Purpose: This study was designed to determine whether there is a methylator phenotype in stage I and II endometrioid endometrial adenocarcinoma, and if so, whether methylation correlates with recurrence.
Experimental design: Bisulfite-converted DNAs from 24 stage I and II primary cancers (12 recurrent and 12 nonrecurrent), and 5 endometrial cancer cell lines were analyzed for methylation in the promoter regions of seven genes. A methylation index (MeI) was calculated for each tumor. Frequent HOXA11 and THBS2 methylation prompted analysis of case-matched bloods and 25 additional nonrecurrent primary cancers. Statistical analysis included Fisher's exact and Student t tests.
Results: Rates of methylation in the initial tumor series were as follows: HOXA11, 70.8%; THBS2, 62.5%; MLH1, 33.3%; CTNNB1, 16.7%; VDR, 4.2%; CDKN2A, 4.2%; and THBS1, 0%. There was no difference in the MeI of recurrent and nonrecurrent cases. However, cell lines had higher mean MeI. High rates of HOXA11 and THBS2 methylation were confirmed in the additional nonrecurrent tumors. None of the 24 case-matched bloods had HOXA11 methylation, whereas three blood DNAs showed THBS2 methylation. There was a statistically significant difference in the rate of HOXA11 methylation in recurrent and nonrecurrent tumors (P = 0.0167).
Conclusions: Endometrial adenocarcinomas have a methylator phenotype. No correlation between MeI and clinicopathologic variables in early stage tumors was observed. High rates of methylation were found in the HOXA11 and THBS2 promoter regions. HOXA11 promoter methylation was significantly more frequent in recurrent than nonrecurrent cases. HOXA11 methylation in early stage endometrial cancer is associated with poor outcome.