Assessment of prescribed herbal medicine on liver function in korea: a prospective observational study

THE JOURNAL OF ALTERNATIVE AND COMPLEMENTARY MEDICINEVolume 14, Number 9, 2008, pp. 000–000 Mary Ann Liebert, Inc.
DOI: 10.1089/acm.2008.0053 Original Papers
Assessment of Prescribed Herbal Medicine on Liver Function in Korea: A Prospective Observational Study Young-Ju Yun, M.D., K.M.D.,1,* Seong-Su Nah, M.D., K.M.D.,2,* Jae-Hyun Park, K.M.D., Ph.D.,3 Seul-Ki Jung, K.M.D.,3 Eun-Kyung Paeck, K.M.D., Ph.D.,3 Jae-Hyung Park, K.M.D.,3 Eui-Joon Lee, K.M.D., Ph.D.,4 Se-Hyun Kim, Ph.D.,1 Hae-Mo Park, K.M.D., Ph.D.,5 Sun-Dong Lee, K.M.D., Ph.D.,5 Ji-Yong Kim, M.D., Ph.D.,6 Seung-Jae Hong, M.D.,7 Bong-Keun Choe, M.D.,7 and Joo-Ho Chung, M.D., Ph.D.7 Abstract
Objective: The aim of this study was to investigate the influence of herbal medicine (HM) prescribed by doc-
tors of Korean medicine (KMD) on liver function in Korea.
Design and interventions: For this multicenter, prospective, observational study, we enrolled patients who
wished to take HM prescribed by KMD for various medical purposes in Korea. One hundred and twenty-two
(122) patients took HM for an average of 20.6 Ϯ 8.4 (mean Ϯ standard deviation) days, and completed ques-
tionnaires.
Outcome measures: Liver function tests (LFTs) were performed before (first test) and after each HM treatment
(second test). For LFT, aspartate aminotransferase, alanine aminotransferase, total bilirubin (t-Bil), direct biliru-
bin, alkaline phosphatase, and ␥-glutamyl transferase were measured.
Results: There were no significant changes in LFT data between the first and second tests, except in the t-Bil
level. However, all data of total bilirubin level in second test were within normal range, except only one pa-
tient. Multivariate analysis did not identify any herb that significantly increased t-Bil; hence no hepatotoxic
herb was found. Twenty-one (21) of the 122 patients were abnormal on first testing, and 10 at the second test-
ing. Of the patients taking herbs, 4 changed from normal to abnormal and 15 from abnormal to normal (p ϭ
0.019).
Conclusion: The current study showed that ingestion of HM prescribed by KMD did not increase the frequency
of abnormal LFTs, at least in the short term.
Introduction
much research. Although various studies have investigatedHM use and the toxicity of some HM,1,2 only a few system- Complementary and alternative medicine (CAM) is more atic studies have examined adverse drug reactions related to
popular today than ever before. Despite the common HM use.3,4 Possible hepatotoxicity is an important issue perception that CAM is natural and therefore safe, many when HM is used.5 Liver injuries associated with the use of studies have demonstrated that CAM is not entirely free of a particular HM or mixed prescriptions have been reported,6 risk. Herbal medicine (HM) has many uses and has attracted and some studies have suggested that HM is an important 1Graduate School of East-West Medical Science, Kyung Hee University, Suwon, Korea.
2Department of Internal Medicine, Soonchunhyang University, College of Medicine, Choenan, Korea.
3Hae-Ma Traditional Korean Medical Clinic, Seoul, Korea.
4Graduate School of Alternative Medicine, Pochon Cha University, Seoul, Korea.
5Department of Preventive Medicine, College of Traditional Korean Medicine, SangJi University, Wonju, Korea.
6Department of Family Medicine, College of Medicine, Dongguk University, Gyungju, Korea.
7Kohwang Medical Research Institute, College of Medicine, Kyung Hee University, Seoul, Korea.
*These two authors contributed equally to this work.
YUN ET AL.
cause of drug-induced liver injury (DILI).6,7 Other studies re-ported that HM-related liver injury occurred rarely, and that prognosis during follow-up was good.8–10 Not only is the real incidence of liver injury unclear, but few data are avail- able on the clinical features and outcomes of liver injuriescaused by HM.
It is important to establish the safety of HM in Korea, be- cause HM have been used very frequently in Korea. Tradi- tional HM in Korea has shared many characteristics with HMin China. Prescribed HM was provided by doctors of Korean medicine (KMD) who had to have graduated traditional medical college for their license. Although some studies haverecently focused on HM in Korea,11,12 considerable dis-agreement remains regarding the safety and efficacy of HM.
A few case reports have described hepatotoxic events causedby the self-purchase and administration of a single herb,13 Enrolled and Complete the 1st Intervention but few clinical studies have examined liver injuries caused by prescribed HM that was provided by KMD. Other stud-ies in Korea have been retrospective, assessed DILI, assumedHM to be the main cause of any liver injury found, and nodistinction was made between self-ingested herbs and pre-scribed HM.14 In view of this lack of previous studies, to evaluate the safety of HM prescribed by KMD in Korea, we prospectively observed changes in liver function test (LFT) data in outpatients treated with HM prescriptions prescribedby KMD, including some herbs that had been assumed to be causes of DILI when given over a short period for various medical purposes. The primary objective of the present studywas to estimate the proportion of abnormal LFTs, and theincidence of DILI, after short-term HM prescribed by KMDtreatment. This study was designed as a pilot study for a na- tional survey aimed at formulating a set of safety guidelines Materials and Methods
This study was multicenter, prospective, and observa- There were no restrictions on the prescribed HM used. All tional. The purpose was to investigate the HM prescribed by prescriptions were provided to the patients as prepared de- a KMD who graduated from an educational institute, college coctions, and the dosing frequency was 3 times a day. The of Korean medicine and is licensed as a doctor by the Ko- quality of and ingredients in HM are regulated by the Ko- rean government. The study was conducted at seven tradi- rea Food and Drug Administration, with 520 herbs registered tional Korean medical clinics and hospitals between Febru- and hence subject to quality control in 2007.15 ary and July 2006. The research protocol was approved bythe Institutional Review Board at University of Sangji Col- lege of Korean Medicine. The subjects were patients whowere taking HM medication for various purposes and who Initial venous sampling was carried out, and patients com- agreed to participate as outpatients. Patients were excluded pleted questionnaires before HM treatment commenced. Af- if they had malignant disorders, active infectious disease, ter HM treatment was complete, follow-up sampling was previous experience of DILI, or liver injury as shown by the performed and patients completed a second questionnaire.
initial LFT. Critically ill patients requiring hospital admis- Serum measurements of aspartate aminotransferase (AST), sion were also excluded. All 187 patients had taken HM pre- alanine aminotransferase (ALT), total bilirubin (t-Bil), direct scriptions chosen by licensed KMD. Among the 187 patients bilirubin (d-Bil), alkaline phosphatase (ALP), and ␥-glutamyl who signed the informed consent form, gave blood initially, transferase (GGT) were made. Venous samples were and answered the questionnaire, 122 completed follow-up processed in the Green Cross Reference Laboratory (an ISO sampling and a final questionnaire; these 122 patients are an- 9002-Certified Laboratory). “Abnormality of liver tests” and alyzed in this paper. There were 65 cases lost to follow-up “liver injury” were defined as agreed at the International because of refusal to give blood a second time, refusal to an- Consensus Meeting of 1990.16 In detail, “liver injury” was swer the questionnaire (for personal reasons), moving used if there was an increase of over twice the upper limit abroad, moving home, or other miscellaneous reasons of the normal range in ALT or d-Bil, or a combined increase(Fig. 1).
in AST, ALP, and t-Bil, provided that at least one measure- PRESCRIBED HERBAL MEDICINE IN KOREA WITH LFT
ment was over twice the upper limit of the normal range.
for categorical data. A paired t-test was used for the one- “Abnormality of liver tests” was shown when there is an iso- sample (pre-HM and post-HM) comparison, and an inde- lated increase, even over twice the upper limit of the normal pendent two-sample t-test was employed for comparison of range, in AST, ALP, or t-Bil. When an increase in ALT, AST, the characteristics in baseline (normal LFT versus abnormal ALP, or t-Bil was between once and twice the upper limit of LFT). All statistical analyses were performed using SAS soft- the normal range, the term “abnormality of liver tests” was ware, version 9.1 (SAS Institute, Cary, NC). Statistical sig- nificance was defined as p Ͻ 0.05 on the basis of a two-sided The first questionnaire gathered basic data on patients, identified the main reasons why they were taking HM,recorded any liver disease, and collected data on drinking habits, smoking habits, and whether any drug or functional medicinal product had been consumed during the previous90 days. The second questionnaire recorded drugs or func- The characteristics of the 122 patients who completed the tional medicinal products taken with HM, and any adverse study are presented in Table 1. The study group was com- events experienced. The data were collected through struc- posed of 40 men and 82 women, with a mean age of 37.3 Ϯ tured interviews between patients and nurses or KMD who 17.0 years. Histories of liver disease including viral hepati- had completed special education for the study. Prescription tis, alcoholic hepatitis, fatty liver, and nonalcoholic steato- records were also verified and collected.
hepatitis were presented by 4.1% of patients. Half (50.0%)had taken HM or an herbal dietary supplement more thanonce during the previous 90 days. The initial LFT data showed that 101 subjects (82.8%) were normal and 21 (17.2%) Data were summarized using means and standard devia- were abnormal. Body mass index was the only characteris- tions for continuous variables, and frequencies and percent- tic significantly related to abnormal LFT (23.1 Ϯ 3.6 [kg/m2] ages for categorical variables. For comparisons using LFT in the normal group, 24.9 Ϯ 3.3 [kg/m2] in the abnormal levels (normal versus abnormal), p values were calculated by group, p ϭ 0.030), while gender, age, drinking, a history of using either the ␹2 test, Fisher’s exact test, or the McNemar liver disease, prior HM ingestion, use of herbal dietary sup- test, depending on the sample size, or paired sample size, plements, and drug medication during the previous 90 days TABLE 1. PATIENT CHARACTERISTICS (N ϭ 122) Other medication during previous 90 days, no. (%) Reason for KHM medication (multiple), no. (%) Duration of KHM medication (mean Ϯ SD day) BMI, body mass index; SD, standard deviation; LFT, liver function test; KHM, Korean YUN ET AL.
TABLE 2. COMPARISON OF LIVER FUNCTION TEST RESULTS FROM THE FIRST AND SECOND TESTS SD, standard deviation; AST, aspartate aminotransferase; ALT, alanine aminotransferase; ALP, alkaline phosphatase; GGT, ␥-glutamyl transferase; T-Bil, total bilirubin; D-Bil, direct bilirubin.
were not. The indication was musculoskeletal disorders normal to normal, so the HM treatments decreased the fre- (27.1%), followed by obesity (18.0%), improvement of health quency of abnormal LFTs (McNemar test, p ϭ 0.019) (Table (14.8%), and neurologic disorders (10.7%). The mean period 3). Of the 101 subjects showing normal LFTs initially, 97 of HM ingestion was 20.6 Ϯ 8.4 days.
(96.0%) remained within the normal range at the second test,while 4 (4.0%) showed abnormal LFTs, but with minimal in-creases in serum liver components, and were not clinically Korean herbs used in prescriptions significant. Of the 21 subjects showing abnormal LFTs ini-tially, 15 showed normal LFTs later, whereas 5 subjects re- All HM were provided as a decoction in a combined herb mained within the abnormal range and 1 subject (0.82% of prescription and were ingested in the hospital. Several herbs the 122 patients) had raised LFT readings indicating liver in- that have been assumed as causes of DILI were also used, jury. Of the 6 subjects who remained within the abnormal including Scutellariae radix (Scutellaria root),6 Bupleuri radix range or developed liver injury, 4 (66.7%) were concurrently (thorowax root), 6 and Psoralea semen (Psoralea fruit).15 taking other drugs (ibuprofen, gabapentine, dimenhydri-nate, eprosartan mesylate).
LFT changes before and after ingestion The single liver injury case was a 48-year-old man who complained of vertigo, and who took HM (the prescription Between the first and second tests, the mean LFT values did not include any suspected herbs) for 20 days. At the sec- showed decreases in ALT and ALP levels, but increases in ond test, all liver parameters were elevated but without any AST, GGT, t-Bil, and d-Bil concentrations. The only signifi- perceived symptoms. The ALT exceeded twice the normal cant change was in t-Bil (p Ͻ 0.015) (Table 2). The case of upper level (ALT 13 Ǟ 166). Other readings were AST 37 Ǟ maximal change of t-Bil (from 0.4 to 1.4 mg/dL) involved a 270, d-Bil 0.2 Ǟ 0.8, t-Bil 0.4 Ǟ 1.4, ALP 117 Ǟ 211, and GGT patient with liver injury. However, no other patients showed 65 Ǟ 481. This patient was a heavy drinker who had been a t-Bil level above the normal range. In an attempt to iden- diagnosed with alcoholic liver disease 18 months previously, tify herbs increasing the mean of t-Bil, we focused on possi- which explained the abnormal level of GGT shown at the ble hepatotoxic herbs. Patients using one or more herbs from first test. He continued alcohol during the course of HM the group, B. radix, Discoreae rhizoma, S. radix, Ephedrae herba, medication, and also took a drug concurrently. This case was Puerariae radix, P. semen, Coptidis rhizome (coptis root), and diagnosed as liver injury due to ingestion of HM with alco- Pleuropterus tuber (Radix polygoni multiflori) did not show hol and another drug in a patient with underlying liver dis- significantly increased t-Bil levels, compared with those of patients not using these herbs (data not shown).
Seventeen (17) of the 122 patients (13.9%) complained of Twenty-one (21) of the 122 patients showed abnormal adverse drug reactions. There were 3 cases of nausea, 5 of LFTs at the first test and 10 at the second test. Four (4) pa- dyspepsia, 3 of abdominal pain, 3 of skin reaction (itching, tients changed from normal to abnormal, and 15 from ab- hives), and 6 of diarrhea (multiple choice). All these adverse TABLE 3. NUMBER OF CHANGES IN LIVER FUNCTION TEST (LFT) RESULTS BETWEEN THE FIRST AND SECOND TESTS PRESCRIBED HERBAL MEDICINE IN KOREA WITH LFT
events were mild and transient, and were not related to ab- thought to be due to the inhibition of metabolism.16,17 Biliru- normal LFT readings at the second test.
bin metabolism may be changed by some herbs only slightlyand transiently, so that DILI does not result. However, inour study, no herbs suspected to increase t-Bil levels were Discussion
used and no drugs associated with possible hepatotoxicity The primary objective of the present study was to inves- were prescribed. Further studies are needed on the influence tigate the incidence of DILI, and the proportion of patients of other herbs and HM prescriptions on bilirubin metabo- with elevated blood levels of liver enzymes, after HM pre- scribed by KMD in Korea. Though 1 case of mild liver in- Although our patients reflect the general use of HM by jury occurred, it could not be assessed as a genuine DILI-in- KMD in Korea, using such a heterogeneous population of duced HM. We found that HM prescriptions seemed not to mixed ages and health conditions and the small sample size change any of the LFT parameters nor increase the propor- in this study limits the generalizability of our results to other tion of abnormal LFT. To the best of our knowledge, this is populations treated with HM. It is likely that a longer and the first study to evaluate prospectively the incidence of DILI more detailed study would allow us to refine the HM pre- Although HM is now widely used throughout the world, the reported frequencies of herb-related liver injuries differ by Conclusion
country and researcher. In Korea, various types of HM ac-counted for 42% of DILI.14 Many cases of liver injury were in- In summary, the current study showed that ingestion of duced, however, by the self-purchase and administration of a HM prescribed by KMD did not increase the frequency of single herb, without a KMD’s prescription.13 To date, only a abnormal LFTs, at least in the short term. Due to the diffi- few studies have evaluated the frequency of adverse effects as- culty in establishing reliable causal links between prescribed sociated with prescribed HM treatment in Korea, and most herbs, whether singly or in combinations, and adverse ef- were limited to retrospective reviews of inpatient medical fects, real-life epidemiological studies will generate more rel- records. One study on children treated with prescribed HM evant results than laboratory toxicological studies. Further- for 3 months also showed no significant change in LFT read- more, rigorous investigations will be needed to determine ings.11 One study in China reported that 24.2% of DILI cases the incidence of herb-induced liver injuries, to identify high- were related to HM, and the other 75.8% to prescribed mod- risk prescriptions, and to define the interactions of herbs in ern medicine.6 Another study in Singapore found that 55% (17 of 31 cases) of DILI cases were related to HM and adulterantswere found in 9 of 17 cases.7 The problem of adulteration in Authors’ Contributions
herbal products was reported by many other studies and the Young-Ju Yun and Seong-Su Nah conceived the entire urgent need for regulation was already emphasized.15 Studies study, evaluated data, and wrote the paper. Jae-Hyun Park, in Germany, England, and Japan reported a low incidence Seul-Ki Jung, Eun-Kyung Paeck, Jae-Hyung Park, and Eui- (0.24–0.9%) of liver injury in HM-treated patients.8–10 Joon Lee participated in the design of the study, enrolled pa- Of the 122 subjects in the present study, 1 with an abnor- tients, and collected and evaluated herb data. Se-Hyun Kim, mal LFT at the first test had higher LFT readings, indicating Hae-Mo Park, Sun-Dong Lee participated in the design of further liver damage, on the second test. This patient had al- the study, performed statistical analysis, evaluated data and coholic liver disease and had taken HM, a prescribed drug, the manuscript at the point of public healthcare. Ji-Yong Kim, and alcohol concurrently. The study subjects had been asked Seung-Jae Hong, Bong-Keun Choe collected LFT data, ana- to give the second blood samples just after completing HM lyzed questionnaire and gave critical comments on the man- medication, and the mean time between the two tests was uscripts. Joo-Ho Chung conceived of the study, evaluated 20.6 Ϯ 8.4 days. This time is sufficient for detection of ab- data, corrected the manuscript, and managed in this study.
normal changes in LFT results, because 77.6% of HM-related All authors read and approved the final manuscript.
ALT elevation was found by 3 weeks after HM treatmentcommenced.10 It is a limitation that the interval between testswas not sufficiently long to allow us to measure delayed tox- Acknowledgments
icity. In 6 patients who showed abnormal LFT data at both This study was partially supported by a grant from the the first and second tests, 4 were concurrently taking another Association of Korean Oriental Medicine (Grant No: KOM drug. Our results are similar to those of a previous study 2006-01). The authors thank Chi-Hong Kim, Bo-Hyun Yun, showing that most patients with LFT readings exceeding Jun-Tae Kim, Ji-Hoon Kim, Suk-Eui Hong, Gyung-Suk Yun, three times the normal upper levels had recently consumed Tae-Hoon Ku, Yeong-Gun Kang, Do-Gyun Nah, Eun-Seu nonsteroidal anti-inflammatory drugs.9 We suggest that pa- Park, Eun-Hee Shon, Man-Ki Hwang, and Yeong-Jin Hur (all tients taking HM and modern pharmaceutical drugs con- KMDs) for participating in this study and providing data.
currently may be at risk because of drug hepatotoxicity orthe possibility of drug interactions. A mild increase in t-Bil Disclosure Statement
was observed in the present study, but only 1patient had at-Bil level above the normal range. Some studies have re- No competing financial interests exist.
ported that berberine, the major ingredient of C. rhizoma, maydecrease the mean serum protein binding of bilirubin be- References
cause of in vivo displacement. Also, a persistent elevation in 1. Schiano TD. Hepatotoxicity and complementary and alter- the steady-state serum concentrations of d-Bil and t-Bil was native medicines. Clin Liver Dis 2003;7:453–473.
YUN ET AL.
2. Farah MH, Edwards R, Lindquist M, et al. International provement of the developmental disordered children. Ko- monitoring of adverse health effects associated with herbal rean J Oriental Physiol Pathol 2006;20:494–497.
medicines. Pharmacoepidemiol Drug Saf 2000;9:105–112.
12. Park KJ, Lee MJ, Kang H, et al. Saeng-Maek-San, a medicinal 3. MacPherson H, Liu B. The safety of Chinese herbal medi- herb complex, protects liver cell damage induced by alco- cine: A pilot study for a national survey. J Altern Comple- hol. Biol Pharm Bull 2002;25:1451–1455.
13. Nam SW, Baek JT, Lee DS, et al. A case of acute cholestatic 4. Hsieh SC, Lai JN, Chen PC, et al. Development of active hepatitis associated with the seeds of Psoralea corylifolia (Boh- safety surveillance system for traditional Chinese medicine: Gol-Zhee). Clin Toxicol 2005;43:589–591.
An empirical study in treating climacteric women. Pharma- 14. Kim JB, Sohn JH, Lee HL, et al. Clinical characteristics of coepidemiol Drug Saf 2006;15:889–899.
acute toxic liver injury. Korean J Hepatol 2004;10:125–134.
5. Itoh S, Marutani K, Nishijima T, et al. Liver injuries induced 15. Pittler MH, Ernst E. Systematic review: Hepatotoxic events by herbal medicine, syo-saiko-to(xiao-chai-hu-tang). Dig Dis associated with herbal medicinal products. Aliment Phar- 6. Li B, Wang Z, Fang JJ, et al. Evaluation of prognostic mark- 16. Yeung CY, Lee FT, Wong HN. Effect of a popular Chinese ers in severe drug-induced liver disease. World J Gastroen- herb on a neonatal bilirubin protein binding. Biol Neonate 7. Wai CT. Tan BH, Chan CL, et al. Drug-induced liver injury 17. Chan E. Displacement of bilirubin from albumin by berber- at an Asian center: A prospective study. Liver Int ine. Biol Neonate 1993;63:201–208.
8. Melchart D, Linde K, Hager S, et al. Monitoring of liver en- zymes in patients treated with traditional Chinese drugs.
Complement Ther Med 1999;7:208–216.
9. Al-Khafaji M. Monitoring of liver enzymes in patients on Chinese medicine. J Chinese Med 2000;62:6–10.
Kohwang Medical Research Institute 10. Mantani N, Kogure T, Sakai S, et al. Incidence and clinical features of liver injury related to Kampo (Japanese herbal) medicine in 2,496 cases between 1979 and 1999: Problems of Dongdaemun-Gu, Hoigi-Dong, 130-701, Seoul the lymphocyte transformation test as a diagnostic method.
11. Paeck EK, Park JH, Park JH, et al. Influence of taking herbal medicine for a period of time on the liver function and im-

Source: http://poori.co.kr/upload/notice/LFT%EC%97%B0%EA%B5%AC_%EB%85%BC%EB%AC%B8_acm.20081.pdf