Session Item

Suitable prognostic biomarkers for clinical outcome are still lacking in Oropharyngeal Cancer (OPSCC). In this scenario, we have recently demonstrated that LINE1 methylation (mLINE1) levels differ between HPV+ and HPV- OPSCC patients (pts). In particular, mLINE1 was lower in HPV- OPSCC pts who relapsed within 2 years, thus indicating the overall level of genomic DNA methylation may have an impact on early OPSCC relapse risk. However, the influence of mLINE1 on OPSCC survival is not yet well defined. 

We retrospectively reviewed a cohort of 156 pts with stage III–IVB OPSCC, and managed with curative intent. Genomic DNA was extracted from OPSCC FFPE tissues. HPV DNA was quantified by qPCR, whereas mLINE1 was evaluated by qMSP. The impact of mLINE1 on progression-free survival (PFS) and overall survival (OS) was assessed using Kaplan-Meier analyses. Optimal cut-points for LINE1 were search through a recursive algorithm which estimates the predictive value for PFS (Harrell's c-index) for each possible couple of LINE1 values; optimal cut-points were defined as those which maximized Harrell's c-index. Multivariate hazard risk (HR) and corresponding confidence intervals (CI) were calculated according to Cox proportional hazard model adjusting for gender, age, TNM stage, and HPV status. 

The median follow up was 30 months. PFS and OS were respectively 64.6% and 71.8% at 2 years, and 49.5% and 53.7% at 5 years. HPV DNA prevalence was 28.8%. Compared to HPV- pts, HPV+ ones had a better outcome in terms of PFS at 2 (81.0% vs 58.2%) and 5 years (70.9% vs 41.6%; p=0.003), and of OS (5-year OS: 74.1 vs 45.9%, respectively; p=0.003). mLINE1 was significantly higher in HPV+ pts than in HPV- ones (median, 66.9% and 48.7% respectively; p<0.001). Three patterns of mLINE1 status were indentified according to both PFS and OS: high ≥55%, medium 35-54%, low <35%. Analyzing all patients, PFS at 2 years was 75.9% in the high mLINE1 group, 62.0% in the intermediate group, and 41.6% in the low mLINE1 group (p≤0.001). Compared to LINE1≥55%, multivariate HR were 1.66 (95% CI: 0.92-2.99) for LINE1 35-54% and 2.84 (95% CI: 1.51-5.37) for LINE1<35%. Moreover, OS at 2 and 5 years was significantly better in the high mLINE1 group (p<0.001), with HRs of death similar to those for PFS. Notably, mLINE1 remained a prognostic factor for both PFS (p= 0.004) and of OS (p=0.011) when we restricted the analysis to HPV- pts. Among these pts, the multivariate HRs for LINE1<35% were 2.87 (95% CI: 1.49-5.52) for PFS and 2.86 (1.43-5.74) for OS. In HPV+ group, respect to pts with LINE1<55%, cases with a LINE1≥55% have had better trend in terms of PFS and OS, but this difference was not statistically significant (p=0.112; p= 0.113), mainly due to small sample size. Only one HPV+ pts showed mLINE1<35%. 

mLINE1 represents a promising prognostic biomarker for OPSCC. However, further validation in a prospective study is needed for its application in daily clinical practice.