AMAP Score Combined With Liver Stiffness Measurement Accurately Assess Liver Fibrosis In Patients With Chronic Hepatitis B Before Or After Antiviral Treatment

Background:Evaluation of liver fibrosis in patients with chronic hepatitis B(CHB)is important for initiation of antiviral therapy as well as evaluation of therapeutic efficacy.The current serum non-invasive liver fibrosis diagnostic models,fibrosis index based on four factors(FIB-4)and aspartate aminotransferase-platelet ratio(APRI)for example,did not show consistent satisfactory performance of liver fibrosis diagnosis in different study populations.In addition,it is still controversial whether the dynamic changes in the liver stiffness measurement(LSM)during antiviral treatment can predict the regression of liver fibrosis within CHB patients.Therefore,new models or indexes are needed to detect liver fibrosis among CHB patients before or after antiviral treatment.Recently built the aMAP score,as an accurate hepatocellular carcinoma risk score,may reflect the degree of liver fibrosis.Objective:We aimed to evaluate the performance of the aMAP score for the diagnosis of cirrhosis and advanced fibrosis in CHB patients;then,built new model to identify cirrhosis and advanced fibrosis in CHB patients after antiviral treatment.Methods:2053 CHB patients with liver biopsy from 2 real-world cohorts and 2 multi-centric randomized controlled trials in China were enrolled,among which 2053 patients were included in the cross-sectional analysis to evaluate and compare the performance of the aMAP score,FIB-4,and APRI for diagnosis of cirrhosis and advanced fibrosis;889 patients with paired liver biopsies before and after 72-or 104-week treatment were included in the longitudinal analysis to construct new models for cirrhosis and advanced fibrosis diagnosis in post-antiviral CHB patients.Dual cut-offs were chose to obtain sensitivity(lower cut-off)and specificity(higher cut-off)of 90%or 95%.Results:In the cross-sectional analysis,the area under the receiver operating characteristic curves(AUROCs[95%CI])of the aMAP score in diagnosing cirrhosis and advanced fibrosis were 0.788[0.770,0.805]and 0.757[0.738,0.775],which were significantly higher or comparable with those of FIB-4(0.763[0.744,0.781],p=0.010;0.753[0.734,0.772],p=0.694)and the APRI(0.607[0.586,0.629],p<0.001;0.613[0.591,0.634],p<0.001),respectively.When applying dual cut-off values,the aMAP score had a trend toward better performance than FIB-4 and APRI with a numerically smaller uncertainty area(UA)and a higher diagnostic accuracy(DA)in diagnosing cirrhosis.Specifically,the UA was 68.3%for the aMAP score with DA of 85.2%in cirrhosis detection(74.2%,84.5%for FIB-4,and 79.0%,78.0%for APRI,respectively).Moreover,compared with stepwise approaches using FIB-4(or APRI)and LSM,the stepwise approach using the aMAP score and LSM also had the smallest UA and satisfactory DA in both cirrhosis(UA:29.7%;DA:82.3%)and advanced fibrosis(UA:46.2%;DA:79.8%)detection.In the longitudinal analysis,we further established a novel model(aMAP-LSM model)comprising the aMAP score at baseline,LSM at week 72(or 104)and the aMAP score at week 72(or 104)by using logistic regression,which had satisfactory performance in diagnosing cirrhosis and advanced fibrosis after anti-HBV treatment and significantly outperformed LSM alone(AUROC[95%CI]:0.839[0.813,0.865]vs.0.799[0.746,0.812],p<0.001 and 0.840[0.815,0.866]vs.0.776[0.745,0.808],p<0.001,respectively).For those with significant decline in LSM after treatment,aMAP-LSM model still presented higher AUROC[95%CI]than LSM alone(cirrhosis:0.828[0.792,0.864]vs.0.748[0.700,0.796],p<0.001;advanced fibrosis:0.825[0.789,0.862]vs.0.750[0.705,0.794],p<0.001).Conclusion:aMAP score is a promising non-invasive tool for diagnosing cirrhosis and advanced fibrosis in CHB patients.The aMAP-LSM model could accurately estimate cirrhosis and advanced fibrosis stage for treated CHB patients.