| Objective:Cervical cancer is one of the most common malignancies in women globally which threatens women’s health.Persistent infection of high-risk human papillomavirus(hr-HPVs)is responsible for the initiation and progression of cervical cancer.As one of the most common subtype of hr-HPVs,HPV16 is the major casual agent of cervical cancer.We previously discovered that there seems to be 2 groups of microRNA(miRNA)which are downregulated in early phase and late phase of HPV16 infection respectively by high throughput microarray analysis.Further prediction via bioinformatic database shows the most probable upstream transcription factors(TF)are c-Rel,p65,p50,STAT1 and ELK 1(termed as HPV related microRNA upstream transcription factor,HPV16rmiR-TF)which may play a vital role in development of HPV positive cervical cancer.This study aims to explore the correlation between HPV16rmiR-TF and initiation of cervical cancer.So the method of immunohistochemistry was used to detect the nuclear expression pattern of HPV16rmiR-TF in various kinds of cervical tissue.Another intention of this study is to investigate the relationship between nuclear HPV16rmiR-TF and time period of HPV16 infection.We then combined the results of HPV typing and pathology examination to rebuild an infection model and tried to figure out the nuclear expression of HPV16rmiR-TF in each period and served it as an evidence for the existence of TF-miRNA regulatory network.The correlation between HPV16rmiR-TF and clinical-pathological parameters of cervical cancer patients and the correlation between nuclear HPV16rmiR-TF in cervical cancer tissue are also the objective of our study.We next analyzed the prognostic significance of HPV16rmiR-TF and other important index including clinical-pathological parameters and serum level of tumor markers.a prognostic model was then established based on the independent prognostic factors which would help us profoundly understand the prognostic significance of HP V16rmiR-TF.This could provide a new insight of the study of prognosis of cervical cancer.Method:Immunohistochemistry was used for detection of HPV16rmiR-TF in non-lesion cervix,LSIL,HSIL and cervical cancer was performed.Theanalysis of correlation between HPV16rmiR-TF and HPV16 infection timephase was based on the nuclear expression of TFs in negative group,earlyphase group and late phase group.To investigate the clinical significance of HPV16rmiR-TF,the collection of important information of cervical cancer patients and the correlation analysis were proceed.After follow-up of enrolled cervical cancer patients,we constructed the database of prognostic information.Univariate and multivariate survival analysis based on Cox proportional hazard model were conducted to identify the independent prognostic factors of post-operation survival/recurrence lifetime of cervical cancer patients.Prognostic Index(PI)scoring system for the prediction of prognosis were then established and evaluated by ROC curve analysis for its accuracy.PI scoring system was compared with the FIGO staging system on the discrimination.The applicability of PI system was assessed by ROC curve analysis on its 1-year,3-year and 5-year accuracy.The baseline hazard function was calculated via interpolation method to establish individualized prediction model.Result:The nuclear expression level of c-Rel,p65,p50 in normal cervix is lower than cervical cancer tissue significantly and the expression pattern is related to the severity of cervical lesions.Nuclear status of STAT1 in cervical cancer is weaker than the status in non-lesion cervix and precancerous lesions.Nuclear ELK1 shows no significant discrepancy between all groups.Nuclear expression of c-Rel,p65,p50 and STAT1 is associated with time phase of HPV16 infection.Nuclear p65 is up-regulatedduring early phase;Nuclear p50 and nuclear STAT1 is down-regulated during early phase;Nuclear c-Rel shows a graduate increase both in earlyand late phase.Nuclear c-Rel and p65 are related with vascular invasion and parametrial extension of HPV16 positive cervical cancer patients.Coxmultivariate analysis indicates that parametrial extension,nuclear p65 and STAT1 are independent prognostic factors of survival while FIGO stage,preoperative SCC-Ag,nuclear p50 and STAT1 are independent prognostic factors of recurrence.With these information the PI scoring system is developed and performs well in prediction of survival and recurrence according to the result of ROC curve analysis.PI scoring system has a better discrimination in prediction than FIGO staging system and capable of prediction of 1-year,3-year and 5-year prognosis.The individualized prediction model for prognosis of cervical cancer patients based on the baseline hazard and survival function was established.Conclusion:The intensity of nuclear c-Rel,p65 and p50 are along with the severity of cervical lesion which suggests they may function as a oncogene in the initiation and progression of cervical cancer.The nuclear STAT1 may be a protector of cervical carcinogenesis.Nuclear expression of c-Rel,p65,p50 and ST AT1 is associated with time phase of HP V16 infection.TF-miRNA regulatory network may consists of c-Rel,p65,p50,STAT1 and specific miRNAs.Nuclear c-Rel,p65 expression is correlated with lymph vascular space invasion,parametrial invasion of HPV16 positive cervical cancer patients,which hints a potential capacity of NF-κB to promote invasion and metastasis of cervical cancer.NF-κB and STAT1 probably show a antagonism in HPV16 positive cervical cancer.Statistical and visible nomogram prognosis model based on nuclear expression of p65,p50 and STAT1 has good discrimination and accuracy. |