Acute superior mesenteric venous thrombosis results in high rates of readmission and morbidity

Abstract

Objective:

Superior mesenteric venous thrombosis (MVT) is a poorly understood clinical entity, and as such, outcomes are poorly described. This study aimed to identify predictors of bowel ischemia after MVT and to compare outcomes for patients treated medically (group 1) with those for patients treated with bowel resection (group 2).

Methods:

This was a retrospective, single-institution study capturing all patients diagnosed with symptomatic acute MVT on computed tomography imaging from 2008 to 2018. Demographics, comorbidities, imaging, laboratory values, and treatment were included. Predictors of bowel resection were analyzed by univariate and multivariate statistics. Outcomes including mortality, readmissions for abdominal pain, and chronic mesenteric venous congestion were compared using χ² test.

Results:

There were 121 patients included in the study; 98 patients were treated medically (group 1), 19 patients were treated with bowel resection (group 2), and 4 patients were treated with endovascular recanalization (group 3). Group 1 and group 2 were compared directly. Patients requiring bowel resection tended to have higher body mass index (P = .051) and a hypercoagulable disorder (P = .003). Patients who required bowel resection were more likely to present with lactic acidosis (P < .001) and leukocytosis (P < .001) with bowel wall thickening on scan (P < .001). On multivariable analysis, a genetic thrombophilia was a strong predictor of bowel ischemia (odds ratio, 3.81; 95% confidence interval, 1.12–12.37). One-year mortality and readmission rates did not differ between groups. However, readmission rates for abdominal pain were high for both groups (group 1, 44.90%; group 2, 57.89%; P = .317), and a significant proportion of patients exhibited chronic mesenteric venous congestion on repeated scan (group 1, 42.86%; group 2, 47.37%; P = .104).

Conclusions:

A genetic hypercoagulable disorder is a predictor of bowel ischemia due to MVT. Regardless of treatment, outcomes after MVT are morbid, with high rates of readmission for abdominal pain. An alternative approach to treat these patients is needed, given the poor outcomes with current strategies.

Acute mesenteric venous thrombosis (MVT) is a poorly understood clinical entity. MVT represents 6% to 9% of all cases of acute mesenteric ischemia. The incidence of MVT is reported to be 2.7 per 100,000 person-years.^1,2 Factors that predispose patients to MVT are well described. They include hypercoagulable states, such as an acquired or inherited thrombophilia; local vessel wall injury, such as inflammation, infection, and malignant disease of the abdominal viscera; recent abdominal surgery; and venous stasis, such as congestive heart failure, splenomegaly, and cirrhosis.³

Patients diagnosed with MVT are typically treated with systemic anticoagulation, bowel rest, and fluid resuscitation. Whereas nearly 95% of patients are adequately managed with anticoagulation alone, there is a subset of patients who develop bowel ischemia and require resection.^4–6 Risk factors for those who fail to respond to conservative management and progress to bowel resection are unknown. Furthermore, long-term outcomes after MVT are not well characterized. There are some case series reporting improved outcomes after percutaneous endovascular treatment of MVT.^7–10 Conclusions are limited by small sample sizes, selection bias, and a poor understanding of appropriate patient selection. In this single-institution series, we sought to identify predictors of bowel infarction due to MVT. Bowel resection after an acute presentation of MVT was used as a surrogate measure of bowel infarction. In addition, we capture patient outcomes after MVT and describe a constellation of signs and symptoms that lead to chronic mesenteric venous congestion.

METHODS

Data source and patient selection.

The data source used for the study was the medical archival system at our institution. This archival system allows all inpatient and outpatient patient encounters, laboratory and pathology results, and imaging reports within the health care system to be queried by Boolean search terms. Inpatient radiology records were queried for all patients who had a computed tomography (CT) scan identifying MVT between January 2008 and December 2018. To identify those patients with acute, symptomatic MVT, a manual chart review was completed. Only patients with acute symptoms (<2 weeks) were included in the study. Patients with chronic pain (defined as >2 weeks), those in whom the MVT was not the source of pain (incidental MVT), patients with MVT due to an active intra-abdominal or retroperitoneal tumor, and patients with end-stage liver disease were excluded from the study. Interestingly, each patient with MVT who also had end-stage liver disease was asymptomatic, and the thrombus was found incidentally on scan; these patients were therefore excluded from the analysis. Incidental MVT was excluded because the acuity of the disease was impossible to reliably capture. MVT associated with an abutting tumor was also excluded as the pathophysio-logic mechanism of the mass effect or erosion of a tumor into the mesenteric venous system is much different from that in the population of patients this study aimed to characterize. Patients who underwent endovascular interventions for the MVT were also excluded from analysis and evaluated separately for descriptive purposes (group 3, n = 4). Patients were separated into two groups: those who did not require bowel resection (group 1) and those who did require bowel resection (group 2) during the index hospital admission for MVT. All patient information was deidentified and patient consent was not required. This study was approved by the University of Pittsburgh Institutional Review Board.

Statistical analysis.

Patients’ demographics, comorbidities, presenting CT scan findings, and anticoagulation details were summarized using descriptive statistics. Comparisons were made using the χ² test and Student t-test with appropriate nonparametric equivalent when appropriate.

The primary outcome of interest was bowel resection after MVT as defined by clinically adjudicated operative report review. Secondary outcomes included bleeding complications, total parenteral nutrition (TPN) dependence, 1-year all-cause mortality, chronic abdominal pain as measured by readmission within 1 year for abdominal pain, and post-thrombotic syndrome. Vitality status was determined by manual chart review and the Social Security Death Index-linked death records. Hospital readmissions included a subsequent admission within any hospital within the large, regional multi-hospital health care system. Post-thrombotic syndrome was captured by the presence of chronic venous congestion and was defined as new venous collaterals, chronic occlusive thrombus, or new portal hypertension noted on follow-up imaging up to 1 year after the initial event.

Multivariable logistic regression was used to identify factors associated with our primary outcome, bowel resection, after accounting for confounding variables. All variables with a P value <.10 were included in the multivariable analysis. All secondary outcomes were compared using χ² test. A P value of <.05 was considered statistically significant. Analyses were completed using Stata SE 15 (StataCorp LLC, College Station, Tex).

RESULTS

Patients’ characteristics.

Fig 1 shows the division of groups. Of the 121 total patients who met all inclusion and no exclusion criteria, 98 did not require bowel resection and were medically managed (group 1); 19 patients underwent bowel resection during the index hospitalization (group 2), and 4 patients underwent endovascular interventions (group 3). Two patients in group 1 underwent an exploratory laparotomy without resection (negative laparotomy) as all bowel was deemed viable intraoperatively by visual inspection.

Fig 1.

Flow diagram demonstrating included and excluded patients in our study cohort. CT, Computed tomography; MVT, mesenteric venous thrombosis; SMV, superior mesenteric vein.

Patients’ demographics, comorbidities, and presenting signs and symptoms are detailed in Table I. Group 1 and group 2 had similar rates of medical comorbidities, such as diabetes, hypertension, hyperlipidemia, and peripheral vascular disease. There were also similar incidences of an active cancer diagnosis, tobacco use, and recent abdominal surgery. There were 20 patients (20.41%) in group 1 and 1 patient (5.26%) in group 2 who had a prior bowel resection. The indications for prior bowel resection in group 1 were as follows: inflammatory bowel disease (n = 5 patients), perforated diverticulitis(n = 4), small bowel obstruction (n = 3), internal hernia after Roux-en-Y gastric bypass (n = 3), perforated small bowel (n = 2), splenic flexure polyp (n = 1), cecal polyp (n = 1), and familial adenomatous polyposis (n = 1). The one patient with a prior history of bowel resection in group 1 had a small bowel obstruction. Group 2 had a higher rate of hypercoagulable disorders compared with group 1 (52.63% vs 20.41%; P = .003). The hypercoagulable states in group 1 were as follows: lupus anticoagulant (n = 9), factor V Leiden (n = 5), paroxysmal nocturnal hemoglobinuria (n = 2), prothrombin gene mutation (n = 2), polycythemia vera (n = 1), and homo-cystinuria (n = 1). The hypercoagulable disorders seen in group 2 were as follows: lupus anticoagulant (n = 4), factor V Leiden (n = 2), antithrombin III deficiency (n = 2), essential thrombocytosis (n = 1), and prothrombin gene mutation (n = 1). Although not statistically significant, group 2 tended to have higher rates of female patients (63.16% vs 40.82%; P = .073) and obesity (52.63% vs 29.59%; P = .051). Group 2 had higher rates of leukocytosis (89.47% vs 34.69%; P < .001), lactic acidosis (47.37% vs 11.22%; P < .001), and bowel wall thickening on initial CT imaging (100% vs 48.98%; P < .001) compared with group 1. Whereas thrombus burden could not be directly calculated, group 1 did have a more extensive anatomic distribution of thrombus on CT imaging (more than two mesenteric veins involved, 27.55% vs 5.26%; P = .037).

Table I.

Baseline demographics, comorbidities, and presenting signs and symptoms of patients treated without bowel resection (group 1) compared with patients treated with bowel resection (group 2)

	Group 1: No bowel resection (n = 98)	Group 2: Bowel resection (n = 19)	P value
Patients’ demographics
Age, years (SD)a	50.75 (14.94)	50.37 (14.05)	.459
Sex, female	40 (40.82)	12 (63.16)	.073
Race, white	85 (86.73)	15 (78.95)	.378
Comorbidities
Obesity	29 (29.59)	10 (52.63)	.051
Diabetes	16 (16.33)	2 (10.53)	.521
Current smoker	36 (36.73)	5 (26.32)	.384
Hypercoagulable disorder	20 (20.41)	10 (52.63)	.003
History of DVT or PE	8 (8.16)	5 (26.32)	.021
Hypertension	37 (37.76)	10 (52.63)	.226
Hyperlipidemia	24 (24.49)	3 (15.79)	.410
End-stage renal disease	3 (3.06)	0 (0.00)	.440
Congestive heart failure	3 (3.06)	0 (0.00)	.440
COPD	7 (7.14)	1 (5.26)	.766
Coronary artery disease	7 (7.14)	1 (5.26)	.776
Active cancer diagnosis	20 (20.41)	2 (10.53)	.313
Inflammatory bowel disease	14 (14.29)	1 (5.26)	.282
Peripheral vascular disease	11 (11.22)	1 (5.26)	.433
Prior splenectomy	2 (2.04)	1 (5.26)	.416
Recent abdominal surgery	18 (18.37)	1 (5.26)	.156
Pancreatitis	13 (13.27)	1 (5.26)	.325
History of bowel resection	20 (20.41)	1 (5.26)	.115
Presenting signs and symptoms
Leukocytosis	34 (34.69)	17 (89.47)	<.001
Elevated lactate	11 (11.22)	9 (47.37)	<.001
Acute kidney injury	7 (7.14)	0 (0.00)	.230
Peritonitis	1 (1.02)	13 (68.42)	<.001
Onset of symptoms in days (SD)	4.30 (4.26)	3.00 (3.32)	.106
Presenting CT findings
Concomitant portal vein thrombus	50 (51.02)	8 (42.11)	.477
Concomitant hepatic vein thrombus	11 (11.22)	1 (5.26)	.433
Concomitant splenic vein thrombus	21 (21.43)	0 (0.00)	.026
Thrombus in >2 veins	27 (27.55)	1 (5.26)	.037
Bowel wall thickening	48 (48.98)	19 (100.00)	<.001

COPD, Chronic obstructive pulmonary disease; CT, computed tomography; DVT, deep venous thrombosis; PE, pulmonary embolism.

Obesity is defined as body mass index >30 kg/m².

^aSD indicates continuous variables with summary measure of mean (standard deviation) and P value from Student t-test. Categorical variables are summarized by number (%), and P values are calculated from the χ² test.

Treatment.

Table II outlines anticoagulation regimens for each group. In group 1, 12.24% of patients were receiving home anticoagulation before presentation. The majority of these patients were taking warfarin (8 of 12 patients). In group 2, two patients were receiving anticoagulation before presentation; one was taking warfarin and one was taking fondaparinux. In group 1, 84 (85.71%) patients were treated initially with a heparin drip. Ten (10.20%) patients were treated with therapeutic low-molecular-weight heparin, and one patient was treated with an argatroban drip. The average time to clinical improvement after initiation of anticoagulation was 2.5 days (standard deviation, 1.89 days). In group 2, 17 (89.47%) patients were treated with a heparin drip, 1 patient was treated with therapeutic low-molecular-weight heparin, and 1 patient was treated with an argatroban drip. Group 2 had a shorter average time from symptom onset to initiation of anticoagulation (3.0 ± 3.32 days) compared with group 1 (4.3 ± 4.26 days). Six of the 19 patients’ bowel resection occurred on the first trip to the operating room. The remaining 13 patients underwent bowel resection at a planned, second-look operation because there were concerns about intestinal viability during the initial operation. The average time to the index operation from initiation of anticoagulation was 0.7 day (±2.19 days). On discharge, 17.35% of patients in group 1 received a direct oral anticoagulant, whereas only 5.26% of patients in group 2 were discharged with a direct oral anticoagulant. The majority of patients were still taking their anticoagulation at 3-month follow-up (58.16% in group 1, 89.47% in group 2). Eleven of the 19 patients in group 2 were receiving lifelong anticoagulation.

Table II.

Anticoagulation profiles of patients treated without bowel resection (group 1) compared with patients treated with bowel resection (group 2)

	Group 1: No bowel resection (n = 98), No. (%)	Group 2: Bowel resection (n = 19), No. (%)	P value
Home anticoagulation
Warfarin	8 (8.16)	1 (5.26)	.541
Rivaroxaban	1 (1.02)	0 (0.00)
LMWH	1 (1.02)	0 (0.00)
Fondaparinux	1 (1.02)	1 (5.26)
Dabigatran	1 (1.02)	0 (0.00)
Initial inpatient anticoagulation
Heparin drip	84 (85.71)	17 (89.47)	.124
Therapeutic LMWH	10 (10.20)	1 (5.26)
Argatroban drip	1 (1.02)	1 (5.26)
Discharge anticoagulation
Warfarin	61 (62.24)	13 (68.42)	.001
Therapeutic LMWH	16 (16.33)	1 (5.26)
Rivaroxaban	15 (15.31)	0 (0.00)
Apixaban	2 (2.04)	1 (5.26)
Fondaparinux	0 (0.00)	3 (15.79)
Dabigatran	1 (1.02)	0 (0.00)
Anticoagulation at 3 months	57 (58.16)	17 (89.47)	.270
Anticoagulation at 6 months	38 (38.78)	11 (57.89)	.757

LMWH, Low-molecular-weight heparin.

P values are calculated from the χ² test.

Outcomes.

The multivariable model using bowel ischemia as an outcome measure is displayed in Table III. Predictors of bowel resection included lactic acidosis on presentation (odds ratio, 6.48; confidence interval, 1.89–22.15; P = .003) and the diagnosis of a hypercoagulable disorder (odds ratio, 3.81; confidence interval, 1.12–12.37; P = .026). However, sex, obesity, and surgical history failed to reach significance.

Table III.

Multivariable analysis of predictors of bowel resection

	OR (95% CI)	P value
Lactate	6.48 (1.89–22.15)	.003
Female	2.92 (0.89–9.60)	.077
Obesity	2.44 (0.73–8.09)	.146
Hypercoagulable disorder	3.81 (1.12–12.37)	.026
Recent abdominal surgery	0.19 (0.01–2.61)	.212
History of small bowel resection	0.04 (0.02–1.87)	.158

CI, Confidence interval; OR, odds ratio.

Number of observations in logistic regression (n = 117).

Outcomes after treatment are displayed as bar graphs in Fig 2. Bleeding events within 1 year of initiation of anticoagulation were similar between groups (15.31% vs 15.79%; P = .957). The majority of these bleeding events were gastrointestinal bleeding. One patient in each group had nonfatal spontaneous intracranial hemorrhage. TPN requirement after MVT was similar between groups (5.10% vs 10.52%; P = .325). In group 1, five patients had a TPN requirement. Two of these patients ultimately required TPN for chronic colitis and inability to tolerate nutrition by mouth, presumably because of mesenteric venous congestion and bowel thickening. Two other patients developed MVT after laparoscopic bowel surgery and were treated with bowel rest and TPN but were ultimately discharged tolerating oral nutrition. Finally, one patient had pancreatitis at the time of MVT and ultimately required TPN. The two patients in group 2 who required TPN did so after development of short gut syndrome from bowel resections. CT imaging at follow-up was assessed for both chronic superior mesenteric vein (SMV) thrombus and chronic mesenteric venous congestion; 40.82% of patients in group 1 had chronic SMV thrombus on repeated scan compared with 26.32% of patients in group 2 (P = .234). Nearly 50% of patients in both groups had signs of post-thrombotic syndrome by mesenteric venous congestion on follow-up scan, but rates were not significantly different between groups (42.86% vs 47.37%; P = .104). Many patients in both groups had chronic abdominal pain with readmissions within 1 year of MVT diagnosis (44.90% vs 57.89%; P = .317). There were three in-hospital deaths in group 1. These patients died of multisystem organ failure, probably secondary to intra-abdominal sepsis. However, because of the decision of the patient or family, they did not undergo an operation. Therefore, in keeping with an intention-to-treat approach, these patients were included in the medically managed group (group 1). At 1 year, 88.78% of patients in group 1 were alive, and 18 of 19 patients in group 2 were alive. One patient died on hospital day 3 after a 90-cm bowel resection. On second look, there was extensive bowel necrosis incompatible with life. Mean follow-up was 562 days; the median follow-up was 254 days.

Fig 2.

Outcomes after mesenteric venous thrombosis (MVT) for patients treated without bowel resection (group 1) compared with patients treated with bowel resection (group 2). A, All-cause bleeding events and total parenteral nutrition (TPN) dependence after MVT for patients treated without bowel resection (group 1) compared with patients treated with bowel resection (group 2). B, Incidence of chronic mesenteric venous congestion and chronic superior mesenteric vein (SMV) thrombus on follow-up imaging for patients treated without bowel resection (group 1) compared with patients treated with bowel resection (group 2). C, Rates of 1-year readmission and 1-year mortality for patients treated without bowel resection (group 1) compared with patients treated with bowel resection (group 2). P values are calculated from the χ² test.

The endovascular group included four patients who underwent endovascular recanalization of the SMV. Suction thrombectomy was performed on all four of these patients. One patient underwent venoplasty and one was treated with a self-expanding covered stent. Two procedures were performed through the trans-splenic route and two were transhepatic. All patients received between 6 and 10 mg of tissue plasminogen activator at time of intervention.

Two of the four patients who underwent endovascular intervention had complete resolution of symptoms. One had persistent abdominal pain requiring an exploratory laparotomy 3 days later, the findings of which were normal; and one had recurrent readmissions for abdominal pain, eventually requiring an exploratory laparotomy 1 year later. Neither of these patients required bowel resection. One patient suffered from rethrombosis of the SMV and chronic mesenteric venous congestion on follow-up imaging. All patients were alive at 1 year, and there were no bleeding complications.

DISCUSSION

This study represents the largest cohort to date describing a single-institution experience managing acute superior MVT both surgically and medically. Of the 121 patients presenting with acute abdominal pain in the setting of MVT, 15.7% developed bowel ischemia requiring a resection. Presenting lactic acidosis and a confirmed genetic thrombophilia were strong predictors of bowel resection. The majority of patients were managed with parenteral heparin initially and discharged with at least 3 months of oral anticoagulation. Regardless of treatment modality, outcomes were morbid.

Presence of a hypercoagulable disorder is a significant predictor of bowel ischemia in the setting of MVT. Previous case reports have demonstrated that mutations such as factor V Leiden and prothrombin mutations may lead to venous thrombosis in unusual areas and at early ages.^11,12 Furthermore, this population of patients is up to three times more likely to form recurrent venous thrombosis.¹³ Similarly, we see that patients with a genetic thrombophilia suffer more aggressive disease. Other traditionally prothrombotic factors, such as malignant disease, tobacco use, and recent abdominal surgery, are not associated with bowel resection in this cohort. Patients with genetic thrombophilias therefore represent a subset of patients for whom MVT portends a poorer prognosis. These patients display more aggressive disease. This is consistent with previous studies,^14,15 including the retrospective analysis conducted by Kumar and Kamath,¹⁴ who concluded that patients with isolated MVT are more likely to have hypercoagulable disorders and more likely to require surgery. Patients with hypercoagulable disorders may benefit from more aggressive treatment of MVT.

Compared with acute arterial mesenteric ischemia, which has reported mortality rates as high as 80%, this cohort had a relatively low mortality rate.¹⁶ One-year mortality rates were <12%, with no significant difference between the two groups. Some studies have reported mortality rates of MVT as high as 25%; however, these are usually in previously sick patients who do not present in the acute setting.^17–20 When patients present acutely and there is prompt treatment, mortality rates have been reported to be as low as 1.25%,¹ consistent with this study. In this series, anticoagulation was initiated promptly. Therefore, swift initiation of anticoagulation, triage, and operative intervention is likely to have a positive impact on mortality rates. The in-hospital mortality rate of 3.31% reported here is similar to that of other single-institution reports.^14,15

Whereas mortality at 1 year was relatively low, readmissions for abdominal pain after MVT are unacceptably high. Overall readmission for abdominal pain was 47.1% at 1 year. Furthermore, nearly half of all patients who present with MVT will have signs of chronic mesenteric venous congestion on follow-up imaging. Few studies have investigated long-term outcomes of MVT. Ageno et al²¹ reported an 11% recurrence rate, whereas secondary portal hypertension has been reported in 11% to 50% of patients after MVT.^20,22 However, chronic MVT has largely been described as an asymptomatic problem, often noted incidentally on follow-up imaging.^22,23 The constellation of symptoms and imaging that is described in this study is most consistent with a post-thrombotic syndrome of the abdominal viscera for which patients seem to suffer recurrent readmissions for abdominal pain. Readmission rates were similar between groups, suggesting that this chronic syndrome persists regardless of whether the affected segment of bowel is resected. Future studies are required to further describe this clinical entity and its cost on the health care system.

Although this was not a prospective study, it was impossible to standardize an anticoagulation protocol for each patient. Duration and type of anticoagulation differed greatly between patients. In general, patients were started on heparin and transitioned to warfarin for at least 3 months. Valeriani et al²⁴ outlined the current treatment paradigms and guidelines for MVT. Whereas data are lacking, the general consensus is to treat MVT with low-molecular-weight heparin bridged to warfarin for 3 to 6 months, assuming there is not an underlying malignant neoplasm or thrombophilia that would require longer treatment. However, as outlined in Table II, many patients in this series did not adhere to treatment recommendations or were lost to follow-up. Such inconsistencies highlight a need for institutional protocols and guidelines for standardized treatment of MVT.

Poor outcomes seen in patients treated with both anticoagulation alone and anticoagulation with bowel resection suggest that an alternative treatment approach may be indicated. In our series, only four patients underwent endovascular recanalization of the SMV. Two patients required subsequent laparotomy and one patient proceeded to develop chronic abdominal pain. This study is limited by a low endovascular sample size. There are some studies that describe promising results with percutaneous techniques to MVT. There are limited series describing single-institution experiences with endovascular recanalization of MVT that describe thrombus resolution in 75% to 100% of patients and resolution of symptoms in 85% to 100% of patients. However, bleeding complication rates vary and are poorly described.^7–10 Yang et al²⁵ suggested that endovascular recanalization is beneficial in the patient with acute MVT in the absence of bowel ischemia. Di Minno et al¹⁰ suggested that percutaneous treatment of MVT reduces the development of portal hypertension by 40%. Further investigation into the benefits of thrombolysis and pharmacomechanical thrombectomy in MVT is required, with chronic abdominal pain and quality of life included as an outcome metric.

This study has a few limitations. First, readmission for abdominal pain was used as a surrogate measure for chronic abdominal pain. The morbidity that is described is therefore likely to be underestimated because only the patients who were admitted with this pain are captured. In addition, because there are no established guidelines for the treatment of MVT, the threshold to take the patient to the operating room is somewhat operator dependent. This could affect the generalizability of the conclusions, as some institutions are more aggressive than others with regard to timing of exploratory laparotomy or long-term anticoagulation. Despite this limitation, this large series does accurately capture the shortcomings of current therapies for MVT and the morbid outcomes regardless of medical or surgical interventions.

CONCLUSIONS

This is the largest single-institution study describing surgical and medical management of MVT. The majority of patients with MVT suffer from a post-thrombotic syndrome of the abdomen with not only venous collateralization and portal hypertension on imaging but also recurrent readmissions for abdominal pain. A genetic hypercoagulable disorder and lactic acidosis on presentation are predictors of bowel resection in the setting of acute MVT. Patients with a hypercoagulable disorder have more aggressive thrombotic disease and are more likely to require a bowel resection. Whereas the outcomes of percutaneous treatment of MVT are still relatively unknown, an alternative approach to treat these patients is needed, given the poor outcomes with current strategies.

ARTICLE HIGHLIGHTS.

• Type of Research: Single-center retrospective cohort study

• Key Findings: Mesenteric venous thrombosis (MVT) in 121 patients resulted in high rates of 1-year mortality and readmissions. A genetic thrombophilia was a strong predictor of bowel ischemia after MVT (odds ratio, 3.81).

• Take Home Message: Outcomes after MVT are morbid, with high rates of readmission for abdominal pain.