Ultrasonographic visualization

of the liver in sites recommended

for blind percutaneous liver biopsy in horses
Sara C. Sammons, DVM, MS; Tracy E. Norman, VMD;
M. Keith Chaffin, DVM, MS; Noah D. Cohen, VMD, MPH, PhD

Objective—To ascertain the frequency of ultrasonographic identification of liver at sites

recommended for blind percutaneous liver biopsy in middle-aged horses and to determine

EQUINE
whether the liver is obscured by other organs or too thin for safe sample collection at rec-
ommended locations.
Design—Prospective case series.
Animals—36 healthy middle-aged (between 6 and 18 years old) Quarter Horses or Quarter
Horse crosses.
Procedures—Blood samples were collected from each horse and submitted for evalua-
tion of liver function. Horses with any indication of liver dysfunction on serum biochemical
analysis were excluded. The region just below a line drawn between the dorsal aspect
of the tuber coxae and the point of the elbow joint in the right 11th, 12th, 13th, and 14th
intercostal spaces (ICSs) was imaged by ultrasonography for the presence of liver. In each
ICS, liver thickness and whether there was partial obstruction in viewing the liver caused by
other abdominal or thoracic organs were recorded.
Results—39% (14/36) of horses had liver imaged on ultrasonographic examination in all of
the 11th to 14th ICSs. None of the 36 horses had liver of adequate thickness (ie, liver thick-
ness ≥ 3.5 cm) for biopsy in all of the imaged ICSs. For 22 horses in which the liver was
not visible on ultrasonographic examination of an ICS, lung was imaged instead in 12 (55%)
horses, intestine in 8 (36%), and both intestine and lung in 2 (9%).
Conclusions and Clinical Relevance—On the basis of the results of this study, the prac-
tice of blind percutaneous liver biopsy in horses is not recommended because of the risk of
serious complications. (J Am Vet Med Assoc 2014;245:939–943)

E valuation of percutaneously obtained liver biopsy

specimens is widely regarded as the most sensitive
and specific antemortem diagnostic test for suspected
hepatopathy in horses.1,2 Evaluation of liver biopsy
specimens may establish the presence or absence of liv-
er disease, provide a specific diagnosis, guide treatment,
and help determine prognosis in cases of suspected liv-
er disease.3 Multiple techniques for obtaining a biopsy
specimen of the liver have been described, including
standing laparoscopic surgery,4 blind percutaneous bi-
opsy, and direct and indirect ultrasound-guided biopsy.
Ultrasonography is also useful to visualize the echo-
genicity, size, and location of the liver. The abdomen
is scanned for the presence of adequate liver tissue, the
biopsy region is marked on the horse, and the area of
interest is prepared for sterile biopsy. This technique
also allows for a brief abdominal ultrasonographic ex-
From the Department of Large Animal Clinical Sciences, College of
Veterinary Medicine and Biomedical Sciences, Texas A&M Univer-
sity, College Station, TX 77843.
Supported by a grant from the Department of Large Animal Clinical
Sciences, College of Veterinary Medicine and Biomedical Sciences,
Texas A&M University.
Presented in abstract form at the 60th Annual American Association
of Equine Practitioners Convention, Salt Lake City, December 2014.
Address correspondence to Dr. Sammons (Ssammons@cvm.tamu.
edu).
CI
ABBREVIATIONS
Confidence interval
ICS Intercostal space
amination following biopsy to ensure that hemorrhage
has not occurred as a result of the procedure.
On ultrasonographic examination, the equine liver
appears as a wedge of weak, homogenously distribut-
ed echoes medial to the diaphragm and ventral to the
lung margin from the 9th to the 16th ICS.5–7 The liver is
most commonly viewed from the right side of the abdo-
men in a healthy horse, but because of the possibility
of atrophy of the right liver lobe in older horses, likely
resulting from pressure of the right dorsal colon and
cecal base, examination of the liver from the left side
may be necessary.7 The left liver lobe can occasionally
be imaged from the left 9th to 11th ICS and ventral to
the lung margin.1,8
The greatest limitation of ultrasonographic organ
localization and biopsy guidance is the inability of
ultrasound to transmit through gas-filled structures
and bone. These structures act as barriers to sound
beam penetration and prevent observation of soft tissue
structures deep to them.1 Even with the liver located in
a normal anatomic position in a healthy horse, it may be
deep to structures that are not able to be penetrated by

JAVMA, Vol 245, No. 8, October 15, 2014 Scientific Reports 939
 ultrasound beams. Therefore, gas-filled structures such
as lung and intestines positioned between the liver and
body wall could inadvertently be penetrated during an
attempted liver biopsy. Pneumothorax, excessive blood
loss into the abdominal or thoracic cavities, or entero-
centesis can occur and require medical intervention.1
Anatomic locations to perform a blind percutane-
ous liver biopsy are published1,9 and referenced to by
many equine veterinarians. The most frequently recom-
mended location for performing a blind percutaneous
liver biopsy has been described as the right 12th to 14th
ICS at approximately the level of a line drawn between
the tuber coxae and the point of the olecranon, direct-
ing the needle toward the contralateral elbow joint.9
EQUINE
Another similar location described is the right side of
the abdomen between the 10th and 11th or the 11th
and 12th ribs just below a line drawn from the point
of the tuber coxae to the point of the shoulder, passing
the biopsy instrument caudad and ventrad through the
intercostal muscles corresponding with expiration to
decrease the likelihood of penetrating the lung.1
In 1 study,9 liver biopsy could not be performed
in 17% (4/24) of study mares because the liver was
not visualized ultrasonographically at the target zone,
which is similar to but encompasses a larger area than
the recommended locations for blind percutaneous
liver biopsy. To the authors’ knowledge, the actual fre-
quency of ultrasonographic identification of the liver
in the suggested locations for blind percutaneous liver
biopsy in horses has not been reported. Furthermore,
we are not aware of any study that has specifically de-
termined the suitability of liver for biopsy at the rec-
ommended blind percutaneous liver biopsy sites in a
specific population of horses. Therefore, the objectives
of the study reported here were to ascertain the fre-
quency of ultrasonographic identification of liver in
the recommended location for blind percutaneous liv-
er biopsy in middle-aged horses and to determine on
ultrasonographic examination whether the liver is ob-
scured by other organs or too thin for safe sample col-
lection at those locations. We hypothesized that the
frequency of identification of liver adequate for biopsy
in the recommended locations for blind percutaneous
liver biopsy is small and that the lack of sufficient liver
tissue in the proposed locations could lead to serious
complications resulting from biopsy including hemor-
rhage, peritonitis, or pneumothorax.
Materials and Methods
Study population and animals—This study’s ani-
mal use protocol was reviewed and approved by the
Texas A&M University Institutional Animal Care and
Use Committee. Horses were either privately owned
and brought to the hospital for routine farrier appoint-
ments or lameness examinations or were part of the
university-owned herd of research mares and geldings.
Written consent was obtained prior to use of each pri-
vately owned horse in the study. Historical information
and physical examination findings were used to identi-
fy healthy horses for inclusion in the study. Horses with
a history of preexisting hepatopathy or those receiving
medications for any ailment were excluded. Thirty-
six middle-aged healthy American Quarter Horses or
940 Scientific Reports
Quarter Horse crosses were included in the study. This
sample size was determined on the basis of a binomial
model, assuming that the proportion of horses in which
liver was not observed within the recommended area
was 30% and that we wanted to estimate the propor-
tion with a margin of error of 15% at a confidence level
of 95%. A review of records indicates that the mean ±
SD age of horses undergoing a liver biopsy at our in-
stitution was 12 ± 6.7 years (median, 11 years; range,
2 months to 28 years). On the basis of these findings,
horses used for this study were between 6 and 18 years
old. Elderly horses (ie, ≥ 20 years) were excluded be-
cause liver atrophy has been documented in older hors-
es.7 Only American Quarter Horses or Quarter Horse
crosses were used for the study to avoid possible breed-
associated anatomic variability.
Experimental design—Blood was drawn into plain
vacuum tubesa and submitted for serum biochemical
analysis to evaluate liver function. Each horse was re-
strained in stocks, and blood was drawn from a jugular
vein to evaluate concentrations of BUN and bile acids
and activities of gamma glutamyltransferase, alkaline
phosphatase, and aspartate aminotransferase. These
specific analytes were chosen because each is associ-
ated with hepatocellular or biliary function in horses.
Bile acids were included as a specific liver function test.
Samples were immediately refrigerated and analyzed
within 4 hours after collection. Horses identified by se-
rum biochemical analysis to have any indication of liver
dysfunction were excluded from the study.
A 3.5- to 8-MHz curvilinear ultrasound trans-
ducerb was used to image the region just below a line
drawn between the dorsal aspect of the tuber coxae
and the point of the elbow joint in the right 11th,
12th, 13th, and 14th ICS for the presence of liver (Fig-
ure 1). Clipping of the hair prior to performing ultra-
sonographic examinations was not necessary because
of the short hair of the horses used in the study. The
dorsal-most aspect of the probe was situated just ven-
tral to the line and parallel to the ribs within each ICS
and directed in an oblique transverse plane. A single
Figure 1—Photograph of a healthy middle-aged Quarter Horse. A
line is drawn on the right side of the horse between the dorsal as-
pect of the tuber coxae and the point of the elbow joint. The 11th,
12th, 13th, and 14th ICSs are designated. The ultrasound probe is
positioned just ventral to this line to image the liver.
JAVMA, Vol 245, No. 8, October 15, 2014
still image was obtained and saved in each ICS (Figure
2). If the liver was identified in a location, the depth
from the superficial to the deep surface of the liver
was measured by means of postacquisition processing
on the ultrasound unit. The presence of other abdom-
inal or thoracic organs that obscured the liver from
view was recorded if the liver itself was not visible.
Because a standard liver biopsy instrument chamber
is 2.5 cm in length, we required that the liver be at
least 3.5 cm thick for an adequate sample that could
safely be obtained without risk of puncture of deeper
structures. All imaging and postimaging evaluation
was performed by the same author (SCS).
The ability to identify the liver at the designated lo-
cation (yes vs no) by ultrasonography was recorded. If
liver was visible, its thickness was measured at the center
of the ultrasound beam to determine whether it was of
adequate thickness for biopsy. If liver was not visible or
if it was partially obscured such that biopsy could not
be performed without penetrating another abdominal
or thoracic organ, the other organs at that location were
noted. If the liver was not visible in all of the 4 ICSs, the
liver was defined as not appropriate for biopsy because
the recommended locations for blind percutaneous liver
biopsy in horses encompass multiple ICSs. If the liver
measured < 3.5 cm in any 1 of the 4 ICSs at the center of
the ultrasound beam, it was also defined as inappropri-
ate for biopsy. Therefore, to be defined as appropriate for
biopsy, the liver needed to be observable, thick enough
for biopsy specimen collection, and accessible if a biopsy
Figure 2—Photograph of a horse demonstrating ultrasound probe
positioning for imaging of the liver between each ICS, parallel to
the ribs, directed in an oblique transverse plane.
JAVMA, Vol 245, No. 8, October 15, 2014
was performed in the center of the ultrasonographic im-
age in each of the 11th to 14th recommended ICSs. Ad-
ditionally, the depth from skin surface to abaxial surface
of the liver and the total depth from skin surface to axial
surface of the liver were recorded.
Data analysis—Data were analyzed through both
descriptive and inferential statistical methods. For de-
scriptive purposes, proportions were used to summa-
rize categorical data and medians, and ranges were used
to summarize continuous data. Confidence limits for
proportions were calculated by means of binomial esti-
mation. The upper bound of the 95% CI was estimated
according to the formula 1 – α1/n, where α is the signifi-
EQUINE
cance level (0.05) and n is the number of horses.
Because observations were not independent, com-
parisons between pairs of ICSs were made by meth-
ods accounting for pairing. For categorical data, the
McNemar test was used, and for continuous data, the
Wilcoxon signed rank test was used. For a given vari-
able, the multiplicity of comparisons was adjusted fol-
lowing the Bonferroni method. A significance level of
adjusted P < 0.05 was used.
Results
The study population comprised 36 Quarter Hors-
es or Quarter Horse crosses. The mean ± SD weight of
the horses was 497 ± 59.3 kg (1,096 ± 130.7 lb), and
all horses had a body condition score of at least 5 of 9.
The heights of the horses were not measured, but all
had typical Quarter Horse body types. The median age
of the population was 11 years (range, 3 to 18 years;
mean, 10.9 ± 4.12 years). Of the 36 horses, 17 (47%)
were females and 19 (53%) were males (15 geldings
[42%] and 4 sexually intact males [11%]). No signifi-
cant difference was found in the age of males (median,
11 years; range, 4 to 18 years) and females (median, 12
years; range, 3 to 16 years).
Of the 36 horses, 14 (39%; 95% CI, 31% to 47%)
had liver visible in all of the 11th to 14th ICSs. None
of the 36 horses (0%; 95% confidence upper limit, 8%)
had liver thickness ≥ 3.5 cm (the thickness deemed ap-
propriate to be sufficient for biopsy collection) in all of
the 11th through 14th ICSs.
The proportion of horses with visible liver and the
median depth from skin surface to abaxial border of
the liver, liver thickness, and depth from skin surface
to axial border of the liver when liver was visible in
each ICS were summarized (Table 1). The proportion
of horses with liver visible was higher for the 13th and
14th ICS than for the 12th and 11th ICS, but there were
no significant differences among proportions. Also, no
significant differences were found among proportions
of horses with adequate liver thickness, although the
11th and 14th ICSs had smaller proportions than the
12th and 13th ICSs.
No significant differences were found between ICSs
for measured depth from skin surface to abaxial surface
of liver except for between the 13th and 14th ICS (Ta-
ble 1). Among the 21 horses that had this depth record-
ed for both 13th and 14th ICSs, the depth from skin
surface to abaxial surface of the liver was significantly
(P ≤ 0.001) greater for the 13th ICS than for the 14th,
Scientific Reports 941
 Table 1—Summary of ultrasonographic identification and measurements of the liver at the recommended ICS locations for blind percu-
taneous liver biopsy in 36 middle-aged Quarter Horses and Quarter Horse crosses.

ICS
Variable 11 12 13 14
No. of horses with liver visualized (% [95% CI]) 18/36 (50 [42–58]) 20/36 (56 [48–64]) 25/36 (69 [61–77]) 25/36 (69 [61–77])
Median (range) liver depth (cm) 4.5 (1.9–5.8)a,b 4.9 (2.4–6.1)a,b 4.8 (1.6–6.0)a 4.1 (1.1–5.7)b
Median (range) liver thickness (cm) 2.4 (0.9–6.0) 3.3 (1.2–7.9) 2.9 (0.6–5.9) 2.6 (0.7–7.5)
No. of horses with adequate liver thickness (% [95% CI]) 5/18 (28 [17–39]) 8/20 (40 [29–51]) 10/25 (40 [30–50]) 7/25 (28 [18–38])
Median (range) skin to axial border (cm) 6.8 (4.0–11.3) 7.6 (4.3–12.9) 7.5 (4.9–10.2) 6.9 (3.7–11.6)

Within a row, values with different superscript letters are significantly (P < 0.05) different.
a,b

Table 2—Agreement in the ability to image the liver between an adequate diagnostic sample can be obtained and to
EQUINE

pairs of ICSs during hepatic ultrasonographic examination in the avoid puncture of lung or intestines.
same horses as in Table 1.
Our results provide more compelling evidence re-
ICS κ 95% CI Overall agreement 95% CI garding the necessity to use ultrasonographic assistance
for liver biopsy than have previous studies. Pearce et al9
11 vs 12 0.78 0.57 to 0.98 89% (32/36) 84% to 94%
11 vs 13 0.50 0.21 to 0.79 75% (27/36) 68% to 82% performed a study of liver biopsy procedures in post-
11 vs 14 0.28 –0.06 to 0.62 64% (23/36) 56% to 72% partum mares in which 17% of study horses did not
12 vs 13 0.71 0.48 to 0.94 87% (31/36) 81% to 93% have liver imaged at the target biopsy zone, which was a
12 vs 14 0.36 0.03 to 0.68 69% (25/36) 61% to 77%
13 vs 14 0.48 0.16 to 0.79 77% (28/36) 70% to 84% triangle formed by joining the tuber coxae to the point
of the olecranon and the scapulohumeral joint between
the 12th and 14th ICSs, an area similar in location
Table 3—Agreement in the presence of adequate liver thickness
to but larger in size than the area used for our study.
between pairs of ICSs during hepatic ultrasonographic examina- Ultrasonography was used to attempt to guide the pre-
tion in the same horses as in Table 1. cise location for liver biopsy procedures within the tar-
get zone and, in all of the 4 cases for which liver was not
ICS κ 95% CI Overall agreement* 95% CI
identifiable, intestine was imaged instead.9 The mares
11 vs 12 0.43 –0.04 to 0.90 77% (13/17) 66% to 87% of that study9 were 11, 11, 13, and 18 years old. The
11 vs 13 0.06 –0.43 to 0.55 59% (10/17) 47% to 71% mares were reexamined ultrasonographically 2 weeks
11 vs 14 –0.36 –0.59 to –0.13 47% (7/15) 34% to 60%
12 vs 13 0.38 –0.04 to 0.79 70% (14/20) 59% to 81% later, and the anatomic situation was the same. Postpar-
12 vs 14 –0.42 –1.00 to 0.18 35% (6/17) 24% to 47% tum mares were used for that study,9 which could cause
13 vs 14 0.22 –0.24 to 0.69 62% (13/21) 51% to 72% a slight difference in anatomic configuration of organs
*The various denominators reflect the number of horses that within the abdomen, compared with that of geldings.
had liver visible in both of the listed ICSs. Our study was designed to provide further infor-
mation about the practice of blind percutaneous liver
biopsy, not to estimate the frequency of complications
and the median difference was 0.3 cm (range, –0.2 to associated with liver biopsy procedures; as such, no
1.1 cm). No significant differences were found in liver liver biopsies were actually performed. Even if ultraso-
thickness among ICSs. Also, no significant difference nography is used to plan the best approach to a liver bi-
was found among ICSs in the distance from skin to the opsy, there are still risks associated with the procedure.
axial border of the liver. A retrospective study10 evaluating the complications as-
Agreement with regard to visibility and presence sociated with liver biopsy procedures in horses identi-
of adequate liver thickness between pairs of ICSs were fied 1 case of hemorrhage resulting from diaphragmatic
examined on the basis of κ statistics and overall agree- hernia. The authors of that study10 advised that ultra-
ment. The κ statistics, 95% CIs, and overall agreement sonographic examination during or directly following
were tabulated (Tables 2 and 3). In general, there was liver biopsy could be used to assess the incidence of
better agreement between adjacent ICSs than more acute bleeding in horses. If severe hemorrhage is identi-
remote spaces. For 22 horses for which the liver was fied ultrasonographically shortly following liver biopsy,
not visible on ultrasonographic examination of an ICS, proper treatment may be prescribed and further com-
lung was imaged instead in 12 (55%) horses, intestine plications prevented.
in 8 (36%), and both intestine and lung in 2 (9%). Research regarding the safety of blind percutaneous
liver biopsy in humans has shown that ultrasonographic
Discussion localization of the liver prior to biopsy substantially di-
minishes the risk of complications in adults.11 In a study12
On the basis of the results of this study, the prac- of percutaneous liver biopsy in children, patients who
tice of blind percutaneous liver biopsy in horses is not developed postprocedural bleeding as a complication
recommended. Fewer than 40% of horses in this study were all subjected to blind percutaneous liver biopsies,
had liver visible in all recommended ICSs for blind per- whereas no hemorrhage occurred in patients undergo-
cutaneous liver biopsy, and no horse had thick enough ing ultrasound-guided liver biopsy procedures. Authors
liver in all of the ICSs. The use of ultrasonography prior of that study10 suspected a decrease in the number of
to liver biopsy is therefore strongly recommended to complications following ultrasound-guided biopsy as a
image the liver and evaluate its thickness to ensure that result of the greater accuracy in liver localization, ability

942 Scientific Reports JAVMA, Vol 245, No. 8, October 15, 2014
to identify and avoid large hepatic vessels and other or-
gans, and the subsequent reduction in number of passes
required to obtain an adequate sample with the biopsy
instrument.12 Ultrasonography was also found to be use-
ful in detection of postbiopsy hematomas, which are an
indication of severe bleeding.12
One limitation of our study was that we evaluated
only the right side of horses. The liver is most commonly
imaged on the right side of the abdomen in healthy hors-
es, but because of the possibility of atrophy of the right
liver lobe in older horses, examination of the liver from
the left side can be more productive in some patients.
Another limitation of this study is that only horses
that were determined to be healthy according to physi-
cal examination findings and results of serum biochem-
ical analysis were used in this study, whereas liver biop-
sy is typically performed in horses with suspected liver
disease. Because hepatopathy can result in enlargement
of the liver or cirrhosis of the liver, results of a study
examining the location and adequacy of liver for biopsy
in patients with liver disease may be different. At our
institution, a small number of patients are admitted
for liver disease; therefore, a multicenter study may be
helpful to achieve adequate case numbers in a reason-
able period.
In this study, because only 1 author performed the
ultrasonographic examinations and took measurements
at a single time point, no interobserver variability was
determined. Because single still images were obtained
and measurements were taken once, intraobserver vari-
ability could not be calculated; this is considered an-
other limitation to our study.
Finally, we used only adult Quarter Horses or
Quarter Horse crosses. The location of the liver within
the abdomen, the dimensions of the abdomen, and the
size of the liver itself may be different depending on the
breed of horse evaluated.
The results of this study indicated a need to use ul-
trasonographic assistance when performing percutane-
JAVMA, Vol 245, No. 8, October 15, 2014
ous liver biopsy procedures in horses. Ultrasound ma-
chines are becoming increasingly affordable and portable
for equine practitioners. With some training and guid-
ance from an experienced ultrasonographer, a practitio-
ner can use ultrasonographic assistance when perform-
ing a percutaneous liver biopsy in horses, thereby reduc-
ing the risk of complications that might otherwise occur.
a. BD Vacutainer Serum, 10-mL plain tube, BD, Franklin Lakes, NJ.
b. Technos ultrasound unit, Esaote, Genoa, Italy.
References
1. Modransky PD. Ultrasound-guided renal and hepatic biopsy
EQUINE
techniques. Vet Clin North Am Equine Pract 1986;2:115–126.
2. Durham AE, Newton JR, Smith KC, et al. Retrospective analysis
of historical, clinical, ultrasonographic, serum biochemical and
haematological data in prognostic evaluation of equine liver dis-
ease. Equine Vet J 2003;35:542–547.
3. Divers TJ, Bernard WB, Reef VB. Equine liver disease and liver
failure—causes, diagnosis, and treatment, in Proceedings. 10th
Bain-Fallon Memorial Lecture 1988;35–46.
4. Fischer AT Jr. Standing laparoscopic surgery. Vet Clin North Am
Equine Pract 1991;7:641–647.
5. Rantanen NW. Ultrasound appearance of normal lung borders
and adjacent viscera in the horse. Vet Radiol 1981;22:217–219.
6. Rantanen NW. Ultrasonic examination of the liver of the horse, in
Proceedings. Annu Conv Am Institute Ultrasound Med 1981;81.
7. Rantanen NW, Timon TC. Ultrasound scan techniques for large
and small animals. Bellevue, Wash: ATL Inc, 1983;14–23.
8. Dyce KM, Sack WO, Wensing CJG. In: Textbook of veterinary
anatomy. Philadelphia: Saunders Co, 1987;503.
9. Pearce SG, Firth EC, Grace ND, et al. Liver biopsy techniques
for adult horses and neonatal foals to assess copper status. Aust
Vet J 1997;75:194–198.
10. Johns IC, Sweeney RW. Coagulation abnormalities and compli-
cations after percutaneous liver biopsy in horses. J Vet Intern
Med 2008;22:185–189.
11. Cadranel JF, Rufat P, Degos F. Practices of liver biopsy in
France: results of a prospective nationwide survey. Hepatology
2000;32:477–481.
12. Nobili V, Comparcola D, Sartorelli MR, et al. Blind and ultra-
sound-guided percutaneous liver biopsy in children. Pediatr Ra-
diol 2003;33:772–775.
Scientific Reports 943