| Objective:1.To evaluate the curative effect of combined therapy(TCM plus entecavir)in HBeAg positive chronic hepatitis B patients based on the research results of the national key project of the 11th-5 years plan named ’the research of Chinese Medicine syndrome in CHB and the integrated medicine’.2.To establish prediction models of curative effect during the combined therapy in HBeAg positive chronic hepatitis B patients.Method:The research was based on prospective,multicenter,randomized,controlled,double-blind clinical trial design method included 596 cases,(301 of treatment group and 295 of control group).Treatment:1.Therapy of treatment group:if the cases were defined as HBeAg>60 S/CO(approximately equal to 14.114 PEIU/ml)in the beginning,then TCM granule A(function:regulating the liver and benefiting the spleen)and placebo of entecavir were given.In the follow-up every 4 weeks for 24 weeks,liver functions and HBV-M were tested,and if the cases were defined as ALT≥8×ULN and HBeAg<200 S/CO(approximately equal to 50.572 PEIU/ml),or HBeAg<60 S/CO,then TCM granule B(function:regulating liver,invigorating spleen and detoxificating)and entecavir were given in week 4-24.If the cases of treatment group were defined as HBeAg<60 S/CO in the beginning,then TCM granule B and entecavir were given in week 0-24.All the cases were given TCM granule B and entecavir in week 25-108.2.Therapy of control group:the cases were given entecavir and placebo of TCM granule in week 0-108.Therapeutic indicators included hepatitis B virus(HBV)serum markers,HBV nucleic acid markers,hepatic functions and TCM symptoms score.According to whether HBeAg loss,stratification research was conducted.We established prediction models of curative effect during the combined therapy in HBeAg positive chronic hepatitis B patients based on binary logistic regression analysis,decision tree and neural network algorithm,then evaluated prediction efficiency of the models.Result:1.Primary outcome indicators:Compared to control group,the rates of HBeAg loss in treatment group were 9.45%higher(35.55%(107 cases)vs 26.1%(77 cases),P<0.05).The levels of HBeAg in treatment group were lower than control group(4.26 S/CO vs 5.84 S/CO,P>0.05).The rates of HBeAg seroconversion were 83.72%(252 cases),84.75%(250 cases)in treatment and control group,respectively.And there was no significant difference between the two groups(P>0.05).The rates of HBV DNA loss were 83.72%(252 cases),84.75%(250 cases)in treatment and control group,respectively.And there was no significant difference between the two groups(P>0.05).2.Secondary outcome indicators:The overall efficiency of improvement of TCM symptoms in treatment group was slightly higher than control group(71.76%vs 71.53%,P>0.05).There was no significant difference in the improvement of single TCM symptom between the two groups(P>0.05).Compared to control group,treatment group was better improvement of hypochondriac pain,drowsiness fatigue,and depressed emotion(P>0.05),especially heavy sensation in limbs at week 24 and 96(P<0.05).Control group was better improvement of dry mouth,jaundice,sticky mouth,red tongue and yellowish fur than treatment group(P>0.05).The rates of ALT normalization were 83.06%,83.39%in treatment and control group,respectively.And there was no significant difference between the two groups(P>0.05).The levels of ALT and AST in treatment group were higher than control group at week 12,24(P<0.05).The levels of GGT in treatment group were higher than control group at week from 12 to 108(P<0.05).The levels of TBIL in treatment group were lower than control group at week 36,48 and 84(P<0.05).The levels of DBIL in treatment group were lower than control group at week 72,84 and 96(P<0.05).The levels of IBIL in treatment group were lower than control group at week 36,48,60 and 96(P<0.05).3.Stratification research:(1)Overall group(596 cases):The levels of HBeAg of subjects who achieved HBeAg loss(group A)were significantly lower than subjects who did not achieve HBeAg loss(group B)(P<0.05).The levels of ALT,AST,GGT,AKP and TP of group A were significantly higher than group B(P<0.05).The levels of age of group A were younger than group B(P<0.05).(2)TCM combined entecavir treatment group(301 cases,treatment group):The levels of HBeAg of subjects who achieved HBeAg loss(group C)were significantly lower than subjects who did not achieve HBeAg loss(group D)(P<0.05).The levels of ALT,AKP and TP of group C were significantly higher than group D(P<0.05).There was no significant difference in the age between the two groups(P>0.05).(3)Entecavir monotherapy group(295 cases,control group):The levels of HBeAg of subjects who achieved HBeAg loss(group E)were lower than subjects who did not achieve HBeAg loss(group F)(P=0.051).The levels of ALT,AST and GGT of group E were significantly higher than group F(P<0.05).There was no significant difference in the age between the two groups(P>0.05).4.Prediction models:The prediction models had good predictive values based on binary logistic regression analysis,decision tree and neural network algorithm.Based on levels of factors at baseline,areas under AUROC were 0.649(P<0.01,95%CI:0.601-0.696),0.663(P<0.01,95%CI:0.615-0.711),0.684(P<0.01,95%CI:0.683-0.729),respectively.Based on levels of factors at week 12,areas under AUROC were 0.696(P<0.01,95%CI:0.651-0.742),0.713(P<0.01,95%CI:0.667-0.758),0.726(P<0.01,95%CI:0.682-0.770),respectively.Based on levels of factors at week 24,areas under AUROC were 0.777(P<0.01,95%CI:0.734-0.82),0.782(P<0.01,95%CI:0.741-0.822),0.798(P<0.01,95%CI:0.759-0.838),respectively.Conclusions:1.The combined therapy(TCM plus entecavir)had higher HBeAg loss rates than monotone entecavir therapy.2.The combined therapy had an advantage in improvement of Liver depression and Spleen deficiency Syndrome,especially in heavy sensation.The monotone entecavir therapy had an advantage in improvement of Liver and Gallbladder Wetness-heat Syndrome.3.Stratification research suggested:The characteristics of subjects who achieved HBeAg loss in overall,treatment and control group were low levels of HBeAg and high levels of transaminase at baseline.4.Compared to prediction models based on baseline levels and treatment response at week 12,prediction efficiency of models based on treatment response at week 24 was better.5.The logistic regression prediction models suggested:TCM intervention,high level of ALT and low level of HBeAg at baseline predicted HBeAg loss.TCM intervention and low level of HBeAg at week 12 and 24 predicted HBeAg loss.6.The decision tree prediction models suggested:The rates of HBeAg in treatment group were higher for the patients with low level of HBeAg at baseline.For the patients with high level of HBeAg at baseline,younger combined with high level of ALT at baseline predicted higher rates of HBeAg loss in treatment and conrol groups,but there was no significant difference between the two groups.For the patients with low level of HBeAg at week 12 and 24 the rates of HBeAg loss in the two groups were higher,but there was no significant difference between the two groups.For the patients with high level of HBeAg at week 12 and 24 the rates of HBeAg loss in the two groups were lower,but the rates of HBeAg loss in treatment group were higher than conrol group. |
PDF Full Text Request