Study On The Origin And Molecular Diagnosis Of Ovarian Low-grade Serous Carcinoma

Ovarian cancer is one of the three malignant tumors from female reproductive system and the most lethal gynecological malignancy.The American Cancer Society estimated a total of 22,240 new cases and 14,070 deaths from the disease in 2018.The five-year survival rate for ovarian cancer was about 46%.There has been no significant progress in the diagnosis and treatment of ovarian cancer in the last 50 years.This dismal state of affairs is mainly due to the fact that the origin and pathogenesis of ovarian cancer are unclear.Ovarian cancer is a highly heterogeneous malignancy,of which epithelial serous carcinomas are the most commom,representing 60-80%of ovarian cancer.In 2004,on the basis of a series of morphologic and molecular genetic studies,Kurman and his colleagues proposed a dualistic model that categorized various types of epithelial ovarian cancer into two groups,type I and II.Ovarian serous carcinoma,the most common type,is divided into high-grade and low-grade serous carcinoma according to this model.Traditionally,it is thought that ovarian serous carcinoma(OSC)is epithelial malignancy originating from ovarian surface epithelium(OSE)or cortical inclusion cyst(CIC),but no convincing precursors were identified in the ovary.In the late 1990s and early 2000s,prophylactic salpingo-oophorectomy was performed in women who were at high risk of developing ovarian cancer on account of either family history or germ-line mutations of BRCA1 and BRCA2.Subsequently,pathologists began to section the fallopian tubes and ovaries meticulously and they found no lesions in ovary but instead discovered occult noninvasive and invasive carcinomas in the fallopian tubes,especially in the fimbria.Subsequently,a study of women,suffered from pelvic high-grade serous carcinomas(HGSC),who were not at high risk of ovarian cancer and didn't harbor BRCA mutations reported tubal intraepithelial carcinomas(STICs)in 48%of the patients.Consequently,investigators begin to question the classical theory that HGSCs originate from ovary.During the last few years,a series of histological and molecular evidence indicates strongly that a large proportion of HGSCs probably originates from the fallopian tube,specifically from STICs and then involving ovary and peritoneum.First,in the patients with HGSC,simultaneous occult noninvaasive and invasive carcinomas were detected in the fallopian tubes,particularly in the fimbriae.Second,STICs shared identical TP53 mutations with HGSCs,suggesting the clonal relationship between them.Third,studies of gene expression demonstrated that the expression profiles and patterns of HGSCs more closely resembled normal falllopian tube epithelium(FTE)than normal OSE.Finally,establishment of an HGSC model transformed from fallopian tube secretory epithelium provided compelling evidence that secretory cells could be the cellular origin of HGSC.To sum up,the vast majority of HGSCs may indeed develop from fallopian tube.Compared to HGSC,the origin of LGSC is less clear so far.The finite number of LGSC cases is the main research obstacle.Morphological and molecular evidence suggests that LGSC evolves in a stepwise fashion,from ovarian epithelial inclusions(OEI)to serous cstadenomas and serous borderline tumors(SBT),and finally to LGSC.Traditionally,it is thought that OEIs develop from invaginations of OSE and then undergo fallopian tubal metaplasia.However,the metaplastic process from OSE to Mullerian epithelium and hybrid type of OEIs with both mesothelial and tubal phenotypes are rarely found in clinical practice.Therefore,we must think about the true cell origin of LGSC and OEI.We speculate that majority of LGSC may originate from fallopian tube.Our pevious study indicated that most(78%)of the OEIs showed a tubal phenotype(calretinin-/PAX8+)and suggested that most OEIs as well as LGSCs are likely to originate from tubal epithelia.Besides,Kurman identified a precursor lesion of SBT,designated as papillary tubal hyperplasia(PTH).At the same time,another study by Laury et al.showed that ciliated cells and secretory cells both existed in SBTs,resembling normal fallopian tubal epithelium.Furthermore,secretory cell outgrowths(SCOUTs)are significantly more common in the fallopian tubes of women with SBTs compared to controls.These morphological and immunologic studies provided powerful evidece for a tubal origin of LGSCs.However,thus far,there are no molecular studies which include investigations of possible tissue origins of LGSCs.Accordingly,we analyzed the gene expression profiles of LGSC,SBT,HGSC,FTE,OSE and human peritoneal mesothelial(HPM)to further explore the origin of LGSC at a molecular level.This study could have important implications on preventative and therapeutic aspects of LGSC.Although numerous studies show that both HGSC and LGSC share the same origin from FTE,they are completely distinct gynecologic tumors rather than different grades of the same neoplasm and there are drastic clinicopathological and molecular differences between the two.HGSC is usually susceptible to platinum-based chemotherapy,but 5-year survival rate is very low.LGSC,with relatively higher 5-year survival rate,develops slowly but shows a poor response to standard chemotherapy.Because of unsuccessful early detection of ovarian cancer,seventy percent of the tumors present typically in advanced stage and it is necessary to receive neoadjuvant chemotherapy for a proportion of patients.In addition,clinical trials specifically targeting the crucial molecular alterations of LGSC such as CI-1040 targeting MAPK pathway,have been launched and is on-going.Therefore,a precise differential diagnosis of HGSC and LGSC is crucial for the decision making for gynecological oncologists prior to any treatment.It may help to optimize therapeutic effectiveness and reduce unnecessary morbidity.At present,the two-tier system for grading serous ovarian carcinoma is based primarily on nuclear atypia and mitotic rate.The system is easy to follow and is reproducible.Nevertheless,subjective judgments on the criteria of nuclear atypia and mitotic rate from individual pathologists could bring difficulty for differential diagnosis.Moreover,finite samples from paracentesis extract or biopsies of pelvic/peritoneal implants can also pose a minor obstacle for differential diagnosis of HGSC and LGSC.Thus,we intend to investigate effective markers and establish objective criteria in order to distinguish HGSC from LGSC efficiently.This could be beneficial for the clinical management,optimizing targeted therapy and thus improving the survival rate of ovarian ccancer.This study includs the following two parts:PART I:Study on the origin of ovarian low-grade serous carcinomaObjective:Ovarian cancer is a highly heterogeneous tumor,of which serous carcinomas are the most common.The dualistic model proposed by Kurman divides serous carcinoma into two types:HGSC and LGSC.Traditionally,it is thought that OSC is one type of epithelial malignancy originating from ovarian surface epithelium.At present,more and more evidence indicates that vast majority of HGSCs originate from FTE.While the cellular origin of LGSC remains unclear so far.Recently,we find that ovarian LGSC may also originate from the fallopian tube through morphological and immunohistochemistry study.However,thus far,there is no molecular evidence confirming the origin of LGSC.Therefore,we compare and analyze the mRNA expression profiles of LGSC,SBT,HGSC,FTE,OSE and HPM.The purpose of the current study was to gain further insight into the origin of LGSC at the molecular level.Material and Method:31 tissue samples were collected in Qilu Hospital of Shandong University from 2014 to 2015,including LGSC(n=6),SBT(n = 6),FTE(n?5),OSE(n = 4),and HPM(n = 4).HGSC cases(n = 6)served as a positive control.RNA-seq analysis was performed on the total of 31 tissue samples to exmain the mRNA expression of them.Then,we performed Kruskal-Wallis contrasts to analyze differential expression of six groups and used unsupervised hierarchical clustering to verify the quality of all samples and determine their overall similarity.On the basis of differentially expressed genes among three normal groups,we performed another unsupervised hierarchical cluster analysis only in LGSC samples and the three control groups to further identify the origin of LGSC.Then,we analyzed the relationship between LGSC and normal control group through rank-sum analysis and Pearson correlation test using the differentially expressed genes.To validate the findings from the gene expression array study,we selected the highly differentially expressed genes(PAX8,CDH1,FOXA2,and ARX)as well as those corresponding proteins and examined their expression levels using qRT-PCR and immunohistochemistry in tissue samples of ovarian serous tumors,fallopian tube,ovarian surface epithelia,and peritoneal mesothelia.Result:Dendrograms produced by unsupervised hierarchical clustering revealed that OSE samples clustered with HPM,while ovarian serous tumors,including LGSC,SBT and HGSC,clustered with FTE.Furthermore,LGSC showed a significantly closer relationship with FTE than with OSE and HPM samples.QRT-PCR indicated that the expression of PAX8,CDH1,FOXA2 and FOLR1 were higher in serous tumors and FTE samples compared to OSE and HPM samples.However,ARX and FCN1 mainly expressed in OSE and HPM samples but not in FTE and serous tumors.The result from immunohistochemistry showed that PAX8,CDH1 and FOXA2 specifically expressed in serous tumors and FTE samples but not in OSE.In contrast,ARX was only found in OSE.Conclusion:The findings of the current study provide further evidence at a molecular level that the fallopian tube is likely the cellular source of LGSC.This finding may facilitate the development of new approaches to prevention,early detection,and allow novel therapies to be tested.PART ?:Study on the molecular diagnosis of ovarian low-grade serous carcinomaObjective:HGSC and LGSC are distinct gynecologic tumors with diverse pathogenesis and characteristic features.They respond differently to same treatment strategy and have dissimilar prognosis.Therefore,it is crucial to obtain accurate differential diagnosis of HGSC and LGSC before clinical treatment.The discrimination between HGSC and LGSC can be challenging due to finite samples from paracentesis extract or biopsy specimen from pelvic/peritoneal implants.The purpose of the study is to explore effective markers to distinguish HGSC from LGSC.Material and Method:Twelve fresh tissue samples were collected in Qilu Hospital of Shandong University from 2014 to 2015,including six HGSCs and six LGSCs.Expression profiles of mRNA were detected using RNA-seq analysis and the differentially expressed genes between the two groups were compared through DESeq analysis.Then,we clustered differentially expressed genes through unsupervised hierarchical clustering method and used qRT-PCR in a large cohort to validate the highly differentially expressed genes(LAMP3?EGFL6?CDKN1A ? AGR3)between the two groups.Anti-TP53 and Anti-AGR3 immunohistochemistry were performed on 145 HGSC and 30 LGSC samples.Result:Dendrogram produced by unsupervised hierarchical clustering showed that 699 differential expression genes(>2 folds difference)between HGSCs and LGSCs distinguished them completely.The anterior gradient homolog 3(AGR3)was verified to be highly upregulated in LGSC compared to HGSC by qRT-PCR and Western blot.Anti-TP53 immunohistochemistry showed that abnormal expression of TP53(0 or?75%positibe expression)was found in 87.6%(127/145)of HGSC and 13.3%(4/30)of LGSC.While,positive staining of AGR3 has a sensitivity of 80.0%(24/30)and a specificity of 89.7%(130/145)for LGSC.TP53 and AGR3 were both found to be efficient in distinguishing HGSC from LGSC(P<0.001 for both).Receiver operating characteristic(ROC)analysis revealed a similar area under the curve for AGR3(0.848)compared to TP53(0.871).Through combination of the two markers(TP53 wild-type pattern and AGR3 positive expression),the accuracy of differential diagnosis reaches up to 93.1%(163/175).Conclusion:These findings provide compelling evidence that differential diagnosis of HGSC and LGSC can be further improved by combined application of these two markers on the basis of conventional histopathologic diagnosis.Furthermore,it could be beneficial for optimizing therapeutic regimen and lowering mortality risk based on robust diagnosis.THE INNOVATION OF THIS STUDY1.It is the first time to confirm that the mRNA expression profiles of ovarian low-grade serous carcinoma resemble those of fallopian tube epithelium at molecular level through high-throughput sequencing,supporting the tubal origin of ovarian low-grade serous carcinoma and providing new approaches to prevention.2.The study firstly combines TP53 and AGR3 to distinguish ovarian high-grade serous carcinoma from low-grade serous carcinoma,improving diagnosic accuracy,guiding clinical treatment and lowering mortality risk.