Journal of Ethnopharmacology 169 (2015) 407–412
Contents lists available at ScienceDirect
Journal of Ethnopharmacology journal homepage: www.elsevier.com/locate/jep
Liver enzyme abnormalities in taking traditional herbal medicine in Korea: A retrospective large sample cohort study of musculoskeletal disorder patients Jinho Lee a, Joon-Shik Shin a, Me-riong Kim a, Jang-Hoon Byun a, Seung-Yeol Lee a, Ye-sle Shin a, Hyejin Kim a, Ki Byung Park a, Byung-Cheul Shin b, Myeong Soo Lee c, In-Hyuk Ha a,n a b c
Jaseng Spine and Joint Research Institute, Jaseng Medical Foundation, 858 Eonju-ro, Gangnam-gu, Seoul, Republic of Korea Busan National University, Yangsan Campus, 49 Busandaehak-ro, Mulgeum-eup, Yangsan-si, Gyeongsangnam-do, Republic of Korea Division of Medical Research, Korea Institute of Oriental Medicine, 1672 Yuseongdae-ro, Yuseong-gu, Daejeon, Republic of Korea
art ic l e i nf o
a b s t r a c t
Article history: Received 10 January 2015 Received in revised form 4 March 2015 Accepted 26 April 2015 Available online 6 May 2015
Ethnopharmacological relevance: The objective of this study is to report the incidence of liver injury from herbal medicine in musculoskeletal disease patients as large-scale studies are scarce. Considering that herbal medicine is frequently used in patients irrespective of liver function in Korea, we investigated the prevalence of liver injury by liver function test results in musculoskeletal disease patients. Materials and methods: Of 32675 inpatients taking herbal medicine at 7 locations of a Korean medicine hospital between 2005 and 2013, we screened for liver injury in 6894 patients with liver function tests (LFTs) at admission and discharge. LFTs included t-bilirubin, AST, ALT, and ALP. Liver injury at discharge was assessed by LFT result classifications at admission (liver injury, liver function abnormality, and normal liver function). In analyses for risk factors of liver injury at discharge, we adjusted for age, sex, length of stay, conventional medicine intake, HBs antigen/antibody, and liver function at admission. Results: A total 354 patients (prevalence 5.1%) had liver injury at admission, and 217 (3.1%) at discharge. Of the 354 patients with liver injury at admission, only 9 showed a clinically significant increase after herbal medicine intake, and 225 returned to within normal range or showed significant liver function recovery. Out of 4769 patients with normal liver function at admission, 27 (0.6%) had liver injury at discharge. In multivariate analyses for risk factors, younger age, liver function abnormality at admission, and HBs antigen positive were associated with injury at discharge. Conclusions: The prevalence of liver injury in patients with normal liver function taking herbal medicine for musculoskeletal disease was low, and herbal medicine did not exacerbate liver injury in most patients with injury prior to intake. & 2015 Elsevier Ireland Ltd. All rights reserved.
Keywords: Herbal medicine Complementary therapies Liver function tests Drug-induced liver injury Risk factors
1. Introduction Complementary and Alternative Medicine (CAM) treatments are most frequently used for musculoskeletal diseases in Korea. The 1st–5th most frequent diseases in inpatient care in National
Abbreviations: CAM, Complementary and Alternative Medicine; DILI, druginduced liver injury; LFT, liver function test; ALT, alanine aminotransferase; AST, aspartate aminotransferase; ALP, alkaline phosphatase; TB, total bilirubin; γ-GTP, γ-glutamyl transpeptidase; HBs, Hepatitis B surface; CIOMS, Council for International Organization of Medical Sciences; Ag, antigen; Ab, antibody; KCD, Korean standard classification of diseases; ICD, International Statistical Classification of Diseases and Related Health Problems n Corresponding author. Tel.: þ 82 2 3218 2188; fax: þ 82 2 3218 2244. E-mail address: [email protected] (I.-H. Ha). http://dx.doi.org/10.1016/j.jep.2015.04.048 0378-8741/& 2015 Elsevier Ireland Ltd. All rights reserved.
Health Insurance reimbursements for Korean medicine in 2013 were all musculoskeletal related (1st–5th most frequent Korean standard classification of diseases: M54 Dorsalgia, M17 Gonarthrosis [arthrosis of knee], S33 Dislocation, sprain and strain of joints and ligaments of lumbar spine and pelvis, M51 Other intervertebral disc disorders, M79 Other and unspecified soft tissue disorders, not elsewhere classified) (Kim and Son, 2014). As CAM use increases worldwide, interest in liver toxicity from herbal medicine is also rising (Wolsko et al., 2003). However, herb-induced liver injury is usually more difficult to detect than drug-induced liver injury (DILI) and incidence of cases is highly variable due to various reasons, such as the fact that many users of herbal products do not discuss them with their physician (Eisenberg et al., 1993, 1998), many herbal preparations are regulated by less stringent standards than conventional drugs
408
J. Lee et al. / Journal of Ethnopharmacology 169 (2015) 407–412
(Licata et al., 2013), many herbal products are available as “overthe counter” drugs or food supplements (Singh et al., 2012), and medical use can be both regular or episodic (Tovar and Petzel, 2009). Chalasani et al. (2008) reported in a retrospective U.S. study that herbal and dietary supplements were implicated in 10% of DILI cases , a prospective study on 22 participants conducted in Singapore reported that 68% were related to DILI (Wai et al., 2007), and data from a Chinese medical center on 30 DILI patients indicates that 40% were herb related causes (Wang et al., 2009). On the other hand, many studies report the incidence rate of herbinduced liver injury at less than 1% and clinical symptoms to be mild (Al-Khafaji, 2000). The discrepancy in previous evidence may be partly attributable to cultural differences regarding herbal products, indiscrimination between prescription herbal medicine, folk medicine, and dietary supplements of uncertain composition and consistency, disparity in research methodology and environment, and small sample sizes. Therefore, to assess the prevalence of liver injury due to herbal medicine, a more specific, standard definition of herbal medicine is called for, and due to relatively low prevalence rates, more large-scale studies are needed. As many herbal medicine practitioners construe herbal medicine to be natural and therefore “safe” (Ye and He, 2010), herbal medicine is frequently used in patients with abnormal liver function test (LFT) results. In order to investigate changes in liver enzymes in inpatients taking herbal medicine prescribed by trained medical professionals for musculoskeletal disease, we retrospectively assessed liver injury at discharge by LFT results at admission and its risk factors.
2.3. Statistical analysis Patients were categorized into 3 groups – “liver injury”, “liver function abnormality”, and “normal liver function” – by LFT results at admission and discharge. Age and γ-GTP results were continuous variables, and sex (male/female), HBs antigen (positive/ negative), HBs antibody (positive/negative), and use of conventional medicine (yes/no) were categorical variables. HBs antigen and antibody results were classified into 4 groups; Ag( þ)/Ab( þ), Ag(þ)/Ab( ), Ag( )/Ab(þ ), and Ag( )/Ab( ). Length of stay was assessed as both continuous and categorical variables (r 20 days, 20–30 days, and 430 days). Simple logistic regression modeling was used to assess influence of age, sex, length of hospital stay, conventional medicine intake, liver function classification at admission, HBs antigen/antibody, and γ-GTP on liver function classification at discharge. Multiple logistic regression modeling was used for factors with a p-value of r0.2. Factors included in the final model were determined in a stepwise manner (included if p-valuer0.05, excluded if p-valueZ0.10). We used stratification of patient characteristics to further assess the association between age and liver function. The statistical package for Social Science for WindowsTM, version 11.0 (SPSS Inc., Chicago, IL, U.S.A.) was used. 2.4. Ethics statement Ethical approval was obtained from the institutional review board of Jaseng Hospital of Korean medicine in Seoul, Korea (SIRB2013-37), and all patients provided informed written consent to use of data for academic means prior to hospitalization. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki and Tokyo.
2. Material and methods 2.1. Subjects We retrospectively reviewed the electronic medical records of 32675 patients who stayed at 7 locations of a musculoskeletal disease-specialty Korean medicine hospital in Korea from December 2005 to December 2013. The total number of patients with blood test results during admission was 29229 and 24% (n ¼7003) received multiple tests at admission and before discharge. Of these patients, we analyzed the results of 6894 patients for alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), and total bilirubin (TB). Inclusion criteria – patients admitted for at least one day who took herbal medicine and at least 2 LFTs during admittance. There were no exclusion criteria. 2.2. Classification of liver enzyme abnormalities LFTs were measured after a 6 h overnight fast. All inpatients received blood tests at admission except those who refused, and follow-up tests were conducted as deemed necessary. The LFTs included in this study were ALT, AST, ALP, TB, γ-glutamyl transpeptidase (γ-GTP), Hepatitis B surface (HBs) antigen, and HBs antibody. The normal range for LFTs was ALTr40 U/L; ASTr40 U/L; ALPr338 U/L; and TBr1.2 mg/dL. According to the modified Council for International Organization of Medical Sciences (CIOMS) criteria, the term “liver injury” was used to describe an increase of over 2 N (upper limit of the normal range) in ALT, or a combined increase in TB, AST, and ALP, provided one of them was above 2 N. Patients with all 4 enzymes within normal range were defined as having “normal liver function”, and those neither in the “normal” nor “liver injury” group as “liver function abnormality”.
3. Results The average length of hospital stay was 26.17 712.31 days, 45% (n ¼3111) were male, and average age was 44.31 714.49 years. We classified the patients into 3 groups (liver injury, liver function abnormality, and normal liver function) according to liver function test results at admission, and subcategorized the groups by liver function test results at discharge to better illustrate the change in distribution of liver function state (total 9 groups). Of 4769 patients with normal liver function at admission, 27 (0.6%) had liver injury at discharge, and out of 354 patients with liver injury at admission, 225 (64%) no longer had liver injury at Table 1 Classification of patients by liver function test results at admission and dischargea. Liver function state at admission
N
Liver injury
354
Liver function abnormality
Normal liver function
Liver function state at discharge
Liver injury Liver function abnormality Normal liver function 1771 Liver injury Liver function abnormality Normal liver function 4769 Liver injury Liver function abnormality Normal liver function
N
129 143 82 61 763 947 27 284 4458
(9) (0) (1) (23) (28) (23) (27) (117) (259)
( ): No. of patients with a twofold increase in at least one blood index of ATP, AST, ALT, or total bilirubin at discharge compared to admission. a The normal range for liver function tests was alanine aminotransferase (ALT) r40 U/L; aspartate aminotransferase (AST)r40 U/L; alkaline phosphatase (ALP) r338 U/L; and total bilirubinr1.2 mg/dL, and the term “liver injury” was used to describe an increase of over 2 N (upper limit of the normal range) in ALT, or a combined increase in total bilirubin, AST, and ALP, provided one of them was above 2 N.
J. Lee et al. / Journal of Ethnopharmacology 169 (2015) 407–412
discharge. The number of patients with liver injury was 354 (total prevalence 5.1%) at admission, and 217 (3.1%) at discharge, indicating a decrease to 61%. Fifty-nine patients, which accounts for 27% of the patients with liver injury at discharge, had liver injury with a twofold increase in ALT, AST, ALP, or TB at discharge compared to admission (Table 1). We compared characteristics and use of conventional and herbal medicine in the 4458 patients who maintained normal liver function at discharge with the 27 patients who had liver injury at discharge out of the total 4769 patients with normal liver function at admission. Age, sex, length of hospital stay, conventional medicine intake, HBs antigen/antibody, γ-GTP, and frequency and patterns of conventional and herbal medicine use were similar in both groups (Table 2). The 27 patients with liver injury at discharge did not display any liver-related clinical symptoms and related data will be reported in a future case series. Chungpa-jun (including variations) was the most frequently prescribed herbal medicine during hospital stay, and the basic composition consists of Ostericum koreanum, Eucommia ulmoides (Hardy rubber tree), Acanthopanax sessiliflorus, Achyranthes bidentata (Ox knee), Psoralea corylifolia (Babchi), Peucedanum japonicum, Cibotium barometz (Golden chicken fern, Woolly fern, Scythian lamb), Lycium chinense (Chinese boxthorn, Chinese wolfberry, Chinese matrimony vine), Boschniakia rossica (Northern groundcone), and Cuscuta chinensis (Chinese dodder seed) plus or minus 2–3 ingre-
409
dients. The formulas of the other herbal prescriptions prescribed in the 27 patients with liver injury at discharge of patients with normal liver function at admission are included in the Appendix (Table A1). Conventional medicine use was mainly comprised of customary musculoskeletal drugs such as anti-inflammatory agents and analgesics, and concomitant drugs such as gastrointestinal system drugs and antidepressants, and is also presented in the Appendix (Table A2). In analyses for risk factors of liver injury prevalence at discharge, age, sex, length of stay, liver function at admission, HBs antigen/ antibody, and γ-GTP showed significant associations. Age, HBs antigen/antibody and liver function at admission were shown to be significantly associated in the multivariate analysis. Liver injury at discharge was more prevalent in younger age, HBs antigen positive, and abnormal liver conditions at admission (Table 3). When stratified to assess the effect of age on liver injury, we found decreasing prevalence of liver injury with increasing age in most factors (Table 4). Diagnostic data was coded both in the Korean standard classification of diseases (KCD), which is the Korean version of the International Statistical Classification of Diseases and Related Health Problems (ICD), and Korean medicine diagnosis and pattern differentiation, which was generated specifically for standard classification of traditional Korean medicine, of which the 3rd revision was incorporated into KCD-6 in 2010. The majority of
Table 2 Comparison of characteristics and conventional and herbal medicine use in patients with normal liver function at admission classified by liver function at discharge (liver injury at discharge vs. normal liver function at discharge).
Age Sex Female Male HBs(Ag)/(Ab)b Ag ( þ )/Ab ( þ) Ag ( þ )/Ab ( ) Ag ( )/Ab (þ ) Ag ( )/Ab ( ) Use of conventional medicine Yes No γ-GTPc Length of hospital stay Classification of conventional medicine by main ingredientd Almagate Eperisone hydrochloride Aceclofenac Naproxen Amitriptyline hydrochloride
Liver injury at discharge N ¼ 27 N(%)/mean 7 SD
Normal liver function at discharge N ¼ 4458 N(%)/mean 7 SD
39.2 714.3
45.4 7 14.7
13 (48.1) 14 (51.9)
2971 (66.6) 1487 (33.4)
0 (0.0) 1 (4.0) 18 (72.0) 6 (24.0)
4 (0.1) 163 (3.8) 2761 (64.5) 1350 (31.6)
10 (37.0) 17 (63.0) 32.8 725.8
1573 (35.3) 2885 (64.7) 27.1 725.8
23.4 78.8
27.2 7 12.5
5 5 3 2 2
562 (14.6) 411 (10.7) 137 (4.7) 347 (9.0) 309 (8.0)
(19.2) (19.2) (11.5) (7.7) (7.7)
Acetaminophen 2 (7.7) Artemisia asiatica 95% ethanol ext. 1 (3.8) Tramadol hydrochloride 1 (3.8)
255 (6.6) 85 (2.2) 80 (2.1)
Limaprost alpha-cyclodextrin clathrate Levosulpiride Glimepiride Simvastatin Clobetasol-17-propionate
1 (3.8)
55 (1.4)
1 1 1 1
50 (1.3) 11 (0.3) 1 (0.0) 1 (0.0)
a
(3.8) (3.8) (3.8) (3.8)
Liver injury at discharge N ¼ 27 N(%) Herbal medicine typea Chung-pa-jun 20 (40.8) Ssang-pae-tang 3 (6.1) Hwal-hyul-ji-tong-tang 2 (4.1) Hwal-lak-tang 2 (4.1) Saeng-gang-gun-bi-tang 2 (4.1) Ssang-hwa-sam-so-eum 1 (2.0) Gwak-hyang-jung-gi-san 1 (2.0) Sook-ji-yang-keun-tang 1 (2.0) Gae-gyung-seo-kyung-tang 1 (2.0) In-sam-yang-wi-tang 1 (2.0) Ma-bal-gwan-jul-tang 1 (2.0) Sam-ryeong-baek-chul1 (2.0) tang Ga-gam-gui-bi-tang 1 (2.0) Pyung-wi-san 1 (2.0) Hyeong-bang-tang Hoe-soo-san Hyang-sa-yuk-gun-ja-tang Yang-hyeol-geo-pung-tang Ga-mi-Yuk-mi-ji-whangtang Shin-tong-chu-uh-tang Hyeong-bang-sa-baek-san Ban-ha-baek-chul-cheonma-tang Man-geum-tang Ga-mi-Yuk-mi Li-gi-geo-poong-san
The pharmacological classification of each conventional medicine is presented in Table A2 in the Appendix. Hepatitis B surface (HBs) antigen (Ag)/HBs antibody (Ab). c γ-glutamyltranspeptidase. d The main ingredients of each herbal medicine are presented as Table A1 in the Appendix. b
1 1 1 1 1
(2.0) (2.0) (2.0) (2.0) (2.0)
Normal liver function at discharge N ¼4458 N(%)
3432 612 122 81 10 563 441 189 122 48 37 28
(42.7) (7.6) (1.5) (1.0) (0.1) (7.0) (5.5) (2.4) (1.5) (0.6) (0.5) (0.3)
22 (0.3) 16 (0.2) 15 13 12 8 5
(0.2) (0.2) (0.1) (0.1) (0.1)
1 (2.0) 1 (2.0) 1 (2.0)
4 (0.0) 3 (0.0) 2 (0.0)
1 (2.0)
2 (0.0)
1 (2.0) 1 (2.0)
1 (0.0)
410
J. Lee et al. / Journal of Ethnopharmacology 169 (2015) 407–412
Table 3 Univariate and multivariate analyses of factors by liver injury at discharge. Univariate analysisa
Age Sex Female Male 0.5 γ-GTPc HBs(Ag)/(Ab)d Ag ( þ)/Ab ( þ ) Ag ( þ)/Ab (-) Ag ( )/Ab ( þ ) Ag ( )/Ab ( ) Use of conventional medicine Yes No Length of hospital stay Liver function test results at admission Liver injury Liver function abnormality Normal liver function
Multivariable analysisb
Total (liver injury)
OR (95% CI)
p-value
OR (95% CI)
p-value
6894 (217)
0.96 (0.95–0.97)
o 0.001 o 0.001
0.97 (0.95–0.98)
o 0.001
3783 (61) 3111 (156)
0.31 (0.23–0.42) (Reference)
4403 (217)
1.01 (1.01–1.01)
8 288 4088 2156
(0) (22) (111) (65)
– 2.66 (1.61–4.39) 0.9 (0.66–1.23) (Reference)
2475 (68) 4419 (149) 6894 (217)
0.81 (0.61–1.08) (Reference) 0.96 (0.95–0.98)
354 (129) 1771 (61) 4769 (27)
100.69 (65.12–155.70) 6.27 (3.97–9.89) (Reference)
o 0.001 o 0.001
o 0.001 – 3.54 (1.62–7.73) 1.05 (0.65–1.70) (Reference)
0.155
o 0.001 o 0.001
0.98 (0.96–1.00)
0.021 o 0.001
73.37 (42.11–127.86) 3.74 (2.03–6.89) (Reference)
a Univariate analysis was conducted on age, γ-GTP, and length of hospital stay as continuous variables, and other factors (sex, HBs antigen/HBs antibody, concurrent conventional medication intake, and liver function test results at admission) as categorical variables. b Factors whose p-value was less than 0.2 in the univariate analysis were selected for multivariable analysis (All 7 factors were selected). Factors included in the final model were determined in a stepwise manner (included if p-value r 0.05, excluded if p-value Z 0.10). c γ-glutamyl transpeptidase. d Hepatitis B surface (HBs) antigen (Ag)/HBs antibody (Ab).
inpatients were musculoskeletal pain patients with lumbar or cervical disc herniation or joint related complaints (details provided in Table A3 in the Appendix).
Table 4 Prevalence ratios with 95% confidence intervals for liver injury at discharge by age. Total (liver injury)
4. Discussion In our study, patients taking herbal medicine for musculoskeletal disease treatment had a lower prevalence of liver injury after intake. A similar tendency was shown in the small number of patients with normal liver function at admission and liver injury at discharge (n ¼27; 0.6%). Though the liver injury subgroup at discharge was too small for statistical analysis of difference in the subgroups at discharge out of the normal liver function group at admission, a comparison of baseline characteristics and intake of herbal medicine and conventional medicine displayed no considerable difference. Analysis of total admitted patients showed that liver dysfunction at admission was associated with a higher likelihood of liver injury at discharge, and younger age was a risk factor for liver injury. The 7 locations of a Korean medicine hospital which was the study setting mainly treat and specialize in musculoskeletal complaints, and was the first Korean medicine hospital to be endorsed by the Korean ministry of health and welfare as a spinal disease specialty hospital. This hospital adopts a conservative integrative treatment model using conventional diagnostic tools (X-rays, CTs, MRIs, blood tests) and traditional Korean medicine (acupuncture, herbal medicine, manual manipulation) as the main treatment intervention with conventional medicine (drugs, injections) utilized as necessary as ancillary treatment. The vast majority of participants suffered from low back pain or neck pain due to disc herniation or joint dysfunction from related diseases, and none were classified as suffering from rheumatologic diagnoses more often seen in younger or female patients, or such pathologies as lupus or myositis which may be related to AST and ALT elevation. The main objectives of herbal medicine use did not differ significantly from that of conventional drugs with Chungpa-jun
Liver function test results at admission Liver injury 113 (65) Liver function abnormality 891 (24) Normal liver function 2984 (19) HBs(Ag)/(Ab)b Ag( þ )/Ab( ) 151 (15) Ag( )/Ab(þ ) 2592 (59) Ag( )/Ab( ) 1240 (34) Length of hospital stay 30o 1219 (22) 20–30 1330 (27) r 20 1439 (59) Conventional medication intake Yes 1356 (37) No 2632 (71) Sex Male 1802 (78) Female 2186 (30) γ-GTPc 73o 465 (45) r 73 3523 (63) a b c
OR (by age)a
95% CI
0.95 0.96 0.98
0.92–0.99 0.93–1.00 0.95–1.02
0.94 0.96 0.97
0.86–1.03 0.93–0.99 0.94–1.00
0.94 0.94 0.98
0.90–0.98 0.90–0.98 0.96–1.01
0.95 0.97
0.92–0.99 0.95–1.00
0.97 0.95
0.95–1.00 0.92–0.98
0.93 0.98
0.90–0.97 0.95–1.00
Age was assessed as a continuous variable. Hepatitis B surface (HBs) antigen (Ag)/HBs antibody (Ab). γ-glutamyl transpeptidase.
used predominantly for spine or joint disorders, Ssangpae-tang, Ssanghwa-samso-eum, and Dokgam-tang prescribed for common cold or flu symptoms and Gwakhyang-junggi-san for gastrointestinal disorders. Other than these, many drugs were regular medications taken for comorbidities (hypertension, diabetes). The composition and dosage of individual herbal medicines are complex and difficult to gauge as they may comprise either a single ingredient or a compound mixture of different herbal ingredients, and though medications can be in liquid, powder, tablet or other forms, we listed only those in liquid form as this was the main
J. Lee et al. / Journal of Ethnopharmacology 169 (2015) 407–412
type of herbal medicine, usually taken for extended periods of time. We did not determine the cumulative exposure dose, instead dividing the prescribed medication by intention of usage for classification purposes. (i.e. Multiple prescriptions of a certain herbal medicine taken by one patient was tallied as 1 prescription of that medicine.) The biggest strength of our study is that it is one of the few large-scale studies to examine use of herbal medicine in musculoskeletal disease patients, who are frequent consumers of herbal medicine. Also, most studies studying potential liver toxicity of herbal medicine only include patients with normal liver function, and assess prevalence after excluding various causes of liver injury (e.g. hepatitis B virus, conventional medicine use, alcohol consumption). However, herbal medicine is frequently used in patients with abnormal LFT results or those with potential causes of liver injury in absence of significant liver injury symptoms as it is considered to be “safe”. Our study assessed the prevalence of liver injury in musculoskeletal patients before and after being prescribed with herbal medicine in inpatient care and found that prevalence of liver injury was lower at discharge. This is one of the few studies to show that herbal medicine does not have a negative impact on liver injury in general musculoskeletal patients (without selective inclusion of only normal liver function patients and with non-exclusion of risk factors for liver injury) in treatment for musculoskeletal disorders. The largest drawback of our study is probably the retrospective study design. We could not draw any conclusions regarding the causality relationship between herbal and conventional medicine use and liver injury. Another limitation is that although direct bilirubin, which is used to assess liver injury in CIOMS criteria, was measured in some inpatients, we excluded the results due to the small sample size, and other indexes that may affect liver function (prothrombin time, anti-HAV IgM, anti-HBc IgM, HCV RNA, anti-HSV, IgM, and anti-EBV IgM) were not measured. Also, as only 24% of total inpatients were followed-up with additional tests, the results of our study may not be widely applicable. However, as the need for follow-up tests was decided by Korean medicine doctors, most patients with no additional tests were those at low risk of liver dysfunction. Therefore it is highly probable that the overall prevalence would be lower. Also when taking into consideration severity of symptoms requiring admission and subsequent distress levels, frequent herbal medicine intake (liquid and/or pill form taken 3 times a day), concurrent use of conventional medicine in 2475 (36%) patients, number of HBV carriers, and other possible liver-related comorbidities, the liver injury prevalence is not high. Twenty-seven patients (0.6%) with normal liver function at admission had liver injury at discharge, which is comparable to other sizable retrospective inpatient studies. Melchart et al., 1999 reported that of 1450 patients treated with traditional Chinese drugs, 14 patients (0.97%) could be considered DILIs showing a twofold increase in ALT , while a study in Japan reported 15 (0.6%) out of 2496 patients (Sheehan et al., 1992), while in a study on 3102 patients, 22 DILI patients were identified; caused by conventional medicine in 15 patients (0.6%) and herbal medicine in 7 (0.23%) (Mitsuma, 2002). A recent study in Korea reported a prevalence of 5 DILI patients (0.56%) of 892. In analyses for risk factors of liver injury at discharge, liver function at admission was most influential. Also, the likelihood of liver injury was higher in younger age. These results differ from previous studies reporting a positive association between age and DILI (Abboud and Kaplowitz, 2007). However, while age is a risk factor for DILI, susceptible age groups differ according to specific medications. For example, older age is a risk factor for DILI related to isoniazid, whereas younger age is a risk factor for valproate and
411
aspirin (Reyes syndrome). Although the exact mechanism of DILI is still largely unidentified, it seems to involve 2 pathways – direct liver toxicity and adverse immune reactions (Holt and Ju, 2006). DILIs from herbal medicines are generally idiosyncratic drug reactions, strongly suggesting a mainly immune response (Ju et al., 2012). Immune responses against foreign antigens and malignant cells are lower in the elderly, and most immune cells involved in innate immunity show decreased function with aging (Tajiri and Shimizu, 2013). This suggests that herbal DILI prevalence may be higher in younger groups as younger patients have stronger immune responses. Another explanation could be that as most symptoms due to herbal medicines are mild and use of herbal medicine increases with age in Korea, the inverse relationship between age and liver injury is due to immune tolerance through repeated sub-clinical immune reactions. There are also consistent reports on how cholestatic liver enzyme patterns are more likely to be seen in liver injury in older patients (Lucena et al., 2009), and it is notable that liver injury related to herbal medicine is generally of hepatocellular type (Jeong et al., 2012).
5. Conclusions In conclusion, prevalence of liver injury in patients taking herbal medicine for treatment of musculoskeletal disease was low, and clinical symptoms were sub-clinical or mild. Liver function at admission was the most prominent risk factor for liver injury, and there was a tendency of increasing prevalence of liver injury with younger age.
Acknowledgments None.
Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.jep.2015.04.048. References Abboud, G., Kaplowitz, N., 2007. Drug-induced liver injury. Drug Saf.: Int. J. Med. Toxicol. Drug Exp. 30, 277–294. Al-Khafaji, M., 2000. Monitoring of liver enzymes in patients on Chinese medicine. J. Chin. Med. 62, 6–10. Chalasani, N., Fontana, R.J., Bonkovsky, H.L., Watkins, P.B., Davern, T., Serrano, J., Yang, H., Rochon, J., Drug Induced Liver Injury Network (DILIN), 2008. Causes, clinical features, and outcomes from a prospective study of drug-induced liver injury in the United States. Gastroenterology 135, 1924–1934 1934.e1-4. Eisenberg, D.M., Davis, R.B., Ettner, S.L., Appel, S., Wilkey, S., Van Rompay, M., Kessler, R.C., 1998. Trends in alternative medicine use in the United States, 1990–1997: results of a follow-up national survey. J. Am. Med. Assoc. 280, 1569–1575. Eisenberg, D.M., Kessler, R.C., Foster, C., Norlock, F.E., Calkins, D.R., Delbanco, T.L., 1993. Unconventional medicine in the United States. Prevalence, costs, and patterns of use. N. Engl. J. Med. 328, 246–252. Holt, M.P., Ju, C., 2006. Mechanisms of drug-induced liver injury. AAPS J. 8, E48–E54. Jeong, T.Y., Park, B.K., Cho, J.H., Kim, Y.I., Ahn, Y.C., Son, C.G., 2012. A prospective study on the safety of herbal medicines, used alone or with conventional medicines. J. Ethnopharmacol. 143, 884–888. Ju, H.Y., Jang, J.Y., Jeong, S.W., Woo, S.A., Kong, M.G., Jang, H.Y., Lee, S.H., Kim, S.G., Cha, S.W., Kim, Y.S., Cho, Y.D., Jin, S.Y., Kim, H.S., Kim, B.S., 2012. The clinical features of drug-induced liver injury observed through liver biopsy: focus on relevancy to autoimmune hepatitis. Clin. Mol. Hepatol. 18, 213–218. Kim, J., Son, MS, 2014. 2013 National Health Insurance Statistical Yearbook. Health Insurance Review and Assessment Service and National Health Insurance Service, Seoul. Licata, A., Macaluso, F.S., Craxi, A., 2013. Herbal hepatotoxicity: a hidden epidemic. Intern. Emerg. Med. 8, 13–22. Lucena, M.I., Andrade, R.J., Kaplowitz, N., Garcia-Cortes, M., Fernandez, M.C., Romero-Gomez, M., Bruguera, M., Hallal, H., Robles-Diaz, M., Rodriguez-
412
J. Lee et al. / Journal of Ethnopharmacology 169 (2015) 407–412
Gonzalez, R.S., Navarro, J.M., Salmeron, J., Martinez-Odriozola, P., Perez-Alvarez, R., Borraz, Y., Hidalgo, R., Spanish Group for the Study of Drug-Induced Liver Disease, 2009. Phenotypic characterization of idiosyncratic drug-induced liver injury: the influence of age and sex. Hepatology 49, 2001–2009. Melchart, D., Linde, K., Hager, S., Kaesmayr, J., Shaw, D., Bauer, R., Weidenhammer, W., 1999. Monitoring of liver enzymes in patients treated with traditional Chinese drugs. Complement. Ther. Med. 7, 208–216. Mitsuma, T., 2002. Adverse reactions encountered during treatment with Japanese oriental (Kampo) medicines. Korean J. Orient. Intern. Med. 2, 9–14. Sheehan, M.P., Rustin, M.H., Atherton, D.J., Buckley, C., Harris, D.W., Brostoff, J., Ostlere, L., Dawson, A., 1992. Efficacy of traditional Chinese herbal therapy in adult atopic dermatitis. Lancet 340, 13–17. Singh, D., Gupta, R., Saraf, S.A., 2012. Herbs-are they safe enough? an overview. Crit. Rev. Food Sci. Nutr. 52, 876–898.
Tajiri, K., Shimizu, Y., 2013. Liver physiology and liver diseases in the elderly. World J. Gastroenterol. 19, 8459–8467. Tovar, R.T., Petzel, R.M., 2009. Herbal toxicity. Disease-a-Month 55, 592–641. Wai, C.T., Tan, B.H., Chan, C.L., Sutedja, D.S., Lee, Y.M., Khor, C., Lim, S.G., 2007. Druginduced liver injury at an Asian center: a prospective study. Liver Int.: Off. J. Int. Assoc. Study Liver 27, 465–474. Wang, Y.P., Shi, B., Chen, Y.X., Xu, J., Jiang, C.F., Xie, W.F., 2009. Drug-induced liver disease: an 8-year study of patients from one gastroenterological department. J. Dig. Dis. 10, 195–200. Wolsko, P.M., Eisenberg, D.M., Davis, R.B., Kessler, R., Phillips, R.S., 2003. Patterns and perceptions of care for treatment of back and neck pain: results of a national survey. Spine 28, 292–297, Discussion 298. Ye, X.F., He, J., 2010. The bright future of Chinese herbal medicine: only after a twisty road. Contemp. Clin. Trials 31, 508–509.
Contents lists available at ScienceDirect
Journal of Ethnopharmacology journal homepage: www.elsevier.com/locate/jep
Liver enzyme abnormalities in taking traditional herbal medicine in Korea: A retrospective large sample cohort study of musculoskeletal disorder patients Jinho Lee a, Joon-Shik Shin a, Me-riong Kim a, Jang-Hoon Byun a, Seung-Yeol Lee a, Ye-sle Shin a, Hyejin Kim a, Ki Byung Park a, Byung-Cheul Shin b, Myeong Soo Lee c, In-Hyuk Ha a,n a b c
Jaseng Spine and Joint Research Institute, Jaseng Medical Foundation, 858 Eonju-ro, Gangnam-gu, Seoul, Republic of Korea Busan National University, Yangsan Campus, 49 Busandaehak-ro, Mulgeum-eup, Yangsan-si, Gyeongsangnam-do, Republic of Korea Division of Medical Research, Korea Institute of Oriental Medicine, 1672 Yuseongdae-ro, Yuseong-gu, Daejeon, Republic of Korea
art ic l e i nf o
a b s t r a c t
Article history: Received 10 January 2015 Received in revised form 4 March 2015 Accepted 26 April 2015 Available online 6 May 2015
Ethnopharmacological relevance: The objective of this study is to report the incidence of liver injury from herbal medicine in musculoskeletal disease patients as large-scale studies are scarce. Considering that herbal medicine is frequently used in patients irrespective of liver function in Korea, we investigated the prevalence of liver injury by liver function test results in musculoskeletal disease patients. Materials and methods: Of 32675 inpatients taking herbal medicine at 7 locations of a Korean medicine hospital between 2005 and 2013, we screened for liver injury in 6894 patients with liver function tests (LFTs) at admission and discharge. LFTs included t-bilirubin, AST, ALT, and ALP. Liver injury at discharge was assessed by LFT result classifications at admission (liver injury, liver function abnormality, and normal liver function). In analyses for risk factors of liver injury at discharge, we adjusted for age, sex, length of stay, conventional medicine intake, HBs antigen/antibody, and liver function at admission. Results: A total 354 patients (prevalence 5.1%) had liver injury at admission, and 217 (3.1%) at discharge. Of the 354 patients with liver injury at admission, only 9 showed a clinically significant increase after herbal medicine intake, and 225 returned to within normal range or showed significant liver function recovery. Out of 4769 patients with normal liver function at admission, 27 (0.6%) had liver injury at discharge. In multivariate analyses for risk factors, younger age, liver function abnormality at admission, and HBs antigen positive were associated with injury at discharge. Conclusions: The prevalence of liver injury in patients with normal liver function taking herbal medicine for musculoskeletal disease was low, and herbal medicine did not exacerbate liver injury in most patients with injury prior to intake. & 2015 Elsevier Ireland Ltd. All rights reserved.
Keywords: Herbal medicine Complementary therapies Liver function tests Drug-induced liver injury Risk factors
1. Introduction Complementary and Alternative Medicine (CAM) treatments are most frequently used for musculoskeletal diseases in Korea. The 1st–5th most frequent diseases in inpatient care in National
Abbreviations: CAM, Complementary and Alternative Medicine; DILI, druginduced liver injury; LFT, liver function test; ALT, alanine aminotransferase; AST, aspartate aminotransferase; ALP, alkaline phosphatase; TB, total bilirubin; γ-GTP, γ-glutamyl transpeptidase; HBs, Hepatitis B surface; CIOMS, Council for International Organization of Medical Sciences; Ag, antigen; Ab, antibody; KCD, Korean standard classification of diseases; ICD, International Statistical Classification of Diseases and Related Health Problems n Corresponding author. Tel.: þ 82 2 3218 2188; fax: þ 82 2 3218 2244. E-mail address: [email protected] (I.-H. Ha). http://dx.doi.org/10.1016/j.jep.2015.04.048 0378-8741/& 2015 Elsevier Ireland Ltd. All rights reserved.
Health Insurance reimbursements for Korean medicine in 2013 were all musculoskeletal related (1st–5th most frequent Korean standard classification of diseases: M54 Dorsalgia, M17 Gonarthrosis [arthrosis of knee], S33 Dislocation, sprain and strain of joints and ligaments of lumbar spine and pelvis, M51 Other intervertebral disc disorders, M79 Other and unspecified soft tissue disorders, not elsewhere classified) (Kim and Son, 2014). As CAM use increases worldwide, interest in liver toxicity from herbal medicine is also rising (Wolsko et al., 2003). However, herb-induced liver injury is usually more difficult to detect than drug-induced liver injury (DILI) and incidence of cases is highly variable due to various reasons, such as the fact that many users of herbal products do not discuss them with their physician (Eisenberg et al., 1993, 1998), many herbal preparations are regulated by less stringent standards than conventional drugs
408
J. Lee et al. / Journal of Ethnopharmacology 169 (2015) 407–412
(Licata et al., 2013), many herbal products are available as “overthe counter” drugs or food supplements (Singh et al., 2012), and medical use can be both regular or episodic (Tovar and Petzel, 2009). Chalasani et al. (2008) reported in a retrospective U.S. study that herbal and dietary supplements were implicated in 10% of DILI cases , a prospective study on 22 participants conducted in Singapore reported that 68% were related to DILI (Wai et al., 2007), and data from a Chinese medical center on 30 DILI patients indicates that 40% were herb related causes (Wang et al., 2009). On the other hand, many studies report the incidence rate of herbinduced liver injury at less than 1% and clinical symptoms to be mild (Al-Khafaji, 2000). The discrepancy in previous evidence may be partly attributable to cultural differences regarding herbal products, indiscrimination between prescription herbal medicine, folk medicine, and dietary supplements of uncertain composition and consistency, disparity in research methodology and environment, and small sample sizes. Therefore, to assess the prevalence of liver injury due to herbal medicine, a more specific, standard definition of herbal medicine is called for, and due to relatively low prevalence rates, more large-scale studies are needed. As many herbal medicine practitioners construe herbal medicine to be natural and therefore “safe” (Ye and He, 2010), herbal medicine is frequently used in patients with abnormal liver function test (LFT) results. In order to investigate changes in liver enzymes in inpatients taking herbal medicine prescribed by trained medical professionals for musculoskeletal disease, we retrospectively assessed liver injury at discharge by LFT results at admission and its risk factors.
2.3. Statistical analysis Patients were categorized into 3 groups – “liver injury”, “liver function abnormality”, and “normal liver function” – by LFT results at admission and discharge. Age and γ-GTP results were continuous variables, and sex (male/female), HBs antigen (positive/ negative), HBs antibody (positive/negative), and use of conventional medicine (yes/no) were categorical variables. HBs antigen and antibody results were classified into 4 groups; Ag( þ)/Ab( þ), Ag(þ)/Ab( ), Ag( )/Ab(þ ), and Ag( )/Ab( ). Length of stay was assessed as both continuous and categorical variables (r 20 days, 20–30 days, and 430 days). Simple logistic regression modeling was used to assess influence of age, sex, length of hospital stay, conventional medicine intake, liver function classification at admission, HBs antigen/antibody, and γ-GTP on liver function classification at discharge. Multiple logistic regression modeling was used for factors with a p-value of r0.2. Factors included in the final model were determined in a stepwise manner (included if p-valuer0.05, excluded if p-valueZ0.10). We used stratification of patient characteristics to further assess the association between age and liver function. The statistical package for Social Science for WindowsTM, version 11.0 (SPSS Inc., Chicago, IL, U.S.A.) was used. 2.4. Ethics statement Ethical approval was obtained from the institutional review board of Jaseng Hospital of Korean medicine in Seoul, Korea (SIRB2013-37), and all patients provided informed written consent to use of data for academic means prior to hospitalization. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki and Tokyo.
2. Material and methods 2.1. Subjects We retrospectively reviewed the electronic medical records of 32675 patients who stayed at 7 locations of a musculoskeletal disease-specialty Korean medicine hospital in Korea from December 2005 to December 2013. The total number of patients with blood test results during admission was 29229 and 24% (n ¼7003) received multiple tests at admission and before discharge. Of these patients, we analyzed the results of 6894 patients for alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), and total bilirubin (TB). Inclusion criteria – patients admitted for at least one day who took herbal medicine and at least 2 LFTs during admittance. There were no exclusion criteria. 2.2. Classification of liver enzyme abnormalities LFTs were measured after a 6 h overnight fast. All inpatients received blood tests at admission except those who refused, and follow-up tests were conducted as deemed necessary. The LFTs included in this study were ALT, AST, ALP, TB, γ-glutamyl transpeptidase (γ-GTP), Hepatitis B surface (HBs) antigen, and HBs antibody. The normal range for LFTs was ALTr40 U/L; ASTr40 U/L; ALPr338 U/L; and TBr1.2 mg/dL. According to the modified Council for International Organization of Medical Sciences (CIOMS) criteria, the term “liver injury” was used to describe an increase of over 2 N (upper limit of the normal range) in ALT, or a combined increase in TB, AST, and ALP, provided one of them was above 2 N. Patients with all 4 enzymes within normal range were defined as having “normal liver function”, and those neither in the “normal” nor “liver injury” group as “liver function abnormality”.
3. Results The average length of hospital stay was 26.17 712.31 days, 45% (n ¼3111) were male, and average age was 44.31 714.49 years. We classified the patients into 3 groups (liver injury, liver function abnormality, and normal liver function) according to liver function test results at admission, and subcategorized the groups by liver function test results at discharge to better illustrate the change in distribution of liver function state (total 9 groups). Of 4769 patients with normal liver function at admission, 27 (0.6%) had liver injury at discharge, and out of 354 patients with liver injury at admission, 225 (64%) no longer had liver injury at Table 1 Classification of patients by liver function test results at admission and dischargea. Liver function state at admission
N
Liver injury
354
Liver function abnormality
Normal liver function
Liver function state at discharge
Liver injury Liver function abnormality Normal liver function 1771 Liver injury Liver function abnormality Normal liver function 4769 Liver injury Liver function abnormality Normal liver function
N
129 143 82 61 763 947 27 284 4458
(9) (0) (1) (23) (28) (23) (27) (117) (259)
( ): No. of patients with a twofold increase in at least one blood index of ATP, AST, ALT, or total bilirubin at discharge compared to admission. a The normal range for liver function tests was alanine aminotransferase (ALT) r40 U/L; aspartate aminotransferase (AST)r40 U/L; alkaline phosphatase (ALP) r338 U/L; and total bilirubinr1.2 mg/dL, and the term “liver injury” was used to describe an increase of over 2 N (upper limit of the normal range) in ALT, or a combined increase in total bilirubin, AST, and ALP, provided one of them was above 2 N.
J. Lee et al. / Journal of Ethnopharmacology 169 (2015) 407–412
discharge. The number of patients with liver injury was 354 (total prevalence 5.1%) at admission, and 217 (3.1%) at discharge, indicating a decrease to 61%. Fifty-nine patients, which accounts for 27% of the patients with liver injury at discharge, had liver injury with a twofold increase in ALT, AST, ALP, or TB at discharge compared to admission (Table 1). We compared characteristics and use of conventional and herbal medicine in the 4458 patients who maintained normal liver function at discharge with the 27 patients who had liver injury at discharge out of the total 4769 patients with normal liver function at admission. Age, sex, length of hospital stay, conventional medicine intake, HBs antigen/antibody, γ-GTP, and frequency and patterns of conventional and herbal medicine use were similar in both groups (Table 2). The 27 patients with liver injury at discharge did not display any liver-related clinical symptoms and related data will be reported in a future case series. Chungpa-jun (including variations) was the most frequently prescribed herbal medicine during hospital stay, and the basic composition consists of Ostericum koreanum, Eucommia ulmoides (Hardy rubber tree), Acanthopanax sessiliflorus, Achyranthes bidentata (Ox knee), Psoralea corylifolia (Babchi), Peucedanum japonicum, Cibotium barometz (Golden chicken fern, Woolly fern, Scythian lamb), Lycium chinense (Chinese boxthorn, Chinese wolfberry, Chinese matrimony vine), Boschniakia rossica (Northern groundcone), and Cuscuta chinensis (Chinese dodder seed) plus or minus 2–3 ingre-
409
dients. The formulas of the other herbal prescriptions prescribed in the 27 patients with liver injury at discharge of patients with normal liver function at admission are included in the Appendix (Table A1). Conventional medicine use was mainly comprised of customary musculoskeletal drugs such as anti-inflammatory agents and analgesics, and concomitant drugs such as gastrointestinal system drugs and antidepressants, and is also presented in the Appendix (Table A2). In analyses for risk factors of liver injury prevalence at discharge, age, sex, length of stay, liver function at admission, HBs antigen/ antibody, and γ-GTP showed significant associations. Age, HBs antigen/antibody and liver function at admission were shown to be significantly associated in the multivariate analysis. Liver injury at discharge was more prevalent in younger age, HBs antigen positive, and abnormal liver conditions at admission (Table 3). When stratified to assess the effect of age on liver injury, we found decreasing prevalence of liver injury with increasing age in most factors (Table 4). Diagnostic data was coded both in the Korean standard classification of diseases (KCD), which is the Korean version of the International Statistical Classification of Diseases and Related Health Problems (ICD), and Korean medicine diagnosis and pattern differentiation, which was generated specifically for standard classification of traditional Korean medicine, of which the 3rd revision was incorporated into KCD-6 in 2010. The majority of
Table 2 Comparison of characteristics and conventional and herbal medicine use in patients with normal liver function at admission classified by liver function at discharge (liver injury at discharge vs. normal liver function at discharge).
Age Sex Female Male HBs(Ag)/(Ab)b Ag ( þ )/Ab ( þ) Ag ( þ )/Ab ( ) Ag ( )/Ab (þ ) Ag ( )/Ab ( ) Use of conventional medicine Yes No γ-GTPc Length of hospital stay Classification of conventional medicine by main ingredientd Almagate Eperisone hydrochloride Aceclofenac Naproxen Amitriptyline hydrochloride
Liver injury at discharge N ¼ 27 N(%)/mean 7 SD
Normal liver function at discharge N ¼ 4458 N(%)/mean 7 SD
39.2 714.3
45.4 7 14.7
13 (48.1) 14 (51.9)
2971 (66.6) 1487 (33.4)
0 (0.0) 1 (4.0) 18 (72.0) 6 (24.0)
4 (0.1) 163 (3.8) 2761 (64.5) 1350 (31.6)
10 (37.0) 17 (63.0) 32.8 725.8
1573 (35.3) 2885 (64.7) 27.1 725.8
23.4 78.8
27.2 7 12.5
5 5 3 2 2
562 (14.6) 411 (10.7) 137 (4.7) 347 (9.0) 309 (8.0)
(19.2) (19.2) (11.5) (7.7) (7.7)
Acetaminophen 2 (7.7) Artemisia asiatica 95% ethanol ext. 1 (3.8) Tramadol hydrochloride 1 (3.8)
255 (6.6) 85 (2.2) 80 (2.1)
Limaprost alpha-cyclodextrin clathrate Levosulpiride Glimepiride Simvastatin Clobetasol-17-propionate
1 (3.8)
55 (1.4)
1 1 1 1
50 (1.3) 11 (0.3) 1 (0.0) 1 (0.0)
a
(3.8) (3.8) (3.8) (3.8)
Liver injury at discharge N ¼ 27 N(%) Herbal medicine typea Chung-pa-jun 20 (40.8) Ssang-pae-tang 3 (6.1) Hwal-hyul-ji-tong-tang 2 (4.1) Hwal-lak-tang 2 (4.1) Saeng-gang-gun-bi-tang 2 (4.1) Ssang-hwa-sam-so-eum 1 (2.0) Gwak-hyang-jung-gi-san 1 (2.0) Sook-ji-yang-keun-tang 1 (2.0) Gae-gyung-seo-kyung-tang 1 (2.0) In-sam-yang-wi-tang 1 (2.0) Ma-bal-gwan-jul-tang 1 (2.0) Sam-ryeong-baek-chul1 (2.0) tang Ga-gam-gui-bi-tang 1 (2.0) Pyung-wi-san 1 (2.0) Hyeong-bang-tang Hoe-soo-san Hyang-sa-yuk-gun-ja-tang Yang-hyeol-geo-pung-tang Ga-mi-Yuk-mi-ji-whangtang Shin-tong-chu-uh-tang Hyeong-bang-sa-baek-san Ban-ha-baek-chul-cheonma-tang Man-geum-tang Ga-mi-Yuk-mi Li-gi-geo-poong-san
The pharmacological classification of each conventional medicine is presented in Table A2 in the Appendix. Hepatitis B surface (HBs) antigen (Ag)/HBs antibody (Ab). c γ-glutamyltranspeptidase. d The main ingredients of each herbal medicine are presented as Table A1 in the Appendix. b
1 1 1 1 1
(2.0) (2.0) (2.0) (2.0) (2.0)
Normal liver function at discharge N ¼4458 N(%)
3432 612 122 81 10 563 441 189 122 48 37 28
(42.7) (7.6) (1.5) (1.0) (0.1) (7.0) (5.5) (2.4) (1.5) (0.6) (0.5) (0.3)
22 (0.3) 16 (0.2) 15 13 12 8 5
(0.2) (0.2) (0.1) (0.1) (0.1)
1 (2.0) 1 (2.0) 1 (2.0)
4 (0.0) 3 (0.0) 2 (0.0)
1 (2.0)
2 (0.0)
1 (2.0) 1 (2.0)
1 (0.0)
410
J. Lee et al. / Journal of Ethnopharmacology 169 (2015) 407–412
Table 3 Univariate and multivariate analyses of factors by liver injury at discharge. Univariate analysisa
Age Sex Female Male 0.5 γ-GTPc HBs(Ag)/(Ab)d Ag ( þ)/Ab ( þ ) Ag ( þ)/Ab (-) Ag ( )/Ab ( þ ) Ag ( )/Ab ( ) Use of conventional medicine Yes No Length of hospital stay Liver function test results at admission Liver injury Liver function abnormality Normal liver function
Multivariable analysisb
Total (liver injury)
OR (95% CI)
p-value
OR (95% CI)
p-value
6894 (217)
0.96 (0.95–0.97)
o 0.001 o 0.001
0.97 (0.95–0.98)
o 0.001
3783 (61) 3111 (156)
0.31 (0.23–0.42) (Reference)
4403 (217)
1.01 (1.01–1.01)
8 288 4088 2156
(0) (22) (111) (65)
– 2.66 (1.61–4.39) 0.9 (0.66–1.23) (Reference)
2475 (68) 4419 (149) 6894 (217)
0.81 (0.61–1.08) (Reference) 0.96 (0.95–0.98)
354 (129) 1771 (61) 4769 (27)
100.69 (65.12–155.70) 6.27 (3.97–9.89) (Reference)
o 0.001 o 0.001
o 0.001 – 3.54 (1.62–7.73) 1.05 (0.65–1.70) (Reference)
0.155
o 0.001 o 0.001
0.98 (0.96–1.00)
0.021 o 0.001
73.37 (42.11–127.86) 3.74 (2.03–6.89) (Reference)
a Univariate analysis was conducted on age, γ-GTP, and length of hospital stay as continuous variables, and other factors (sex, HBs antigen/HBs antibody, concurrent conventional medication intake, and liver function test results at admission) as categorical variables. b Factors whose p-value was less than 0.2 in the univariate analysis were selected for multivariable analysis (All 7 factors were selected). Factors included in the final model were determined in a stepwise manner (included if p-value r 0.05, excluded if p-value Z 0.10). c γ-glutamyl transpeptidase. d Hepatitis B surface (HBs) antigen (Ag)/HBs antibody (Ab).
inpatients were musculoskeletal pain patients with lumbar or cervical disc herniation or joint related complaints (details provided in Table A3 in the Appendix).
Table 4 Prevalence ratios with 95% confidence intervals for liver injury at discharge by age. Total (liver injury)
4. Discussion In our study, patients taking herbal medicine for musculoskeletal disease treatment had a lower prevalence of liver injury after intake. A similar tendency was shown in the small number of patients with normal liver function at admission and liver injury at discharge (n ¼27; 0.6%). Though the liver injury subgroup at discharge was too small for statistical analysis of difference in the subgroups at discharge out of the normal liver function group at admission, a comparison of baseline characteristics and intake of herbal medicine and conventional medicine displayed no considerable difference. Analysis of total admitted patients showed that liver dysfunction at admission was associated with a higher likelihood of liver injury at discharge, and younger age was a risk factor for liver injury. The 7 locations of a Korean medicine hospital which was the study setting mainly treat and specialize in musculoskeletal complaints, and was the first Korean medicine hospital to be endorsed by the Korean ministry of health and welfare as a spinal disease specialty hospital. This hospital adopts a conservative integrative treatment model using conventional diagnostic tools (X-rays, CTs, MRIs, blood tests) and traditional Korean medicine (acupuncture, herbal medicine, manual manipulation) as the main treatment intervention with conventional medicine (drugs, injections) utilized as necessary as ancillary treatment. The vast majority of participants suffered from low back pain or neck pain due to disc herniation or joint dysfunction from related diseases, and none were classified as suffering from rheumatologic diagnoses more often seen in younger or female patients, or such pathologies as lupus or myositis which may be related to AST and ALT elevation. The main objectives of herbal medicine use did not differ significantly from that of conventional drugs with Chungpa-jun
Liver function test results at admission Liver injury 113 (65) Liver function abnormality 891 (24) Normal liver function 2984 (19) HBs(Ag)/(Ab)b Ag( þ )/Ab( ) 151 (15) Ag( )/Ab(þ ) 2592 (59) Ag( )/Ab( ) 1240 (34) Length of hospital stay 30o 1219 (22) 20–30 1330 (27) r 20 1439 (59) Conventional medication intake Yes 1356 (37) No 2632 (71) Sex Male 1802 (78) Female 2186 (30) γ-GTPc 73o 465 (45) r 73 3523 (63) a b c
OR (by age)a
95% CI
0.95 0.96 0.98
0.92–0.99 0.93–1.00 0.95–1.02
0.94 0.96 0.97
0.86–1.03 0.93–0.99 0.94–1.00
0.94 0.94 0.98
0.90–0.98 0.90–0.98 0.96–1.01
0.95 0.97
0.92–0.99 0.95–1.00
0.97 0.95
0.95–1.00 0.92–0.98
0.93 0.98
0.90–0.97 0.95–1.00
Age was assessed as a continuous variable. Hepatitis B surface (HBs) antigen (Ag)/HBs antibody (Ab). γ-glutamyl transpeptidase.
used predominantly for spine or joint disorders, Ssangpae-tang, Ssanghwa-samso-eum, and Dokgam-tang prescribed for common cold or flu symptoms and Gwakhyang-junggi-san for gastrointestinal disorders. Other than these, many drugs were regular medications taken for comorbidities (hypertension, diabetes). The composition and dosage of individual herbal medicines are complex and difficult to gauge as they may comprise either a single ingredient or a compound mixture of different herbal ingredients, and though medications can be in liquid, powder, tablet or other forms, we listed only those in liquid form as this was the main
J. Lee et al. / Journal of Ethnopharmacology 169 (2015) 407–412
type of herbal medicine, usually taken for extended periods of time. We did not determine the cumulative exposure dose, instead dividing the prescribed medication by intention of usage for classification purposes. (i.e. Multiple prescriptions of a certain herbal medicine taken by one patient was tallied as 1 prescription of that medicine.) The biggest strength of our study is that it is one of the few large-scale studies to examine use of herbal medicine in musculoskeletal disease patients, who are frequent consumers of herbal medicine. Also, most studies studying potential liver toxicity of herbal medicine only include patients with normal liver function, and assess prevalence after excluding various causes of liver injury (e.g. hepatitis B virus, conventional medicine use, alcohol consumption). However, herbal medicine is frequently used in patients with abnormal LFT results or those with potential causes of liver injury in absence of significant liver injury symptoms as it is considered to be “safe”. Our study assessed the prevalence of liver injury in musculoskeletal patients before and after being prescribed with herbal medicine in inpatient care and found that prevalence of liver injury was lower at discharge. This is one of the few studies to show that herbal medicine does not have a negative impact on liver injury in general musculoskeletal patients (without selective inclusion of only normal liver function patients and with non-exclusion of risk factors for liver injury) in treatment for musculoskeletal disorders. The largest drawback of our study is probably the retrospective study design. We could not draw any conclusions regarding the causality relationship between herbal and conventional medicine use and liver injury. Another limitation is that although direct bilirubin, which is used to assess liver injury in CIOMS criteria, was measured in some inpatients, we excluded the results due to the small sample size, and other indexes that may affect liver function (prothrombin time, anti-HAV IgM, anti-HBc IgM, HCV RNA, anti-HSV, IgM, and anti-EBV IgM) were not measured. Also, as only 24% of total inpatients were followed-up with additional tests, the results of our study may not be widely applicable. However, as the need for follow-up tests was decided by Korean medicine doctors, most patients with no additional tests were those at low risk of liver dysfunction. Therefore it is highly probable that the overall prevalence would be lower. Also when taking into consideration severity of symptoms requiring admission and subsequent distress levels, frequent herbal medicine intake (liquid and/or pill form taken 3 times a day), concurrent use of conventional medicine in 2475 (36%) patients, number of HBV carriers, and other possible liver-related comorbidities, the liver injury prevalence is not high. Twenty-seven patients (0.6%) with normal liver function at admission had liver injury at discharge, which is comparable to other sizable retrospective inpatient studies. Melchart et al., 1999 reported that of 1450 patients treated with traditional Chinese drugs, 14 patients (0.97%) could be considered DILIs showing a twofold increase in ALT , while a study in Japan reported 15 (0.6%) out of 2496 patients (Sheehan et al., 1992), while in a study on 3102 patients, 22 DILI patients were identified; caused by conventional medicine in 15 patients (0.6%) and herbal medicine in 7 (0.23%) (Mitsuma, 2002). A recent study in Korea reported a prevalence of 5 DILI patients (0.56%) of 892. In analyses for risk factors of liver injury at discharge, liver function at admission was most influential. Also, the likelihood of liver injury was higher in younger age. These results differ from previous studies reporting a positive association between age and DILI (Abboud and Kaplowitz, 2007). However, while age is a risk factor for DILI, susceptible age groups differ according to specific medications. For example, older age is a risk factor for DILI related to isoniazid, whereas younger age is a risk factor for valproate and
411
aspirin (Reyes syndrome). Although the exact mechanism of DILI is still largely unidentified, it seems to involve 2 pathways – direct liver toxicity and adverse immune reactions (Holt and Ju, 2006). DILIs from herbal medicines are generally idiosyncratic drug reactions, strongly suggesting a mainly immune response (Ju et al., 2012). Immune responses against foreign antigens and malignant cells are lower in the elderly, and most immune cells involved in innate immunity show decreased function with aging (Tajiri and Shimizu, 2013). This suggests that herbal DILI prevalence may be higher in younger groups as younger patients have stronger immune responses. Another explanation could be that as most symptoms due to herbal medicines are mild and use of herbal medicine increases with age in Korea, the inverse relationship between age and liver injury is due to immune tolerance through repeated sub-clinical immune reactions. There are also consistent reports on how cholestatic liver enzyme patterns are more likely to be seen in liver injury in older patients (Lucena et al., 2009), and it is notable that liver injury related to herbal medicine is generally of hepatocellular type (Jeong et al., 2012).
5. Conclusions In conclusion, prevalence of liver injury in patients taking herbal medicine for treatment of musculoskeletal disease was low, and clinical symptoms were sub-clinical or mild. Liver function at admission was the most prominent risk factor for liver injury, and there was a tendency of increasing prevalence of liver injury with younger age.
Acknowledgments None.
Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.jep.2015.04.048. References Abboud, G., Kaplowitz, N., 2007. Drug-induced liver injury. Drug Saf.: Int. J. Med. Toxicol. Drug Exp. 30, 277–294. Al-Khafaji, M., 2000. Monitoring of liver enzymes in patients on Chinese medicine. J. Chin. Med. 62, 6–10. Chalasani, N., Fontana, R.J., Bonkovsky, H.L., Watkins, P.B., Davern, T., Serrano, J., Yang, H., Rochon, J., Drug Induced Liver Injury Network (DILIN), 2008. Causes, clinical features, and outcomes from a prospective study of drug-induced liver injury in the United States. Gastroenterology 135, 1924–1934 1934.e1-4. Eisenberg, D.M., Davis, R.B., Ettner, S.L., Appel, S., Wilkey, S., Van Rompay, M., Kessler, R.C., 1998. Trends in alternative medicine use in the United States, 1990–1997: results of a follow-up national survey. J. Am. Med. Assoc. 280, 1569–1575. Eisenberg, D.M., Kessler, R.C., Foster, C., Norlock, F.E., Calkins, D.R., Delbanco, T.L., 1993. Unconventional medicine in the United States. Prevalence, costs, and patterns of use. N. Engl. J. Med. 328, 246–252. Holt, M.P., Ju, C., 2006. Mechanisms of drug-induced liver injury. AAPS J. 8, E48–E54. Jeong, T.Y., Park, B.K., Cho, J.H., Kim, Y.I., Ahn, Y.C., Son, C.G., 2012. A prospective study on the safety of herbal medicines, used alone or with conventional medicines. J. Ethnopharmacol. 143, 884–888. Ju, H.Y., Jang, J.Y., Jeong, S.W., Woo, S.A., Kong, M.G., Jang, H.Y., Lee, S.H., Kim, S.G., Cha, S.W., Kim, Y.S., Cho, Y.D., Jin, S.Y., Kim, H.S., Kim, B.S., 2012. The clinical features of drug-induced liver injury observed through liver biopsy: focus on relevancy to autoimmune hepatitis. Clin. Mol. Hepatol. 18, 213–218. Kim, J., Son, MS, 2014. 2013 National Health Insurance Statistical Yearbook. Health Insurance Review and Assessment Service and National Health Insurance Service, Seoul. Licata, A., Macaluso, F.S., Craxi, A., 2013. Herbal hepatotoxicity: a hidden epidemic. Intern. Emerg. Med. 8, 13–22. Lucena, M.I., Andrade, R.J., Kaplowitz, N., Garcia-Cortes, M., Fernandez, M.C., Romero-Gomez, M., Bruguera, M., Hallal, H., Robles-Diaz, M., Rodriguez-
412
J. Lee et al. / Journal of Ethnopharmacology 169 (2015) 407–412
Gonzalez, R.S., Navarro, J.M., Salmeron, J., Martinez-Odriozola, P., Perez-Alvarez, R., Borraz, Y., Hidalgo, R., Spanish Group for the Study of Drug-Induced Liver Disease, 2009. Phenotypic characterization of idiosyncratic drug-induced liver injury: the influence of age and sex. Hepatology 49, 2001–2009. Melchart, D., Linde, K., Hager, S., Kaesmayr, J., Shaw, D., Bauer, R., Weidenhammer, W., 1999. Monitoring of liver enzymes in patients treated with traditional Chinese drugs. Complement. Ther. Med. 7, 208–216. Mitsuma, T., 2002. Adverse reactions encountered during treatment with Japanese oriental (Kampo) medicines. Korean J. Orient. Intern. Med. 2, 9–14. Sheehan, M.P., Rustin, M.H., Atherton, D.J., Buckley, C., Harris, D.W., Brostoff, J., Ostlere, L., Dawson, A., 1992. Efficacy of traditional Chinese herbal therapy in adult atopic dermatitis. Lancet 340, 13–17. Singh, D., Gupta, R., Saraf, S.A., 2012. Herbs-are they safe enough? an overview. Crit. Rev. Food Sci. Nutr. 52, 876–898.
Tajiri, K., Shimizu, Y., 2013. Liver physiology and liver diseases in the elderly. World J. Gastroenterol. 19, 8459–8467. Tovar, R.T., Petzel, R.M., 2009. Herbal toxicity. Disease-a-Month 55, 592–641. Wai, C.T., Tan, B.H., Chan, C.L., Sutedja, D.S., Lee, Y.M., Khor, C., Lim, S.G., 2007. Druginduced liver injury at an Asian center: a prospective study. Liver Int.: Off. J. Int. Assoc. Study Liver 27, 465–474. Wang, Y.P., Shi, B., Chen, Y.X., Xu, J., Jiang, C.F., Xie, W.F., 2009. Drug-induced liver disease: an 8-year study of patients from one gastroenterological department. J. Dig. Dis. 10, 195–200. Wolsko, P.M., Eisenberg, D.M., Davis, R.B., Kessler, R., Phillips, R.S., 2003. Patterns and perceptions of care for treatment of back and neck pain: results of a national survey. Spine 28, 292–297, Discussion 298. Ye, X.F., He, J., 2010. The bright future of Chinese herbal medicine: only after a twisty road. Contemp. Clin. Trials 31, 508–509.
