AIMS: Ductal carcinoma in situ (DCIS) and invasive ductal carcinoma (IDC) exhibit frequent RUNX3 inactivation by promoter hypermethylation and protein mislocalization. The aim of this study was to analyse columnar cell lesions (CCLs) to further characterize RUNX3 involvement in breast carcinogenesis. METHODS AND RESULTS: RUNX3 expression and methylation was analysed by immunohistochemistry and methylation-specific polymerase chain reaction (PCR), respectively, in 75 CCLs. Our previously reported DCIS and IDC data were also included. Consistent with terminal duct lobular units (TDLUs) (73 of 75, 97%), active nuclear RUNX3 protein was observed in 73 of 75 (97%) CCLs [columnar cell change, 46 of 48 (96%); columnar cell hyperplasia, 12 of 12 (100%) and flat epithelial atypia, 15 of 15 (100%). In contrast to matched TDLUs from cancer specimens [four of 40 (10%)] and CCLs, significantly inactivated RUNX3 expression was detected in DCIS [17 of 20 (85%)] and IDC [18 of 20 (90%)] (all P < 0.001). RUNX3 methylation was more frequent in DCIS [15 of 20 (75%)] and IDC [16 of 20 (80%)] than CCLs [(none of 20 (0%)] and matched TDLUs [one of 10 (10%)] from cancer patients (all P < 0.001). CONCLUSIONS: RUNX3 inactivation occurs specifically in DCIS and IDC cells. In addition, RUNX3 inactivation may not be a common association between CCLs and breast carcinomas.
AIMS: Ductal carcinoma in situ (DCIS) and invasive ductal carcinoma (IDC) exhibit frequent RUNX3 inactivation by promoter hypermethylation and protein mislocalization. The aim of this study was to analyse columnar cell lesions (CCLs) to further characterize RUNX3 involvement in breast carcinogenesis. METHODS AND RESULTS:RUNX3 expression and methylation was analysed by immunohistochemistry and methylation-specific polymerase chain reaction (PCR), respectively, in 75 CCLs. Our previously reported DCIS and IDC data were also included. Consistent with terminal duct lobular units (TDLUs) (73 of 75, 97%), active nuclear RUNX3 protein was observed in 73 of 75 (97%) CCLs [columnar cell change, 46 of 48 (96%); columnar cell hyperplasia, 12 of 12 (100%) and flat epithelial atypia, 15 of 15 (100%). In contrast to matched TDLUs from cancer specimens [four of 40 (10%)] and CCLs, significantly inactivated RUNX3 expression was detected in DCIS [17 of 20 (85%)] and IDC [18 of 20 (90%)] (all P < 0.001). RUNX3 methylation was more frequent in DCIS [15 of 20 (75%)] and IDC [16 of 20 (80%)] than CCLs [(none of 20 (0%)] and matched TDLUs [one of 10 (10%)] from cancerpatients (all P < 0.001). CONCLUSIONS:RUNX3 inactivation occurs specifically in DCIS and IDC cells. In addition, RUNX3 inactivation may not be a common association between CCLs and breast carcinomas.