Prognostic Significance of p16^INK4a Hypermethylation in Non-small Cell Lung Cancer is Evident by Quantitative DNA Methylation Analysis

Abstract

Background: p16^INK4a is a tumor suppressor gene frequently inactivated by aberrant promoter hypermethylation. In the present study, p16^INK4a methylation was evaluated in non-small cell lung cancer (NSCLC) using a quantitative assay and the clinical significance of the methylation was explored. Materials and Methods: A total of 244 tumor samples from formalin-fixed paraffin-embedded archives were examined in this study. p16^INK4a methylation was analyzed by the fluorescence-based, real-time methylation-specific PCR assay, MethyLight. The quantitative methylation value was expressed as the percentage of methylated reference (PMR). Results: The median level of p16^INK4 methylation was 0.55 PMR (range 0.00-503.4). The p16^INK4 methylation value was significantly higher in males (p=0.005) and in squamous cell carcinoma (p=0.018). Prognostic analysis using the Cox proportional hazard model showed that the p16^INK4a methylation value was a significant prognostic factor (odds ratio, 1.005; 95%CI, 1.003 to 1.008; p<0.0001). The p16^INK4a methylation value remained a significant prognostic factor (p=0.0004) in multivariate analysis including age, gender, histological type and clinical stage. Specimens were then classified into hypermethylated or non-hypermethylated groups based on the p16^INK4a methylation value using various cut-offs from 1 to 100 PMR. There was no significant difference in prognosis between the two groups using a cut-off value of 1 PMR. On the other hand, there was a significant difference using 6 PMR or more as the cut-off value (p<0.01). Conclusion: These results provide clear evidence for the prognostic significance of p16^INK4a methylation in NSCLC using quantitative DNA methylation analysis. Careful assessment of DNA methylation is needed because qualitative methylation analysis may overestimate low levels of methylation, which have less clinical significance.