) and specificity (negative percent agreement) for T790M detection when compared with the next-generation sequencing reference method: positive percent agreement of 91% versus 89% and negative percent agreement of 97% versus 98%. CONCLUSIONS: In both trials, the rate of detection of T790M mutation in patients with advanced NSCLC was approximately 63% and was unaffected by immediately prior treatment with an EGFR TKI or ethnicity.
INTRODUCTION: Osimertinib is an oral, central nervous system-active, EGFR tyrosine kinase inhibitor (TKI) for the treatment of EGFR T790M-positive advanced NSCLC. Here we have evaluated EGFR mutation frequencies in two phase II studies of osimertinib (AURA extension and AURA2). METHODS: After progression while receiving their latest line of therapy, patients with EGFR mutation-positive advanced NSCLC provided tumor samples for mandatory central T790M testing for the study selection criteria. Tumor tissue mutation analysis for patient selection was performed with the Roche cobas EGFR Mutation Test (European Conformity-in vitro diagnostic, labeled investigational use only) (Roche Molecular Systems, Pleasanton, CA). Patients should not have been prescreened for T790M mutation status. The cobas test results were compared with those of the MiSeq next-generation sequencing system (Illumina, San Diego, CA), which was used as a reference method. RESULTS: Samples from 324 and 373 patients screened for AURA extension and AURA2, respectively, produced valid cobas test results. The T790M detection rates were similar between AURA extension and AURA2 (64% and 63%, respectively). The pooled T790M rate was 63%, with no difference by ethnicity (63% for Asian and non-Asian patients alike) or immediately prior treatment with an EGFR TKI (afatinib, 69%; erlotinib, 69%; and gefitinib, 63%). A higher proportion of patients had T790M detected against a background of exon 19 deletions versus L858R mutation (73% versus 58% [p = 0.0002]). In both trials the cobas test demonstrated high sensitivity (positive percent agreement) and specificity (negative percent agreement) for T790M detection when compared with the next-generation sequencing reference method: positive percent agreement of 91% versus 89% and negative percent agreement of 97% versus 98%. CONCLUSIONS: In both trials, the rate of detection of T790M mutation in patients with advanced NSCLC was approximately 63% and was unaffected by immediately prior treatment with an EGFR TKI or ethnicity.
Authors: Diego Cortinovis; Umberto Malapelle; Fabio Pagni; Alessandro Russo; Giuseppe Luigi Banna; Elisa Sala; Christian Rolfo Journal: Transl Lung Cancer Res Date: 2021-07