Académique Documents
Professionnel Documents
Culture Documents
H E PAT O L O G Y
Key words
acute hepatitis B, alpha-fetoprotein, chronic
hepatitis B, differential diagnosis, hepatitis B
virus.
Accepted for publication 24 February 2008.
Correspondence
Dr Yongnian Han, Research Unit of Liver
Disease, Shanghai No. 8 Peoples Hospital,
8 Caobao Road, Shanghai 200235, China.
Email: hanyn88@163.com
Abstract
Background and Aim: In areas with high or intermediate endemicity for chronic hepatitis
B virus (HBV) infection, it is difficult to distinguish acute hepatitis B (AHB) from chronic
hepatitis B with an acute flare (CHB-AF) in patients whose prior history of HBV infection
has been unknown. The present study aimed to screen laboratory parameters other than
immunoglobulin M antibody to hepatitis B core antigen (IgM anti-HBc) to discriminate
between the two conditions.
Methods: A retrospective and prospective study was conducted in patients first presenting
clinically as HBV-related acute hepatitis to sort out acute self-limited hepatitis B (ASLHB). Then, clinical and laboratory profiles were compared between patients with ASL-HB
and CHB-AF. Parameters closely associated with ASL-HB were chosen to evaluate sensitivity, specificity, accuracy, positive predictive values and negative predictive values for
diagnosing AHB.
Results: There were significant differences between patients with ASL-HB and CHB-AF
in relation to clinical and laboratory aspects, with many outstanding differences in levels of
serum HBV-DNA, hepatitis B e antigen (HBeAg) and alpha-fetoprotein (AFP) as well as
IgM anti-HBc. In particular, there was a greater difference between the two groups in low
levels of HBeAg (ratio of the optical density of the sample to the cut-off value [S/CO] <20)
than in negativity for HBeAg (42.7% and 13.5% vs 49.3% and 45.9%). 1:10 000 IgM
anti-HBc had a sensitivity and specificity of 96.2% and 93.1%, respectively, for predicting
ASL-HB. Combining it with AFP, HBeAg or HBV-DNA could improve diagnostic power.
A combination of IgM anti-HBc, HBV-DNA and HBeAg had a predictive value of 98.9%
and a negative predictive value of 100.0%, similar to that of a combination of IgM anti-HBc
and HBV-DNA. Adding AFP to the combinations of IgM anti-HBc and HBV-DNA or
HBeAg could further heighten the positive predictive value. The positive predictive value
and negative predictive value of the combination of IgM anti-HBc, HBV-DNA and AFP
were both 100.0%.
Conclusions: (i) There are significant differences with respect to clinical, biochemical,
immunological and virological aspects between ASL-HB and CHB-AF. (ii) Of several
diagnostic combinations, IgM anti-HBc jointing HBV-DNA is most effective and most
practicable in distinguishing ASL-HB from CHB-AF. (iii) A low HBeAg level is more
useful than negative HBeAg in differential diagnosis between ASL-HB and CHB-AF. (iv)
In those patients with a high level of IgM anti-HBc, serum AFP level >10 upper reference
limit could rule out a probability of ASL-HB.
Introduction
Hepatitis B virus (HBV) is a peculiar virus, leading to both an
acute infection and a chronic one. HBV-related acute hepatitis may
be a true episode of acute hepatitis B (AHB) or an acute flare of
chronic hepatitis B (CHB) hitherto unknown. It is important to
distinguish AHB patients, a great number of whom will have a
1728
Y Han et al.
Methods
Patients and study design
A retrospective investigation (from January 2003 to March 2005)
and a prospective study (from April 2005 to June 2006) were put
into practice in patients with their first attack of HBV-related acute
hepatitis whose status of prior HBV infection had been unknown.
HBV-related acute hepatitis was defined according to: (i) biochemical parameters: levels of serum alanine aminotransferase
(ALT) >10-fold the upper reference limit (URL), or total serum
bilirubin (TBil) >5 URL, and serum alkaline phosphatase (ALP)
<3 URL;11 and (ii) HBV markers: positive for serum hepatitis B
surface antigen (HBsAg), or serum antibodies to HBsAg, hepatitis
B e antigen (HBeAg), and hepatitis B core antigen (HBcAg).
Hepatitis A, C, D and E, and non-viral causes, such as drugs,
alcohol, pregnancy, ischemia, etc. that can lead to the same biochemical profiles, were all excluded. Patients who showed ultrasonographic evidence of cirrhosis were also excluded. They would
be examined retrospectively or followed up prospectively. Acute
self-limiting hepatitis B (ASL-HB) was defined as HBsAg clearing from serum after the disease remitted, or positivity for IgM
anti-HBc and for antibodies to HBsAg, HBeAg and HBcAg, with
infections with other non-hepatotropic viruses, such as human
cytomegalovirus (CMV), EpsteinBarr virus (EBV) and enteric
viruses being excluded. Alternatively, if serum HBsAg had
persisted for at least 12 months after the onset of clinically acute
hepatitis (HBsAg occasionally became undetectable until 1 year
after acute HBV infection,12 and one of our patients with ASL-HB
cleared HBsAg from serum at the 12th month after the onset of the
illness), the condition was classified as CHB.
Meanwhile, known chronically HBV-infected patients with the
same biochemical profiles as described above hospitalized during
the same period were named as CHB-AF, recruited as a control
group when other types of viral hepatitis and the above non-viral
causes were excluded, and compared with ASL-HB patients with
regard to clinical, biochemical, immunological and virological
profiles. Then, parameters with great differences and little overlap
in values were selected to assess their sensitivity, specificity,
accuracy, positive predictive value and negative predictive value
for diagnosing AHB or ASL-HB.
Data collection
Data were collected including clinical symptoms, physical signs
and laboratory examinations. The latter included: (i) biochemical
tests reflecting hepatocytic damage, such as serum TBil, ALT,
g-glutamyl transferase (GGT), prealbumin, all assayed by a colorimetric method (CSL Behring, King of Prussia, PA, USA), and
prothrombin time (PT), done as described by the manufacturers
instruction (Trinity Biotech, Wicklow, Ireland); (ii) HBV markers,
such as HBV antigens and antibodies, detected by commercially available enzyme immunoassays (Shanghai Kehua Bioengineering Co., Shanghai, China) and HBV-DNA, determined
by a fluorescent quantifying polymerase chain reaction (PCR)
method with a low limit of detection of 103 copies/mL (PG Biotech,
Shenzhen, China); (iii) alpha-fetoprotein (AFP), measured by
the electrochemiluminescence immunoassay (Roche Diagnostics,
Mannheim, Germany).
For the patients in the retrospective subgroup, HBV immunological markers were recorded on the basis of inpatient and outpatient information stored in our hepatic clinic, and outpatient
information in other hospitals acquired by questionnaire or telephone visits with copies of test reports being sent to us. Those who
had no results of HBV markers were called to our clinic to undergo
the examination. For patients in the prospective subgroup, HBV
markers were determined monthly during the first 6 months or
bimonthly during the next 6 months.
1729
Y Han et al.
was less than 5, Fishers exact test was used for the data suitable
for a fourfold table and MannWhitneys U-test was used for those
unsuitable for a fourfold table. All tests were two-sided and all
analyses were carried out with the SPSS version 11.5 software
package (SPSS Inc., Chicago, IL, USA).
Sensitivity, specificity, accuracy, positive predictive values, and
negative predictive values of each or combinations of parameters
screened out were calculated, respectively.
Results
Outcomes of HBV-related acute hepatitis
Of the 219 patients with HBV-related acute hepatitis, 27 of 141
in the retrospective subgroup were not located, but none of 78 in
the follow-up subgroup was lost. Of 192 patients studied, 138 were
confirmed to have ASL-HB, diagnosed by the disappearance of
serum HBsAg, whereas the remaining 54 were shown to remain
HBsAg positive after remission of the disease, diagnosed as
having CHB, including CHB to which AHB evolved, and real
CHB with acute exacerbation that had not been found before,
which is a common phenomenon in China.
ASL-HB
(n = 138)
36.6 10.2
113/25
120 (87.0)
18 (13.0)
23
109
102
69
34
16
29
95
(16.7)*
(79.0)*
(73.9)
(50.0)
(24.6)*
(11.6)
(21.0)
(68.8)
129 (93.5)**
18 (13.1)
24 (17.5)*
1730
Table 2
Variable
TBil
Normal (%)
Elevated (mmol/L):
n/mean SD
ALT (URL)
GGT
Normal (%)
Elevated (URL):
n/mean SD
PT
Normal (%)
Elevated (INR):
n/mean SD
AFP
Normal (%)
Elevated (ng/mL):
n/mean SD
Prealbumin (mg/mL)
CHB-AF (n = 133)
ASL-HB (n = 138)
15 (11.3)
118/68.1 71.2
3 (2.2)*
135/109.1 82.1**
18.2 9.4
32.4 19.9**
9 (6.8)
124/3.1 2.1
5 (3.6)
133/4.1 3.7*
105 (78.9)
28/2.0 1.0
127 (92.0)*
11/2.3 1.4
49 (36.8)
84/171.7 328.2
102 (73.9)**
36/13.9 8.3**
153.6 88.7
199.4 106**
Y Han et al.
Table 3
ASL-HB
Test was not done in seven patients with ASL-HB and in three patients
with CHB-AF in the retrospective group.
anti-HBe, antibody to hepatitis B e antigen; IgM anti-HBc, immunoglobulin M antibody to hepatitis B core antigen; ASL-HB, acute self-limited
hepatitis B; CHB-AF, chronic hepatitis B with acute flare; HBeAg, hepatitis B e antigen; HBsAg, hepatitis B surface antigen; HBV, hepatitis B
virus; PCR, polymerase chain reaction; S/CO, ratio of optical density of
samples read at 450 nm to cut-off value.
Discussion
Diagnostic power of laboratory parameters
According to the above analyses, four parameters, IgM anti-HBc,
HBeAg, HBV-DNA and AFP, all with great differences and little
1731
Table 4
Y Han et al.
Laboratory tests
Sensitivity
Specificity
Accuracy
Positive
predictive
value
Negative
predictive
value
96.2
75.4
86.6
100.0
100.0
98.9
100.0
100.0
100.0
93.1
79.2
57.1
89.4
90.7
99.0
97.9
89.5
100.0
94.6
77.0
66.8
97.1
97.0
99.0
99.3
98.2
100.0
93.3
78.4
63.8
96.2
95.8
98.9
98.9
98.3
100.0
96.0
75.7
75.0
100.0
100.0
99.0
100.0
100.0
100.0
sion for HBV in China.14,15 Thus, we have been taught that AHB is
rare in China after HBV infection has prevailed for centuries. The
lifetime risk of HBV infection, as judged by the peak prevalence of
HBV markers in elderly people, is approximately 80% in China,
but the rate of chronic HBV carriage is only 8.511.1%,14 suggesting that a great majority of HBV-infected individuals have spontaneously cleared the HBV, particularly in adults. Taking this into
consideration, we paid special attention to the follow-up care of
adult patients with a first episode of HBV-related acute hepatitis
whose previous history of HBV infection had not been known.
During the follow-up period, most of them cleared HBV from their
serum and were confirmed as ASL-HB. Sequentially, we located
and retrospectively investigated the patients who were admitted
with signs, symptoms and biochemical tests suggestive of acute
hepatitis, and were diagnosed as CHB due to evidence of HBV
infection when they were discharged from our hospital from
January 2003 to March 2005. A prospective arm on similar patients was conducted simultaneously. As we suspected, sporadic
ASL-HB is not rare but common now in China, opposite to what
has generally been accepted. So ensues a practical problem of
distinguishing AHB from CHB-AF. Here, we first compared
ASL-HB patients with CHB-AF patients in relation to clinical,
biochemical, immunological and virological aspects, and then
sought the methods to discriminate between the two diseases.
Comparison of patients with ASL-HB and CHB-AF had shown
that ASL-HB patients were more symptomatic, had a more severe
necroinflammatory liver but a more functionally compensated
liver. This finding could be explained by the fact that clinically
acute hepatitis occurred in a healthy and injured liver, respectively,
in the two conditions. Unfortunately, none of these biochemical
parameters at initial presentation could differentiate between
patients with ASL-HB and CHB-AF, which is in keeping with
Kumar et al.s observation.7
The cause of elevated AFP in patients with non-tumor liver
disease is unclear. Our findings that the peak production of a great
majority of the two groups of patients with elevated AFP levels
was observed at early phase of the disease, when liver damage was
the severest, but not at the convalescent phase (data not shown),
when hepatocytic regeneration occurred, suggest that elevations of
serum AFP in acute and chronic liver diseases may not be due to
1732
subsequent hepatocyte regeneration induced by hepatic inflammation. The literature on the mechanism of serum AFP elevation in
AHB patients is not available at the present. Longitudinal studies
showed that elevations of serum AFP levels at baseline in CHB
patients confirmed by liver biopsy had been proven to be associated with a higher risk of decompensated cirrhosis, hepatocellular
carcinoma (HCC),16,17 implying that patients with elevated serum
AFP had more advanced liver disease than did those with normal
levels. Very high levels of serum AFP suggestive of the possibility
of HCC were occasionally found in patients with chronic HBV
infection, especially those with cirrhosis, but no occurrence of
HCC.18,19 In the setting of chronic hepatitis C, the mean serum
AFP value was significantly greater in patients with more marked
fibrosis.20 All these observations put forward the hypothesis that
marked fibrosis or cirrhosis, a state of significant altered hepatocyte architecture, may be the underlying cause of increased serum
AFP and, just at the presence of fibrosis or cirrhosis, hepatocyte
necroinflammation can trigger elevations of AFP. This can explain
why AFP elevations have frequently been found in CHB patients,
with remarkable AFP elevations being associated with exacerbations of the underlying liver disease,16,21 whereas normal AFP
levels are found in a great majority of patients with ASL-HB and
only low levels of AFP in a minority of them, although they had
more severe liver necroinflammation.
As shown by other researchers,7,10 ASL-HB patients had a
higher level of serum IgM anti-HBc, lower levels of serum HBVDNA, HBeAg and HBsAg than those with CHB-AF. These phenomena can be explained by a rapid clearance of serum HBVDNA as a result of a coordinated response of innate and adaptive,
humoral and cellular immune systems in AHB.15,22 The reason why
the proportion of patients with negative HBeAg in the two groups
is similar (49.3% vs 45.9%) is that some CHB patients are infected
with HBeAg-negative variants of HBV. The finding that a low level
of serum HBeAg was observed in more patients with ASL-HB
than in patients with CHB-AF suggests that a low HBeAg level is
more useful than negative HBeAg in the differential diagnosis
between ASL-HB and CHB-AF.
Very important differences in HBV-DNA, HBeAg and AFP, as
well as IgM anti-HBc, between the two groups have been observed
in our study. When used singly, the diagnostic power for ASL-HB
Y Han et al.
References
1 Lok ASF, McMahon BJ. Chronic hepatitis B. Hepatology 2007; 45:
50739.
2 Dufour DR, Lott JA, Nolte FS, Gretch DR, Koff RS, Seeff LB.
Diagnosis and monitoring of hepatic injury. II. Recommendations for
use of laboratory tests in screening, diagnosis, and monitoring. Clin.
Chem. 2000; 46: 205068.
3 Decker RH. Diagnosis of acute and chronic hepatitis B. In:
Zuckerman AJ, Thomas HC, eds. Viral Hepatitis. Singapore:
Harcourt Asia, 2000; 20116.
4 Lemon SM, Gates NL, Simms TE, Bancroft WH. IgM antibody to
hepatitis B core antigen as a diagnostic parameter of acute infection
with hepatitis B virus. J. Infect. Dis. 1981; 143: 8039.
5 Paraevangelou G, Poumeliotou-Karayannis A, Tassopoulos N,
Stathopoulou P. Diagnostic value of anti-HBc IgM in high HBV
prevalence area. J. Med. Virol. 1984; 13: 3939.
1733