| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Report |
1
Department of Radiology, University of North Carolina School of Medicine,
Campus Box 7510, Rm. 2016, Old Clinic Bldg., Manning Dr., Chapel Hill, NC
27599-7510.
2
Division of Hematology, Department of Medicine, University of North Carolina
School of Medicine, Chapel Hill, NC 27599-7510.
Received February 22, 2001;
accepted after revision April 6, 2001.
Presented at the annual meeting of the American Roentgen Ray Society,
Seattle, April-May, 2001.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CONCLUSION. The most common predisposing factors in superior mesenteric vein thrombosis with radiologically occult cause are recent abdominal surgery, infection, and hypercoagulable states. Although no correlation was noted between risk factor and outcome, the presence of bowel wall thickening and mesenteric congestion on CT or MR imaging was associated with the development of bowel ischemia. Prognosis is good in this group of patients, with a mortality of only 7%, although bowel ischemia was noted in 21%.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Clinical information and laboratory results were obtained by review of the patient's chart. Platelet count was recorded if obtained within 48 hr of diagnosis. Other laboratory data were recorded if drawn at any time in the patient's workup. An inherited or acquired hypercoagulable abnormality was deemed present if the platelet count was greater than or equal to 800 x 106/mL; if protein S, protein C, or antithrombin III activity levels were less than 60% of normal; if anticardiolipin antibodies or homocysteine levels were elevated to greater than twice normal; or if the factor V Leiden, prothrombin G20210A mutation, or lupus anticoagulant were present. Results of the sugar water test and CD59 assay for paroxysmal nocturnal hemoglobinuria were also listed. The presence of bowel ischemia or infarction was determined from observation at surgery or autopsy.
Imaging studies were evaluated if they were performed within 3 days of diagnosis, before any catheter intervention, and could be located (n = 38). All CT and MR imaging examinations were performed with IV contrast material. CT and MR scans were evaluated by consensus interpretation of three experienced abdominal radiologists. Acute thrombosis was defined as complete nonopacification or as a sharply marginated filling defect visible in a normal or enlarged vein. The clot was classified as either occlusive or central if a central thrombus was present or if the vein was completely nonopacified and sufficient time had elapsed for contrast material to have filled the vein. The clot was characterized as eccentrically located if it was on one side of the vein and there was flow around it. Occlusive and central clots were considered together because it was not always possible to distinguish between possible peripheral flow around a central clot and marked venous wall enhancement. The clot was judged to be continuous if no uninvolved vein was present between involved segments. The remainder of the abdomen was evaluated for the presence of mesenteric congestion and thickened bowel wall.
Basic descriptive statistics (mean, range) were calculated for clinical data as well as for imaging findings. Comparison of continuous data between groups with and without ischemia was performed with an unpaired t test. Comparison of categoric data between groups with and without ischemia was performed using a 2 x 2 contingency table analysis and Fisher's exact test. A p value equal to or less than 0.05 was required for rejection of the null hypothesis. StatView 5 software (SAS Institute, Cary, NC) was used for the statistical analysis.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Forty patients had abdominal or systemic symptoms at the time of diagnosis; however, in three patients the symptoms were not believed to be related to thrombosis. Abdominal pain was the most common symptom and was present in 84% of patients. Nausea or vomiting was present in 56%, as was fever or chills. Diarrhea was present in 23%, with blood noted in the stool in 23%. The mean duration of symptoms before diagnosis was 9.1 days (range, 1-42 days). Only six patients had a history of prior deep venous thrombosis.
Predisposing Factors
Twenty patients had a history of surgery within the 3 months before
diagnosis (splenectomy [n = 5], ileocolectomy [n = 3], right
hemicolectomy [n = 3], total colectomy [n = 3], partial
colectomy with pelvic mass excision [n = 1], cholecystectomy
[n = 2], cholecystectomy and ileocolectomy [n = 1], excision
of ovarian cyst [n = 1], lysis of adhesion [n = 1], and
small-bowel resection [n = 1] (Figs.
1,2A,2B,2C,3).
One patient had two separate surgeries during this time. The mean time from
surgery to clot diagnosis in these 20 patients was 26.7 days (range, 3-82
days).
|
|
|
|
|
An infection was noted in the 6 weeks before clot diagnosis in 20 patients (Fig. 4A,4B). This consisted of appendicitis with abscess (n = 4), diverticulitis with abscess (n = 1), cholecystitis (n = 1), postoperative infection (wound or surgical bed [n = 5], infected ascites [n = 1]), urinary tract infection (n = 3), pelvic inflammatory disease (n = 1), sepsis (n = 1), subacute bacterial endocarditis (n = 1), and pneumonia (n = 1). One patient had been diagnosed with both Clostridium difficile colitis and a urinary tract infection.
|
|
Fourteen inherited or acquired hypercoagulable abnormalities were identified in 12 patients. The G to A point mutation at nucleotide position 1691 in the factor V gene (factor V Leiden) was evaluated in 19 patients. Two patients were noted to be heterozygous for this mutation. The G to A point mutation at position 20210 in the prothrombin gene was evaluated in 11 patients. Two patients were noted to be heterozygous for this mutation (prothrombin polymorphism).
Protein S activity was low in three of the 18 patients in whom it was checked. A platelet count was drawn within 48 hr of diagnosis in 36 patients and was greater than or equal to 800 x 106/mL in six patients. Homocysteine was evaluated in 11 patients and was elevated to greater than twice normal in one patient. No abnormality of protein C, antithrombin III, anticardiolipin antibodies, or lupus anticoagulant was noted in the patients tested (protein C [n = 26], antithrombin III [n = 27], anticardiolipin antibodies [n = 17], and lupus anticoagulant [n = 17]). Paroxysmal nocturnal hemoglobinuria was evaluated by the sugar water test in 13 patients and by a CD59 assay in one patient, but no positive results were noted.
Oral contraceptive use was noted in three patients. One patient was using medroxyprogesterone acetate injectable suspension (Depo-Provera; Pharmacia and Upjohn, Kalamazoo, MI). Five additional patients were using either oral or transdermal estrogen.
Five patients had a history of inflammatory bowel disease (Crohn's [n = 3], ulcerative colitis [n = 2]). A myeloproliferative disorder was noted in three patients (polycythemia rubra vera [n = 1], essential thrombocytosis [n = 1], and myelofibrosis [n = 1]). One patient had sickle cell disease. One patient was pregnant at the time of superior mesenteric vein thrombosis. Six patients had a history of malignancy diagnosed or treated in the 6 months before thrombosis. Two patients had known residual disease.
Nine patients (21%) had none of these predisposing factors (Fig. 5). A single risk factor was identified in 12 patients (28%), with two factors noted in 11 patients (26%). Seven patients (16%) had three risk factors noted, and four patients (9%) had four predisposing factors identified.
|
Appearance on CT or MR Imaging
Clot involved only the superior mesenteric vein in 11 patients and was
continuous in 31 patients. The clot was completely occlusive or centrally
positioned in 33 patients (87%). Mesenteric congestion was seen in 20
patients. Small-bowel wall thickening was noted in 10 patients. All patients
with bowel wall thickening had occlusive or central clot and mesenteric
congestion.
Therapy and Outcome
Systemic anticoagulation was used in 36 patients. Transcatheter clot lysis
was attempted in 13 patients. In four patients, clot lysis was performed by
superior mesenteric artery infusion only. In five patients, a transhepatic
transportal approach was used, and in four patients, both approaches were
used. Lytic therapy was given for a mean of 2 days (range, 1-3 days).
Follow-up studies after transcatheter therapy showed no considerable change in
five patients. Seven patients showed radiographic evidence of either partial
or complete clearing of clot from the superior mesenteric vein. In one
patient, no followup studies were undertaken; however, at autopsy, this
patient had persistent clot in the superior mesenteric vein. Two patients had
intraperitoneal bleeding as a complication of the procedure.
Nine of the 43 patients developed bowel ischemia. All but three patients were alive 60 days after thrombosis. The three patients who succumbed all had necrotic bowel. The outcome correlation is summarized in Table 1.
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In general, superior mesenteric vein thrombosis has an acute or subacute presentation. It may mimic the findings in acute mesenteric arterial ischemia. Commonly, however, the presentation is less classic and severe. In a study by Rhee et al. [1], 75% of patients had been symptomatic for longer than 48 hr at the time of presentation. Most patients in our subset (86%) were also symptomatic at the time of diagnosis, although the symptoms were often nonspecific and nonlocalizing. Diagnosis was also frequently delayed. These findings are consistent with the variable nature of both the clot itself and its impact on the bowel circulation.
Our patient sample showed a variety of risk factors, with recent abdominal surgery and infection being the most common. Thrombosis associated with surgery may be a result of transient hypovolemia or a release of tissue thromboplastins at surgery. Eleven of our cases occurred after partial or total colectomies, which would have resulted directly in less flow in the splanchnic circulation. Splenectomy for both trauma and chronic hematologic conditions has also been associated with transient thrombocytosis and thrombosis in the splenic, portal, and superior mesenteric veins [5, 6]. Similarly, in our study, platelet counts greater than or equal to 800 x 106/ mL were noted in four of the five patients who had undergone recent splenectomy. In patients with infection, thrombosis is presumably associated with an ascending phlebitis [7]. This association emphasizes both the importance of evaluating the entire abdomen and pelvis for infection once superior mesenteric vein thrombosis has been found, and the converse, looking for superior mesenteric vein thrombosis in patients with intraabdominal infection. This is particularly relevant in an era in which focused CT examinations without IV contrast material are becoming popular. Interestingly, in seven of our patients, extraperitoneal infection was present. In these cases, the pathogenesis of the thrombosis is less clear but may be a result of circulating endotoxin or other thrombogenic factors.
An inherited or acquired hypercoagulable abnormality was the third most common risk factor. Deficiencies of protein S, protein C, and antithrombin III, as well as mutations in factor V and prothrombin, have all been cited as risk factors in superior mesenteric vein thrombosis. Thrombocytosis, paroxysmal nocturnal hemoglobinuria, sickle cell disease, myeloproliferative disorders, and homocystinemia have also been implicated [1, 4, 8]. In our data, as well, a variety of defects were noted. We suspect that our data underestimate the true prevalence of coagulation abnormalities because a substantial number of our patients were not fully evaluated.
In approximately half our patients, two or more risk factors were present, suggesting that the thromboses noted in this series were often the result of multiple insults. This belief is substantiated in the literature in which studies have shown, for example, that the fourfold risk of deep venous thrombosis associated with oral contraceptive use alone is increased to a 30-fold risk when the person has factor V Leiden genotype [9].
Prior published series have shown a substantial idiopathic rate ranging from 19% to 46% [1,2,3], similar to that noted in our subset (21%). This persistent idiopathic rate may be explained by the fact that our understanding and evaluation of coagulation remains incomplete or that thrombosis in some cases is a result of the summation of multiple subclinical events.
Although surgery is indicated when bowel infarction with peritoneal signs is present, the treatment of mesenteric venous thrombosis without bowel infarction is less certain. Suggested treatment has included thrombectomy [10], mechanical and pharmacologic percutaneous clot lysis [11, 12], and simple anticoagulation. To our knowledge, no controlled trials comparing these therapeutic options have been published. The relative lack of radiologic success with transcatheter lysis in this study may reflect either rapid organization of thrombus or the possibility that partial thrombosis may have been longstanding, with the patient's clinical presentation occurring only with complete occlusion. Despite this lack of objective success with transcatheter therapy, only one of the 13 patients treated went on to develop bowel ischemia.
Although our sample size is small, ischemia was not statistically more common with any particular risk or demographic subgroup. However, imaging findings of mesenteric congestion and bowel wall thickening were associated with ischemia, and their presence should alert the clinician to a possible negative outcome. Previous studies have generally shown a worse prognosis than noted in this series [1, 3, 4]. The difference may reflect the inclusion of patients with nonoccult causes, such as pancreatic malignancy. The increasing use of CT in the evaluation of vague abdominal pain may also have allowed the diagnosis of patients in our series that would earlier have been overlooked.
Although this paper provides a reasonable snapshot of our present understanding of mesenteric venous thrombosis with radiologically occult cause, there are limitations inherent in a retrospective study. It is difficult to reconstruct a patient's exact presentation and course from a chart review. Therapy is undertaken on a case-by-case basis rather than systematic basis. Laboratory evaluation is not uniform. Bias can also be introduced by case selection methodologies. For example, patients diagnosed only at surgery or at autopsy or those with falsely negative imaging studies would not have been included in this cohort.
In summary, the most common predisposing factors in superior mesenteric vein thrombosis with radiologically occult cause are recent abdominal surgery, infection, and hypercoagulation. Prognosis is fairly good in this subset of patients, with a mortality of only 7%, although bowel ischemia was noted in 21%.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  M. E. Baker, F. Remzi, D. Einstein, M. Oncel, B. Herts, E. Remer, and V. Fazio CT Depiction of Portal Vein Thrombi after Creation of Ileal Pouch-Anal Anastomosis Radiology, April 1, 2003; 227(1): 73 - 79. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. S. Kashyap, S. Kashyap, D. M. Warshauer, J. K. T. Lee, M. A. Mauro, and G. C. White II Mesenteric Venous Thrombosis Am. J. Roentgenol., October 1, 2002; 179(4): 1072 - 1073. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
