Study Of Epigenetic Modification In Gynecologic Oncology Diagnosis

Epigenetics is the study of heritable changes in gene activity that are not caused by changes in the DNA sequence, these changes can be stable heredity in development and cell proliferation and have potential to be reversed. There are two parts of epigenetics, one is a selective transcriptional regulation of gene expression, including DNA methylation, chromatin remodeling, histone modification, genomic imprinting, gene silence, nucleolar dominance and dormant transposon activation. The other part is post-transcriptional regulation, including no-encoding RNA, MicroRNA, antisense RNA, intron and riboswitches. Comparing to classical genetics which studies the effects of the biological function of the gene sequences, the main task of epigenetic is the establishment and maintain mechanisms of these "epigenetic phenomenon". As the supplement to classical genetics, epigenetics make us have more comprehensive understanding of the mechanisms of gene regulation. Recent studies have confirmed that epigenetic modifications played an important role in stem cell differentiation and early embryonic development and a variety of diseases (cancer, autoimmune diseases and metabolic diseases, etc.), especially in the field of cancer research, such as oncogene activation, tumor suppressor gene inactivation, DNA damage repair defects and cancer stem cell differentiation.The ultimate goal for the study of epigenetic mechanism is to be applied to clinical diagnosis and preventive treatment. The studies of DNA methylation and MicroRNA have been well established. Fujiwara and other studies have shown that the specificity and positive predictive values of cell-free serum DNA methylation in five genes were85%and75%in the early diagnosis of lung cancer. Many studies also indicated that MicroRNA expression profiling can be used for classification of breast cancer (such as estrogen and progesterone receptor status, tumor stage, metastasis and Her2status, etc.). The main task of this research is to explore the potential application of DNA methylation and MicroRNA spectrum in early epithelial ovarian cancer diagnosis and uterine smooth muscle tumor classification.Part Ⅰ:A multiplex methylation-specific PCR assay for the detection of early-stage ovarian cancer using cell-free serum DNABackground and PurposeEpithelial ovarian cancer (EOC) is the third leading type of cancer among female gynecological system carcinomas, with the highest mortality rate. Due to the lack of early diagnosis methods and no obvious symptoms,85%of patients are first diagnosed at advanced stages (FIGO Ⅱ-Ⅳ), and the survival rates for advanced-stage patients are confined to30%-44%. However, if diagnosed at the localized stage (FIGO I), the5-year survival rate is93%. In addition to exploring the tumorigenesis of ovarian cancer and generating novel effective chemotherapies, the establishment of effective early diagnosis methods is paramount to improving the survival of women with ovarian cancer.There is no effective method for ovarian cancer screening. Clinically, CA125is a routinely used serum marker; however, it is elevated in less than50%of stage I ovarian cancers. Moreover, the specificity of CA125is poor, and a large number of benign and malignant conditions (such as endometriosis, pregnancy, pelvic inflammatory disease, lung cancer and colon cancer) may result in falsely elevated CA-125values. The positive predictive value for detection is only10%-35%. Therefore, the traditional methods for early diagnosis (such as serum CA125, trans-vaginal ultrasound probe and magnetic resonance imaging), which exhibit high false-negative rates, as well as lower sensitivity and specificity for clinical needs, have not demonstrated the capacity to reduce the population morbidity and/or mortality.DNA methylation is an important epigenetic phenomenon that affects gene expression without changing the DNA sequence. Aberrant hypermethylation occurring in promoter CpG islands is a significant mechanism of tumor suppressor gene (TSG) silencing and, in certain cases, may be the only mechanism. Recent studies have shown that aberrant DNA methylation, which usually occurs before patients develop clinical manifestations and radiographic evidence, provides a new molecular approach for the early diagnosis of cancer. A large number of genes have been identified as hypermethylated and have been associated with molecular, clinical and pathological features of ovarian carcinomas. The levels of cell-free serum DNA are abnormally high in both early and advanced-stage tumors. Two primary mechanisms proposed for this phenomenon include the following:1) cells in cancer tissues in situ undergo apoptosis and/or necrosis, and2) cells detach and extravasate into the bloodstream and undergo lysis. Tumor-specific aberrant methylation of cell-free serum DNA has been detected in patients with various tumor types, such as prostate, lung, breast and colon cancer tumors, and altered methylation has also been confirmed to be an independent prognostic marker of progression-free and overall survival. Despite its potential clinical applications, the methylation detection of cell-free serum DNA has several limitations, such as the extremely small amounts of available cell-free serum DNA, missing bisulfite-conversions and the low sensitivity conferred by a single marker. In the present study, we developed a novel Multiplex-MSP assay to detect the methylation status of cell-free serum DNA. We confirmed that this novel assay had higher efficiency in the detection of EOC, especially at early stages, compared to the conventional tumor marker CA125.Materials and Methods1. Screening for epithelial ovarian cancer-specific methylated genes.2. Collecting ovarian tumors and normal control serum and tissue samples3. Designing and optimization of Multiplex-MSP assay4. Detecting methylation statues of serum and tissue samples by Multiplex-MSP assay5. Comparison of the methylation status in tissue and serum of EOC6. Retrospective study of serum samples to determine the sensitivity, specificity and accuracy of the multiplex MSP assay for the detection of ovarian cancer7. Ovarian cancer screening using multiplex MSP assay in patients with pelvic massesResult1. A PubMed search of the English literature published between January1,2000, and January1,2012, was performed with the keywords "methylation" and "epithelial ovarian cancer" and finally seven tumor suppressor genes (APC, RASSF1A, CDH1, RUNX3, TFPI2, SFRP5and OPCML) with hypermethylation rate were selected to construct the Multiplex-MSP assay.2. Successfully combining multiplex PCR, nested PCR and methylation-specific PCR together to construct the multiplex-MSP assay3. Compared the methylation status in tissue and serum of EOC and found the serum methylation profiles were included in the tissue methylation profiles in both early and advanced-stage EOC.4. An evaluation of the ability to differentiate EOC from benign pelvic masses showed that the sensitivity, specificity and accuracy of the Multiplex-MSP assay were90.57%,89.66%and91.26%, respectively, all of which were higher than the respective values for CA125. When dividing patients in advanced-stage EOC, no significant superiority of the multiplex-MSP assay compared to CA125was found (P=0.5817). However, the specificity, sensitivity and accuracy of the Multiplex-MSP assay were significantly higher than CA125in early-stage EOC (P=0.0036)5. In screening test, the specificity, sensitivity and accuracy of the multiplex-MSP assay compared to CA125were83.33%vs.50.00%,82.76%vs.72.41%and91.27%vs.89.70%, respectively.Conclusion1. The status of cell-free serum DNA methylation can accurately reflect the DNA methylation status of tumor tissue and can be used as a marker of tumor earluy diganosis.2. The multiplex-MSP assay serves as a new powerful tool to identify the methylation status of cell-free serum DNA.3. The multiplex-MSP assay is confirmed to be sensitive and specific enough to detect ovarian cancers at an early stage, thereby holding great promise for the possible early detection of other cancers. Background and PurposeUterine smooth muscle tumors (USMTs) range from conventional leiomyomas (ULM) to highly aggressive and fully malignant leiomyosarcomas (LMS). Conventional leiomyomas are characterized by bland cytology and mitotic inactivity (usually<10MF/10HPF). Leiomyomas may often demonstrate hyaline (ischemic type) necrosis but lack classic coagulative tumor cell necrosis. In contrast, uterine leiomyosarcomas are morphologically defined by the presence of at least2of the following3features:(1) cytologic atypia,(2) increased mitotic rate (usually>10MF/10HPF), and (3) coagulative tumor cell necrosis.Between conventional leiomyomas and leiomyosarcoma, there are several categories of uterine smooth muscle tumors of intermediate types. They show some but not all of the characteristics of malignant tumors, including the following categories:uncertain malignant potential (STUMP), atypical leiomyoma (ALM), mitotically-active leiomyoma (MALM) and cellular leiomyoma (CLM). The diagnostic criteria are based on WHO and Stanford (Bell) scheme. In addition to the typical uterine fibroids, other types of uterine smooth muscle tumors have very low incidence. The studies were mainly on the clinic and pathology, less on molecular analysis.There are two conventional views about USMTs:1. LMS arise de novo, rather than from any precursor lesions;2. ALM is a benign variant of ULM. But with advances in molecular research in USMTs, these two views have been challenged. The current World Health Organization (Stanford) schema uses the term atypical leiomyoma; older synonymous names include pleomorphic, symplastic, and bizarre leiomyoma. There is no clinical guideline treatment. According to the literature, ALM often had a high cure rate and a low recurrence rate. But rare exceptions to this may exist, a small case series reported that3of18women with cellular or atypical variants died of their disease, of note, there was a longer latency between diagnosis and death in these patients compared with those with LMS (>6years versus median of2.1years). In clinical practice, we have seen very few ALMs progress to LMSs in a short time. Chromosome lp deletion can also be found in some ALM, which is more similar to LMS than ULM.In the previous study, we collected167cases of the six different USMT variants, then reviewed and confirmed the diagnoses based on the WHO and Stanford criteria. We conducted molecular and genetic analyses of these cases using a group of gene markers that are most relevant to USMT. These included the P53, MED12and PTEN genes. The gene signatures and the genetic fingerprints of each tumor variants were further analyzed. MED12mutations were high at75%in ULM cases (30/40). MED12mutations were significantly low in ALM (10%,4/42) and LMS (10%,4/38)(P>0.05). P53mutation rate in ALM and LMS were12%and24%, respectively, while no mutations were found in ULM. PTEN loss detection also showed ALMs and LMSs had similar frequency (P>0.05). A total of17oncogenic markers were selected for next study. They were all relevant to LMS in the functional pathways involving in tumor growth, differentiation, and tumor stem cells, including Bcl-2, CD24, C-Kit, EGFR, ER, Fascin, HMGA2, Ki-67, PR, p16, p21, p53, pAKT, PTEN, Rb, PS6, and RBL2. Overall, the immunoprofile of these17markers clearly separated LMS from other variants. The IHC signature for LMS requires ER, PR, Bcl-2, P16, P21, P53, and Ki-67. When we removed ER and PR, more than70%of ALM aggregated with LMS. Interestingly, ALM shared several oncogenic markers with LMS, including P16, P53, Bcl-2, and Fascin. Unsupervised cluster analysis showed ALM and LMS had similar protein expression profiles. Based on the above basis, the research group assumed ALM may be the precursor lesion of LMS.This research focuses the MicroRAN profiling and DNA methylation analysis on USMTs, to seek more evidence to support the hypothesis ALM may be the precursor lesion of LMS.Materials and Methods1. We reviewed the pathology database from1993to2003in two institutions (Northwestern Memorial Hospital, Northwest University and Qilu Hospital, Shandong University) to retrieve160patients with a diagnosis of usual type leiomyomas (ULM), cellular leiomyomas (CLM), atypical leiomyomas (ALM), Uterine smooth muscle tumors of uncertain malignant potential (STUMP) and leiomyosarcomas (LMS). Among160patients,4CLMs,15ALMs and5STUMPs were from Qilu Hospital. Each case was reviewed by at least two pathologists to confirm the diagnosis based on Stanford scheme and WHO criteria (2003).2. Extracted RNA from8tumor samples from each of5different tumor types (LMS, STUMP, ALM, CLM and ULM). RNA was examined on custom designed miRNA chip of FirePlexTM.3. Extracted DNA from18ULM and paired myometrium. Genomic methylation analysis was performed by Infinium Human Methylation27K BeadChip.4. Used Sequenom Methylation MassArray to detect the selected genes{KLF11、 DLEC1and RUNX3) methylation status in30LMS,25ALM,12ULM,6MM.Result1. Unsupervised clustering analysis showed ALM and LMS had similar miRNA expression profiling.2. Two significantly hypermethylated genes (KLF11and DLEC1) were found by Infinium Human Methylation27K BeadChip in ULM (P<0.001).3. ALM showed similar methylation pattern in selected genes (KLF11、DLEC1and RUNX3) with LMS.4. Principal component analysis which showed the3D distance of all type uterine smooth muscle tumors found that ALM was much closer to LMS than other types. Conclusion1. Rather than ULM, ALM had similar miRNA and methylation profiles with LMS. This suggested that ALM may be precursor lesion of LMS.2. Because of ALM shared similar epigenetic and other molecule characteristics with LMS, the clinical treatment of ALM patients should be different from ULM, should be closely followed up.