Molecular profiling of non-small-cell lung cancer patients with or without brain metastases included in the randomized SAFIR02-LUNG trial and association with intracranial outcome

•Intracranial response to lung cancer systemic treatments is inconsistent.•Lung cancer brain metastases harbor a specific molecular profile.•A 24-gene signature of primary tumor or metastasis tissue is associated with lung cancer brain metastasis.•EGFR amplification plays a major role in the brain dissemination process, even in lung cancers without EGFR mutations.•Intracranial disease control appears to be lower with immune checkpoint inhibitors than with other systemic treatments. Lung cancer remains the most frequent cause of brain metastases (BMs) and is responsible for high morbidity and mortality. Intracranial response to systemic treatments is inconsistent due to several mechanisms: genomic heterogeneity, blood–tumor barrier, and the brain-specific microenvironment. We conducted a study using data from the SAFIR02-LUNG trial. The primary objective was to compare the molecular profiles of non-small-cell lung cancer (NSCLC) with or without BMs. The secondary objective was to explore central nervous system (CNS) outcomes with various maintenance treatment regimens. In total, 365 patients harboring interpretable molecular data were included in this analysis. Clinical and biological data were collected. Genomic analyses were based on array-comparative genomic hybridization and next-generation sequencing (NGS) following the trial recommendations. Baseline genomic analyses of copy number variations identified a 24-gene signature specific to lung cancer BM occurrence, all previously known to take part in oncogenesis. NGS analysis identified a higher proportion of KRAS mutations in the BM-positive group (44.3% versus 32.3%), especially G12C mutations (63% versus 47%). Protein interaction analyses highlighted several functional interactions centered on EGFR. Furthermore, the risk of CNS progression was decreased with standard pemetrexed maintenance therapy. The highest rate of CNS progression was observed with durvalumab, probably because of the specific intracranial immune microenvironment. This work identified a 24-gene signature specific to lung cancer with BM. Further studies are needed to precisely determine the functional implications of these genes to identify new therapeutic targets for the treatment of lung cancer with BM.