Long interspersed nuclear element 1 hypomethylation has novel prognostic value and potential utility in liquid biopsy for oral cavity cancer

New biomarkers are urgently needed to improve personalized treatment approaches for head and neck squamous cell carcinoma (HNSCC). Global DNA hypomethylation has wide-ranging functions in multistep carcinogenesis, and the hypomethylation of long interspersed nucleotide element-1 (LINE-1) is related to increased retrotransposon activity and induced genome instability. However, little information is available regarding LINE-1 hypomethylation and its prognostic implications in HNSCC. In this study, we analyzed LINE-1 hypomethylation levels in a well-characterized dataset of 317 primary HNSCC tissues and 225 matched pairs of normal mucosa tissues, along with five oral cavity cancer (OCC) circulating tumor DNA (ctDNA) samples using quantitative real-time methylation and unmethylation PCR. The analysis was performed according to various clinical characteristics and prognostic implications. The results demonstrated that LINE-1 hypomethylation levels were significantly higher in the HNSCC tissues than in corresponding normal tissues from the same individuals (P < 0.001). Univariate analysis revealed that high levels of LINE-1 hypomethylation were correlated with poor disease-free survival (DFS; log-rank test, P = 0.038), whereas multivariate analysis demonstrated that they were significant independent prognostic factor for DFS (hazard ratio: 2.10, 95% confidence interval: 1.02–4.36; P = 0.045). Moreover, samples with high LINE-1 hypomethylation levels exhibited the greatest decrease in 5-hydroxymethylcytosine (5-hmC) levels and increase in tumor-suppressor gene methylation index (P = 0.006 and P < 0.001, respectively). Further, ctDNA studies also showed that LINE-1 hypomethylation had high predictive ability in OCC. LINE-1 hypomethylation is associated with a higher risk of early OCC relapse, and is hence, a potential predictive biomarker for OCC. Furthermore, 5-hmC levels also exhibited predictive potential in OCC, based on their inverse correlation with LINE-1 hypomethylation levels. LINE-1 hypomethylation analysis, therefore, has applications in determining patient prognosis and real-time surveillance of disease recurrence, and could serve as an alternative method for OCC screening.