Associations between single-nucleotide polymorphisms in the PI3K-PTEN-AKT-mTOR pathway and increased risk of brain metastasis in patients with non-small cell lung cancer.

PURPOSE Non-small cell lung cancer (NSCLC) metastasizes fairly often to the brain, but identifying which patients will develop brain metastases is problematic. The phosphoinositide 3-kinase (PI3K)-AKT-mTOR signaling pathway is important in the control of cell growth, tumorigenesis, and cell invasion. We hypothesized that genotype variants in this pathway could predict brain metastasis in patients with NSCLC. METHODS We genotyped 16 single-nucleotide polymorphisms (SNP) in five core genes (PIK3CA, PTEN, AKT1, AKT2, and FRAP1) by using DNA from blood samples of 317 patients with NSCLC, and evaluated potential associations with the subsequent development of brain metastasis, the cumulative incidence of which was estimated with Kaplan-Meier analysis. Multivariate Cox regression analysis was used to analyze correlations between genotype variants and the occurrence of brain metastasis. RESULTS In analysis of individual SNPs, the GT/GG genotype of AKT1: rs2498804, CT/TT genotype of AKT1: rs2494732, and AG/AA genotype of PIK3CA: rs2699887 were associated with higher risk of brain metastasis at 24-month follow-up [respective HRs, 1.860, 95% confidence interval (CI) 1.199-2.885, P = 0.006; HR 1.902, 95% CI 1.259-2.875, P = 0.002; and HR 1.933, 95% CI 1.168-3.200, P = 0.010]. We further found that these SNPs had a cumulative effect on brain metastasis risk, with that risk being highest for patients carrying both of these unfavorable genotypes (P = 0.003). CONCLUSIONS Confirmation of our findings, the first to indicate that genetic variations in PI3K-AKT-mTOR can predict brain metastasis, in prospective studies would facilitate stratification of patients for brain metastasis prevention trials.