The paper presented by Magnuson et al. is a multiinstitutional retrospective study looking to the treatment of brain metastases among patients with tyrosine kinase inhibitor naive epidermal growth factor mutant non-small cell lung cancer: they concluded that deferral of radiotherapy is associated with inferior survival and the best approach is radiosurgery (RS) followed by EGFR-TKI to avoid the neurocognitive toxicity of whole brain radiotherapy (WBRT) (1). This is currently a hot question in the literature with those favoring a combined approach and those favoring only TKI. This is certainly not an easy question as there are many possible endpoints to evaluate the efficacy of the treatment: local response, time to brain progression, quality of life, toxicity and overall survival, the ultimate endpoint. The later will directly be influenced not only by the brain tumor control but also by the extracranial metastatic disease and the response to the systemic treatment which may explain the differences reported in the literature. We also must point out the fact that all studies are retrospective with all the possible biases due to this approach: the choice of treatment is often related to the practitioner both for radiotherapy and the type of treatment (WBRT or RS) and the treatment timing. In a recent paper, Doherty et al. has observed a longer time to intracranial progression (TTIP) with WBRT compared to RS or TKI but no difference in survival suggesting that WBRT may be deferred until cranial progression: nevertheless, the rate of patients with brain symptomatic lesions was higher in the WBRT group, 52% vs. 16% for RS and 11% for the TKI (2). A similar observation in Byeon series was in favor of only TKI (3). Other series may include only asymptomatic patients such as in Liu's paper (4).A first comment is certainly the changing pattern of outcome for this group of patients: median overall survival for the three cohorts of patients range from 25 months for the TKI only group to 30 and 46 months for the combined approach with WBRT or RS. This is major improvement in contrast to the results observed in the study of Gaspar et al. with patients included in three randomized trials of the RTOG: using the recursive partitioning analysis (RPA) including four variables (age, performance status, control of primary disease and extracranial metastases); a median survival time of 7.7 months was observed for the best group and less than 3 months for the worse (5). In our experience with the Leksell Gamma Knife RS and using the basic score for brain metastases which includes the performance status, the presence or absence of extracranial disease, the control or not of the primary disease, the median survival rose from 2.4 months for the worse subgroup to more than 30 months for the best subgroup (6).This major improvement in patient's survival is partially related to the patient selection but also outline the efficacy of the current systemic treatment, especially the targeted agents in a population with EGFR mutant adenocarcinoma...