B-MODE AND DOPPLER ULTRASOUND IMAGING OF THE SPLEEN WITH

CANINE SPLENIC TORSION: A RETROSPECTIVE EVALUATION

H. MARKSAUNDERS, J. NEATH,BSc, BVetMed, DANIEL

VMD, MS, PRUDENCE BVSc
J. BROCKMAN,

The ultrasonographic appearance of splenic torsion has been described; the splenic parenchyma can be
normal, hypoechoic or anechoic with interspersed linear echoes (coarse/"lacy" appearance). The ul-
trasonographic parenchymal appearance of 15 dogs in this report with splenic torsion varied: mottled
hypoechoic regions (n = 2), diffusely hypoechoic (n = 11) and normal (n = 2). Because splenic torsion
causes vascular congestion due to splenic vein compression and eventual thrombosis, visible splenic vein
intraluminal echogenicities compatible with thrombi were seen in 13 dogs using B-mode. Using spectral
Doppler and color Doppler imaging of the splenic veins, no measurable flow velocities were detected in
any of the 15 dogs. The varied B-mode ultrasonographic appearance of the splenic parenchyma with
splenic torsion necessitates B-mode evaluation of the splenic veins for intraluminal echoes and spectral
or color Doppler evaluation for absent velocity flow. Veterinary Radiology & Ultrasound, Vol. 39, No. 4,
1998, pp. 349-353.

Key words: canine, splenic torsion, venous thrombosis, ultrasound, Doppler.

Introduction ing large anechoic areas and splenic vein enlargement at

the splenic h i l u ~ .These
~ ultrasonographic findings are

S PLENIC TORSION occurs when the spleen rotates around

the gastrosplenic and phrenosplenic ligaments,' Result-
ing occlusion of the splenic vessels leads to vascular throm-
also similar to that seen with splenic necrosis secondary to
infarction.6 Although splenic vein enlargement has been
seen ultra~onographically,'~~ the results of Doppler assess-
bosis, congestion, hemorrhage and necrosis of the spleen.' ment of these veins has not been reported. This report de-
Splenic torsion is most common in large, deep-chested scribes the use of B-mode, spectral and color Doppler im-
breed dogs and can occur either as an isolated process or aging to evaluate the spleen in dogs with splenic torsion.
secondary to gastric dilation-volvulus.' The clinical presen-
tation, physical examination findings and clinical pathology
Methods
results in dogs with splenic torsion are non-specific.3
The described radiographic abnormalities seen with isolated
The medical records of the Veterinary Hospital of the
splenic torsion can also be non-spe~ific.'.~Decreased
University of Pennsylvania (VHUP) from 1988 to 1996
abdominal detail, displacement of other abdominal organs,
were searched for dogs with confirmed splenic torsion. In-
absence of the splenic body in the left cranial abdominal
clusion in the retrospective study required surgical or patho-
quadrant, abnormal splenic location or shape, spleno-
logic confirmation of splenic torsion and a complete ab-
megaly, and splenic gas may be seen on abdominal radio-
dominal ultrasonographic study report with hard copy im-
graph~.~
ages. The breed, sex, age and weight of the dogs were
The ultrasonographic abnormalities associated with
obtained from the records. The ultrasound study written
splenic torsion in the dog have also been described.' Splenic
report and images were reviewed for splenic size, paren-
congestion, hemorrhage and necrosis were character-
chymal echogenicity, presence of visible splenic vein in-
ized ultrasonographically by splenomegaly, diffusely
traluminal echogenicities and integrity of splenic vein blood
hypoechoic splenic parenchyma with linear echoes separat-
flow. Presence or absence of splenic vein blood flow was
determined from spectral and color Doppler evaluation of
From the Department of Clinical Studies, School of Veterinary Medi- these vessels. The ultrasound written report and images
cine, University of Pennsylvania, 3850 Spruce Street, Philadelphia, PA were also reviewed for splenic artery visualization and pres-
1 9 104-6010. ence or absence of arterial flow. Confirmation of splenic
Address correspondence and reprint requests to H.M. Saunders, VMD,
MS at the above address. torsion and presence of vascular intraluminal thrombi were
Received June 4,1997; accepted for publication August 2, 1997. obtained from the surgery and histopathology reports.

349
 SAUNDERS
ET AL 1998

TABLEI . B-Mode and Douoler Ultrasound Examination Results

Splenic ultrasound findings*
Age Hypoechoic Lacy Intraluminal Absent
Breed (yrs) Sex parenchyma parenchyma venous echoes venous flow
1 . German Shepherd 6 MC + + +
2. Great Dane 5 MC + + +
3. Saint Bernard 3 M + + +
4. Mixed breed 7 M + + +
5 . Mixed breed 8 M - + +
6. German Shepherd 7 FS + (mottled) + +
7. Great Dane 1.5 F + + +
8. Briard 11 MC + + +
9. Great Dane 4 M + + +
10. Mixed breed 4.5 MC + - +
1 1. German Shepherd 6 M + + +
12. German Shepherd 3.5 M - - +
13. Bulldog 0.5 M + + +
14. German Shepherd 3.5 M + (mottled) + +
15. Great Dane 4 M + + +
*+ = Present, - = Absent.

Results Splenomegaly was recorded on the abdominal ultrasono-

Patient Profile
Fifteen dogs with confirmed splenic torsion and complete
abdominal ultrasound examinations were identified. The
dogs were primarily young or middle age (mean = 5.0
years, median = 4.5 years) and male (n = 13). They were
predominantly large breed dogs (body weight: mean = 44
kg, median = 40 kg) with two breeds, the German Shep-
herd dogs (n = 5) and Great Danes (n = 4), accounting for
the majority of the 15 dogs (Table 1). Males, German Shep-
herds and Great Danes were found to be at increased risk for
developing splenic torsion compared with the hospital
population used as a contr01.~
Ultrasound Examinations
Abdominal ultrasonographic examinations were per-
formed using commercially available diagnostic machines."
B-mode and spectral Doppler imaging were performed in all
dogs; color Doppler imaging was performed in six dogs.
Doppler imaging parameters (spectral and color gains, pulse
repetition frequency, frequency filtration) had been opti-
mized for venous velocity flow.' Accurate Doppler assess-
ment of blood flow was limited to those splenic veins ac-
cessible to imaging at an appropriate angle (<60") with the
sample volume. B-mode images were recorded on film?
using matrix cameras;$ color Doppler images were recorded
on print films using a color video printer.((
"Aloka 650, 5.0 MHz convex array, 7.5 MHz mechanical, Corometrics
Medical Systems, Inc., Wallingford, CT; ATL Ultramark 9 HDI, C7-4
MHz curved array, L10-5 MHz linear array, Advanced Technology Labo-
ratories, Inc. Bothell, WA.
./.Kodak NMB, DuPonVSterling Cronex MPF 33.
$Video Imagcr I0 10-6, Matrix Instruments; Aspect Mini-Imager.
SSony UPC-1010.
I/Sony UP 1850 MD Color Video Printer, Sony Corp., Japan.
graphic report in all 15 dogs. On the B-mode images in 13
dogs, the splenic parenchyma was mottled with hypoechoic
regions (n = 2) or diffusely hypoechoic (n = 11) when
compared with the normal hepatic and renal cortical
echogenicity (Table 1). The two remaining dogs had normal
splenic parenchymal echogenicity. In nine of the 11 dogs
with a diffusely hypoechoic spleen, the parenchyma also
had multiple linear hyperechoic foci (Fig. 1). This appear-
ance has been previously described as a coarse/' 'lacy" pat-
tern.6 On B-mode images, 13 dogs had visible intraluminal
echogenicies in the splenic veins; the echogenicies were
seen in both the intrasplenic and mesenteric portions of the
veins (Fig. 2).
Spectral Doppler imaging in all dogs and color Doppler
imaging in six dogs were used to characterize blood flow in
the splenic veins at the level of the splenic hilar border.
Doppler interrogation of appropriately located splenic veins
failed to detect blood flow in all 15 dogs (Fig. 3). In the two
dogs with no visible venous thrombi on B-mode imaging,
no splenic venous flow was detected with Doppler imaging.
Absent splenic venous flow was detected using Doppler
imaging in the four dogs with hypoechoic, non-lacy splenic
parenchyma.
No information concerning the status of the splenic ar-
teries was provided on review of the ultrasonographic re-
ports. The hilar and intraparenchymal splenic arteries were
not seen on review of the B-mode images in all IS dogs. On
the color Doppler images of one dog, there was no evidence
of blood flow in the splenic artery adjacent to the occluded
splenic vein.
Surgical and Histopathology Results
Splenic torsion was confirmed at surgery in all 15 dogs;
splenectomies were performed in all dogs. Splenic tissue
VOL. 39, No. 4 DOPPLER
WITH SPLENIC
TORSION
Fro. I . B-mode image of the spleen (dog 11) curving across the near
field. The splenic parenchyma was hypoechoic with interspersed hyper-
echoic linear foci characteristic of the coarse/‘‘lacy” pattern. The larger
hyperechoic structures within the spleen (arrowheads) represent splenic
veins adjacent to, and exiting, the hilus. Hyperechoic mesentery is deep to
the spleen. The image was obtained with a curved array multifrequency (4-7
MHz) transducer; the depth scale in centimeters is to the right of the image.
samples were obtained for histopathology in ten dogs. His-
topathology findings of splenic congestion, hemorrhage and
necrosis where consistent with splenic torsion. Splenic ar-
terial and venous thrombi were identified in fjve dogs at
surgery (n = 2), on histopathology (n = 2) and both (n =
1). These five dogs with thrombi all had intraluminal splenic
vein echogenicities visible on B-mode ultrasound images.
In the remaining ten dogs, no mention of vascular throm-
FIG.2. B-modc image of the spleen (dog 8) obtained with a curved array
5.0 MHz transducer; centimeter depth scale is to the left of the image. The
ultrasonographic appearance of the spleen was hypoechoic with inter-
spersed hyperechoic linear foci (coarse/“lacy” pattern). Enlarged splenic
venous rami at the hilus (arrowheads) were seen joining into the larger
splenic vein (arrows). Static intraluminal echoes were seen in the veins
consistent with venous thrombi.
35 1
FIG.3. Duplex image (B-mode on the left, spectral Doppler on the right)
of the same area of the spleen shown in Fig. 2 obtained with a curved array
5.0 MHz (3.0 MHz Doppler) transducer. In the B-mode image, the Doppler
sample volume (2-mm width) was positioned approximately at a 20 degree
angle with the splenic vein branch (arrowheads) at the hilus. As seen on the
spectral Doppler image, no flow velocity was recorded compatible with
venous obstruction due to the suspected thrombus. Spectral Doppler ve-
locity scale = 0.32 meterslsec.
bosis was made on the surgery (n = 10) or histopathology
(n = 7) reports.
Discussion
Causes of splenomegaly are numerous and are catego-
rized as inflammatory (splenitis), hyperplastic, congestive
and infiltrative forms.’ Splenic torsion should be considered
in dogs presented with splenomegaly. The pathogenesis of
splenic torsion is unknown; splenic torsion is hypothesized
to cause, or is a sequele of, gastric volvulus.’oTorsion of the
spleen around its pedicle produces vascular congestion due
to compression of the thin-wall splenic veins whereas the
’
thicker-wall arteries remain patent. Ultimately, blockage
of the artery occurs.2 Therefore, thrombosis of the splenic
veins and eventually the splenic arteries occur with splenic
torsion.2 Splenic congestion, hemorrhagic infarction and ne-
crosis are histopathology findings with splenic torsion.’
Abdominal radiography is warranted in dogs with pal-
pable organomegaly. Splenic torsion should be considered
when splenomegaly and abnormal splenic location are seen
on radiograph^.',^ Because radiographic findings of spleno-
megaly are non-specific, abdominal ultrasonography is in-
dicated to further define the abdominal abnormality.
The ultrasonographic abnormalities characteristic of
splenic congestion, hemorrhage, necrosis and infarction
have been described in a total of six dogs; in four dogs the
changes were secondary to splenic t o r ~ i o n .In
~ ’the
~ remain-
ing two dogs, splenic infarction and necrosis were due to an
unknown cause and trauma. Of the four dogs with splenic
torsion, three dogs had splenic parenchyma which ultra-
352 SAUNDERS
ET AL
sonographically appeared diffusely hypoechoic with linear
echoes separating large anechoic regions. This splenic pa-
renchymal appearance was characterized as a coarse/
“lacy” pattern which histopathologically was due to
splenic infarction and necrosk6 Two of the three dogs also
had enlarged splenic veins at the hilus; thrombosis of the
splenic vasculature was found at surgery in one d ~ g .The ~,~
one remaining dog with splenic torsion had a homogenous
splenic parenchymal echogenic appearance and enlarged
splenic veins.’
The varied ultrasonographic appearance of the splenic
parenchyma in patients with suspected splenic torsion has
caused the author to examine the splenic vasculature with
both B-mode imaging for evidence of thrombi and Doppler
imaging for the presence or absence of blood flow. The size,
number and distribution of splenic venous rami penetrating
the spleen along the ridge-like hilus varies. l 1 Large venous
rami tend to course parallel with the hilus before passing
through the splenic capsule and enter the splenic paren-
chyma obliquely. Smaller venous rami enter perpendicu-
larly. Splenic arterial rami, although paired with venous
rami, are seldom observed at the hilus using B-mode imag-
ing.’
B-mode detection of vascular intraluminal thrombi is de-
pendent upon thrombus age.I2 The echogenicity of an early
in vitro intravascular thrombus was found to be indistin-
guishable from the surrounding static blood.’* With time,
the relative echogenicity of the retracted thrombus increased
from the surrounding blood. Unfortunately, these studies
used static rather than moving blood. In clinical studies it
has been shown that fresh clotted blood may not be detected
ultrasonographically due to inherent lack of clot echogenic-
it^.'^,'^ The vessel may appear anechoic and thus patent
despite the presence of an occlusive clot.12,‘3
Echogenicity of flowing blood is caused by red blood cell
aggregation.14 Red blood cell aggregation is determined by
hematocrit, erythrocyte membrane condition, plasma mac-
romolecules and shear rate.I4 Shear rate, rather than flow
velocity, determines echogenicity. Low shear rate blood
flow is more echogenic than high shear rate flow.I4 Venous
flow is characterized by low shear rates and veins are there-
fore more likely to have intraluminal echogenicities. Upon
cessation of flow, liquid whole blood has been found to
become echogenic within a few seconds to 3 minutes
following stasis.15 This rapid onset of echogenicity was
seen in B-mode, real-time imaging using 7.5 MHz, but not
3.5 MHz transducers.” Conversely, in the same study,
whole clotted blood was echogenic with both 3.5 and 7.5
MHz transducers. The development of echogenicity re-
quired the physical layering of erythrocytes plus fibrinogen
or its products.I5
Thirteen of the 15 dogs reported here had visible static
splenic intravascular echogenicities using B-mode imaging.
Based on vessel diameter, these intravascular echogenicities
1998
were in splenic veins. The intraluminal splenic vein echoge-
nicities represented either formed thrombi or static
echogenic blood. In normal patent splenic veins, it is com-
mon with B-mode imaging to see moving intraluminal
echogenicities due to the lower shear rate of venous blood
flow. Five of the 13 dogs with visible static intravascular
echogenicities had splenic vein thrombi confirmed at sur-
gery or on histopathology. In six dogs, there was no mention
of vascular thrombi at surgery and on histopathology. In the
remaining two dogs in which intravascular echogenicities
were seen, the histopathology report stated that no arterial
and venous thrombi were seen in the submitted splenic
samples. Therefore, either the suspected thrombosed vessels
were not sampled for histopathology or intravascular
echogenicities were present due to static blood prior to
thrombus formation.
B-mode evaluation of the splenic vasculature is limited to
assessment of vessel size and intraluminal thrombus detec-
tion. Complete hemodynamic evaluation requires spectral
or color-flow Doppler imaging to assess blood flow. Inter-
pretation of the spectral Doppler waveform provides quali-
tative (flow presence, direction, characteristics) and quan-
titative (velocity, pressure gradient and acceleration) infor-
mation. l 6 But, valid pulsed-wave spectral Doppler
information is only obtained from visible vessels, where the
sample volume can be accurately positioned and where the
ultrasound beam-vessel angle is less than 60 degrees.l6
Compared with spectral Doppler, color-flow Doppler is
easier to perform. Color Doppler imaging permits simulta-
neous anatomic, mean blood flow velocity and direction
information; the presence and direction of flow can be rap-
idly assessed.
Using spectral Doppler evaluation of multiple, visible
splenic veins, no venous flow was present in all 15 dogs.
Lack of venous flow occurred in two dogs in whom B-mode
imaging failed to detect intravascular echogenicities com-
patible with venous thrombi or static blood. It is unlikely
that all veins exiting the spleen hilus were sampled and
therefore it is possible that some veins were patent. Absence
of flow was determined by lack of measurable flow veloc-
ities with spectral Doppler. Detection of low-velocity ve-
nous flow requires appropriate Doppler instrument settings.
The ultrasound beam should be near parallel to the vessel
and the pulsed Doppler sample volume should span the
vessel width. l6 Low pulse repetition frequency (low veloc-
ity scale), low filtration and high Doppler gain should be
selected.’
The availability of color Doppler imaging in six dogs
permitted more rapid assessment of the splenic vasculature.
Accurate detection of venous flow with color Doppler re-
quires similar instrument setting to that of spectral Doppler:
low pulse repetition frequency, low filtration and high color
Doppler gain.’ In the six dogs in the study reported here,
color Doppler imaging complemented spectral Doppler; no
VOL. 39, No. 4 DOPPLER
WITH SPLENIC TORSION
measurable flow velocity was detected in the splenic veins
using either modality. Color Doppler imaging was per-
formed on the two dogs with no visible intraluminal echoge-
nicities; the absence of detectable velocities using color
Doppler prompted further evaluation with spectral Doppler.
Because the splenic arteries are not normally seen with
B-mode imaging, they were not assessed in all dogs. Color
Doppler allows the evaluation of flow in vessels which are
not seen with B-mode imaging. When evaluating the splenic
hilus with color Doppler, the splenic artery adjacent to the
vein will be included in the sampling area. The ultrasound
reports failed to mention the status of the splenic artery flow
in all 15 dogs, but in the color Doppler images of one dog,
there was no evidence of blood flow in the splenic artery
adjacent to the occluded vein. The status of splenic arterial
blood flow with splenic torsion may be dependent on the
degree of rotation and duration of the torsion. With the
increasing availability of color Doppler imaging, investiga-
tions into splenic arterial blood flow integrity should be
undertaken.
Splenic thrombosis and infarction is not pathognomonic
REFERENCES
I . O’Brien TR. Radiographic diagnosis of abdominal disorders in the
dog and cat, 2nd ed. Philadelphia: WB Saunders Co, 1978.
2. Valli VEO. The hematopoietic system. In: Jubb KVF, Kennedy PC,
Palme N. Pathology of domestic animals, 4th ed. San Diego: Academic
Press Inc., 1993.
3 . Lipowitz AJ, Blue J. Spleen. In: Slatter D. Textbook of small animal
surgery, 2nd ed. Philadelphia: WB Saunders Co, 1993.
4. Stickle RL. Radiographic signs of isolated splenic torsion in dogs:
eight cases (1980-1987). J Am Vet Med Assoc 1989;194:103-106.
5. Konde LJ, Wrigley RH, Lebel JL, Park RD, Pugh C, Finn S. Sono-
graphic and radiographic changes associated with splenic torsion in the
dog. Vet Radiol 1989;30:41-45.
6. Schelling CG, Wortman JA, Saunders HM. Ultrasonic detection of
splenic necrosis in the dog-three case reports of splenic necrosis second-
ary to infarction. Vet Radiol 1988;29:227-233.
7. Neath PJ, Brockman DJ, Saunders HM. Retrospective analysis of
nineteen animals suffering isolated torsion of the splenic pedicle. J Small
Anim Pract 1997;38:387-392.
8. Pozniak MA, Zagzebski JA, Scanlan KA. Spectral and color Dopp-
lcr artifacts. RadioGraphics 1992;12:35-44.
9. Couto CG, Hammer AS. Diseases of the lymph nodes and the
spleen. In: Ettinger SJ, Feldman EC (eds). Textbook of veterinary internal
medicine, 4th ed. Philadelphia: WB Saunders, 1995.
353
for splenic torsion. Thrombosis of the splenic arteries and
veins can occur in hypercoagulable conditions secondary to
hepatic disease, renal disease, hyperadrenocorticism, hem-
orrhagic pancreatitis and immune hemolytic anemia.2217
Splenic thrombosis and infarction is more likely in any dis-
ease process which causes splenomegaly. For example,
thrombosis and infarction may occur with splenic enlarge-
ment due to lymphoma or 1 e ~ k e m i a . l ~
Due to the varied ultrasonographic appearance of the
spleen with torsion, B-mode and Doppler assessment of the
splenic vasculature becomes important. The presence of
splenomegaly with a hypoechoic, coarse/‘ ‘lacy” parenchy-
mal appearance with B-mode imaging is highly suggestive
of splenic torsion. With B-mode imaging, the splenic veins
may be normal or enlarged with or without intraluminal
echoes suggestive of thrombi. Because thrombus age can
determine echogenicity, an occluding thrombus may be
present despite an anechoic venous lumen. Therefore, spec-
tral and/or color Doppler imaging of the splenic vasculature
can be crucial to the ultrasonographic diagnosis of vascular
occlusion and splenic torsion.
10. Barton CL. The spleen: pathophysiology of disease. In: Bojrab MJ
(ed). Pathophysiology in small animal surgery. Philadelphia: Lea & Fe-
biger, 1981.
11. Wood AKW, McCarthy PH, Angles JM. Ultrasonographic-
anatomic correlation and imaging protocol for the spleen in anesthetized
dogs. Am J Vet Res 1990;51:1433-1438.
12. Raghavendra BN, Horii SC, Hilton S, Subramanyam BR, Rosen RJ,
Lam S. Deep venous thrombosis: detection by probe compression of veins.
J Ultrasound Med 1986;5:89-95.
13. Anderson JC, Baltaxe HA, Wolf GL. Inability to show clot: one
limitation of ultrasonography of the abdominal aorta. Radiology 1979;132:
693-696.
14. Sigel B, Machi J, Beitler JC, Justin JR. Red cell aggregation as a
cause of blood-flow echogenicity. Radiology 1983;148:799-802.
15. Sigel B, Coelho JC, Spigos DG, Flanigan DP, Schuler JJ, Kasprisin
DO, Nyhus LM, Capek V. Ultrasonography of blood during stasis and
coagulation. Invest Radiol 1981;16:71-76.
16. Burns PN. Interpretation and analysis of Doppler signals. In: Taylor
KJW, Burns PN, Wells, PNT. Clinical applications of Doppler ultrasound.
New York: Raven Press, 1988.
17. Hardie EM, Vaden SL, Spaulding K, Malarkey DE. Splenic infarc-
tion in 16 dogs: a retrospective study. J Vet Intern Med 1995;9:141-148.