部分中药肝毒

Traditional Chinese medicine causing hepatotoxicity in patients with chronic hepatitis B infection: a 1-year prospective study

http://onlinelibrary.wiley.com/doi/10.1111/j.1365-2036.2006.03111.x/full

• First published: 20 September 2006Full publication history
• DOI: 10.1111/j.1365-2036.2006.03111.xView/save citation
• Cited by: 30 articles

Dr M.-F. Yuen, Associate Professor, Department of Medicine, The University of Hong Kong, Queen Mary Hospital, Pokfulam Road, Hong Kong.
E-mail: [email protected]

Abstract

Summary

Background

Safety of traditional Chinese medicine in patients with chronic hepatitis B is unknown.

Aim

To study the clinical outcome of traditional Chinese medicine-induced hepatotoxicity in chronic hepatitis B patients.

Patients and methods

All
chronic hepatitis B patients in 2004 with liver dysfunction requiring
hospitalization were screened prospectively for traditional Chinese
medicine intake. The hepatotoxicity of individual traditional Chinese
medicine elements was determined by extensive search of both English and
Chinese publications.

Results

Of
45 chronic hepatitis B patients, the liver dysfunction in seven (15.6%)
was attributable to traditional Chinese medicine. All had liver
dysfunction pattern resembling those of acute exacerbation of chronic
hepatitis B. Three patients had adverse outcomes (two deaths, one liver
transplantation). One patient had accelerated course of cirrhosis now
awaiting liver transplantation. The identified hepatotoxic components
were Polygonum multiflorum Thunb, Cassia obtusifolia L, Melia toosendan Sieb., Rheum palmatum L., Scolopendra subspinipes mutilans L, Alisma orientale Juzepe, Glycyrrhiza uralensis Fisch. and Mentha haplocalyx Briq. One traditional Chinese medicine formula was adulterated with a highly hepatotoxic agent, N-nitrosofenfluramine.

Conclusions

Traditional
Chinese medicine-related hepatotoxicity resulted in high mortality in
chronic hepatitis B patients. Prospective randomized-controlled trials
with the same stringent criteria as western medicine clinical trials are
required for Chinese medicines, to document their efficacies and safety
before they can be advocated for the treatment of patients.

Background

Despite
the lack of controlled studies to validate the clinical efficacies and
safety profiles of traditional Chinese medicine (TCM), TCM has become
more frequently used recently, both in Asian and Caucasian populations.^1, 2 One study has shown that more than 40% Americans take TCM.³
One of the main reasons for taking TCM is that TCM are perceived to
have much less side effects and more ‘harmonizing and balancing’ power
to the body compared with western medicine. However, there is increasing
number of reports showing that several TCM can lead to toxicity.^2, 4, 5
Systematic and comprehensive study of the toxicities related to TCM is
difficult to conduct because of several reasons. These include
incomprehensive understanding of the modes of action, lack of
standardization in the preparatory procedure with different
formulations, lack of quality control and high complexity of the
ingredients, particularly those from herbal extracts.

Since the
currently approved agents for the treatment of chronic hepatitis B (CHB)
are usually unable to completely eradicate the hepatitis B virus (HBV),
many CHB patients opt for alternative therapies such as TCM for the
protection and optimization of liver function. One report finds that 39%
of patients with chronic liver diseases admitted to using some form of
complementary and alternative medicine.⁶
Another report shows that 21% and 13% of patients attending liver
clinics have taken TCM previously or are currently using TCM to treat
the liver condition, respectively.⁷ According to a recent study from Hong Kong, 32% of patients with CHB infection admitted a history of TCM usage.⁸
There are two important issues concerning TCM in patients with CHB.
First, being the major organ for drug biotransformation, the liver is at
a unique and the highest risk for TCM toxicity of the body. According
to the reports of the National Poisons Information Service in the United
Kingdom, hepatotoxicity is the most frequent adverse event among the
possible or confirmed TCM-associated toxicity.⁹
Secondly, the hepatotoxic effects of TCM may be augmented in CHB
patients because of the background HBV infection, just as the increased
hepatotoxicity of isoniazid in CHB patients.¹⁰ These patients may therefore be at a higher risk for liver decompensation.

The present study aimed to study the clinical outcome of TCM-induced hepatotoxicity in patients with CHB.

Patients and methods

From
January to December 2004, all patients with known CHB infection
[defined as positivity of hepatitis B surface antigen (HBsAg) for more
than 6 months] admitted to Queen Mary Hospital, The University of Hong
Kong, Hong Kong because of liver biochemistry derangement were
prospectively screened. The inclusion criteria were HBsAg positivity,
intake of TCM within 6 months prior to admission, elevated bilirubin
levels of more than two times upper limit of normal (ULN) or elevated
level of at least one of the liver enzymes including alanine
aminotransferase (ALT), aspartate aminotransferase (AST) to more than
five times ULN, or alkaline phosphatase (ALP) and γ-glutamyltranspeptidase
(GGT) to more than two times ULN. The exclusion criteria included: (i)
evidence of HBV exacerbation defined by elevated ALT levels associated
with HBV-DNA level more than 10⁶ copies/mL with or without
positive IgM antibody to hepatitis B core antigen (anti-HBc IgM); (ii)
evidence of other acute or chronic hepatitis virus infections including
hepatitis A, C, D and E [confirmed by antibodies against the hepatitis A
virus (anti-HAV IgM), the hepatitis C virus (anti-HCV IgG), the
hepatitis D virus (anti-HDV IgG) and the hepatitis E virus (anti-HEV
IgM)]; (iii) evidence of recent severe sepsis; (iv) intake of western
medication with known hepatotoxicity; (v) history of significant alcohol
intake (>20 g/day for women and >30 g/day for men); (vi) history
of ultrasonographic evidence of steatohepatitis and (vii) evidence of
autoimmune hepatitis, Wilson‘s disease and primary biliary cirrhosis.
The likelihood of TCM-induced hepatotoxicity was decided with reference
to the international consensus criteria.¹¹

For
patients with TCM-induced hepatotoxicity, details of the TCM including
the drug formulae prescribed by the herbalists, the doses and the
frequency of intake were analysed. Extensive literature search through
PubMed; Medline; Electronic Knowledge Gateway of the Hospital Authority,
Hong Kong; Chinese Medicine Library of the Baptist University, Hong
Kong; books on TCM written in Chinese and English^12, 13
were performed to determine the hepatotoxicity associated with
individual TCM element. Complete blood count, liver biochemistry, renal
function test and prothrombin time were monitored. Hepatic
decompensation is defined as prolonged prothrombin time to more than 5 s
of controls and elevated bilirubin level to more than two times ULN, or
development of ascites and/or hepatic encephalopathy.

Results

Between
January and December 2004, 45 patients with CHB were admitted with
derangement of liver biochemistry. Of these, seven patients (15.6%) had
liver derangement attributable to the intake of TCM. The causes of liver
biochemical derangement for the remaining 38 patients are as follows:
HBV exacerbation (n = 28; eight had TCM intake which were not known to have hepatotoxicity); acute hepatitis A infection (n = 2); acute hepatitis E infection (n = 3); liver toxicities because of other medications including carbimazole and isoniazid (n = 3) and acute alcoholic hepatitis (n = 2). The clinical information and the progress of the seven patients with TCM-associated liver derangement are as follows.

Patient 1

A
33-year-old female (CSM) HBV carrier was admitted with 5 days history
of jaundice, associated with nausea and anorexia. She had been taking
two capsules of a Chinese proprietary medicine three times daily for
weight reduction for 6 weeks. Physical examination was normal except for
jaundice. The liver biochemistry on presentation was as follows:
albumin 35 g/L (normal 42–54), bilirubin 173 μm
(normal 7–19), ALT 3002 U/L (normal 5–31), AST 2371 U/L (normal 12–28),
ALP 133 U/L (normal 34–104) and GGT 110 U/L (normal 9–42). The
prothrombin time was 15.7 s (control 13.2 s). She was positive for the
antibody to hepatitis B e antigen (anti-HBe) but negative for IgM
antibody to the hepatitis B core antigen (anti-HBc IgM). HBV-DNA was
undetectable with the Digene Hybrid II Capture assay (Gaithersburg, MD,
USA; lower limit of detection 1.42 × 10⁵ copies/mL).
Ultrasonogram of abdomen did not show evidence of cirrhosis with normal
liver and spleen size. The liver function continued to deteriorate with
prothrombin time reaching 53 s and bilirubin level 710 μm.
In view of the irreversible hepatic decompensation, living-related
liver transplantation was performed 4 weeks after admission. The
explanted liver revealed confluent necrosis with marked cholestasis and
sparse lymphocytic aggregates without any significant fibrosis. The
histological features were compatible with drug-induced injury. The
Chinese propriety medicine contained 12 herbal items, two of which,
Heshouwu and Juemingzi, listed in Table 1 contain anthraquinones (approximately 1% by weight) are known to be hepatotoxic on prolonged intake.^14–17
In addition, with further laboratory analysis, the prescribed TCM
formula is found to be adulterated with N-nitrosofenfluramine which is
highly hepatotoxic.^18, 19 The patient recovered well after the liver transplantation.

Table 1.   Details of the elements of the traditional Chinese medicines for the first five patients

Patient 2

A
45-year-old man (LTS) with known CHB complaining of epigastric
discomfort. The liver biochemistry was normal (albumin 39 g/L, bilirubin
13 μm, ALT 23 U/L) 4 weeks
before presentation to our hospital. Ultrasonogram of abdomen did not
show evidence of cirrhosis with normal liver and spleen size. He
consulted an herbalist for epigastric discomfort and was prescribed five
different formulae of TCM. One month later, he developed ascites with
derangement of liver function: albumin 30 g/L, bilirubin 450 μm (conjugated bilirubin 372 μm),
ALT 27 U/L, ALP 247 U/L and GGT 190 U/L. The prothrombin time was
prolonged to 24.2 s. He was positive for anti-HBe and negative for
anti-HBc IgM. The HBV-DNA level was undetectable by bDNA assay (Bayer
HealthCare LLC, Tarrytown, NY, USA; lower limit of detection
2000 copies/mL). He also developed acute renal failure [creatinine level
990 μm (normal 82–126), urea level 45.6 mm (normal 3.2–7.5)] requiring supportive haemodialysis. The platelet count decreased from 147 to 36 × 10⁹/L.
Bone marrow examination revealed reactive change only. The prothrombin
time and bilirubin level continued to increase and the renal function
failed to improve. He finally died of sepsis and multiorgan failure
2 weeks after admission. Of the five TCM formulae, several identifiable
elements were potentially hepatotoxic and nepthrototoxic (Table 1): Chuanlianzi contains a hepatotoxic alkaloid, toosendanin.²⁰ Dajuang contains a hepatotoxic anthraquinone (0.7–4.2% by weight) as in patient 1 (anthraquinone is also nepthrotoxic).²¹
Wugong from dried powder of centipedes, causes fatal cases of
multiorgan failure (reported in Chinese literature only). There are at
least four others (Quanxie, Baihuagshe, Wushaoshe, Jiangcan) that
contain animal toxins that are potentially hepatotoxic.

Patient 3

A
59-year-old man (CTK) with HBeAg-positive liver cirrhosis presented to
our hospital with gum bleeding and bilateral lower limbs petechiae.
Baseline liver biochemistry during routine follow-up 2 months previously
was: albumin 30 g/L, bilirubin level 32 μm, ALT level 113 U/L, ALP level 122 U/L. The platelet count was 72 × 10⁹/L. The creatinine level was 74 μm.
Ultrasonogram of abdomen showed small-size cirrhotic liver and
splenomegaly. He also had oesophageal varices, which had been
endoscopically banded prophylactically. He took eight doses of a Chinese
herbal formula daily for leg cramps 3 weeks prior to admission. On
admission, the liver biochemistry was grossly deranged: albumin 21 g/L,
bilirubin level 484 μm (conjugated bilirubin 339 μm),
ALT level 41 U/L, ALP level 123 U/L and GGT level 54 U/L. He developed
severe coagulopathy with extremely low platelet count of 2 × 10⁹/L,
prolonged prothrombin time of 15.1 s and activated partial
thromboplastin time of 52.8 s, and increased D-dimer level of 1971 ng/mL
(normal <500). Peripheral blood film showed occasional fragmented
red cells. Direct Coomb‘s test was positive. He also developed acute
renal failure with elevation of creatinine level to 503 μm. His HBV-DNA level was 3.2 × 10³ copies/mL
(measured by bDNA assay). Bone marrow examination revealed a
hypocellular marrow with the presence of ring siderblasts suggestive of
recent drug/toxin exposure. The overall clinical picture was compatible
with drug-induced systemic toxicity involving the haematological
(disseminated intravascular coagulopathy, haemolytic anaemia), hepatic
and renal systems. He finally succumbed 4 weeks after admission. There
were 11 different herbal elements in the drug formula. One of the
elements, Zexie listed in Table 1, has been reported to have hepatotoxicity on liver and other organs in studies with mice in the Chinese literature.

Patient 4

A
45-year-old man (CKS) with HBeAg-positive CHB infection was first seen
4 months prior to admission. The liver biochemistry in the first clinic
visit was: albumin 42 g/L, bilirubin 27 μm,
ALT 70 U/L, AST 44 U/L, ALP 102 U/L and GGT 180 U/L. One month later,
he started to take a TCM formula named ‘Lingyang Qingfei Keli’ for its
‘cooling’ effect on the lungs and relief of pharyngeal discomfort. He
took 6 g three times daily till the onset of jaundice 3 months later.
Physical examination revealed ascites. The liver biochemistry was found
to be grossly deranged: albumin 24 g/L, bilirubin 288 μm,
ALT 414 U/L, AST 495 U/L, ALP 135 U/L and GGT 178 U/L. The prothrombin
time was 22.2 s. Ultrasonogram of abdomen showed small-size cirrhotic
liver, splenomegaly and ascites. The HBV-DNA level was low at
290 copies/mL (by Cobas Amplicor HBV Monitor test, Roche Diagnostics,
Branchburg, NJ, USA). Anti-HBc IgM was negative. The liver biochemistry
gradually improved after the cessation of the TCM and became static with
albumin level of 29 g/L, bilirubin level of 94 μm
and prothrombin time of 20 s after 3 months. The patient was
subsequently put on the waiting list for liver transplantation. There
were 12 elements identified in the TCM formula ‘Lingyang Qingfei Keli’.
Of these, three are known to be hepatotoxic (Table 1).
These include Zudahuang that is a processed Dahuang (mentioned in
patient 2), Gancao with its active ingredient, glycyrrhizin, being
potentially hepatotoxic on prolonged intake and Bohe (Mentha haplocalyx) a kind peppermint known to be hepatotoxic in Chinese literature. Pennyroyal oil derived from the same genus, Metha pulegium
contains several monoterpenes, principally pulegone, known to have
toxic effects on the liver and lungs. Several reports of hepatitis
because of pennyroyal oil have been reported.^22–25

Patient 5

A
46-year-old man (CCK) had known history of IgA nephropathy and
HBeAg-positive CHB infection followed up in the renal clinic for
8 years. The liver biochemistry was completely normal all along. He
complained of malaise and loss of appetite, and took two doses of a TCM
formulae prepared by his friend in China 2 days before the onset of
jaundice. The liver biochemistry on admission was: albumin 36 g/L,
bilirubin level 132 μm, ALT
level 1324 U/L, AST level 1599 U/L, ALP level 144 U/L and GGT level
165 U/L. The prothrombin time was mildly elevated to 18.1 s. HBV-DNA
level was 4.7 × 10⁵ copies/mL (measured by bDNA assay).
Anti-HBc IgM was negative. Ultrasonogram of abdomen revealed a small
liver with features suggestive of cirrhosis, and mild splenomegaly. The
formulae contained 11 elements in which Gancao (Table 1;
mentioned in patient 4) is likely the toxic agent for the
hepatotoxicity. The liver biochemistry gradually returned to normal
2 months later.

Patient 6

A
37-year-old man (STW) with anti-HBe-positive CHB infection had liver
biochemistry on presentation as follows: albumin level 49 g/L, bilirubin
24 μm, ALT 403 U/L, AST 204 U/L, ALP 87 U/L and GGT 23 U/L. The HBV-DNA level was 3.6 × 10⁵ copies/mL.
One month later, he was admitted because of a significant increase in
the ALT level to 851 U/L with a normal albumin level of 47 g/L and
bilirubin level of 19 μm. The prothrombin time was also normal. He had been taking Phyllanthus urinaria extracts for 5 months for its supposed anti-HBV effect. HBV-DNA level remained static at 2 × 10⁵ copies/mL.
Anti-HBc IgM was negative. Ultrasonogram of abdomen did not show
evidence of cirrhosis with normal liver and spleen size. The liver
biochemistry returned to normal 2 months after stopping of the drug. The
possibility of spontaneous HBV reactivation was unlikely because of the
relative low and static HBV-DNA levels, checked on two occasions
1 month apart. Though P. urinaria extracts have been reported to have anti-HBV effect,²⁶ meta-analysis disproves this.²⁷ Several studies using P. amarus (same genus but different species) have been conducted to determine its efficacy for treating CHB.^28–32 The findings are conflicting and inconsistent, but largely indicate the lack of efficacy for CHB.

Patient 7

A
38-year-old man (TKF) with CHB had regular follow-up in our clinic with
documented HBeAg seroconversion 6 years prior to admission. The liver
biochemistry remained completely normal until he presented with
tea-colour urine, malaise and right upper quadrant discomfort in 2004.
He took Ji Gu Cao (Abrus cantonienesis Hance) for 3 days 1 week
prior to admission. Physical examination revealed jaundice. The liver
biochemistry on admission was: albumin 39 g/L, bilirubin 197 μm,
ALT 3070 U/L, AST 1470 U/L, ALP 145 U/L and GGT 173 U/L. Prothrombin
time was 17.5 s. He was negative for anti-HBc IgM. The HBV-DNA level was
4.3 × 10⁵ copies/mL by the Digene Hybrid Capture II assay.
Ultrasonogram of abdomen did not show evidence of cirrhosis with normal
liver and spleen size. After reaching the peak bilirubin level of 404 μm
6 weeks after admission, the liver function became normal after
5 months. The possibility of spontaneous HBV exacerbation was unlikely
because of the persistent normal ALT levels for 6 years after stable
HBeAg seroconversion, relatively low HBV-DNA level and anti-HBc IgM
negativity on admission. Abrus cantonienesis Hance is prepared from the dried stems and roots of a plant named Abrus leguminosiae.
Contamination with the seeds of these plants, which contain a
toxoalbumin, Abrin, and a toxic glycoside, Abric acid, which are highly
toxic to the liver and kidney, may occur during the preparatory
procedures.

The clinical features and outcome of these seven patients are summarized in Table 2.

Table 2.   Summary of the clinical features of the seven patients with TCM-induced hepatotoxicity

Discussion

Of
the 45 admissions because of liver derangement in patients with CHB
infection, 15.6% were attributed to the TCM-induced hepatotoxicity. This
figure is likely to be an underestimation at first, because patients do
not always own up to the use of herbal medicines. Secondly, liver
derangement in CHB patients is often simply attributed to hepatitis B
exacerbation without detailed enquiry into the drug history. Thirdly,
even with a documented history of herbal intake, doctors are often
uncertain as to the culpability of the herbs because of the lack of
detailed information on Chinese herbs. It is also not always easy to
differentiate HBV exacerbation from herb-induced hepatotoxicity in CHB
patients. Detailed history of drugs or herbs intake including dietary
supplements to identify the offending elements, for example, those
listed in Table 1,
and tests for diagnosing acute exacerbation of HBV including HBV-DNA
levels and anti-HBc IgM may be of great value. Involvement of other
systemic organs in the initial stage of the disease as shown in patients
2 and 3 favours the diagnosis of drug-induced hepatotoxicity.

In
patients 6 and 7, the diagnosis of herb-induced hepatotoxicity was based
on the temporal relationship of the intake of herbs and the onset of
liver derangement together with the absence of convincing evidence of
acute exacerbation of the CHB, although specific liver toxic elements
were not found in the herbal formulae. There may be several reasons for
this. First, it is difficult to get the complete herbal formulae as
patients may not recall correctly or fail to disclose fully their herbal
intake. Secondly, impurities adulterating the herbs during the process
of preparation remain a concern because of the lack of quality control
and standardization during the preparatory procedure (as illustrated in
patient 1). Thirdly, the amount of the herbs taken by patients may far
exceed the safety limit, particularly in patients who take the TCM
continually as daily supplements. Fourthly, interactions between
different elements of the TCM and also between western medications may
occur when they are taken at the same time. Finally, the possibility of
synergistic hepatotoxicity of herbal preparation and HBV is largely
unknown.

In the present study, five patients (patients 1, 4, 5, 6
and 7) had marked elevation of ALT levels ranging from 414 to 3070 U/L
and all except patient 6 were followed by marked cholestasis. The
remaining two patients (patients 2 and 3) presented with marked
cholestasis (bilirubin levels of 450 and 484 μm
respectively) with relatively normal ALT levels (27 and 41 U/L
respectively). All seven patients had either normal or only mildly
elevated levels of ALP (range: 102–247 U/L) and GGT levels (range:
54–190 U/L). The predilection for elevation of parenchymal enzymes in
TCM-induced liver derangement makes it difficult to differentiate from
HBV exacerbation in CHB patients. Cases of TCM-induced hepatotoxicity
have been clinically confused with acute viral hepatitis.³³

Of
the seven CHB patients reported, three had hepatic decompensation
resulting in two deaths and one liver transplant. In patients with
pre-existing cirrhosis as in patient 3, any insult to the liver; however
mild, can easily lead to hepatic decompensation and mortality. In
addition, the threshold for hepatotoxicity of different elements of TCM
may be lower in CHB patients. Finally, some TCM may contain steroids and
immunosuppressives that will enhance the HBV replication leading to
severe HBV flares once they are taken off. To determine whether the
clinical outcome of CHB patients with TCM toxicity are more severe
compared with non-HBV subjects require further studies.

In patient
4, though the liver function gradually improved after cessation of the
offending TCM, it did not return to baseline level. The liver function
was irreversibly worsened after the added insult of the TCM. This
suggests that TCM can potentially accelerate the cirrhotic process in
patients with CHB infection.

In conclusion, several elements of
TCM had been identified to be associated with hepatotoxicity, often with
other systemic toxicity. TCM-related hepatotoxicity resulted in high
mortality/requirement for liver transplantation in patients with CHB
infection. Randomized-controlled trials with the same stringent criteria
as trials in western medicine are required to document the efficacies
and safety of various TCM before they can be advocated for the treatment
of diseases. In addition, strict quality control for the preparation of
TCM should be universally adopted to avoid adulteration with other
toxic elements.

Acknowledgement

This study was funded by a grant from the Hepatology Research Fund, The University of Hong Kong.