The past decades have witnessed the distinct oncogenic outcome in OC patients, which spans less than 5 months to over 10 years. However, no individualized treatment was used in those cases with significantly poor long-term survival for a lack of predictive models helping to identify them accurately. Consequently, it is urgently needed to construct a prognostic tool which could accurately identify those patients with poor survival. In this study, a novel prognostic classifier based on 26 epigenetic genes was developed to improve the prediction of OS for OC patients after surgical resection. By applying the epigenetic signature to the TCGA discovery set patients, a clear separation was observed in the survival curves between low and high-risk patients. And it was internally validated in the validation series of TCGA patients and the external cohorts of GSE14764 and GSE26712, indicating the good reproducibility of this signature in OC. Stratified by FIGO stage, debulking status, and tumor grade, the epigenetic signature remains a good prognostic model, implying that the signature can be used to refining the current staging and risk evaluation system. In addition, the time-dependent ROC and DCA curve at 5 year suggested that this epigenetic signature has the most significant accuracy and clinical utility in predicting mortality after initial resection of OC. Therefore, our study identified an epigenetic signature that could help identify patients with high risk of mortality and guide individualized treatment of patients with OC, which is credible to be applied to clinic.
The epigenetic modifiers have been highlighted for critical functions in cancer initiation and progression. Accumulating evidence indicated that agents targeting epigenetic alterations have promising outcomes in retarding OC progression. However, no previous study has assessed a subset of epigenetic genes that collectively form an epigenetic signature and serve as an independent prognostic factor for OC. To date, numerous studies have developed gene expression signatures to stratify OC survival based on different cohorts. However, none of them has been incorporated into clinical practice might due to several limitations. First, the sample size was usually small that lacked sufficient validation to prove model stability. Second, most of them generate the prognostic score based on the whole transcriptome profiling, which easily leaded to overfitting and ignored other casual genes. In this present study, we focused on the epigenetic gene set that is more generic than the normal signature.
Most of genes included in the signature have been experimentally demonstrated to be linked with OC. BRD9 which was previously reported to regulate chromatin remodeling and transcription was found overexpressed in ovarian cancer and depleting BRD9 sensitizes cancer cells to olaparib and cisplatin[21]. PHF20 is a transcription factor, which was originally identified in glioma patients and its phosphorylation plays an important role in tumorigenesis via regulating of p53 mediated signaling in OC[22]. PADI4 is an enzyme that converts both histone arginine and mono-methyl arginine residues to citrulline. It was proved to be able to regulate the proliferation, apoptosis, invasion and migration of ovarian cancer cells[23]. EHMT1 and EHMT1 have roles in epigenetic silencing of gene expression and disrupting EHMT1/2 will sensitizes OC cells to PARPi. DDB2 is a kind of DNA damage-binding protein[24]. Previous studies suggested that DDB2 can repress OC cell dedifferentiation by suppressing ALDH1A1 expression[25]. CARM1, an arginine methyltransferase was proved as an informative prognostic biomarker for predicting resistance to chemotherapy for ovarian cancer[26]. CHD2, chromodomain helicase DNA binding protein could bind to microRNA-144 and partially inhibited its activity, thereby promoting the proliferative and migratory abilities of OC cells[27]. APOBEC3 DNA cytosine deaminase family members normally defend against viruses and transposons and it expression was found to be correlated with T-Cell Infiltration and Improved Clinical Outcomes in OC based on public genetic data[28]. As for the rest epigenetic genes integrated in our signature, further clinical and basic research should be conducted to explore their value in OC.
Though this is the first study investigating the prognostic role of epigenetic genes and successfully developing a epigenetic signature in OC, some limitations are Inevitable in our study. Firstly, our study was based on the data from public-available datasets without testing prospectively in an independent cohort. Furthermore, the information of several other important clinicopathological features, like histology, was not available in these datasets. Finally, mechanisms of the identified epigenetic genes on the progression of OC are still needed to be further investigated.