BioMed Central
BMC Gastroenterology
Open Access
Research article
The accuracy of the report of hepatic steatosis on ultrasonography
in patients infected with hepatitis C in a clinical setting: A
retrospective observational study
Matthew J Hepburn*1, Jeffrey A Vos1, Eric P Fillman2 and Eric J Lawitz1
Address: 1Departments of Medicine Brooke Army Medical Center, Fort Sam Houston Texas, USA and 2Department of Pathology, Brooke Army
Medical Center, Fort Sam Houston Texas, USA
Email: Matthew J Hepburn* - matthew.hepburn@us.army.mil; Jeffrey A Vos - jeffrey.vos@amedd.army.mil;
Eric P Fillman - eric.fillman@amedd.army.mil; Eric J Lawitz - lawitz@alamomedicalresearch.com
* Corresponding author
Published: 13 April 2005
Received: 22 November 2004
Accepted: 13 April 2005
BMC Gastroenterology 2005, 5:14
doi:10.1186/1471-230X-5-14
This article is available from: http://www.biomedcentral.com/1471-230X/5/14
© 2005 Hepburn et al; licensee BioMed Central Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abstract
Background: Steatosis is occasionally reported during screening ultrasonography in patients with
hepatitis C virus (HCV). We conducted a retrospective observational study to assess the factors
associated with steatosis on ultrasonography and the relationship between steatosis on ultrasound
versus biopsy in patients infected with HCV in a clinical setting. Our hypothesis was
ultrasonography would perform poorly for the detection of steatosis outside of the context of a
controlled study, primarily due to false-positive results caused by hepatic fibrosis and inflammation.
Methods: A retrospective review of ultrasound reports was conducted on patients infected with
HCV in a tertiary care gastroenterology clinic. Reports were reviewed for the specific
documentation of the presence of steatosis. Baseline clinical and histologic parameters were
recorded, and compared for patients with vs. without steatosis. Multiple logistic regression analysis
was performed on these baseline variables. Liver biopsies were reviewed by two pathologists, and
graded for steatosis. Steatosis on biopsy was compared to steatosis on ultrasound report, and the
performance characteristics of ultrasonography were calculated, using biopsy as the gold standard.
Results: Ultrasound reports were available on 164 patients. Patients with steatosis on ultrasound
had a higher incidence of the following parameters compared to patients without steatosis: diabetes
(12/49 [24%] vs. 7/115 [6%], p < 0.001), fibrosis stage >2 (15/48 [31%] vs. 16/110 [15%], p = 0.02),
histologic grade >2 (19/48 [40%] vs. 17/103 [17%], p = 0.002), and ALT (129.5 ± 89.0 IU/L vs. 94.3
± 87.0 IU/L, p = 0.01). Histologic grade was the only factor independently associated with steatosis
with multivariate analysis. When compared to the histologic diagnosis of steatosis (n = 122),
ultrasonography had a substantial number of false-positive and false-negative results. In patients
with a normal ultrasound, 8/82 (10%) had >30% steatosis on biopsy. Among patients with steatosis
reported on ultrasound, only 12/40 (30%) had >30% steatosis on biopsy review.
Conclusion: Steatosis on ultrasound is associated with markers of inflammation and fibrosis in
HCV-infected patients, but does not consistently correlate with steatosis on biopsy outside of the
context of a controlled study. Clinicians should be skeptical of the definitive diagnosis of steatosis
on hepatic ultrasonography.
Page 1 of 8
(page number not for citation purposes)

---

BMC Gastroenterology 2005, 5:14
http://www.biomedcentral.com/1471-230X/5/14
Background
ciated with radiology reports of steatosis on ultrasound in
Hepatic steatosis is commonly detected on both radio-
HCV-infected patients and to describe the association
logic and histologic examination in patients infected with
between steatosis on ultrasound and steatosis on liver
the hepatitis C virus (HCV). In particular, steatosis is often
biopsy. We specifically wanted to clarify the implications
observed in genotype 3 infection [1]. The underlying
of the identification of steatosis on an ultrasound report
mechanism of steatosis in hepatitis C is not completely
for the practicing clinician. Our purpose was not to assess
understood, and is most likely multifactorial. Animal
the performance characteristics of hepatic ultrasonogra-
models [2,3] and in vitro experimentation [4] suggest that
phy under controlled conditions. Rather, we were inter-
a virologic effect induces steatosis. Clinical observational
ested in the utility of these reports in routine clinical
data suggest that a possible direct virologic effect occurs
practice, with different radiologists over a three-year
with hepatitis C genotype 3 infection [1,5]. However, stea-
period. The first part of our analysis was to examine all
tosis in genotype 1 and 2 HCV infections is more closely
patients who had hepatic ultrasonography performed,
related to risk factors that are known to be associated with
including patients who did not have a liver biopsy. The
non-alcoholic steatohepatitis (i.e. diabetes mellitus and
second part of our analysis (comparing steatosis on ultra-
obesity)[5,6]. The assessment of steatosis may have clini-
sound report to steatosis on biopsy) included only
cal relevance, as some studies have suggested that the
patients who had hepatic ultrasonography and liver
detection of steatosis can be an independent predictor of
biopsy. We observed that ultrasound reports of hepatic
response to therapy [7].
steatosis were particularly associated with histologic
inflammation, as well as fibrosis, and the sensitivity and
Clinicians routinely order hepatic ultrasonography on
specificity of steatosis on ultrasound was poor, when
patients with hepatitis C as an initial screening test for
compared to steatosis on biopsy.
hepatocellular carcinoma or anatomic abnormality. In
ordering this test, the radiologist may also detect and
Methods
mention steatosis. However, the diagnosis of steatosis on
Patients
ultrasound can be problematic, as the ultrasonic appear-
A retrospective review was performed of all patients who
ance of steatosis may be that of a 'bright liver,' which can
underwent screening hepatic ultrasonography seen in the
also be observed with hepatic fibrosis and/or inflamma-
hepatitis research clinic at our tertiary care, military aca-
tion [8]. Even though these overlap syndromes exist, we
demic medical center over a three-year period. Patients
observed a number of hepatic ultrasonography reports
were those eligible for care at a military treatment facility,
identifying steatosis in our clinical practice, as opposed to
including active duty military personnel, their spouses,
echogenic or 'bright liver'.
and military retirees (serving 20 or more years on active
duty). All patients had hepatitis C infection, confirmed by
The sensitivity and specificity of ultrasonography for the
serum HCV RNA PCR testing. Patients were referred from
detection of steatosis may be very high in the hands of an
primary care clinics at our institution, and military pri-
expert radiologist who consistently applies a particular cri-
mary care clinics throughout the region (Texas, Louisiana,
teria. However, many clinicians in the United States
Oklahoma, Kansas, Missouri and Colorado) for further
employ the services of a radiology department or group,
assessment of their hepatitis C infection and for discus-
in which many radiologists with varying levels of experi-
sion of treatment options. The Institutional Review Board
ence actually interpret the images. We were interested in
at Brooke Army Medical Center approved the study.
the utility of a report of steatosis in this context.
Demographic and Clinical Characteristics: Age, gender,
Only one prior study has compared steatosis on ultra-
body mass index (BMI) and ethnicity were recorded on
sound to histology in HCV-infected patients [9], and no
each patient and utilized for analysis. The following labo-
correlation was observed between histologic and radio-
ratory parameters were noted: alanine aminotransferase
logic results. However, the conclusions of this study are
(ALT), alpha-fetoprotein (AFP), and HCV genotype.
limited since the diagnosis of HCV was based on the
Patients were considered to have diabetes if they were tak-
detection of HCV antibodies, rather than a determination
ing medication for diabetes treatment. Liver biopsy
of viremia with polymerase chain reaction (PCR) technol-
reports included a Metavir stage (0=no fibrosis to 4=cir-
ogy. Additionally, this study was not performed in the
rhosis) [10] and histologic grade (0–4, with 0=no inflam-
context of clinical practice.
mation to 4=severe inflammation), which were also
included in the analysis. Each of these parameters was
We conducted a retrospective observational study to
converted into a binomial variable, with 0–2 and 3–4 as
assess the radiology reports of hepatic ultrasonography of
the resultant two categories.
HCV-infected patients. Our purpose was to determine
demographic, clinical and laboratory characteristics asso-
Page 2 of 8
(page number not for citation purposes)

---

BMC Gastroenterology 2005, 5:14
http://www.biomedcentral.com/1471-230X/5/14
Ultrasonography
3) As a binomial, comparing significant steatosis (>30%)
All patients underwent hepatic ultrasonography during
vs. not significant steatosis (<30%).
their initial evaluation if they did not have an ultrasound
report from a prior assessment. Computerized records
Analysis
were reviewed for the ultrasonography reports. Ultra-
Descriptions of the 'operative characteristics' (sensitivity,
sonography was primarily performed at Brooke Army
specificity, positive predictive value, negative predictive
Medical Center, the same institution as our hepatitis
value) of steatosis on ultrasound in our clinical setting
research clinic. The machines utilized in the Radiology
were calculated, with the assumption that the gold stand-
Department were the ATL Ultramark HDI 3000® Ultra-
ard for the diagnosis of hepatic steatosis was histologic
sound System and the ATL Ultramark HDI 5000® Ultra-
examination. These calculations were performed for both
sound System. A small number of patients in the hepatitis
the determination of any steatosis and significant steatosis
research clinic (<10%) are referred from Wilford Hall Air
on histologic examination. Any steatosis was defined as
Force Medical Center, also located in San Antonio, Texas.
more than 2% steatosis on biopsy, while significant stea-
If the patient had their ultrasound performed at the Air
tosis was defined as >30% steatosis on biopsy. Prior liter-
Force hospital, they were included in the analysis. The
ature suggests that ultrasonography performs well in the
Radiology Departments at these two institutions (Brooke
diagnosis of significant steatosis [11].
Army Medical Center and Wilford Hall Air Force Medical
Center) maintain a combined accredited teaching pro-
Demographic, clinical and histologic characteristics were
gram for radiology residents. These reports were scored as
compared between patients with steatosis on biopsy ver-
a binomial variable. If the ultrasound report mentioned
sus patients without steatosis on biopsy. For continuous
steatosis as a finding, it was designated positive. If the
variables, an independent sample t-test was utilized if
ultrasound report did not mention steatosis, it was
their distribution was normal. For continuous variables in
labeled negative. Equivocal studies containing phrases
which the distribution of values was not normal, the
such as "possible steatosis" or "inflammation or steatosis"
Mann-Whitney rank sum test was used. For categorical
were excluded from the final analysis. Reports did not dis-
variables, a Pearson's chi-square test was employed. To
tinguish between diffuse or focal fatty liver.
determine the association between steatosis on ultra-
sound and steatosis on biopsy, a Spearman's rank correla-
Histologic Examination
tion coefficient was calculated. A kappa-statistic was
A liver biopsy was offered to patients during their initial
calculated to assess interobserver variability between stea-
evaluation in order to assess the patient's severity of dis-
tosis percentages observed by the two pathologists. Agree-
ease. Liver biopsies were generally performed within 1–2
ment between pathologists was determined to be +/- 5%
months of hepatic ultrasound. There were no liver biop-
for the liver biopsies examined.
sies in this study utilized that were > 6 months after ultra-
sonography. The initial histologic examination of the
A multiple logistic regression model, with stepwise back-
biopsy focused on the histologic stage (fibrosis) and grade
ward elimination of non-significant variables, was devel-
(inflammation). For our study, two pathologists (J.A.V.
oped to determine factors that were associated with the
and E.P.F.) retrospectively reviewed these biopsies and a
detection of steatosis on ultrasound. Steatosis on ultra-
percentage of steatosis was assigned to each biopsy speci-
sound was the dependent variable, with the independent
men. Any steatosis was denoted as 1%, while a complete
variables including: age, gender, histologic stage (as a
absence of steatosis was classified as 0%. If steatosis was
binomial variable), histologic grade (as a binomial varia-
considered >2%, it was recorded in 5% increments. These
ble), ALT, genotype (genotype 3 compared to all other
pathologists were blinded to ultrasound results, as well as
genotypes), body mass index, and a history of diabetes.
any clinical characteristics of the patients. In order to
All variables were included in the model.
explore the relationship between the report of steatosis on
ultrasound compared to biopsy, the biopsies were catego-
Results
rized three different ways:
Pre-treatment hepatic ultrasound reports were available
on 171 patients. Seven reports were excluded because of
1) grouped into 5 categories (based on the percentage of
equivocal descriptions (see Methods section, reports of
steatosis on biopsy): 0–2%, 2–10%, 10–30%, 30–60%,
"inflammation or fatty infiltration," or "steatosis vs. fibro-
and >60%.
sis" were excluded). A total of 49/164 (30%) of patients
had steatosis documented in the ultrasound report. The
2) As a binomial, comparing no steatosis (<2%) vs. any
comparison of baseline and clinical characteristics
steatosis (>2%).
between patients with versus without steatosis on ultra-
sound indicated significant differences in ALT and AFP, a
Page 3 of 8
(page number not for citation purposes)

---

BMC Gastroenterology 2005, 5:14
http://www.biomedcentral.com/1471-230X/5/14
Table 1: Comparison of demographic, clinical and laboratory characteristics of patients with versus without steatosis on hepatic
ultrasound. Legend: ALT- alanine aminotransferase; AFP- alpha fetoprotein; BMI- body mass index.
Parameter
Steatosis
No Steatosis
p value
Gender (% male)
31/49 (63%)
77/115 (67%)
0.68
Genotype (3 vs. others)
4/46 (9%)
5/110 (5%)
0.31
Diabetes (% with DM)
12/49 (24%)
7/115 (6%)
0.001
Age
47.9 ± 10.3
46.3 ± 10.3
0.38
ALT (IU/L)
129.5 ± 89.0
94.3 ± 87.0
0.01
AFP (ng/ml)
13.1 ± 25.5
8.0 ± 12.3
0.03
BMI (kg/m2)
28.6 ± 5.1
27.4 ± 4.8
0.17
Stage (% stage 3–4)
15/48 (31%)
16/110 (15%)
0.02
Grade (% stage 3–4)*
19/48 (40%)
17/103 (17%)
0.002
*Seven histology reports scored only stage, and did not report a histologic grade.
history of diabetes mellitus, and advanced histologic stage
A total of 136 patients were included in the multivariate
and grade (Table 1).
logistic regression model to determine factors associated
with steatosis on ultrasound (see Methods for variables
Liver biopsy specimens were available from 122 patients
included in the model). The other 28 patients were not
for our pathologists to assess the amount of steatosis. The
included due to a missing variable (genotype not per-
remaining 42 patients either did not have an initial liver
formed, liver biopsy not performed, etc). The only statis-
biopsy (n = 6) or their liver biopsy slides were not availa-
tically significant factor associated with steatosis on
ble at our institution (n = 36), as they had been returned
ultrasound was histologic grade (OR 3.6, 95% CI 1.3–
to the patient's referring institution. The kappa statistic for
9.8). A history of diabetes mellitus approached statistical
interobserver reliability of steatosis between the two
significance (OR 3.8, 95% CI 0.96–14.70). No other fac-
pathologists was calculated to be 0.87. A relationship
tors were independently associated with steatosis on
between steatosis on ultrasound and steatosis on biopsy
ultrasound.
was detected (Figure 1). The Spearman's rank correlation
coefficient was rs = 0.27 (p = 0.003, n = 122). Of the 122
Discussion
patients that had both histologic review and ultrasound,
The detection of steatosis on ultrasound in a clinical set-
40 (33%) had steatosis noted on their ultrasound report.
ting appears to be generally associated with steatosis on
A substantial number of false-positive and false-negative
biopsy, but also with hepatic inflammation and fibrosis.
results were observed. For example, 10% (8/82) of
In particular, steatosis on ultrasound was independently
patients with a normal ultrasound had >30% steatosis on
associated with moderate-severe histologic inflammation.
biopsy. Among patients with steatosis reported on ultra-
The ability of ultrasonography to accurately detect hepatic
sound, only 12/40 (30%) had >30% steatosis on biopsy
steatosis is questionable outside of a controlled research
review.
setting, as both the sensitivity and specificity of this imag-
ing technique were unacceptably low in our study. An
We determined the test characteristics for steatosis on
ultrasound report of steatosis was as predictive of fibrosis
ultrasound (in the clinical setting of our study) compared
as it was predictive of >30% steatosis, which reflects the
to biopsy as the gold standard, which were calculated for
unreliability of this imaging modality in differentiating
the detection of any steatosis (>2%), or for significant
fibrosis and steatosis. Based on the results of our study,
steatosis (>30%) (Table 2). For either of these calcula-
the clinician should understand that an ultrasound report
tions, the sensitivity and specificity of ultrasonography
of steatosis could mean the patient has fibrosis, inflam-
were not high. We also examined whether or not the diag-
mation, significant steatosis, or a combination of these
nosis of steatosis on ultrasound was more useful in detect-
conditions. Alternatively, the patient may have none of
ing substantial fibrosis (Metavir stage 3–4) on biopsy. The
these pathologic findings. Practitioners should rely on
sensitivity of steatosis on ultrasound predicting fibrosis
other diagnostic modalities to assess the liver for steatosis.
on biopsy was 48%, while the specificity was 74%. These
Magnetic resonance imaging or computerized tomogra-
performance characteristics for assessing fibrosis were very
phy are two potential techniques that may be clinically
similar to the effectiveness of ultrasonography to correctly
useful in the diagnosis of hepatic steatosis, pending fur-
identify significant (>30%) steatosis on biopsy.
ther study [12].
Page 4 of 8
(page number not for citation purposes)

---

BMC Gastroenterology 2005, 5:14
http://www.biomedcentral.com/1471-230X/5/14
80
on
70
60
teatosis
S
50
ith
w
40
Ultrasound 30
ubjects
20
S
%
10
0
0-2%
2-10%
10-30%
30-60%
>60%
(n=57)
(n=37)
(n=8)
(n=10)
(n=10)
% Steatosis on Biopsy Categories
Comparis
Figure 1 on of steatosis on biopsy versus ultrasound
Comparison of steatosis on biopsy versus ultrasound. For each of the five categories of amount of steatosis on biopsy,
the percentage of patients in those categories with a positive ultrasound for steatosis is displayed. rs = 0.27, p = 0.003.
Table 2: Performance characteristics of the detection of steatosis on ultrasound (n = 122). PPV = positive predictive value;
NPV=negative predictive value. Gold standard was steatosis on histologic examination.
Histologic Examination
Sensitivity
Specificity
PPV
NPV
Any Steatosis (>2%)
43%
79%
70%
55%
Steatosis >30%
60%
73%
30%
90%
Hepatic steatosis, detected by histologic examination,
hypertriglyceridemia [13]. We observed that steatosis on
appears to have a multifactorial etiology in patients
ultrasonography was associated with factors representa-
infected with hepatitis C. Steatosis may be a result of a
tive of inflammation (histologic grade on biopsy and
direct virologic effect, particularly in patients infected with
ALT) and fibrosis (histologic stage on biopsy and alpha-
genotype 3 [1,5]. Additionally, steatosis in HCV-infected
fetoprotein levels [15]) as well as diabetes mellitus. We
patients may be associated with accepted steatosis risk fac-
did not observe an obvious association between genotype
tors, including obesity [6,13,14], diabetes mellitus [6] and
or body mass index with steatosis on ultrasound. Lack of
Page 5 of 8
(page number not for citation purposes)

---

BMC Gastroenterology 2005, 5:14
http://www.biomedcentral.com/1471-230X/5/14
association of genotype and BMI with steatosis in our
'bright liver', instead of definitively reporting the observa-
study is likely related to factors confounding the diagnosis
tion of steatosis.
of steatosis on ultrasound, such as the additional presence
of fibrosis and inflammation.
For the clinician, ultrasound could conceivably be utilized
in the documentation of an echogenic liver only. Simi-
Increased echogenicity is the characteristic ultrasono-
larly, any report of steatosis, fibrosis, or inflammation
graphic finding that identifies hepatic steatosis. The
could be understood by the clinician as consistent with
increased echogenicity is often compared to the spleen
the presence of liver pathology. However, we also
and kidney [16,17]. A loss of definition of the hemi-dia-
observed false negatives in our study, in which patients
phragm and decreased detail of the intrahepatic architec-
with significant steatosis had normal ultrasound reports.
ture (particularly the portal veins) may be supportive
Additionally, the prognostic significance for clinical
findings [11]. The performance characteristics of ultra-
course and response to therapy may be very different for
sonography in the detection of steatosis vary considerably
steatosis compared to fibrosis, and therefore the signifi-
among studies. In the examination of patients with
cance of an echogenic liver on ultrasound may vary sub-
mostly alcoholic liver disease, a specificity of 84% and
stantially between patients.
sensitivity of 94% for the detection of steatosis on ultra-
sound was described [18]. Using multiple criteria to diag-
Certain limitations of the study should be mentioned. The
nose steatosis, positive predictive values can be as high as
patients in this study had been referred for specialty eval-
94% in high-prevalence populations [19]. Performance
uation in a hepatitis C clinic in which therapeutic clinical
characteristics tend to improve with the diagnosis of mod-
trials are emphasized. These patients may be different
erate and severe steatosis [11]. One study suggested that
than the general population with HCV infection, limiting
33% steatosis seen on biopsy was an optimal threshold
the generalizability of our results. Selection bias that is
for the radiographic detection of steatosis [20].
inherent in retrospective studies may limit the applicabil-
ity of our results. Also, for some patients, there was time
However, concomitant liver pathology may complicate
between the dates in which the ultrasound and biopsy
the diagnosis of steatosis on ultrasound. Echogenicity on
were performed. It is possible that small changes in liver
ultrasound may be consistent with either fibrosis or stea-
histology could have occurred, if the patient had signifi-
tosis, and ultrasonography may not effectively differenti-
cant weight gain or loss, for example. Ethanol consump-
ate between these two conditions [8]. Due to the overlap
tion was not measured in our study, which could have
in appearance of fibrosis and steatosis, some radiologists
impacted some of the baseline variables which were com-
opt to utilize the terms "fatty-fibrotic" [16,21] or "steatofi-
pared between patients with or without steatosis on
brosis" [22] when describing this echogenic pattern.
ultrasound.
Fibrosis has been demonstrated to be independently asso-
ciated with steatosis in hepatitis C patients [6,14,23].
The retrospective nature of this study may introduce
Some authors do not recommend the use of ultrasound as
potential bias in data collection which could limit the
a screening tool for hepatic steatosis due to questions
clinical applicability of the findings. For example, some of
regarding sensitivity and specificity of this test [24].
the ultrasounds were not performed within the same 1–2
weeks as the liver biopsy. In addition, incorporating the
Our results suggest that ultrasound is an unreliable predic-
radiologic interpretations of multiple radiologists (in con-
tor of steatosis when described on a routine ultrasound
trast to a single radiologist) has certain advantages and
report in HCV-infected patients. These findings are con-
disadvantages. While multiple radiologists potentially
sistent with a prior study of hepatic steatosis which docu-
introduce significant variability in interpretation, they
mented no correlation between biopsy and
more accurately simulate a realistic clinical scenario as
ultrasonography in 64 patients with a positive HCV anti-
opposed to the artificial framework of a study utilizing a
body [9]. Our results differed slightly from this previous
single expert radiologist. Since our study focused on the
report in that some correlation between ultrasound and
utility of ultrasonography results for clinicians, examining
biopsy was demonstrated (rs = 0.27). However, both false-
this question in the context of multiple radiologists over
positives and false-negatives were observed in our study.
time seemed most appropriate. A similar review in other
A report of steatosis was equally likely to indicate signifi-
clinical systems, such as medical systems from other coun-
cant steatosis or fibrosis. Therefore, we conclude that hep-
tries, would be useful. It is possible that in other systems,
atitis C produces significant liver pathology that may
radiology do not attempt to define steatosis, but rather
confound the diagnosis of steatosis on liver ultrasound in
only remark on the presence and degree of an echogenic
a clinical setting. It may be advisable for radiologists to
liver. It would also be interesting to survey radiologists to
report 'echogenic liver', 'possible steatosis vs. fibrosis' or
attempt to assess the range of techniques and criteria for
the assessment of steatosis on ultrasonography.
Page 6 of 8
(page number not for citation purposes)

---

BMC Gastroenterology 2005, 5:14
http://www.biomedcentral.com/1471-230X/5/14
Additionally, liver biopsy itself is not always a definitive
Authors' contributions
test, as sampling error can occur with this procedure. Dif-
MH and EL developed the initial concept of the study. All
ferences in the assessment of histologic stage and grade
authors participated in study design and data analysis. JV
have been observed in patients with hepatitis C who have
and EF conducted the histologic review of the slides. MH
undergone simultaneous liver biopsies in different lobes
conducted the review of ultrasound reports and organiza-
[25]. This error may account for some of the false-positive
tion of the data. All authors contributed to the manuscript
ultrasound reports in the assessment of steatosis in our
preparation and completion.
study.
Acknowledgements
Conclusion
The opinions or assertions contained herein are those of the authors and
Unfortunately, routine hepatic ultrasonography does not
are not to be construed as official policy or as reflecting the views of the
provide an accurate non-invasive mechanism for the diag-
Department of the Army or the Department of Defense. The authors are
nosis of hepatic steatosis in HCV-infected patients in the
employees of the U.S. Government. This work was prepared as part of
their official duties and, as such, there is no copyright to be transferred.
clinical context of our study. These findings should be
This study was presented in part as a podium presentation at Digestive Dis-
examined in other clinical settings, perhaps in other coun-
eases Week, Orlando FL, May 19–22, 2003.
tries. Clinicians should interpret a report of steatosis on
ultrasound with caution, and also consider that this report
References
could suggest a combination of inflammation, steatosis
1.
Ramalho F: Hepatitis C virus infection and liver steatosis. Anti-
and/or fibrosis. Even patients who did not have a liver
viral Res 2003, 60:125-7.
2.
Moriya K, Yotsuyanagi H, Shintani Y, Fujie H, Ishibashi K, Matsuura Y,
biopsy were included in the analysis of associations with
Miyamura T, Koike K.: Hepatitis C core protein induces hepatic
steatosis, and the description of steatosis on ultrasound
steatosis in transgenic mice. J Gen Virol 1997, 78:1527-1531.
was associated with factors reflective of hepatic inflamma-
3.
Perlemuter G, Sabile A, Letteron P, Vona G, Topilco A, Chretien Y,
Koike K, Pessayre D, Chapman J, Barba G, Brechot C: Hepatitis C
tion (such as ALT). Conversely, lack of ultrasonographic
virus core protein inhibits microsomal triglyceride transfer
evidence of steatosis does not definitively exclude the
protein activity and very low density lipoprotein secretion: a
model of viral-related steatosis. FASEB J 2002, 16:185-194.
presence of steatosis as shown on biopsy. Until a non-
4.
Shi ST, Polyak SJ, Tu H, Taylor DR, Gretch DR, Lai MM: Hepatitis C
invasive modality is proven to be superior in a clinical set-
virus NS5A colocalizes with the core protein on lipid drop-
ting, liver biopsy remains the optimal diagnostic proce-
lets and interacts with apolipoproteins. Virology 2002,
292:198-210.
dure for the determination of steatosis in patients infected
5.
Kumar D, Farrell GC, Fung C, George J: Hepatitis C virus geno-
with hepatitis C.
type 3 is cytopathic to hepatocytes: reversal of hepatic stea-
tosis after sustained therapeutic response. Hepatology 2002,
36:1266-1272.
List of abbreviations
6.
Monto A, Alonzo J, Watson JJ, Grunfeld C, Wright TL: Steatosis in
AFP Alpha-fetoprotein
chronic hepatitis C: relative contributions of obesity, diabe-
tes mellitus, and alcohol. Hepatology 2002, 36:729-736.
7.
Bjoro K, Bell H, Hellum KB, Skaug K, Raknerud N, Sandvei P, Doske-
ALT Alanine aminotransferase
land B, Maeland A, Lund-Tonnesen S, Myrvang B: Effect of com-
bined interferon-α induction therapy and ribavirin on
chronic hepatitis C virus infection: a randomized multicen-
BMI Body mass index
tre study. Scand J Gastroenterol 2002, 37:226-232.
8.
Fontana RJ, Lok AS: Noninvasive monitoring of patients with
chronic hepatitis C. Hepatology 2002, 36:S57-S64.
CI Confidence interval
9.
Kutcher R, Smith GS, Sen F, Gelman SF, Mitsudo S, Thung SN, Reinus
JF: Comparison of sonograms and liver histologic findings in
HCV Hepatitis C virus
patients with chronic hepatitis C virus infection. J Ultrasound
Med 1998, 17:321-325.
10.
The French METAVIR Cooperative Study Group: Intraobserver
OR Odds ratio
and interobserver variations in liver biopsy interpretation in
patients with chronic hepatitis C. Hepatology 1994, 20:15-20.
11.
Joseph AE, Saverymuttu SH, al-Sam S, Cook MG, Maxwell JD: Com-
PCR Polymerase chain reaction
parison of liver histology with ultrasonography in assessing
diffuse parenchymal liver disease. Clin Radiol 1991, 43:26-31.
12.
Saadeh S, Younossi ZM, Remer EM, Gramlich T, Ong JP, Hurley M,
RNA Ribonucleic acid
Mullen KD, Cooper JN, Sheridan MJ: The utility of radiological
imaging in nonalcoholic fatty liver disease. Gastroenterology
US Ultrasonography
2002, 123:745-50.
13.
Czaja AJ, Carpenter HA, Santrach PJ, Moore SB: Host- and disease-
specific factors affecting steatosis in chronic hepatitis C. J
Competing interests
Hepatol 1998, 29:198-206.
14.
Hourigan LF, Macdonald GA, Purdie D, Whitehall VH, Shorthouse C,
The author(s) declare that they have no competing
Clouston A, Powell EE: Fibrosis in chronic hepatitis C corre-
interests.
lates significantly with body mass index and steatosis. Hepa-
tology 1999, 29:1215-1219.
15.
Bayati N, Silverman AL, Gordon SC: Serum alpha-fetoprotein
levels and liver histology in patients with chronic hepatitis C.
Am J Gastroenterol 1998, 93:2452-2456.
Page 7 of 8
(page number not for citation purposes)

---

BMC Gastroenterology 2005, 5:14
http://www.biomedcentral.com/1471-230X/5/14
16.
Joy D, Thava VR, Scott BB: Diagnosis of fatty liver disease: is
biopsy necessary? Eur J Gastroenterol Hepatol 2003, 15:539-543.
17.
Kurtz AB, Middleton WD: Ultrasound. The Requisites New York:
Mosby; 1996:16-18.
18.
Saverymuttu SH, Joseph AE, Maxwell JD: Ultrasound scanning in
the detection of hepatic fibrosis and steatosis. Brit Med J 1986,
292:13-15.
19.
Mathiesen UL, Franzen LE, Aselius H, Resjo M, Jacobsson L, Foberg U,
Fryden A, Bodemar G: Increased liver echogenicity at ultra-
sound examination reflects degree of steatosis but not fibro-
sis in asymptomatic patients with mild-moderate
abnormalities of liver transaminases. Digest Liver Dis 2002,
34:516-522.
20.
Saadeh S, Younossi ZM, Remer EM, Gramlich T, Ong JP, Hurley M,
Mullen KD, Cooper JN, Sheridan MJ: The utility of radiological
imaging in nonalcoholic fatty liver disease. Gastroenterology
2002, 123:745-750.
21.
Needleman L, Kurtz AB, Rifkin MD, Cooper HS, Pasto ME, Goldberg
BB: Sonography of diffuse benign liver disease: accuracy of
pattern recognition and grading. Am J Roentgenol 1986,
146:1011-1015.
22.
Debongnie JC, Pauls C, Fievez M, Wibin E: Prospective evaluation
of the diagnostic accuracy of liver ultrasonography. Gut 1981,
22:130-135.
23.
Castera L, Hezode C, Roudot-Thoraval F, Bastie A, Zafrani ES, Paw-
lotsky JM, Dhumeaux D: Worsening of steatosis is an independ-
ent factor of fibrosis progression in untreated patients with
chronic hepatitis C and paired liver biopsies. Gut 2003,
52:288-292.
24.
Siegelman ES, Rosen MA: Imaging of hepatic steatosis. Semin Liver
Dis 2001, 21:71-80.
25.
Regev A, Berho M, Jeffers LJ, Milikowski C, Molina EG, Pyrsopoulos
NT, Feng ZZ, Reddy KR, Schiff ER: Sampling error and intraob-
server variation in liver biopsy in patients with chronic HCV
infection. Am J Gastroenterol 2002, 97:2614-2618.
Pre-publication history
The pre-publication history for this paper can be accessed
here:
http://www.biomedcentral.com/1471-230X/5/14/pre
pub
Publish with Bio Med Central
and every
scientist can read your work free of charge
"BioMed Central will be the most significant development for
disseminating the results of biomedical researc h in our lifetime."
Sir Paul Nurse, Cancer Research UK
Your research papers will be:
available free of charge to the entire biomedical community
peer reviewed and published immediately upon acceptance
cited in PubMed and archived on PubMed Central
yours — you keep the copyright
Submit your manuscript here:
BioMedcentral
http://www.biomedcentral.com/info/publishing_adv.asp
Page 8 of 8
(page number not for citation purposes)

---

Document Outline

• Abstract
  • Background
  • Methods
  • Results
  • Conclusion
• Background
• Methods
  • Patients
  • Ultrasonography
  • Histologic Examination
  • Analysis
    • Table 1
• Results
  • Table 2
• Discussion
• Conclusion
• List of abbreviations
• Competing interests
• Authors' contributions
• Acknowledgements
• References
• Pre-publication history

---