Portal Venous Thrombosis in a Backpacker: The Role of Exercise

Mark Walsh, MD; James Moriarity, MD; Janice Peterson, MD; George Friend, MD; Rochelle Chodock; Michael Rogan, MD

If your browser does not support tables click here.

In Brief: Effort thrombosis usually afflicts an extremity and is caused by compression. This case report, in contrast, involves superior mesenteric and left portal vein septic thrombosis in a backpacker following prolonged hiking and abdominal straining. The condition may have been caused by localized splanchnic venous ischemia, erosion of the bowel-blood barrier, and release of bacterial endotoxin in this dehydrated and detrained athlete. Diagnosis of this disorder is aided by noting characteristic abdominal pain, fever, nausea, and vomiting, as well as by imaging with MRI, CT, or duplex ultrasonography. Antibiotics and anticoagulants are key to treatment.

Deep vein thrombosis of the extremities sometimes occurs in active people, notably in weight lifters. This condition is often thought to stem from mechanical compression of a vein due to muscle hypertrophy (1). However, we report a case of septic thrombophlebitis of the portal venous system in a young woman following exhaustive hiking. Primary care physicians should be aware of this potentially serious complication of physical activity.

Case History

After 2 1/2 days of vigorous hiking with a 50-lb backpack in hot and humid weather, a recently detrained 19-year-old female bicyclist had a strenuous bowel movement in a squatting position. She had hiked 25 miles each of the previous 2 days. That morning she had covered 11 miles.

Very shortly, she experienced severe periumbilical pain and vomiting, both of which continued intermittently through the day. The next day (the fourth day of the trip), she awakened shivering and vomiting. However, she hiked another 8 miles until she collapsed. On the fifth day, she was unable to continue and complained of sharp bouts of epigastric pain with chills and fever. She was taken off the mountain by horse. She noted that jostling caused more pain. She was then driven 15 hours to her college infirmary for initial evaluation.

History and physical. Because she had drunk unfiltered stream water on the trip and had a temperature of 104°F(40°C) with a disproportionately slow pulse of 60 per minute, a tentative diagnosis of salmonella septicemia was made.

Her history was remarkable for an emergency-room evaluation 4 months previous for intense abdominal pain following a 50-mile bicycling competition. Workup that included a complete blood count, urinalysis, and plain films had been negative. The patient was treated with injectable narcotics and released with a diagnosis of intestinal spasm. She had no family history of clotting abnormalities and was not taking oral contraceptives.

Physical exam revealed an acutely ill patient with a temperature of 103.6°F (39.8°C), a pulse of 72 per minute, respiration rate of 24/min, and blood pressure of 110/80 mm Hg. Her neck was supple, lungs clear, and heart normal. Her abdomen was soft, yet moderately tender in the right upper quadrant and epigastrium, and epigastric fullness was noted. There was no palpable spleen tip. Bowel sounds were present, and no rebound tenderness was evident. Pelvic exam was normal, and the rectal exam was negative for occult blood. Neurologic and musculoskeletal exams were normal. There was no rash.

The patient was begun on ceftriaxone sodium 2 g every 12 hours after appropriate cultures were drawn. She was then hospitalized for further evaluation.

Lab and radiology tests. Significant normal laboratory data included urinalysis, blood urea nitrogen, creatinine, electrolytes, phosphorous, calcium, bilirubin, lactate dehydrogenase, alanine aminotransferase, globulin, cholesterol, and triglycerides. Other laboratory values included gamma-glutamyl transpeptidase 80 IU/L, aspartate aminotransferase 50 IU/L, albumin 3.2 g/dL, hemoglobin 12 g/dL, hematocrit 37%, and white blood cell count (WBC) of 7,640 with 40 neutrophils, 41 bands, 17 lymphocytes, 1 atypical lymphocyte, 1 eosinophil, normal platelets, and marked toxic change of the neutrophils. The erythrocyte sedimentation rate was 88 mm/hr. The lab values were thought to correlate with an acute infectious process.

A chest x-ray revealed an interstitial infiltrate on the right. Computed tomography (CT) of the abdomen (figure 1) revealed partial obstruction of the superior mesenteric vein, total occlusion of the left portal vein, and swelling of the left lobe of the liver. A thrombosis was suspected in view of these CT findings.

Hospital course. In the afternoon of the second day of hospitalization, blood cultures grew Bacteroides fragilis. The patient's antibiotic was changed to intravenous imipenem-cilastatin 500 mg every 6 hours; she later was given intravenous metronidazole 500 mg every 6 hours. Heparin infusion was also started, followed by oral warfarin sodium.

On the fourth day of hospitalization, magnetic resonance imaging (MRI) confirmed the diagnosis of superior mesenteric vein and portal vein thrombosis (figure 2). A repeat chest x-ray was normal, confirming the suspicion that the infiltrate seen earlier signified incipient adult respiratory distress syndrome, which resolved after the septicemia was treated with antibiotics. An echocardiogram showed no abnormality. Indium-platelet-tagged WBC scan, ventilation-perfusion lung scan, and colonoscopy were all normal.

The patient was treated for all 10 hospital days with intravenous metronidazole and for another 2 weeks with oral metronidazole. Anticoagulation with warfarin was continued. Blood saved from before the administration of heparin was evaluated for a hypercoagulable state. All results, with the exception of the mildly elevated fibrin/fibrinogen degradation product, were within normal limits (table 1).

Table 1. Coagulation Profile of a 19-Year-Old Female Backpacker With Abdominal Pain and Fever
Test Performed	Result	Normal Range
Antithrombin III activity	106%	80%-120%
Fibrinogen	250 mg/dL	190-420 mg/dL
Fibrin split products	greater than 40 micrograms/dL	less than 40 micrograms/dL
Protein C, total	90%	70%-100%
Protein C, functional	70%	70%-130%
Protein C, antigen	100%	70%-130%
Heparin cofactor II	101%	70%-130%
Lupus anticoagulant	none detected	none detected
Prothrombin time	12.7 sec	11.9 sec control
Protein S	75%	60%-130%
Activated partial thromboplastin time	31.1 sec	29.7 sec control
Tissue plasminogen activator (TPA)	.91 micrograms/mL	1-12 micrograms/mL
TPA inhibitor	7.2 IU/mL	2.5-20 IU/mL

Follow-up. Anticoagulation was continued for 3 months. CT, MRI, and Doppler ultrasound of the patient's abdomen at 3 months demonstrated persistent partial occlusion of the superior mesenteric vein and total occlusion of the left portal vein. The patient was limited to mild exercise for 4 months following hospital discharge. She occasionally had symptoms of postprandial epigastric fullness that were reminiscent of the pain on admission.

At 7 months, Doppler ultrasound revealed complete clearing of the superior mesenteric vein obstruction but persistent total occlusion of the left portal vein. The patient was allowed to progress to her usual vigorous training. She developed no symptoms of abdominal fullness. A repeat MRI scan after 1 year revealed complete opening of the mesenteric and portal systems.

Thrombosis Characteristics

Mesenteric vein thrombosis remains an elusive diagnosis because of its insidious presentation. Characteristically, symptoms begin with abdominal pain, fever, nausea, and vomiting. Guarding and rebound tenderness develop later. The pain is most often out of proportion to the physical findings, and the patient appears more ill than the physical examination indicates (2). Most often, the diagnosis is made noninvasively with the help of MRI, CT, or duplex ultrasonography (3-5) or confirmed at the time of surgery. We chose CT with our patient because of the anatomic detail it provides of the liver, pancreas, kidneys, and lower chest.

This rare disease is associated with other conditions (table 2) (6). Frequently, no cause is found for mesenteric vein thrombosis, but this frequency is declining as the number of recognized cases of hypercoagulable states increases (7). Most of the world literature on this entity predates the discovery of antithrombin III, protein S, protein C, and circulating lupus anticoagulant antibody. The identification of these factors has improved our ability to trace the causes of deep vein thrombosis.

Table 2. Conditions Associated With Mesenteric Vein Thrombosis Abdominal trauma Anticardiolipin antibodies Antithrombin III deficiency Congestive heart failure Dysfibrinogenemia (mutated fibrinogen) Inflammatory bowel disease Intra-abdominal surgery Migratory thrombophlebitis Neoplastic disease Oral contraceptives Pelvic abscess Polycythemia vera Portal hypertension Liver transplant Pregnancy Protein C deficiency Protein S deficiency Thrombocytosis Tissue plasminogen activator deficiency

The Exercise Connection

It has been noted that during strenuous exercise, splanchnic blood flow decreases 80% (8). Exercise in the heat further reduces central blood flow because of thermoregulatory shunting of blood from the core to the skin.

Moreover, volume depletion from dehydration results in additional hemoconcentration and relative increases in plasma proteins. Endotoxemia has been proven to increase systemic hematocrit (9). The combination of exercise, heat, dehydration, and endotoxin-induced hemoconcentration contributes to vascular sludging. Disruption of the integrity of the intestinal mucosa as a result of sluggish blood flow and relative tissue hypoxia might allow intestinal bacteria to penetrate the vascular supply (10). Ultimately, the thrombotic sequence can be initiated.

Exercise has been noted to cause hemorrhagic colitis associated with a measured reduction of the splanchnic blood flow (11-13). Beaumont and Teare (14) describe a detrained, 42-year-old elite distance runner who had segmental ischemic colitis that required subtotal colectomy. The runner experienced abdominal pain, bloody diarrhea, vomiting, and abdominal tenderness within 3 days after running her personal best half marathon on a hot, humid day with no race hydration. By 10 days postrace, she had developed abdominal ascites, renal failure, and disseminated intravascular coagulation. Laparotomy revealed a diffusely inflamed transverse and descending colon that required surgical removal.

Dehydration from exercise contributes to the development of a hypercoagulable state at a local level in the bowel by the release of endotoxin. Brock-Utre et al (15) showed an increase of circulating endotoxin in 80% of ultramarathon runners who completed a 100-km race. Notably, the less fit marathoners-those who took longer to finish-had higher levels of endotoxin. This may be significant in our case because the patient was detrained for 2 months before she embarked on her exhaustive hike.

In contrast, elevated endotoxin levels did not correlate with symptoms in a recent study by Moore et al (16), who looked at endotoxin release as a possible cause of exercise-associated vomiting, nausea, diarrhea, rigors, and syncope in cyclists. However, none of the cyclists in this study was described as being significantly ill, and none required hospitalization.

Tracing the Etiology

Although a search found no reported cases of exercise-induced mesenteric vein thrombosis, evidence in the literature supports a pathophysiologic chain of events that could have precipitated our patient's thrombosis. The detrained athlete embarked on a long hike laden with a heavy backpack in hot weather. Exhaustion and dehydration precipitated the release of endotoxin, focal ischemia of bowel with microscopic hemorrhage, and release of Bacteroides fragilis into the vascular splanchnic bed.

An increase in intra-abdominal pressure from straining at stool further eroded the bowel-blood barrier. Localized bacteremia ensued. The endotoxins released caused an intense sensation of cold, goose bumps, and shivering (17). Septicemia followed. Bacteremia, extreme exercise with dehydration, and abdominal straining all contributed to sluggish splanchnic venous flow with resultant thrombosis of the superior mesenteric vein and subsequent extension to the left portal vein.

Treatment Issues

The treatment of septic thrombosis requires antibiotic therapy and anticoagulation in the hospital setting. Heparin is the anticoagulant of choice for initial treatment because of its immediate action; warfarin takes 3 to 5 days to reach its maximum anticoagulation effect. Surgery may be required if the patient has an infarcted bowel.

The duration of anticoagulation in our patient was an issue because at 3 months there was still radiographic evidence of partial obstruction of the superior mesenteric vein and total occlusion of the left portal vein. The appropriate duration of anticoagulation is not defined, and the literature offers little recommendation for treatment in patients who have residual thrombus after anticoagulation therapy (2,8).

The decision was made to discontinue warfarin on the assumption that good collateral blood flow had developed in the hepatic circulation. The patient's youth and active lifestyle favored cessation of the drug. The return to athletic activity may have hastened the dissolution of the clot in the left portal vein, since moderate conditioning programs enhance natural fibrinolysis (18) Eichner (19) noted that exercise-induced fibrinolysis was first described 50 years ago. Fibrinolysis increases after as little as 5 minutes of bicycling and can last up to 90 minutes (20).

Summing Up

This case report demonstrates that superior mesenteric vein thrombosis is a most insidious disease that may easily elude even the most astute clinician. The diagnostic studies of choice are mesenteric angiography, CT, MRI, and duplex ultrasonography. Anticoagulant therapy beyond 3 months may not be necessary for dissolution of clots and revascularization in patients who have no underlying predisposing conditions. Finally, return to exercise, with its enhancement of fibrinolysis, may even enhance the dissolution of the clot.

References

1. Mitchell PB Jr, MacGillivray DC, Almond MD: Effort thrombosis: recognition and management while underway. Mil Med 1991;156(10):561-563
2. Boley SJ, Kaleya RN, Brandt LJ: Mesenteric venous thrombosis. Surg Clin North Am 1992;72(1):183-201
3. Hricak H, Amparo E, Fisher MR, et al: Abdominal venous system: assessment using MR. Radiology 1985;156(2):415-422
4. Matos C, Van Gansbeke D, Zalcman M, et al: Mesenteric vein thrombosis: early CT and US diagnosis and conservative management. Gastrointest Radiology 1986;11(4):322-325
5. Kitchens CS: Evolution of our understanding of the pathophysiology of primary mesenteric venous thrombosis. Am J Surg 1992;163(3):346-348
6. Font VE, Hermann RE, Longworth DL: Chronic mesenteric venous thrombosis: difficult diagnosis and therapy. Cleve Clin J Med 1989;56(8):823-828
7. Case Records of the Massachusetts General Hospital: weekly clinicopathological exercises: case 9-1991: a 61-year-old woman with recurrent bouts of abdominal pain and fever. N Engl J Med 1991;324(9):613-623
8. Clausen JP: Effect of physical training on cardiovascular adjustments to exercise in man. Physiol Rev 1977;57(4):779-815
9. Lee JS: Microvascular volume changes induced by exercise, heat exposure, or endotoxin injection. Am J Physiol 1994;267(3 pt 2):H1142-H1150
10. Zhi-Yong S, Dong YL, Wang XH: Bacterial translocation and multiple system organ failure in bowel ischemia and reperfusion. J Trauma 1992;32(2):148-153
11. Moses FM, Brewer TG, Peura DA: Running-associated proximal hemorrhagic colitis. Ann Intern Med 1988;108(3):385-386
12. Heer M, Repond F, Hany A, et al: Acute ischemic colitis in a female long distance runner. Gut 1987;28(7):896-899
13. Stewart JG, Ahlquist DA, McGill DB, et al: Gastrointestinal blood loss and anemia in runners. Ann Intern Med 1984;100(6):843-845
14. Beaumont AC, Teare JP: Subtotal colectomy following marathon running in a female patient. J Royal Soc Med 1991;84(7):439-440
15. Brock-Utre JG, Gaffin SL, Wells MT, et al: Endotoxaemia in exhausted runners after a long-distance race. S Afr Med J 1988;73(9):533-536
16. Moore GE, Holbein MEB, Knochel JP: Exercise-associated collapse in cyclists is unrelated to endotoxemia. Med Sci Sports Exerc 1995;27(9):1238-1242
17. Cooper KE: Some responses of the cardiovascular system to heat and fever. Can J Cardiol 1994;10(4):444-448
18. Keber D, Stegnor M, Keber I, et al: Influence of moderate and strenuous daily physical activity on fibrinolytic activity of blood: possibility of plasminogen activator stores depletion. Thromb Maemost 1979;41(4):745-755
19. Eichner ER: Coagulability and rheology: hematologic benefits from exercise, fish, and aspirin: implications for athletes and nonathletes. Phys Sportsmed 1986;14(10):102-110
20. Marsh NA, Gaffney PJ: Exercise-induced fibrinolysis: fact of fiction? Thromb Haemost 1982;48(2):201-203

Dr Walsh is the director of medical education at Saint Joseph Medical Center in South Bend, Indiana, and a fellow of the American College of Emergency Physicians. Dr Moriarity is a team physician at the University of Notre Dame in South Bend and a member of the American College of Sports Medicine. Dr Peterson is the director of the Ethiopian Relief Fund in Addis Ababa, Ethiopia. Dr Friend is a surgeon at Saint Joseph Medical Center in South Bend and a fellow of the American College of Surgeons. Ms Chodock is a medical student at Brown University School of Medicine in Providence, Rhode Island. Dr Rogan is a resident in the Department of Emergency Medicine at Methodist Hospital in Indianapolis. Address correspondence to James Moriarity, MD, University Health Services, University of Notre Dame, Notre Dame, IN 46556-5693; e-mail to James.M.Moriarity@ND.edu

RETURN TO DECEMBER 1996 TABLE OF CONTENTS