Objective: While several prognostic models have been presented in NSCLC patients with brain metastasis, none of these models have included molecular markers as an index. The aim of our study was to evaluate the prognostic value of EGFR mutations and to integrate these EGFR mutations into the prognostic index in NSCLC patients with brain metastasis.
Materials and methods: We analyzed retrospectively 292 lung adenocarcinoma patients with brain metastasis. Clinico-pathological features and overall survival (OS) were compared between patients with EGFR mutations and patients with EGFR wild type. EGFR mutation status was integrated with lung specific graded prognostic assessment (GPA) score.
Results: Among 292 patients, EGFR mutation status was tested in 183 patients. One hundred and five patients (57.4%) had EGFR activating mutations, 14 (7.7%) had EGFR non-activating mutations and 64 (35.0%) had EGFR wild type. OS was significantly longer in patients with EGFR activating mutations than in those with EGFR wild type patients (20.4 vs. 10.1 months, p = 0.002). However, patients with EGFR non-activating mutations did not show superior OS compared with EGFR wild type patients (14.6 vs. 10.1 months, p = 0.83). Multivariate analysis revealed that the presence of EGFR activating mutation is an independent positive prognostic factor for OS (adjusted hazard ratio 0.56, p = 0.002).
Conclusions: EGFR activating mutations have a prognostic role in lung adenocarcinoma patients with brain metastasis that is independent of other known prognostic factors. The frequency of EGFR mutation was higher than expected. The presence of EGFR activating mutations should be included as an index in the prognostic models for lung adenocarcinoma patients with brain metastasis.
Keywords: Brain metastasis; EGFR mutation; GPA score; Lung cancer; adenocarcinoma; prognosis.
Copyright © 2014 Elsevier Ireland Ltd. All rights reserved.