The effect of body mass index on the molecular landscape of endometrial cancer

Abstract

Objectives: Obesity-associated endometrial cancers follow different clinical courses than those not associated with obesity, regardless of histology. While prior studies have examined the link between discrete molecular features and obesity in endometrial cancer, no work has performed an integrated analysis of the association of body mass index (BMI) on the full molecular landscape of endometrial cancer. Methods: Using data from The Cancer Genome Atlas (TCGA), we characterized the effect of BMI on the epigenome, genome, and transcriptome in 547 endometrial cancer samples via multivariate modeling. Results: Increasing BMI was associated with relative hypomethylation and increased RNA abundance in endometrial cancer. We identify 19 genes that show both differential methylation and RNA abundance in relation to BMI (Figure 1) and some of these predict overall survival. Amongst known cancer driver genes, DICER1 is more likely to harbor point mutations in lower BMI tumors. Increasing BMI is also positively associated with microRNA abundance. Intriguingly, there were no significant associations between copy number aberrations and BMI. Conclusions: The molecular landscape of endometrial cancer changes significantly with alterations in BMI. Some of these features may be important in explaining the prognostic differences between obese endometrial cancer patients and non-obese endometrial cancer patients, which may ultimately affect the consideration of adjuvant therapy for these patients. Obesity-associated endometrial cancers follow different clinical courses than those not associated with obesity, regardless of histology. While prior studies have examined the link between discrete molecular features and obesity in endometrial cancer, no work has performed an integrated analysis of the association of body mass index (BMI) on the full molecular landscape of endometrial cancer. Using data from The Cancer Genome Atlas (TCGA), we characterized the effect of BMI on the epigenome, genome, and transcriptome in 547 endometrial cancer samples via multivariate modeling. Increasing BMI was associated with relative hypomethylation and increased RNA abundance in endometrial cancer. We identify 19 genes that show both differential methylation and RNA abundance in relation to BMI (Figure 1) and some of these predict overall survival. Amongst known cancer driver genes, DICER1 is more likely to harbor point mutations in lower BMI tumors. Increasing BMI is also positively associated with microRNA abundance. Intriguingly, there were no significant associations between copy number aberrations and BMI. The molecular landscape of endometrial cancer changes significantly with alterations in BMI. Some of these features may be important in explaining the prognostic differences between obese endometrial cancer patients and non-obese endometrial cancer patients, which may ultimately affect the consideration of adjuvant therapy for these patients.