Prognostic Factors Of Ovarian Serous Cystadenocarcinoma

There are several histologic subtypes of ovarian tumors,and the clinical features and prognosis vary widely among subtypes.Ovarian serous cystadenocarcinoma is the most common ovarian malignancy,accounting for40%-50% of cases.Therefore,this study focused on this type of cancer,using the Surveillance,Epidemiology,and End Results(SEER)database and the Cancer Genome Atlas Ovarian(TCGA-OV)database of the National Cancer Institute(NCI)to explore clinical features and gene expression profile characteristics of patients with ovarian serous cystadenocarcinoma in relation to their prognosis.We also explored the survival patterns and established more precise reference values for survival time groupings.Part One A Nomogram Predicts the Survival in Patients with Ovarian Serous CystadenocarcinomaObjective: Ovarian serous cystadenocarcinoma is a common malignancy of the female reproductive system,and a variety of clinical and pathological factors may affect the prognosis of patients.The purpose of this study was to develop a nomogram model of the clinical characteristics of patients with ovarian serous cystadenocarcinoma and to validate its predictive accuracy and reliability.The conclusions will provide a basis for predicting patient prognosis and lay the foundation for our subsequent studies.Methods:1.Cases of ovarian serous cystadenocarcinoma diagnosed between 1992 and 2013 in 20 US states were searched in the Surveillance,Epidemiology,and End Results(SEER)database of the National Cancer Institute.Information on cases with complete data was extracted,including clinical features such as age at diagnosis,year of diagnosis,race,clinical stage,tumor laterality,treatment method,histological grade,survival time,vital status,and cause of death.2.Kaplan-Meier survival curves were plotted for patients grouped by clinical characteristics,and differences in survival time between groups were compared using Log-rank tests.3.Univariate Cox regression analysis was used to analyze the correlation between patients' outcome events(death)and clinical variables,and variables that were statistically different in the univariate analysis(P<0.05)were included in the multivariate Cox proportional hazard regression analysis.A nomogram was constructed based on the multivariate Cox proportional hazard regression model.4.Bootstrap method was used to internally validate the efficacy of the nomogram model.The calibration curve was obtained by resampling the data1000 times to visualize the agreement between predicted and actual survival rates,while avoiding overfitting of the model;the concordance index(C-index)was used to assess the consistency of the model.Results:1.A total of 10680 patients with ovarian serous cystadenocarcinoma were extracted from the SEER database,with average age of 62.0±12.078years(range from 17 to 96 years).Median of overall survival(OS)time was4.2 years and longest OS was over 24.8 years.1-year,5-year and 10-year survival probabilities were 77%,39% and 25%,respectively.2.The results of the Log-rank test suggested that age at diagnosis,year of diagnosis,clinical stage,tumor laterality,treatment method,and histological grade were associated with the prognosis of patients with ovarian serous cystadenocarcinoma(P<0.05).3.Univariate and multivariate Cox proportional hazard regression analysis suggested that age at diagnosis over 62 years and late clinical stage were poor prognostic factors,year of diagnosis between 2003 and 2013,single side involved,surgical treatment,and low histological grade were favorable prognostic factors.Meanwhile,race was not a prognostic factor.4.The nomogram model included 6 independent clinical variates,including age at diagnosis,year of diagnosis,clinical stage,tumor laterality,treatment method,and histological grade.The C-index was 0.663,suggesting moderate predictive accuracy of the model.The calibration plot showed good agreement between predicted survival and actual survival,but its accuracy in predicting 1-year,3-year,and 5-year varied.Conclusions:1.There existed long-term survivors in patients with ovarian serous cystadenocarcinoma and the longest survival time is over 24.8 years.2.The nomogram model included 6 independent factors affecting the prognosis of ovarian serous cystadenocarcinoma: age at diagnosis,year of diagnosis,clinical stage,tumor laterality,treatment method,and tissue grade.3.The nomogram models based on clinical characteristics showed moderate agreement between predicted and actual patient survival rates.It is suggested that other factors,such as gene expression profiles,may exist in addition to clinical variables that may affect prognosis,motivating us to proceed to the next part of the study.Part Two Survival Pattern and Subgroup Reference for Ovarian Serous CystadenocarcinomaObjective: The majority of patients with ovarian serous cystadenocarcinoma have a survival period of less than 5 years,while a small number of patients can survive for a long time.However,there is no uniform standard about long-term survival,and current relevant studies have selected a certain number of years as the basis for grouping,such as 3,5,7,or 10 years.In order to explore the factors affecting their long-term survival,reliable and accurate cutoffs for overall survival time need to be clarified.In this study,we investigated the survival patterns of ovarian serous cystadenocarcinoma patients,hoping to establish precise survival time grouping reference values on this basis and lay the foundation for further exploration of genetic differences.Methods:1.The cases in this study were retrieved from the SEER 18 Regs dataset in the Surveillance,Epidemiology,and End Results(SEER)database of the National Cancer Institute.Inclusion criteria:(1)Year of diagnosis=“1992 to 2013”;(2)Histologic type ICD-O-3=“8441”;(3)Site recode ICD-O-3/WHO 2008=“ C56.9”;(4)Behavior recode for analysis=“malignant”.Cases with incomplete survival information were excluded.2.Survival time ridge curves were plotted separately according to each clinical characteristic of the patients,including age at diagnosis,year of diagnosis,race,tumor laterality,regions of residence,clinical stage,tissue grade,treatment method,and cause of death.Based on these results,we explored the survival patterns of patients with ovarian serous cystadenocarcinoma.3.Based on the principle of Tukey's test for outliers,box plots were drawn with the survival time of patients who had an event(death),and outliers were long-term survivors.The reference range of long-term survivors was defined as >(Q3+1.5×IQR),and the reference range of ordinary survivors was between Q1 and Q3.4.The subgroup reference were applied to choose cases from the 13929 patients with ovarian serous adenocarcinoma into long-term and normal-term groups.Results:1.A total of 13929 patients with ovarian serous adenocarcinoma were included with a median age of 63 years.Survival time ridge curves were plotted separately according to each clinical characteristic of the patients.2.The survival ridge curves showed a bimodal distribution pattern,with the two peaks corresponding to normal-term survivors and long-term survivors,respectively.The patient's age at diagnosis had a great effect on the shape of the left-hand peak which represents the normal-term survivor group.The survival time reference values for long-term survivors and normal-term survivors decreased with increasing age: the lower boundary for long-term survivors decreased from 9.83 to 4.3 years;the lower boundary for normal-term survivors decreased from 2.08 to 0.08 years,while the upper boundary decreased from 5.17 to 1.83 years.3 The grouping effect of the grouping reference value was validated in a cohort of 13929 patients with ovarian serous adenocarcinoma,and 45.5% of patients could be assigned to the long-term survival group(1288 cases)or the normal-term survival group(4908 cases).Conclusions:1.The survival ridge curves showed a bimodal distribution pattern,with the two peaks corresponding to normal-term survivors and long-term survivors,respectively.2.The application of the Tukey abnormality test allowed the determination of reference values,taking into account the effect of age at diagnosis.3.The study cohort had a high data utilization rate of 44.5% when applying the new subgroup reference values.Part Three Long-term Survival Related Genes Identification and Functional Analysis for Ovarian Serous CystadenocarcinomaObjective: Some patients with ovarian serous cystadenocarcinoma have one or more negative prognostic factors,but still become long-term survivors,suggesting that other factors such as genetic factors may have a decisive impact on prognosis.We applied a new set of grouping reference values established in Part II to screen long-term survivors and normal survivors from patients with ovarian serous cystadenocarcinoma,identify their long-term survival-related gene expression profiles,and perform functional analysis.Methods:1.Cases obtained from the Cancer Genome Atlas Ovarian Cancer(TCGA-OV)database were divided into long-term survivors(LTS)and normal-term survivors(NTS)based on survival cutoffs defined by the previous study of part two.2.The gene expression matrices of LTS and NTS were downloaded,normalized,and filtered.Differentially expressed genes(DEGs)were identified by Edge R analysis pipeline with criteria ?log FC??1 and FDR?0.001.Gene Ontology(GO)functional and Kyoto Encyclopedia of Genes and Genomes(KEGG)pathway enrichment analyses were performed using DAVID Bioinformatics Resources.DEG-related protein-protein interactions(PPI)were extracted from the STRING database and hub genes were identified by Cyto Hubba package in the Cytoscape program.Results:1.In this study,19 LTS cases and 105 NTS cases were collected from 374 cases in the TCGA-OV data set according to the new survival cutoffs.There was no clinical baseline difference between the two groups.2.Differential gene expression analysis of tumor tissues from patients in the LTS and NTS groups identified 157 DEGs associated with long-term survival,including 155 upregulated genes and 2 downregulated genes.3.Upregulated genes were statistically enriched in 80 GO terms and 11 KEGG pathways related to energy and substrate metabolism,such as protein absorption,digestion,and metabolism as well as signaling pathways,including chromatin silencing,regulation of ERK1 and ERK2 cascade,and regulation of MAPKKK.ALB and POMC were the common hub genes.Conclusions: Protein anabolism may crucial to long-term survival of ovarian serous cystadenocarcinoma,regulated by activation of the MAPK/ERK signaling pathway and chromatin silencing.Comprehensive understanding of the molecular mechanisms via further exploration may contribute toward an effective treatment for ovarian cancer.