Structure-based classification predicts drug response in EGFR-mutant NSCLC

Epidermal growth factor receptor (EGFR) mutations typically occur in exons 18–21 and are established driver mutations in non-small cell lung cancer (NSCLC)^1,2,3. Targeted therapies are approved for patients with ‘classical’ mutations and a small number of other mutations^4,5,6. However, effective therapies have not been identified for additional EGFR mutations. Furthermore, the frequency and effects of atypical EGFR mutations on drug sensitivity are unknown^1,3,7,8,9,10. Here we characterize the mutational landscape in 16,715 patients with EGFR-mutant NSCLC, and establish the structure–function relationship of EGFR mutations on drug sensitivity. We found that EGFR mutations can be separated into four distinct subgroups on the basis of sensitivity and structural changes that retrospectively predict patient outcomes following treatment with EGFR inhibitors better than traditional exon-based groups. Together, these data delineate a structure-based approach for defining functional groups of EGFR mutations that can effectively guide treatment and clinical trial choices for patients with EGFR-mutant NSCLC and suggest that a structure–function-based approach may improve the prediction of drug sensitivity to targeted therapies in oncogenes with diverse mutations.

Patients with classical EGFR mutations (L858R or exon 19 deletions (Ex19del)) show marked improvements in clinical outcomes when treated with first-, second- or third-generation tyrosine kinase inhibitors (TKIs)^4,5,6,11. Other EGFR mutations in the kinase domain (exons 18−21) have also been established as oncogenic drivers of NSCLC⁸. Patients with atypical EGFR mutations show heterogeneous and reduced responses to EGFR inhibitors^1,3,7,8,9,10, including osimertinib¹². Atypical EGFR mutations with US Food and Drug Administration (FDA)-approved treatments are EGFR S768I, L861Q and G719X, for which afatinib was deemed effective on the basis of retrospective studies^13,14,15, and the EGFR/MET bispecific antibody amivantamab for exon 20 insertions (Ex20ins)¹⁶. There are no clear established guidelines for EGFR TKI treatment for patients with atypical EGFR mutations without an FDA-approved TKI, often resulting in patients receiving chemotherapy. Clinical trial design and treatment of patients with atypical EGFR mutations often rely on mutated-exon location to predict treatment, although heterogeneity in drug sensitivity across a single exon has been observed^1,8,17. Therefore, there is an unmet clinical need to establish a system for classifying EGFR mutations that is predictive of drug sensitivity and for more robust clinical trial design.

To characterize the molecular landscape of EGFR-mutant NSCLC, we used 5 patient databases with genomic profiling (Methods), representing 16,715 patients with EGFR-mutant NSCLC. There were 11,619 patients in whom primary and/or co-occurring mutations were recorded on a per-patient basis (Extended Data Fig. 1a). Among those patients, 67.1% had classical EGFR mutations (L858R and/or Ex19del with or without T790M); 30.8% had atypical EGFR mutations, including Ex20ins (9.1%), atypical mutations (12.6%), or a complex mutation including an atypical mutation (9.1%); and 2.2% had a classical mutation with T790M and an atypical mutation (Fig. 1a, Extended Data Fig. 1b). Atypical EGFR mutations (n = 7,199) occurred primarily in exons 18 (23.7%) and 20 (20.9% insertions and 19.2% point mutations; Fig. 1b). Prevalent hotspots for atypical mutations were the P-loop (L718–V726, 13.6%) and the C-terminal loop of the αC-helix (A767–G779, 29.4%, Fig. 1c).