| Background:Herbal medicine has a long history in the world,and is thought to be safe and non-toxic to human's health for a long time.But more and more reports show that Traditional Chinese Medicine(TCM),Natural Medicine(NM),Health Products(HP)and Dietary Supplements(DS),which are also called as herbal medicines and dietary supplements(HDS)in America and European countries,can cause various patterns of liver injury.With the economic development and the life style changes,non infectious liver diseases such as non-alcoholic fatty liver disease(NAFLD),alcoholic liver disease(ALD)and autoimmune liver disease(AILD)become more and more common.Drug-induced liver disease(DILI)on the basis of these chronic liver disease has become one of the research hotspots.The relationship between drugs and other liver diseases is very complicated.Some drugs can both treat the primary liver disease and cause a new liver injury.With the changes of life style,the increasing of environmental pollution and various diseases,application of herbs becomes more and more popular.The constituent ratio of herbs-induced liver injury(HILI)in DILI is getting higher and higher.However,the clinical characteristics,diagnosis and differential diagnosis of HILI,especially the differential diagnosis between HILI and non-infectious liver disease need further study.These studies will raise the awareness of HILI in public and clinician and improve the prevention,diagnosis,treatment and prognosis of HILI.Methods:Clinical data of hospitalized patients diagnosed as DILI between January 2012 and November 2016(a total of 4 years and 11 months)in the Liver Disease Center of 81st Hospital,Nanjing University of Chinese Medicine,were retrospectively collected and analyzed in detail.The search keywords included not only "DILI",but also "drug-induced hepatitis","drug-induced liver failure","drug-induced hepatocirrhosis","sinusoidal obstruction syndrome","veno-occlusive disease" or other diagnostic terms which may indicate the existence of DILI.By further manual search,only the cases with suspected HILI were included in this study;the cases whose liver injury may be caused by western medicine,or combination of chinese and western medicine,or those whose crucial clinical data were incomplete,were ruled out from our study.Then we recorded the clinical data such as general information,medication history,clinical manifestation,laboratory and pathological examination,imaging examination,treatment and prognosis of the patients.The causality between herbs and liver injury was evaluated by two editions of RUCAM scale.The demographic data,components of suspected herbs,clinical features,diagnosis and differential diagnosis between HILI and non-infectious liver disease,and prognosis of patients with suspected HILI were analysed.Results:243 cases confirmed as suspected HILI by manual search of 528 cases of suspected DILI were collected in this study,accounting for 46.20%of DILI,including 34 cases of sinusoidal obstruction syndrome(SOS)and 209 cases of non vascular HILI(nvHILI).The 209 cases of nvHILI can be devided into group without chronic liver disease(simple nvHILI,n=150/209,71.77%)and group with coexistent chronic liver disease(CLD+nvHILI,n=59/209,28.23%).The 59 patients with coexistent chronic liver disease can be furtherly devided into NAFLD+nvHILI(n=26/59,44.07%),ALD+nvHILI(n=14/59,23.73%)and AILD+nvHILI(n=19/59,32.20%).The median age of 243 patients with HILI was 51(16-88)years,and 48(16-88)years in simple nvHILI,51(32-82)years in CLD+nvHILI.Both in group without CLD and in group with CLD,the highest median age were in the range of 40-59 years.In addition,the ratios of female to male were 1:1.31 in 243 cases of HILI,1:2 in 209 cases of nvHILI,1:1.4 in 26 cases of NAFLD+nvHILI,14:0 in 14 cases of ALD+nvHILI and 1:8.5 in 19 cases of AILD+nvHILI respectively.The definite specific culprit herbs were not available in 47.33%(115/243)of the HILI cases.Among the confirmed culprit herbs,Gynura segetum(33/243,13.58%),Polygonum multiflorum(20/243,8.23%)and health products(12/243,4.94%)were listed as the most often factors leading to DILI,which mainly used for the treatment of bone fractures(34/243,13.99%),digestive diseases(27/243,11.11%),skin diseases(26/243,10.70%)and health care purposes(21/243,8.64%).There was no significant difference of the sorts of suspected herbs between group with CLD+nvHILI and group with simple nvHILI.The median latency of simple nvHILI,NAFLD+nvHILI,ALD+nvHILI,AILD+nvHILI and SOS were 30(0-3622),31(0-2374),42(0-346),54(0-1338)and 76(0-3579)days respectively.There was no significant difference in latency between simple nvHILI and CLD+nvHILI(Mann-Whitney test,Z=-1.847,p=0.065),but had a significant difference in latency between simple nvHILI and SOS(Mann-Whitney test,Z=-3.046,p=0.002).The main clinical symptoms were dark urine(153/209,73.2%),fatigue(148/209,70.8%)and anorexia(133/209,63.6%)in the nvHILI group.The range of median values of liver function tests(LFTs)and coagulation tests of the first time in each nvHILI subgroup were as follows,TBIL:53.7-126.7umol/L,DBIL:41.6-79.6umol/L,ALT:688-1048U/L,AST:558.5-698U/L,GGT:195-274.8U/L,ALP:139-254U/L,PTA:91.4-103.1%,INR:0.97-1.03.There were no significant difference of the levels of TBIL(Kruskal Wallis test,?2=5.231,p=0.156),DBIL(Kruskal Wallis test,?2 =4.132,p=0.248),ALT(Kruskal Wallis test,?2=1.325,p=0.723),AST(Kruskal Wallis test,?2=1.150,p=0.765),GGT(Kruskal Wallis test,?2=3.173,p=0.366),ALP(Kruskal Wallis test,?2=2.420,p=0.490),PTA(Kruskal Wallis test,?2=2.666,p=0.446),INR(Kruskal Wallis test,?2=3.063,p=0.382)among the subgroups.What's more,there were no any specific features in imaging examination.A total of 3 patients with nvHILI received liver biopsy,and the pathological reports were DILI,NAFLD+DILI and chronic DILI,respectively.The clinical data of 141 patients,especially the initial LFTs values were relatively complete in the nvHILI and can be used to calculate the R values.Based on the R values,the patterns of liver injury can be divided into hepatocellular injury(n=125),cholestatic injury(n=7)and mixed liver injury(n=9).126 of the 141 patients had definite medical history,thus were evaluated with RUCAM scales,with the results of 15 "possible" cases,91 "probably" cases and 20 "highly probably" cases.There was no difference between the 1993 and 2015 editions of the RUCAM scale(?2=0.000,p=1.000).Among 204 nvHILI cases with perfect hospitalized examination data,the severity of liver injury can be divided into mild degree in 78 cases(78/209,37.32%),moderate degree in 42 cases(42/209,20.10%),severe degree in 77 cases(77/209,36.84%)and liver failure in 7 cases(7/209,3.35%).The average hospital stay of patients with mild,moderate,severe liver injury or liver failure were 14(2-56),21(4-65),29(4-116)and 19(1-54)days,respectively.There was significant difference in the hospital days among patients with mild,moderate and severe liver injury(F=21.953,p=0.000).The respective average stay in simple nvHILI,NAFLD+nvHILI,ALD+nvHILI and AILD+nvHILI were 21(1-93),22(3-116),24(7-46)and 21(10-46)days,there have no statistical difference among the groups(F=0.223 p=0.925).The effective rates of treatment in mild,moderate,severe and liver failure patients were 96.15%(75/78),97.62%(41/42),93.51%(72/77)and 14.29%(1/7),respectively.Compared with that of mild(?2=45.476,p=0.000<0.008),moderate(?2=34.028,p=0.000<0.008)and severe(?2=35.386,p=0.000<0.008)liver injury,patients with liver failure had significantly lower effective rate of treatment.The effective rates of treatment in subgroups with simple nvHILI,NAFLD+nvHILI,ALD+nvHILI or AILD+nvHILI were 94%(141/150),92.31%(24/26),85.71%(12/14)and 89.47%(17/19),respectively;and there were no significant difference among those subgroups(?2=1.704,p=0.636).In the SOS group,the median age of the patients was 64(45-79)years old,and the ratio of male to female patients was 4.7:1.The main clinical symptoms were abdominal distension(30/34,88.2%),fatigue(20/34,58.8%)and anorexia(19/34,55.9%).The levels of TBIL(38.65vs89.3,Z=-3.493,p=0.000),DBIL(23.40vs61.25,Z=-3.348,p=0.001),ALT(177vs939,Z=-5.619,p=0.000),AST(146vs698,Z=-4.891,p=0.000)and PTA(66.15vs91.4,Z=-2.757,p=0.006)in SOS group were significantly lower than that in patients with simple nvHILI.The features of imaging showed patch(maps)-like changes,inhomogeneous enhancement in the liver,and the hepatic veins were unclear or disappeared.For the treatment of SOS,29 patients received symptomatic treatment,2 patients received symptomatic and anticoagulant treatment,and 3 patients were treated with TIPS.The severity of SOS was improved in 18 patients and showed no improvement in 13 patients.Three patients were loss for follow-up.There was no significant difference of response to the treatment among the three treatment strategies(?2=0.146,p=0.930).Discussion and Conclusion:This study shows that the incidence of HILI is increasing and herbs have become one of the main causes of DILI,especially in middle-aged women.Gynura segetum,Polygonum multiflorum and HP are the main factors that can induce liver injury,and are used in the bone fractures,digestive diseases,skin diseases and health purposes commonly.Patients with CLDs such as NAFLD,ALD and AILD did not take in specific herbal medicines,but the coexistent CLDs really contributed to the difficulty of diagnosis of HILI.On the other hand,carelully investigation of the medical history,laboratory results,radiographic examination and liver biopsy will give help to identify HILI occurred on the CLDs effectively.There is no difference in the evaluation results between the 1993-and 2015-editions of RUCAM scale in the study.Because the 2015 RUCAM has a more friendly operation interface,well-defined evaluation elements,and relatively lower intra-and inter-observators inconsistency,thus is recommended as the new causality assessment tools of DILI/HILI.It's reminded that the RUCAM is more suitable for the prospective evaluation than the retrospective studies of DILI/HILI,unless the patients in retrospective studies have all of the information that can be used to definitely rule out any other typs of liver diseases.In this study,hepatocellular injury was the main pattern of liver injury and the prognosis was good.The average length of hospital stay and effectiveness of treatment were related to the degree of liver injury,while the CLDs such as NAFLD,ALD and AILD seemed to have no significant influence on the latency,average hospitalization time,degree of liver injury and prognosis of HILI.This needs to be furtherly investigated in more patients with variable severity of underlying liver disease in the future.SOS was induced by Gynura segetum,which was usually used to treat bone fracture diseases.It had a longer incubation period,especially in aged persons.The main clinical manifestations were abdominal distension,and the levels of serum bilirubin,transaminase and coagulation were lower than that of simple nvHILI.But its imaging features will help make the diagnosis of SOS definitely.Patients with SOS should be treated combinedly with symptomatic therapy,anticoagulation and TIPS,but the overall prognosis is poor.HILI can mimic any other type of acute and chronic liver disease,including DILI caused by non-herbal factors.Currenty,due to the lacking of specific diagnostic markers,the diagnosis of HILI is mainly based on the exclusion of other factors that can lead to liver injury.Another concern is that the components of suspicious herbs are very complex,and their chemical composition and toxicological mechanism are not clear.Accordingly,the nomenclature of HILI and classification of culprit herbs are various in different countries and regions.Furthermore,the differential diagnosis of HILI from coexistent CLD is usually complicated and uncertain,and these also increase the difficulty of clinical diagnosis of HILI.The futural,research of HILI needs the strategic cooperation clinicians,chemists,toxicologists,medical scientists and drug administrators,including large multi-center prospective research of DILI/HILI,identification of chemical constituents and toxicological analysis of the herbs with modern technologies,investigation of the pathogenesis of HILI and exploration of the HILI-specific biomarkers through various "omics" ways.Thus the chemists and toxicologists are able to separate and identify the potential hepatotoxic herbal components,and clinicians can recognize the relevant herbs that induce liver injury in order to diagnose HILI accurately.It also will provide evidence for the food and drug administration to strengthen and improve the relevant regulations and adverse event reporting system about TCM-NM-HP-DS in order to ensure public health. |
PDF Full Text Request