Brain MRI imaging characteristics predict treatment response and outcome in patients with de novo brain metastasis of EGFR-mutated NSCLC

Research output: Contribution to journalArticle


Patients with non-small cell lung cancer (NSCLC) and de novo brain metastasis (BM) have poor prognosis. We aim to investigate the characteristic of brain magnetic resonance (MR) imaging and the association with the treatment response of epidermal growth factor receptor-tyrosine kinase inhibitors (EGFR-TKIs) for lung cancer with BM.EGFR-mutated NSCLC patients with BM from October 2013 to December 2017 in a tertiary referral center were retrospectively analyzed. Patient's age, sex, cell type, EGFR mutation status, treatment, and characteristics of BM were collected. Survival analysis was performed using Kaplan-Meier method. The efficacy of different EGFR-TKIs were also analyzed.Among the 257 eligible patients, 144 patients with Exon 19 deletion or Exon 21 L858R were included for analysis. The erlotinib group had the best progression free survival (PFS) (median PFS 13 months, P = .04). The overall survival (OS) revealed no significant difference between three EGFR-TKI groups. Brain MR imaging features including tumor necrosis, rim enhancement and specific tumor locations (frontal lobe, putamen or cerebellum) were factors associated with poor prognosis. Patients with poor prognostic imaging features, the high-risk group, who received erlotinib had the best PFS (median PFS 12 months, P < .001). However, the OS revealed no significant difference between 3 EGFR-TKI groups. The low risk group patients had similar PFS and OS treated with three different EGFR-TKIs.In NSCLC patients with common EGFR mutation and de novo BM, those with poor prognostic brain MR characteristics, erlotinib provided better PFS than afatinib or gefitinib.

Original languageEnglish
Article numbere16766
JournalMedicine (United States)
Issue number33
Publication statusPublished - 2019 Aug 1


All Science Journal Classification (ASJC) codes

  • Medicine(all)

Cite this