The primary aim of this study was
The primary aim of this study was to evaluate DAPK methylation as a tumor marker which is not only detected in the primary tumor but also in the serum of patients. We used matched pre-operative serum samples to determine the correlation of tumor and serum results. Despite an equal methylation frequency of 56% (13/23 serum samples) in comparison to the primary tumors (50% MF) we did not observe a high agreement at the patient level. Only half of the patients (5/10) having primary tumors with DAPK methylation and evaluable serum samples (beta-actin positive) also showed methylated DAPK fragments in serum whereas 8 patients were positive for DAPK hypermethylation in serum only. Missing methylation in serum could be caused by absent or low levels of circulating tumor DNA. However it Dabigatran should also be considered that late stage ovarian cancers are polyclonal in nature with heterogeneous characteristics both at the genetic and transcriptional level in primary tumors and metastases , . Such polyclonality is also reflected by epigenetic heterogeneity and thereby could contribute to differences between primary tumor and serum data. Importantly this can contribute to both false positive and false negative serum results. Further analyzes regarding tumor heterogeneity should be included in future studies. Interestingly all 3 DAPK positive tumors in the study of Caceres et al. were stage III tumors and both Collins et al. and the herein presented data originate from late stage tumors , . Possibly, this points to DAPK methylation as a late event susceptible to effects of polyclonal heterogeneity. To validate the specificity of methylated DAPK in serum for the presence of ovarian cancer we analyzed not only sera from healthy women but also from uterine leiomyoma patients. Leiomyoma are benign tumors occurring at a high incidence but often without clinical symptoms  thus potentially interfering with ovarian cancer screening. We have found DAPK methylation in 35.3% and 23.8% of primary tissue and serum of leiomyoma patients respectively. The methylation frequency in tissue is comparable to the ovarian cancer samples (35.3% vs. 50%) whereas fewer leiomyoma patients exhibit methylated DAPK fragments in serum (23.8% vs. 56%; p<0.05). Nevertheless, the high prevalence of leiomyomas compromises the specificity of hypermethylated DAPK as serum marker for the detection of OvCa similar to serum marker CA125 which can be elevated by benign conditions such as endometriosis. Matched and evaluable samples from tissue and serum were available for 5 leiomyoma patients only. Contrary to ovarian cancer patients we observed an agreement between primary tissue and serum in leiomyoma patients. However, these results are limited by the low number of patients (n=5). Importantly these data demonstrate the theoretical presence of DAPK methylation in serum from OvCa patients caused by occult or clinically apparent leiomyoma. Therefore we reviewed the medical records of analyzed OvCa patients. As expected from the generally high incidence of leiomyoma, in 28.1% of women with ovarian cancer leiomyoma were present in the removed tissue at the time of primary surgery. Indeed two of eight patients who showed DAPK methylation in serum but not in primary ovarian cancer had histopathologically confirmed leiomyoma. The leiomyoma tissue of these patients was not available for analyses. Thus differences in DAPK methylation between primary tissue and serum of OvCa patients may be due to the presence of leiomyoma and/or the aforementioned ovarian cancer heterogeneity. However we could also detect at least partially methylated DAPK fragments in 4 of 8 healthy women (IVF patients). This points to an additional source of methylated DAPK sequences in serum of healthy women and potentially in serum of ovarian cancer/leiomyoma patients too. Reddy et al. could show 0.003 to 1.181% DAPK methylation in DNA from peripheral blood cells of 75/143 healthy women by sensitive real-time MSP . In particular they could assign this methylation to IgM-negative B-lymphocytes. Therefore we could not exclude the detection of DAPK methylation caused by this B-cell population. However DAPK methylation was not identified by a genome wide screen for methylated DNA in lymphocytes . Moreover we observed differences between healthy women and OvCa/leiomyoma patients regarding the methylation density of amplified MSP products (Fig. 3). Contrary to fully methylated DAPK sequences in ovarian cancer and leiomyoma patients we detected only partially methylated fragments in healthy women. This difference could be used to clarify the source of methylated DAPK in serum of diseased and healthy women.