Epigenetic profiling reveals etiologically distinct patterns of DNA methylation in head and neck squamous cell carcinoma

Abstract

Head and neck squamous cell carcinomas (HNSCCs) represent clinically and etiologically heterogeneous tumors affecting >40 000 patients per year in the USA. Previous research has identified individual epigenetic alterations and, in some cases, the relationship of these alterations with carcinogen exposure or patient outcomes, suggesting that specific exposures give rise to specific types of molecular alterations in HNSCCs. Here, we describe how different etiologic factors are reflected in the molecular character and clinical outcome of these tumors. In a case series of primary, incident HNSCC (n = 68), we examined the DNA methylation profile of 1413 autosomal CpG loci in 773 genes, in relation to exposures and etiologic factors. The overall pattern of epigenetic alteration could significantly distinguish tumor from normal head and neck epithelial tissues (P < 0.0001) more effectively than specific gene methylation events. Among tumors, there were significant associations between specific DNA methylation profile classes and tobacco smoking and alcohol exposures. Although there was a significant association between methylation profile and tumor stage (P < 0.01), we did not observe an association between these profiles and overall patient survival after adjustment for stage; although methylation of a number of specific loci falling in different cellular pathways was associated with overall patient survival. We found that the etiologic heterogeneity of HNSCC is reflected in specific patterns of molecular epigenetic alterations within the tumors and that the DNA methylation profiles may hold clinical promise worthy of further study.

Fig. 1.

(A) Unsupervised hierarchical clustering and heat map of methylation beta values for 1250 most variable loci across all samples. (B) Recursive partitioning mixture model classification of normal and tumor head and neck tissues using all methylation beta values resulting in eight classes whose average methylation beta values are represented in the heat map. Distribution of normal and tumor samples within each class is depicted in pie charts on the right.

Fig. 2.

Recursive partitioning mixture model classification of HNSCCs (A) resulting in six classes with average methylation beta values across loci depicted in the heat map. (B) Average age of (C) lifetime average packs of cigarettes smoked per day by, (D) distribution of tumor location of and (E) lifetime average alcoholic drinks per week consumed by patients whose samples are members of the distinct methylation classes depicted in (A).