Detection Of High-Risk Individuals Of Autoimmune Liver Disease By Health Check-up And Study Of Genetic Susceptibility In Family

Background and AimsAutoimmune liver disease(AILD)is a chronic autoimmune disease caused by genetic and environmental factors.With the prevalence increased,AILD has been one of the most important etiology of liver cirrhosis.Previous studies have showed that early detection of AILD is crucial for the prognosis of patients.However,identifying high-risk individuals and early diagnosis of AILD is still a challenge due to the lack of screening methods.In addition,heredity plays a key role in the pathogenesis of AILD,especially the internationally recognized human leukocyte antigen haplotypes.Over the last few years,genome-wide association studies and related genetic association studies found some susceptible genes associated with AIH and PBC and tried to analyze their pathogenesis through the susceptibility genes,but family study was still rare.Primary biliary cholangitis-autoimmune hepatitis overlap syndrome(PBC-AIH OS)is a rare subtype of AILD with only a few reported cases,and its genetic research is still blank.The current study aim to: 1)analyze the predictive value of several indicators in routine health check-up for AILD,and establish prediction models for identifying high-risk individuals of AILD;2)Explore the common low frequency mutations among the three PBC-AIH OS families through the whole exome sequencing(WES),and the candidate germline variations were screened by bioinformatics methods to discover the pathogenesis of PBC-AIH OS.Method Part I: Detection of High-Risk Individuals of Autoimmune Liver Diseases by Health Check-upA total of 581 patients with AILD diagnosed between 2001 and 2017 were enrolled,and 1000 control subjects were collected from health check-up in the same center.We analyzed the cause of AILD patients referring and the proportion of cirrhosis at diagnosis over the years,and analyzed the clinical and laboratory information at first diagnosis.Factors like age,gender,liver function tests and extrahepatic autoimmune diseases in routine health check-up were compared betweenAILD patients and controls.Correlation,regression analyses and cross-validation,classification and regression tree(CART)of machine learning were performed to select indicators and establish prediction models.Part II:Whole exon sequencing for familial PBC-AIH OS gene mutationWe have performed WES in 3 autosomal dominant PBC-AIH OS families including 7 patients and 2 normal family members.We further used bioinformatics analysis to select the common variants,and family analysis,gene-phenotype-function analysis were also performed to enrich the candidate germline variations and obtain the most likely potential pathogenic variations.Result Part I: Detection of High-Risk Individuals of Autoimmune Liver Diseases by Health Check-up1.From 2000 to 2017,the number of AILD patients diagnosed in our department increased over the years,and most of the patients had no obvious symptoms at presentation.Abnormal liver function in routine health check-up became the main reason of diagnosis with AILD.2.The proportion of AILD patients who had liver cirrhosis at presentation gradually decreased from 49%(before 2006)to 20%(2017),and the decreasing trend was confirmed by statistic tests(?2=19.36,P<0.0001).3.Compared to control subjects in health check-up,eight paramethers including age,gender,globulin,alanine aminotransferase(ALT),?-glutamyltransferase(GGT),peri-hepatic lymph node enlargement,extrahepatic autoimmune diseases and familial autoimmune history were highly associated with AILD(all P value <0.001).4.In the logistic regression model,globulin(OR,1.28)was the strongest predictor followed by ALT(OR,1.04)and GGT(OR,1.01).Further validation showed excellent predictive power(AUC=0.98),sensitivity(96%)and specificity(92%).5.In order to identify AILD from people with abnormal liver function tests,we further excluded two direct liver or cholangetic injury parameters(ALT and GGT),another regression model and a CART model were established with the remaining 6indicators.Both of them should be applied under the premise of abnormal liverfunction tests and also had good predictive abilities(AUC 0.97 vs 0.91)in external validation.Part II:Whole exon sequencing for familial PBC-AIH OS gene mutation1.The 3 PBC-AIH OS families were autosomal dominant inheritance.Through WES and bioinformatics analysis of the 7 patients and 2 healthy family members,a total of237 low-frequency mutations were selected from patients in family 1 and 327 in family 2.In family 3,there were 203 low frequency mutations were found in the patients which were not found in the control subjects.2.Among the above mentioned candidate mutations,the terminating mutation rs200988634(OTOA)is a common SNV site in three families,and Ch2:18755905(A> AACC)on FAM171 B gene is a non-frameshift insertion mutation of three families.The two variants were co-separated with PBC-AIH OS in the 3 families.3.We evaluated the candidate variations by pathway enrichment and gene-phenotypic association analysis,and found that TNPO3(rs760634627),C1S(rs147971272),PDE8B(Ch5:76506914),TYK2(rs368109068),SLC10A2(rs201206937)were also potential pathogenic variations of PBC-AIH OS,which laid a foundation for exploring the target of diagnosis and treatment of PBC-AIH OS.Conclude1.The majority of patients with AILD were asymptomatic at diagnosis whether there were cirrhosis or not.Identifying high-risk individuals during health check-up can significantly reduce the proportion of cirrhosis at presentation.Several parameters in routine health check-up were associated with the risk of AILD.The prediction model based on these indicators could identify high-risk individuals of AILD,which provides a way for early detection of AILD.2.The mutations rs200988634(OTOA),rs760634627(TNPO3),rs147971272(C1S),Ch5:76506914(PDE8B),rs368109068(TYK2),rs201206937(SLC10A2)in PBC-AIH OS family may be correlated to the pathogenesis of PBC-AIH OS.