Association between CNVs of BMI-related gene and survival status of LUAD and LUSC patients.

Abstract

e20524 Background: In recent decades, many GWAS studies help to discover the gene related to BMI or obesity. FTO, BDNF, LEPR, MC4R, SH2B1, GNPDA2, INSIG2, KCTD15, LEP, MTCH2, NEGR1, NPC1, PCSK1, POMC, TMEM18 and UCP2 are 16 genes showing most significant association with BMI. Those genes impact BMI by either alternating sex hormone metabolism, influencing insulin and insulin-like growth factor, or mediating inflammatory and adipokines. Those pathways impact cancer cells also. Here we report the association between copy number variables (CNV) of those genes and outcome of lung adenocarcinoma (LUAD) and lung squamous cell cancer (LUSC). Methods: 522 individuals in LUAD dataset of TCGA and 504 individuals in LUSC dataset of TCGA are included in this study. CNVs of each gene are main predictors, the survival time is the main outcome in the study, and gender, tumor stage, smoking status are chosen to adjust model. Multivariable Cox models are generated for LUAD patients and LUSC patients respectively. To control false discovery rate (FDR), Benjamini-Hochberg procedure is employed. Results: For LUAD patients, tumor stages impact survival status significantly, but age, gender and smoking status don’t impact survival time significantly. At the same time, CNV of TMEM18 has hazard ratio of 0.0353 (95% CI: 0.0032, 0.3910), which means the gain of genetic materials on the position of TMEM18 gene will protect patients to die after diagnosis of LUAD, and the effect is extremely large. The p-value of 0.006 is less than 0.00625, which is derived with Benjamini-Hochberg procedure with false discovery rate of 0.1. For LUSC patients, only serious tumor stages, tumor stage III and tumor stage IV, impact survival status significantly, and old age will cause patients die earlier significantly. Neither gender nor smoking status don’t impact survival status significantly. When adjusting age, gender, smoking status, and tumor stage, gain of genetic materials on the position of NEGR1 gene does increase the risk of death significantly, and the hazard ratio of CNV of NEGR1 is 9.920 (95% CI: 1.237, 79.552). The p-value of 0.031 is large than 0.00625, which is derived with Benjamini-Hochberg procedure with fase discovery rate of 0.1. Conclusion: Copy number variable of TMEM18 gene impact outcome of LUAD patients significantly. The gain of TMEM18 genetic materials protects patients to die, meanwhile the loss of TMEM18 genetic materials cause patients to have shorter overall survival time. TMEM18 affect the central nervous system to imbalance energy metabolism in previous murine models, and here we found TMEM18 has significant impact on survival status of LUAD patients, so the transmembrane protein gene TMEM18 may be a key factor to outcome of LUAD.