The American Journal of Surgery 181 (2001) 20 –23

Scientific papers

A retrospective study of diagnosis and management of mesenteric vein

thrombosis
Celia M. Divino, M.D.a,*, In Soon Park, M.D.a, Luz P. Angel, M.D.a, Sharif Ellozy, M.D.a,
Renee Spiegel, M.D.b, Unsup Kim, M.D.b
a
Department of Surgery, Division of Surgical Oncology, Mount Sinai Medical Center, One Gustave L. Levy Place, Box 1259, New York, NY 10029, USA
b
Department of Surgery, Elmhurst Hospital Center, Elmhurst, NY 11373, USA

Received February 2, 2000; revised manuscript August 21, 2000

Abstract
Background: Mesenteric vein thrombosis (MVT) is an uncommon type of intestinal ischemia associated with significant mortality and
morbidity because of its delay in diagnosis.
Methods: A retrospective analysis of 9 patients treated surgically for MVT during 1982 to 1997 was performed.
Results: Nine patients underwent surgical therapy for intestinal ischemia due to MVT. The most common presenting symptom was
abdominal pain with bloody diarrhea in 3 patients; preoperative diagnosis of MVT was suspected in 2. Radiologic tests included plain
roentgenograms, computed axial tomography, and ultrasound. Time to surgery ranged from 3 hours to 7 days after admission. All patients
underwent resection of infarcted bowel with primary anastomosis and immediate postoperative anticoagulation. No patient underwent a
second-look operation. The postoperative morbidity and mortality rates were 55% and 11%, respectively.
Conclusion: Diagnosis of intestinal ischemia from MVT is often delayed, and strong clinical suspicion and aggressive treatment are
necessary in its management. © 2001 Excerpta Medica, Inc. All rights reserved.

Keywords: Mesenteric venous thrombosis; Radiographic findings; Surgery

Introduction eases, use of estrogen-containing compounds, polycythemia

Mesenteric vein thrombosis (MVT) is a rare but well-
recognized cause of bowel infarction. In contrast to mesen-
teric infarction due to arterial occlusion, infarction due to
venous thrombosis is subacute in its presentation and in-
volves a limited, usually well-demarcated segment of
bowel. Early surgical intervention with postoperative anti-
coagulant therapy results in a favorable outcome. However,
because mesenteric vein thrombosis has a vague clinical
presentation and lacks specific diagnostic tests, early diag-
nosis is rare. The clinical presentation of MVT can range
from abdominal pain of insidious onset without physical
findings to peritonitis with bloody diarrhea and leukocytosis
[1–3]. MVT may occur in association with portal hyperten-
sion, abdominal neoplasm, trauma, inflammatory bowel dis-
* Corresponding author. Tel.: ⫹1-212-241-6591; fax: ⫹1-212-534-
2654.
E-mail address: celia.divino@mountsinai.org.
vera, and hypercoagulable states resulting from deficiencies
in the intrinsic plasma anticoagulant system, ie, antithrom-
bin III, protein C, protein S, and factor V Leiden [4].
Methods
A retrospective analysis of clinical presentation, diagnos-
tic modalities, operative findings, and surgical outcome of
MVT was performed. From 1982 to 1997, 12 patients were
treated for MVT at Elmhurst Medical Center; 9 underwent
surgical exploration owing to intestinal infarction as a result
of MVT. Three additional asymptomatic patients with inci-
dental computed axial tomography (CT) findings of MVT
were managed with anticoagulation by the medical service
and excluded from this study. Except for the 1 postoperative
mortality, the 8 patients who survived were followed up
either as outpatients in clinic or by telephone contact. Fol-
low-up time ranged from 2 to 10 years after surgery.

0002-9610/01/$ – see front matter © 2001 Excerpta Medica, Inc. All rights reserved.
PII: S 0 0 0 2 - 9 6 1 0 ( 0 0 ) 0 0 5 3 2 - 8
 C.M. Divino et al. / The American Journal of Surgery 181 (2001) 20 –23 21

Table 1
Preoperative diagnosis of patients with mesenteric vein thrombosis

Preoperative diagnosis Number of patients

Mesenteric vein thrombosis 2

Small bowel obstruction 3
Hepatitis 1
Appendicitis 1
Peritonitis 2

Results

Ages of the patients (7 men, 2 women) ranged from 31 to

74 years. All 9 patients initially presented with abdominal
pain of insidious onset associated with nausea and vomiting;
3 had bloody diarrhea. The symptoms’ duration ranged
from 2 days to 3 weeks. Upon admission, 6 patients had
minimal abdominal findings, whereas 3 had signs of acute
peritonitis. MVT was suspected in 2 patients preoperatively
because of known prothrombotic conditions; preoperative
diagnoses in the remaining seven are listed in Table 1. Time
to laparotomy ranged from 3 hours to 7 days, with 3 patients
undergoing emergency surgery because of signs of perito-
nitis. The primary cause of delay in surgery was inability to
establish the diagnosis.
Risk factors
MVT was suspected in 2 patients preoperatively because
of known prothrombotic conditions; 1 had polycythemia
vera and another had protein C deficiency. A third patient
had factor V Leiden deficiency diagnosed postoperatively
on hematologic evaluation. The cause of MVT in the re-
maining 6 patients was thought to be idiopathic.
Diagnostic tests and laboratory findings
All patients had leukocytosis ranging from 12,000 to
29,000 cells per cubic millimeter, and evidence of intravas-
cular depletion with admission hematocrit ranging from
47% to 57%. Plain abdominal roentgenograms were per-
formed in all patients, with nonspecific findings ranging
from scarcity of bowel gas to partial small-bowel obstruc-
tion (Fig. 1). An upper gastrointestinal series performed in
1 patient showed rigid small-bowel loops with edematous
bowel wall. CT scan of the abdomen was performed in 2
patients, which showed thrombus within the superior mes-
enteric vein (SMV) and dilated, fixed, thick-walled small-
bowel loops (Fig. 2). In 1 of these patients, thrombus was
visualized in the portal, superior mesenteric, and splenic
veins. Duplex ultrasound with color flow showed a throm-
bosed portal vein in another patient.
Fig. 1. Radiologic studies obtained preoperatively.
foration or peritonitis. The infarcted segment ranged from 1
1/2 feet to 12 feet. The mesentery was thickened with edema
and diffuse petechiae. Multiple thrombi were seen in the
mesenteric veins (see Fig. 3). Good arterial pulsations, and
varying amounts of hemorrhagic ascites were noted in all
patients. The infarcted segment of bowel was resected with
generous margins, and primary anastomosis was performed
in all patients. Anticoagulation was started immediately
postoperatively with conversion to oral warfarin at appro-
priate times.
Postoperative course
One of the 9 (11%) patients died. A 31-year-old man
who had an extensive evaluation at another institution was
explored 3 days after admission when he failed to improve
on nonoperative measures. At exploration, 10 feet of small
bowel from the midjejunum to proximal ileum showed
hemorrhagic infarction. Although the patient tolerated the
procedure, he arrested in the recovery room and the cause of
death was attributed to hypoxia secondary to premature
extubation.
The overall morbidity rate was 55%. Ascites responsive
to diuretic therapy developed in 3 patients; 1 developed
short-gut syndrome secondary to extensive bowel resection.

Operative findings and surgical procedures

All 9 patients had segmental infarction of the small Fig. 2. Blank arrow directed at the thrombus within the superior mesen-
intestine (Fig. 3). No patient was found to have frank per- teric vein. Smaller arrows point to fixed, thickened small bowel.
22 C.M. Divino et al. / The American Journal of Surgery 181 (2001) 20 –23
Fig. 3. Infarcted small bowel with multiple thrombi within mesenteric
veins.
Treatment with antispasmodics and elemental diet resulted
in relief of symptoms. One patient developed portal hyper-
tension 6 months after surgery. She subsequently bled from
esophageal varices, necessitating emergency surgery with
devascularization of the lower esophagus, ligation of the
azygous-coronary system, and splenectomy. Her postoper-
ative course was complicated but had a successful outcome
eventually. One patient underwent serial sonographic exam-
inations to evaluate the effect of anticoagulation on the
patency of the vessels. After 8 months of anticoagulation
therapy, no dissolution of the thrombus was noted, but
extensive venous collateralization around the SMV and por-
tal vein was present.
Comments
Since Warren and Eberhard first described mesenteric
vein thrombosis as a distinct cause of bowel infarction in
1935 [5] the overall outcome of this disease has not im-
proved significantly. The mortality rate remains approxi-
mately 20%. The principle cause of mortality is the diffi-
culty in diagnosing this disease, which results in the delay of
surgical intervention [1]. The condition is frequently mis-
diagnosed initially as partial small-bowel obstruction or
gastroenteritis, and a definitive diagnosis is usually made at
laparotomy or autopsy.
Clinical signs and symptoms of intestinal ischemia due
to MVT are nonspecific. Clinical presentation includes
vague, often crampy abdominal pain of insidious onset
lasting days to weeks, and bloody diarrhea associated with
abdominal pain with minimal physical findings. A high
white cell count was noted in all the patients; however, this
was not helpful in diagnosis because an increased white cell
count itself is not specific to this disease. Plain roentgeno-
grams or small-bowel contrast studies can be helpful when
they show either a scarcity of gas or distended, rigid, fluid-
filled small-bowel loops [2,6]. These findings, however, are
subtle and often overlooked or misinterpreted.
More sensitive imaging modalities are currently avail-
able to evaluate MVT [7–11]. For example, in a CT scan of
the abdomen with contrast, the lumen of the thrombosed
vein does not enhance in comparison to the surrounding
structures. Occasionally, periportal enhancement is present,
possibly reflecting the proliferation of the vasa vasorum.
Other CT findings include ascites, bowel dilatation, wall
thickening, mesenteric stranding, submucosal edema and
hemorrhage (areas of varying attenuation), and pneumatosis
intestinalis. Rhee and colleagues [12] noted a 100% sensi-
tivity rate for CT scan in detecting any abnormalities asso-
ciated with MVT. Similarly, both CT scans performed in the
2 patients in this study resulted in significant findings.
Duplex sonography of the abdomen, with careful evaluation
of the portal and mesenteric vessels, has also proved to be
useful in some patients with MVT [10,11]. Acute thrombus
may be anechoic; however, on color flow Doppler, a flow
void is seen. The lack of flow, despite adequate scan angle
and adequate access, is diagnostic of mesenteric vein throm-
bosis. Magnetic resonance imaging (MRI) is comparable to
CT scan in evaluating MVT, and may be more useful when
surgical clips obscure the field. On T1-weighted images,
thrombus is hyperintense relative to the liver. Angiography,
whereas more invasive than the studies mentioned previ-
ously, could assess patency of the vessels, collateral flow
pathways, nonoccluding thrombus, and direction of flow
accurately [10]. These imaging modalities should be used
early, when MVT is suspected.
Segmental resection of the involved bowel with primary
anastomosis is accomplished easily, because the hemor-
rhagic infarction associated with MVT is limited. Although
Pavel and coworkers [13] advocated a second-look opera-
tion and enterostomy in some patients, none of the patients
in this study required either procedure. If resection of the
infarcted bowel is followed by immediate anticoagulant
therapy, progression of the disease is uncommon and a
second-look operation is rarely indicated. Resection of the
involved bowel with wide margins and involved mesentery
to assure proper healing of the anastomosis and prevent
propagation of thrombi from the mesentery is critical.
Although all the patients were treated surgically, throm-
bolytic therapy may have been another treatment option for
some. Poplausky and colleagues [14] reported successful
treatment of MVT with the use of localized thrombolytic
therapy. After infusion of urokinase into the superior mes-
enteric artery in a patient with an acute symptomatic MVT,
both clinical and radiologic improvement were noted within
24 hours, and surgery was avoided. Thrombolytic therapy
used through the transjugular and transhepatic vein has also
resulted in favorable outcomes [15,16]. However, for
thrombolytic therapy to be an effective treatment option,
diagnosis of MVT must first be established firmly, which
proved to be difficult in this study. Correct diagnosis was
suspected in only 2 patients. Furthermore, the patient must
 C.M. Divino et al. / The American Journal of Surgery 181 (2001) 20 –23
be stable clinically with no evidence of infarcted bowel and
no contraindication to thrombolytic therapy. Therefore,
thrombolytic therapy is an effective treatment in a limited
number of patients with MVT.
The role of anticoagulant therapy in MVT is well estab-
lished [13,17–19]. Matthews and White [20] reported far
better survival rates among patients who received anticoag-
ulant therapy as compared with those who did not (63%
versus 44%). Anticoagulant therapy was shown to prevent
propagation of thrombi to the portomesenteric venous sys-
tem and promote formation of collaterals [1]. Although
dissolution of mesenteric vein thrombosis with anticoagu-
lant therapy has been reported [11], no information exists in
the literature regarding the fate of the mesenteric thrombus
with anticoagulation. The potential use of serial sono-
graphic duplex studies to evaluate evolution of the thrombus
is unknown. The optimal duration of anticoagulant therapy
has not been defined; however, recommended is that anti-
coagulation should be continued indefinitely, because it
reduces the incidence of recurrence from 30% to 40% to 3%
to 5% [19].
Although only 3 patients in this study had signs of
peritonitis, surgical exploration was necessary in the other
patients because of lack of diagnosis and worsening abdom-
inal symptoms. However, surgical management of MVT is
not always necessary. Incidental asymptomatic thrombus in
mesenteric vein should be managed medically. In cases of
mild MVT, a trial of nonoperative management— consist-
ing of fluid resuscitation, anticoagulation, and thromboly-
sis—may be acceptable in clinically stable patients with
early diagnosis. However, if patients show clinical signs of
peritonitis or deteriorates on medical management, prompt
surgical intervention is warranted.
Conclusions
Mesenteric vein thrombosis remains a difficult problem.
The mortality rate remains high because of delay in diag-
nosis. A careful history with attention to possible risk fac-
tors, supported by radiographic studies— especially with
CT scan and color flow Doppler—aid in making the diag-
nosis. Early surgical intervention with resection of infarcted
bowel with good margins and involved mesentery and
prompt anticoagulant therapy postoperatively are critical for
a favorable outcome.
23
Acknowledgments
The authors thank Dr. Morris Kerstein for helpful com-
ments, and Gae O. Decker-Garrard for editorial assistance.
References
[1] Abdu RA, Zakhour BJ, Dallas DJ. Mesenteric venous thrombosis
1911 to 1984. Surgery 1987;101:383– 8.
[2] Grendell JH, Ockner RK. Mesenteric venous thrombosis. Gastroen-
terology 1982;82:358 –721.
[3] Kirshner PA. Occlusion of the mesenteric arteries and veins with
infarction of the bowel. Ann Mt Sinai Hosp 1955;21:307–17.
[4] Bontempo FA, Hassett AC, Faruki H, et al. The factor V Leiden
mutation: spectrum of thrombotic events and laboratory evaluation. J
Vasc Surg 1997;25:271– 6.
[5] Warren S, Eberhard TP. Mesenteric venous thrombosis. Surg Gy-
necol Obstet 1935;61:102–21.
[6] Witte CL, Brewer ML, Witte MS, et al. Protean manifestations of
pyelothrombosis. Ann Surg 1935;202:191–202.
[7] Rosen A, Korobkin M, Silverman PM, et al. Mesenteric vein throm-
bosis: CT identification. Am J Radiol 1984;143:84 – 6.
[8] Sommer A, Jaschke W, Georgi M. CT diagnosis of acute mesenteric
vein thrombosis with intestinal infarction. Akt Radiol 1994;4:344 –7.
[9] Kim JY, Byun JY, Lee JM, et al. Intestinal infarction secondary to
mesenteric venous thrombosis: CT pathologic correlation. J Comput
Assist Tomogr 1993;17:382–5.
[10] Abbitt PL. Portal vein thrombosis: imaging features and associated
etiologies. Curr Probl Diagn Radiol 1992;21:115– 47.
[11] Verbanck JJ, Rutgeerts LJ, Haerens MH, et al. Partial splenoportal
and superior mesenteric venous thrombosis. Early sonographic diag-
nosis and successful conservative management. Gastroenterology
1984;86:940 –52.
[12] Rhee RY, Gloviczki P, Mendonca CT, et al. Mesenteric venous
thrombosis: still a lethal disease in the 1990s. J Vasc Surg 1994;20:
688 –97.
[13] Pavel J, Levy MM, Krausz JM. The role of second-look procedure in
improving survival time for patients with mesenteric vein thrombosis.
Surg Gynecol Obstet 1990;170:287–91.
[14] Poplausky MR, Kaufman JA, Geller SC, Waltman AC. Mesenteric
venous thrombosis treated with urokinase via the superior mesenteric
artery. Gastroenterology 1996;110:1633–5.
[15] Yankes JR, Uglietta JP, Grant J, Braun SD. Percutaneous transhepatic
recanalization and thrombolysis of the superior mesenteric vein. AJR
Am J Roentgenol 1991;151:289 –90.
[16] Rivitz SM, Geller SC, Hahn C, Waltman AC. Treatment of acute
mesenteric venous thrombosis with transjugular intramesenteric
urokinase infusion. J Vasc Interv Radiol 1995;6:219 –28.
[17] Sack J, Aldrete JS. Primary mesenteric venous thrombosis. Surg
Gynecol Obstet 1982;154:205– 8.
[18] Nelson LE, Kremen AJ. Experimental occlusion of the superior
mesenteric vessels with special reference to the role of intravascular
thrombosis and its prevention with heparin. Surgery 1950;154:205– 8.
[19] D’Abreu F, Humble JG. Mesenteric vein thrombosis. Recovery after
resection with heparin. Lancet 1946;1:354 – 6.
[20] Matthews JE, White TP. Primary mesenteric venous occlusive dis-
ease. Am J Surg 1972;122:579 – 83.