Case 25-2005 — A 40-Year-Old Man with Prolonged Fever and Weight Loss
http://www.100md.com
《新英格兰医药杂志》
Presentation of Case
A 40-year-old man was admitted to this hospital because of persistent fever, night sweats, anorexia, and weight loss.
The patient had been in excellent health until eight and a half weeks before admission, when fever, chills, severe headache, and diarrhea developed the day after he and his family ate in a fast-food restaurant. Others in the family had similar symptoms. The headaches and diarrhea resolved after five days, but the fever persisted, with nightly temperatures of 38.9°C to 39.4°C associated with drenching night sweats. The fever responded to ibuprofen and did not occur during the day. Anorexia developed, with a weight loss of 6.4 kg, and the patient felt some pressure in the right upper abdominal quadrant. He noted a brownish-orange color to his urine. Evaluation by his primary care physician two and a half weeks after the onset of symptoms (six weeks before admission to this hospital) revealed normal vital signs and normal findings on physical examination; laboratory-test results are shown in Table 1. Computed tomographic (CT) scanning of the abdomen and thorax showed no abnormalities except for patchy, nonspecific heterogeneous attenuation in the liver, which was thought to suggest fatty infiltration.
Table 1. Hematologic Laboratory Values.
Two weeks later, four weeks before admission, persistent symptoms prompted referral to a second physician for consultation. The results of the physical examination were again normal. Serum levels of electrolytes, calcium, phosphorus, glucose, urea nitrogen, and creatinine were normal. The results of tests for C-reactive protein, alpha1-antitrypsin, alpha-fetoprotein, ceruloplasmin, and thyrotropin were normal. Assays for the presence of IgG antibodies to Epstein–Barr virus (EBV) latent membrane protein and EBV nuclear antigen were positive; those for IgM anti-EBV antibodies, hepatitis B antigen and antibody, and antimitochondrial antibodies were negative. Antinuclear antibodies were present at a titer of 1:32 in a speckled pattern. Urinalysis was positive for urobilinogen (+) and was otherwise normal. Other results from this second round of laboratory tests are shown in Table 1.
Repeated CT scanning of the abdomen, performed after the administration of intravenous contrast material as well as oral contrast material, showed inflammatory changes around the cecum and possibly the terminal ileum. The liver showed diffuse low attenuation in the center without biliary obstruction or a focal mass. The gallbladder was absent. The spleen, pancreas, adrenal glands, and kidneys were normal. A biopsy specimen of bone marrow showed normal hematopoiesis and increased iron storage.
Colonoscopy, esophagogastroduodenoscopy, and endoscopic retrograde cholangiopancreatography performed during the following week showed no abnormalities except for a few erosions in the gastric antrum; the pancreatic ducts and bile ducts had no filling defects. A biopsy specimen from the liver was reported to show a nonspecific increase in inflammatory cells. An upper gastrointestinal series with small-bowel follow-through showed that the terminal ileum was never well distended and had some mucosal irregularity. About six weeks after the onset of his illness, the patient was admitted to another hospital for exploratory laparoscopy. An inflamed retrocecal appendix was removed; there was evidence of periappendiceal and pericolonic inflammation. Pathological examination of the appendix was reported to show inflammation and cancer.
Within 24 hours, the patient reported improvement in his symptoms, and he was discharged to his home. He felt well for about one week; then the fever and night sweats recurred. Repeated CT scanning of the abdomen and pelvis at that time, after the administration of oral and intravenous contrast material, showed no opacification of the main portal and superior mesenteric veins and stranding of the peripancreatic and periportal fat. The patient was referred to this hospital the next day, and he was admitted.
The patient had undergone laparoscopic cholecystectomy two years earlier. He did not smoke, drink alcohol, or use illicit drugs, and he drank minimal amounts of coffee. He was married, with two children; all three family members were well. He worked as a mechanic but had been unable to work for the past month because of his illness. An uncle and a grandmother had had lung cancer, and the patient's father had died at 64 years of age of lung disease related to asbestosis and emphysema. His mother and three siblings were in good health. He had not traveled outside the United States for 13 years. He had no pets, no exposure to animals, and no known tick bites. The family's drinking water came from a well. He had no nausea or vomiting, diarrhea, hematochezia, hematemesis, muscle weakness, joint problems, or neurologic symptoms. He took no medications.
On examination, the blood pressure was 115/70 mm Hg, the pulse 105 beats per minute, the respiratory rate 18 breaths per minute, the temperature 37.3°C, and the oxygen saturation 100 percent while the patient was breathing room air. He was in no acute distress. The skin was pale, with no rashes, lesions, or jaundice. An examination of the head, eyes, ears, neck, and chest revealed no abnormalities. The heart rate and rhythm were regular, with no murmurs, gallops, or rubs. The abdomen was soft, nontender, and nondistended, with normal bowel sounds. The liver and spleen were not palpable. The extremities appeared normal. Neurologic examination revealed no abnormalities. Specimens of blood were drawn for culture. Laboratory-test results are shown in Table 2.
Table 2. Blood-Chemistry and Immunologic Laboratory Values.
Magnetic resonance imaging (MRI) of the liver on the day after admission (before and after administration of intravenous gadolinium) showed a nonenhancing thrombus in the main portal vein that extended into both the right and left intrahepatic branches. No mass lesion, focal abnormality, or dilatation of the bile ducts was present in the liver. The hepatic arteries, hepatic veins, and splenic vein were patent.
A diagnostic test result was reported.
Differential Diagnosis
Dr. Dennis L. Kasper: This patient and his family probably had food-borne infectious diarrhea, and this patient's subsequent problems may well have resulted from complications of this primary infection.
The differential diagnosis of food-borne infections is wide, with possible pathogens ranging from bacterial to viral to protozoan. These causes may be clinically indistinguishable; stool cultures would have been useful and might have been diagnostic earlier in the course of the illness. The persistence of this man's symptoms for weeks, with fever and subsequent complications, limits the scope of the diagnostic possibilities. Overall, this complex clinical course is most consistent with a bacterial cause of diarrhea. Salmonella, campylobacter, and yersinia would be at the top of my list of possible causes.1 In the United States, salmonella and campylobacter are much more common food-borne pathogens than yersinia.2
This case begins to distinguish itself from routine infectious diarrhea in its subsequent course. The weight loss and pressure in the right upper quadrant that the patient had after the diarrheal illness are unusual and suggest other problems that may have been initiated by the original gastrointestinal infection.
Persistent Symptoms with Abnormal Results on Liver-Function Tests
The increased serum level of alanine aminotransferase and a borderline low serum level of albumin, along with the weight loss and pressure in the right upper quadrant, suggest the evolution of a new process involving the liver or perhaps the presence of underlying liver disease that predisposed the patient to more severe infection. The incubation period was too short for hepatitis to have been acquired during the patient's recent contact with apparently contaminated food. The workup for hepatitis B and EBV-related hepatitis was negative. A workup for hepatitis A was not done.
Liver disease due to iron overload or hemochromatosis would predispose a patient to yersiniosis. Yersiniae exploit host iron and use it to enhance their virulence. The detection by CT of nonspecific heterogeneous liver attenuation suggests a diffuse hepatic process. Heterogeneous attenuation, as compared with homogeneous attenuation, is less consistent with infiltrative disease due to iron deposition and more consistent with a vascular process. The patient's serum iron levels and total iron-binding capacity make hemochromatosis an unlikely diagnosis, and no aspect of his history suggests iron overload.
Deterioration of Liver Function
Over the next few weeks, the results of tests of liver function became increasingly abnormal. Alkaline phosphatase rose to a very high level, and the level of lactate dehydrogenase was high, with ongoing elevation of hepatocellular enzymes. The erythrocyte sedimentation rate continued to rise, suggesting an ongoing inflammatory process. A process that connects the diarrheal disease to the abnormal results on liver-function tests must be considered. Although pyogenic liver abscesses would have been high on my diagnostic list, repeated CT two weeks after the initial CT did not support this diagnosis. It did show ongoing inflammatory change around the cecum and possibly the terminal ileum, suggesting that the initial infection may not have resolved. The liver continued to show low-level heterogeneous attenuation, with no abscesses or biliary obstruction.
A liver biopsy showed a nonspecific increase in inflammatory cells. Without definitive pathological findings in the liver, vascular disruption causing diffuse heterogeneous liver changes must be strongly considered. The development of liver disease after infectious diarrhea, the subsequent radiologic studies, and the continuing fevers all suggest the portal venous system as a site where infection might arise, where organisms might disseminate from the bowel to the systemic circulation, and where vascular occlusion may develop, possibly as a result of portal-vein thrombosis.
Ileocecitis and Periappendicitis Associated with Food-Borne Pathogens
The upper gastrointestinal series with small-bowel follow-through showed a lack of distention in the terminal ileum, as well as mucosal irregularities. At this point, a possible link between the diarrheal illness and the continuing mucosal irregularities in the small bowel becomes more apparent. Yersinia, salmonella, and campylobacter have been associated with ileocolitis, periappendiceal inflammation, and appendicitis following acute infection.3,4,5,6,7,8,9,10,11,12 Clinically, these gastrointestinal infections may be indistinguishable once appendicitis develops.5 Of these three organisms, yersinia is most often associated around the world with subsequent inflammation of the appendix, periappendiceal areas of the ileum, or mesenteric lymph nodes.4,13 In the United States, Yersinia enterocolitica has reportedly been found in up to 9 percent of appendixes removed because of symptoms of appendicitis.4 It is not clear whether Y. enterocolitica is the causative agent of appendicitis in these cases. Y. pseudotuberculosis is a pathogen that is found primarily in animals and rarely in humans.8 Y. enterocolitica is usually transmitted to humans from domestic and wild animal reservoirs by way of food consumption. Although self-limited enterocolitis is much more commonly caused by Y. enterocolitica than by Y. pseudotuberculosis, the latter is more often associated with mesenteric lymphadenitis and terminal ileitis.8
Salmonella, which can cause pseudoappendicitis or even an illness mimicking inflammatory bowel disease,10 may have initiated the disease process in this patient. In the United States, several studies indicate that diarrheal disease due to campylobacter is more common than that due to salmonella and shigella combined.6 Campylobacter jejuni is the principal diarrheal pathogen belonging to this genus; it causes tissue injury in the jejunum, ileum, and colon and has also been associated with pseudoappendicitis.6
During an exploratory laparotomy, an inflamed retrocecal appendix was removed. Periappendiceal and pericolonic inflammation was reported. The disease process in the ileal and appendiceal areas was probably associated with the original infectious cause of diarrhea. In spite of an initial postoperative improvement, fever and night sweats recurred.
A workup at this hospital included what turned out to be an unremarkable medical history and an essentially normal physical examination. However, MRI of the liver with gadolinium contrast material yielded findings of interest. May we review these studies?
Dr. Dushyant Sahani: Contrast-enhanced CT of the abdomen, performed in arterial and portal venous phases of liver enhancement on the day before admission, showed heterogeneous enhancement of the liver parenchyma in the arterial phase (Figure 1A). However, in the portal venous phase, the liver enhanced homogeneously. An occlusive thrombus was identified in the left main branch of the portal vein and a nonocclusive thrombus in the superior mesenteric vein (Figure 1B). In addition, there was stranding in the fat, which indicated inflammation in the right lower quadrant, along the medial aspect of the cecum (Figure 1C). There was no ascites.
Figure 1. Contrast-Enhanced CT Images of the Liver Obtained One Day before Admission.
The arterial-phase image (Panel A) shows heterogeneous enhancement of the liver parenchyma at the level of the porta hepatis (arrows). In the subsequent image from the portal venous phase (Panel B), a filling defect, or thrombus, is evident in the main portal vein and the left branch (arrow). In an image obtained at the level of the right lower quadrant (Panel C), there is stranding in the pericecal fat (arrows), indicating inflammation.
One day after admission, a contrast-enhanced MRI of the liver confirmed the finding of a nonenhancing, occlusive thrombus in the portal vein. When this scan was compared with the CT scan, progression of the thrombus in the main portal vein, intrahepatic branches, and the superior mesenteric vein (Figure 2) was evident.
Figure 2. A Dynamic Contrast-Enhanced MRI Scan of the Liver Obtained on the Second Hospital Day.
An occlusive thrombus is present in the superior mesenteric vein (arrow).
Pylephlebitis Following Appendicitis
Dr. Kasper: The imaging tests and the case presentation are consistent with pylephlebitis (septic thrombophlebitis of the portal veins). This condition appears to have developed as a result of the appendicitis and periappendicitis, which were probably due to one of the diarrheal pathogens I have already mentioned. It is less likely, but possible, that these food-borne organisms were responsible for the pylephlebitis. Only rarely do case reports describe pylephlebitis as having been caused by salmonella14; yersinia and campylobacter are also unlikely suspects. Appendicitis itself is a well-known cause of pylephlebitis.15 The venous drainage from the appendiceal area and terminal ileum flows directly into the portal system (Figure 3). Bacteroides species are components of the normal ileal and cecal flora.16 In appendicitis, both vascular flow to the appendix and vascular leakage of bacteria are increased. In one series of appendicitis cases in Croatia,17 portal-vein bacteremia was found in 19 of 50 cases, but patients had systemic bacteremia in only 3 of the cases. Bacteroides and Escherichia coli predominated among the isolates.
Figure 3. The Pathogenesis of Pylephlebitis and Liver Abscess after Acute Appendicitis.
Infection in the appendix can result in bacterial leakage into the portal venous system. Bacteroides species, particularly B. fragilis, have unique virulence factors that contribute to thrombosis.
In a 1995 review of 19 cases of pylephlebitis reported in the literature after 1979,15 appendicitis was second only to diverticulitis as a predisposing condition for pylephlebitis. The organisms most frequently cultured from the blood were bacteroides (especially Bacteroides fragilis) and E. coli. Pylephlebitis often leads to inflammation and, not uncommonly, to abscess of the liver.
Bacteroides and Thrombotic Disease
Anaerobic infections are strongly associated with thrombotic diseases. Thrombophlebitis is found in as many as 5 to 12 percent of cases of anaerobic bacteremia.18 The responsible mechanism is not fully understood, but several factors may contribute. Bacteroides species elaborate enzymes that degrade heparin.19 Bacteroides can accelerate coagulation in mice, as shown by decreased clotting time; this acceleration is secondary to a cell-wall component.20 Surface components of bacteroides substantially enhance fibrin clotting, probably in tandem with a capsular component that activates macrophages and initiates the clotting cascade.21 B. fragilis expresses capsular polysaccharides that have structural characteristics that make them fibrinogenic and thus capable of facilitating abscess formation22,23; the underlying mechanism is probably quite similar to that which enables the organism to cause thrombophlebitis.21 The unique immunologic properties found in the capsular zwitterionic polysaccharides of B. fragilis promote this procoagulant activity.23 Although I would usually be inclined to favor a unifying diagnosis, in this case I think the probability of two more common associations — food-borne infection causing appendicitis and B. fragilis causing pylephlebitis — is greater. Therefore, I believe that the most definitive diagnostic test in this case was the blood culture, which I suspect grew B. fragilis (although it may have grown yersinia, salmonella, or campylobacter).
Dr. Nancy Lee Harris (Pathology): Dr. Ivers, you were the infectious-diseases fellow who cared for this patient; can you give us your clinical impressions?
Dr. Louise Ivers (Infectious Diseases): We reviewed the patient's radiology images from previous admissions with radiology staff members on the day of admission. On the basis of the finding of portal-vein thrombosis in combination with the history of subacute appendicitis and prolonged fever, our leading differential diagnosis was septic portal-vein thrombosis. Since the patient was clinically stable, we recommended that antibiotics be withheld for the first 24 hours after admission while blood cultures were obtained.
Dr. Harris: Would stool cultures have been of any use in determining what caused the patient's original diarrhea?
Dr. Kasper: The patient still had signs of active ileocecitis, as seen in the CT scan. I would like to know whether stool cultures were done.
Clinical Diagnosis
Pylephlebitis.
Dr. Dennis L. Kasper's Diagnosis
Ileocecitis with appendicitis in association with infection by yersinia, salmonella, or campylobacter, leading to pylephlebitis with B. fragilis bacteremia.
Pathological Discussion
Dr. Joseph Misdraji: The specimen obtained during liver biopsy is perhaps most remarkable for the absence of pathological findings in the portal vein, given the imaging studies that showed complete occlusion of this vessel. The portal tracts are edematous and the bile ducts appear reactive and injured. Inflammatory cells are sparse, but a few neutrophils infiltrate the bile duct epithelium (Figure 4A). Although these changes are not well developed, they raise the possibility of some degree of bile-duct obstruction. In the setting of portal-vein thrombosis, this may be due to a condition known as portal biliopathy, in which extrahepatic biliary obstruction results from an engorgement of collateral veins around the bile duct.24
Figure 4. Biopsy Specimens from the Liver and the Appendix (Hematoxylin and Eosin).
The specimen from the liver biopsy (Panel A) contains an edematous portal tract. The bile-duct epithelium shows reactive changes with nuclear enlargement, irregular nuclear placement, and cytoplasmic eosinophilia. A very few inflammatory cells are seen infiltrating the duct epithelium. The specimen from the appendix (Panel B) shows patchy mucosal ulceration consistent with appendicitis. A vein in the periappendiceal fat (Panel C) reveals a vein occluded by loose fibrous tissue, indicating an organized thrombus.
The appendix showed areas of mucosal ulceration (Figure 4B), associated with hemosiderin deposition, and periappendiceal fat necrosis, indicating an acute and chronic inflammatory process. In the periappendiceal fat, there was a thrombosed vein (Figure 4C), which is of interest in the context of the patient's portal-vein thrombosis. As Dr. Kasper predicted, a blood culture yielded B. fragilis, establishing the diagnosis of pylephlebitis as a result of infection with B. fragilis.
Pylephlebitis begins with thrombophlebitis of the small veins draining the infected site, which in this case was the appendix. The involvement, by extension, of larger veins leads to septic thrombophlebitis of the portal vein, and further extension to the large mesenteric veins may result in bowel ischemia — a rare complication that carries a high risk of death. Embolization of an infected thrombus results in liver abscesses.25 The organisms that are most often cultured from the blood are E. coli and B. fragilis. Although the role of these organisms in precipitating appendicitis is controversial, their role in abdominal infectious complications of appendicitis has been well established; these two organisms are those most frequently recovered from inflamed, gangrenous appendixes or the peritoneal fluid in cases of appendicitis.26,27,28,29,30,31,32,33 Clinical trials have shown that the use of antibiotics with activity against B. fragilis in the treatment of complicated appendicitis can reduce the rate of postoperative infectious complications.34,35
Discussion of Management
Dr. Harris: Dr. Kasper, could you discuss your recommendations for the management of bacteroides pylephlebitis?
Dr. Kasper: In a case such as this one, until culture results are known, metronidazole and a quinolone agent should be administered to treat bacteroides and food-borne infection, respectively. If the blood yielded only a food-borne pathogen, metronidazole would not be needed. Anticoagulant therapy in this clinical setting has not been proved essential but has been associated with improvement in the condition of some patients, with a decrease in septic embolization to the liver from infected portal thrombi and consequent prevention of liver abscesses. Many patients respond to antibiotics alone, whether or not they receive heparin.15
Dr. Harris: Dr. Ivers, would you tell us how you treated this patient's condition and how he is doing?
Dr. Ivers: Anticoagulation with intravenous heparin was begun. When we received the positive results of the blood cultures, we started treatment with cefepime and metronidazole and switched to levofloxacin and metronidazole after the identification of B. fragilis. The patient's fever immediately resolved. Subsequent blood cultures were negative. Although the infectious-diseases consultants had requested stool cultures, these were not done. A transthoracic echocardiogram did not suggest the presence of valvular vegetations. Esophagogastroduodenoscopy revealed grade 1 esophageal varices. The heparin was ultimately discontinued and oral warfarin was begun. A workup to evaluate the presence of a hypercoagulable state revealed a mildly elevated titer of anticardiolipin IgM, which was to be followed up after discharge. The patient was discharged to his home. He was well and preferred to follow up with his physician in his home city, who reported that the patient was doing well. The testing for hypercoagulability has not been repeated.
Dr. Sahani: Follow-up CT scans performed at another hospital two and a half months after the patient was discharged showed a persistent portal-vein thrombus, with the development of venous collaterals or cavernous transformation of the portal vein, suggestive of portal hypertension, as well as evidence of chronic liver disease, with enlargement of the caudate lobe (Figure 5).
Figure 5. A Contrast-Enhanced CT Scan Obtained Two and a Half Months after the Patient's Discharge.
There are signs of chronic portal hypertension, including the development of venous collaterals around the thrombosed portal vein (arrow) and hypertrophy of the caudate lobe of the liver (arrowhead).
Dr. Kenneth McIntosh (Pediatrics, Children's Hospital): Had blood cultures been done before the patient's admission to this hospital?
Dr. Ivers: No cultures were done before his admission here. It appeared from the record that the primary concern of the physicians caring for him before the laparoscopy was that he had an occult malignant condition — in particular, lymphoma.
Dr. Nesli Basgoz (Infectious Diseases): Should people with a thrombotic complication of infection be investigated for a hypercoagulable state? Is there any reason to expect that such patients are uniquely susceptible to thrombotic conditions because of preexisting hypercoagulable states?
Dr. Kasper: Not to my knowledge. It is the infection with B. fragilis that predisposes the patient to hypercoagulation. I know of no association of the organism with a preexisting hypercoagulable state or the subsequent development of thrombotic disease.
Anatomical Diagnosis
Portal-vein thrombosis due to B. fragilis infection.
Dr. Kasper reports having received consulting fees from Chiron Vaccines and reports holding patents in several areas of group B streptococcal vaccines and in zwitterionic polysaccharides as immunomodulators. Dr. Sahani reports having received grant support from Bracco Diagnostics.
Source Information
From the Department of Medicine, Brigham and Women's Hospital (D.L.K.); the Departments of Radiology (D.S.) and Pathology (J.M.), Massachusetts General Hospital; and the Departments of Medicine (D.L.K.), Radiology (D.S.), and Pathology (J.M.), Harvard Medical School — all in Boston.
References
Butterton JR, Calderwood SB. Acute infectious diarrheal diseases and bacterial food poisoning. In: Kasper DL, Fauci AS, Longo DL, Braunwald E, Hauser SL, Jameson LJ, eds. Harrison's principles of internal medicine. 16th ed. New York: McGraw-Hill, 2005:754-9.
Preliminary FoodNet data on the incidence of infection with pathogens transmitted commonly through food -- selected sites, United States, 2003. MMWR Morb Mortal Wkly Rep 2004;53:338-343.
Arda IS, Ergin F, Varan B, Demirhan B, Aslan H, Ozyaylali I. Acute abdomen caused by Salmonella typhimurium infection in children. J Pediatr Surg 2001;36:1849-1852.
Bennion RS, Thompson JE Jr, Gil J, Schmit PJ. The role of Yersinia enterocolitica in appendicitis in the southwestern United States. Am Surg 1991;57:766-768.
Blakelock RT, Beasley SW. Infection and the gut. Semin Pediatr Surg 2003;12:265-274.
Blaser MJ. Infections due to Campylobacter and related species. In: Kasper DL, Fauci AS, Longo DL, Braunwald E, Hauser SL, Jameson LJ, eds. Harrison's principles of internal medicine. 16th ed. New York: McGraw-Hill, 2005:907-9.
Chandler ND, Parisi MT. Yersinia enterocolitica masquerading as appendicitis. Arch Pediatr Adolesc Med 1994;148:527-528.
Dennis DT, Campbell GL. Plague and other Yersinia infections. In: Kasper DL, Fauci AS, Longo DL, Braunwald E, Hauser SL, Jameson LJ, eds. Harrison's principles of internal medicine. 16th ed. New York: McGraw-Hill, 2005:921-9.
Deutsch A, Wasserman D, Ruchelli E, Johnson J, Broussard DL. An uncommon presentation of Salmonella. Pediatr Emerg Care 1996;12:285-287.
Lesser CF, Miller SI. Salmonellosis. In: Kasper DL, Fauci AS, Longo DL, Braunwald E, Hauser SL, Jameson LJ, eds. Harrison's principles of internal medicine. 16th ed. New York: McGraw-Hill, 2005:897-902.
Naktin J, Beavis KG. Yersinia enterocolitica and Yersinia pseudotuberculosis. Clin Lab Med 1999;19:523-536.
Puylaert JB, Van der Zant FM, Mutsaers JA. Infectious ileocecitis caused by Yersinia, Campylobacter, and Salmonella: clinical, radiological, and US findings. Eur Radiol 1997;7:3-9.
Dionisio D, Belli A, Dionisio A, et al. Appendicitis: microbial interactions and new pathogens. Recenti Prog Med 1992;83:330-336.
Knobel B, Sommer I, Menashes Z. A rare presentation of systemic salmonellosis. Infection 1985;13:70-72.
Plemmons RM, Dooley DP, Longfield RN. Septic thrombophlebitis of the portal vein (pylephlebitis): diagnosis and management in the modern era. Clin Infect Dis 1995;21:1114-1120.
Hentges DJ. Anaerobes as normal flora. In: Finegold SM, George WL, eds. Anaerobic infections in humans. San Diego, Calif.: Academic Press, 1989:37-53.
Juric I, Primorac D, Zagar Z, et al. Frequency of portal and systemic bacteremia in acute appendicitis. Pediatr Int 2001;43:152-156.
Gorbach SL, Bartlett JG. Anaerobic infections. N Engl J Med 1974;290:1289-1294.
Gesner BM, Jenkin CR. Production of heparinase by Bacteroides. J Bacteriol 1961;81:595-604.
Bjornson HS, Hill EO. Bacteroidaceae in thromboembolic disease: effects of cell wall components on blood coagulation in vivo and in vitro. Infect Immun 1973;8:911-918.
Rosenthal GA, Levy G, Rotstein OD. Induction of macrophage procoagulant activity by Bacteroides fragilis. Infect Immun 1989;57:338-343.
Tzianabos AO, Onderdonk AB, Rosner B, Cisneros RL, Kasper DL. Structural features of polysaccharides that induce intra-abdominal abscesses. Science 1993;262:416-419.
Choi Y-H, Roehrl MH, Kasper DL, Wang JY. A unique structural pattern shared by T-cell-activating and abscess-regulating zwitterionic polysaccharides. Biochemistry 2002;41:15144-15151.
Perego P, Cozzi G, Bertolini A. Portal biliopathy. Surg Endosc 2003;17:351-352.
Slovis TL, Haller JO, Cohen HL, Berdon WE, Watts FB Jr. Complicated appendiceal inflammatory disease in children: pylephlebitis and liver abscess. Radiology 1989;171:823-825.
Martirosian G, Bulanda M, Wojcik-Stojek B, et al. Acute appendicitis: the role of enterotoxigenic strains of Bacteroides fragilis and Clostridium difficile. Med Sci Monit 2001;7:382-386.
Mosdell DM, Morris DM, Fry DE. Peritoneal cultures and antibiotic therapy in pediatric perforated appendicitis. Am J Surg 1994;167:313-316.
Rautio M, Saxen H, Siitonen A, Nikku R, Jousimies-Somer H. Bacteriology of histopathologically defined appendicitis in children. Pediatr Infect Dis J 2000;19:1078-1083.
Heseltine PN, Yellin AE, Appleman MD, et al. Perforated and gangrenous appendicitis: an analysis of antibiotic failures. J Infect Dis 1983;148:322-329.
Bennion RS, Baron EJ, Thompson JE Jr, et al. The bacteriology of gangrenous and perforated appendicitis -- revisited. Ann Surg 1990;211:165-171.
Pieper R, Kager L, Lindberg AA, Nord CE. Acute appendicitis and Bacteroides fragilis. Scand J Infect Dis Suppl 1979;19:92-97.
Pieper R, Kager L, Weintraub A, Lindberg AA, Nord CE. The role of Bacteroides fragilis in the pathogenesis of acute appendicitis. Acta Chir Scand 1982;148:39-44.
Elhag KM, Alwan MH, Al-Adnani MS, Sherif RA. Bacteroides fragilis is a silent pathogen in acute appendicitis. J Med Microbiol 1986;21:245-249.
Wilson SE, Boswick JA Jr, Duma RJ, et al. Cephalosporin therapy in intraabdominal infections: a multicenter randomized, comparative study of cefotetan, moxalactam, and cefoxitin. Am J Surg 1988;155:61-66.
Bennion RS, Thompson JE, Baron EJ, Finegold SM. Gangrenous and perforated appendicitis with peritonitis: treatment and bacteriology. Clin Ther 1990;12:Suppl C:31-44.(Dennis L. Kasper, M.D., D)
A 40-year-old man was admitted to this hospital because of persistent fever, night sweats, anorexia, and weight loss.
The patient had been in excellent health until eight and a half weeks before admission, when fever, chills, severe headache, and diarrhea developed the day after he and his family ate in a fast-food restaurant. Others in the family had similar symptoms. The headaches and diarrhea resolved after five days, but the fever persisted, with nightly temperatures of 38.9°C to 39.4°C associated with drenching night sweats. The fever responded to ibuprofen and did not occur during the day. Anorexia developed, with a weight loss of 6.4 kg, and the patient felt some pressure in the right upper abdominal quadrant. He noted a brownish-orange color to his urine. Evaluation by his primary care physician two and a half weeks after the onset of symptoms (six weeks before admission to this hospital) revealed normal vital signs and normal findings on physical examination; laboratory-test results are shown in Table 1. Computed tomographic (CT) scanning of the abdomen and thorax showed no abnormalities except for patchy, nonspecific heterogeneous attenuation in the liver, which was thought to suggest fatty infiltration.
Table 1. Hematologic Laboratory Values.
Two weeks later, four weeks before admission, persistent symptoms prompted referral to a second physician for consultation. The results of the physical examination were again normal. Serum levels of electrolytes, calcium, phosphorus, glucose, urea nitrogen, and creatinine were normal. The results of tests for C-reactive protein, alpha1-antitrypsin, alpha-fetoprotein, ceruloplasmin, and thyrotropin were normal. Assays for the presence of IgG antibodies to Epstein–Barr virus (EBV) latent membrane protein and EBV nuclear antigen were positive; those for IgM anti-EBV antibodies, hepatitis B antigen and antibody, and antimitochondrial antibodies were negative. Antinuclear antibodies were present at a titer of 1:32 in a speckled pattern. Urinalysis was positive for urobilinogen (+) and was otherwise normal. Other results from this second round of laboratory tests are shown in Table 1.
Repeated CT scanning of the abdomen, performed after the administration of intravenous contrast material as well as oral contrast material, showed inflammatory changes around the cecum and possibly the terminal ileum. The liver showed diffuse low attenuation in the center without biliary obstruction or a focal mass. The gallbladder was absent. The spleen, pancreas, adrenal glands, and kidneys were normal. A biopsy specimen of bone marrow showed normal hematopoiesis and increased iron storage.
Colonoscopy, esophagogastroduodenoscopy, and endoscopic retrograde cholangiopancreatography performed during the following week showed no abnormalities except for a few erosions in the gastric antrum; the pancreatic ducts and bile ducts had no filling defects. A biopsy specimen from the liver was reported to show a nonspecific increase in inflammatory cells. An upper gastrointestinal series with small-bowel follow-through showed that the terminal ileum was never well distended and had some mucosal irregularity. About six weeks after the onset of his illness, the patient was admitted to another hospital for exploratory laparoscopy. An inflamed retrocecal appendix was removed; there was evidence of periappendiceal and pericolonic inflammation. Pathological examination of the appendix was reported to show inflammation and cancer.
Within 24 hours, the patient reported improvement in his symptoms, and he was discharged to his home. He felt well for about one week; then the fever and night sweats recurred. Repeated CT scanning of the abdomen and pelvis at that time, after the administration of oral and intravenous contrast material, showed no opacification of the main portal and superior mesenteric veins and stranding of the peripancreatic and periportal fat. The patient was referred to this hospital the next day, and he was admitted.
The patient had undergone laparoscopic cholecystectomy two years earlier. He did not smoke, drink alcohol, or use illicit drugs, and he drank minimal amounts of coffee. He was married, with two children; all three family members were well. He worked as a mechanic but had been unable to work for the past month because of his illness. An uncle and a grandmother had had lung cancer, and the patient's father had died at 64 years of age of lung disease related to asbestosis and emphysema. His mother and three siblings were in good health. He had not traveled outside the United States for 13 years. He had no pets, no exposure to animals, and no known tick bites. The family's drinking water came from a well. He had no nausea or vomiting, diarrhea, hematochezia, hematemesis, muscle weakness, joint problems, or neurologic symptoms. He took no medications.
On examination, the blood pressure was 115/70 mm Hg, the pulse 105 beats per minute, the respiratory rate 18 breaths per minute, the temperature 37.3°C, and the oxygen saturation 100 percent while the patient was breathing room air. He was in no acute distress. The skin was pale, with no rashes, lesions, or jaundice. An examination of the head, eyes, ears, neck, and chest revealed no abnormalities. The heart rate and rhythm were regular, with no murmurs, gallops, or rubs. The abdomen was soft, nontender, and nondistended, with normal bowel sounds. The liver and spleen were not palpable. The extremities appeared normal. Neurologic examination revealed no abnormalities. Specimens of blood were drawn for culture. Laboratory-test results are shown in Table 2.
Table 2. Blood-Chemistry and Immunologic Laboratory Values.
Magnetic resonance imaging (MRI) of the liver on the day after admission (before and after administration of intravenous gadolinium) showed a nonenhancing thrombus in the main portal vein that extended into both the right and left intrahepatic branches. No mass lesion, focal abnormality, or dilatation of the bile ducts was present in the liver. The hepatic arteries, hepatic veins, and splenic vein were patent.
A diagnostic test result was reported.
Differential Diagnosis
Dr. Dennis L. Kasper: This patient and his family probably had food-borne infectious diarrhea, and this patient's subsequent problems may well have resulted from complications of this primary infection.
The differential diagnosis of food-borne infections is wide, with possible pathogens ranging from bacterial to viral to protozoan. These causes may be clinically indistinguishable; stool cultures would have been useful and might have been diagnostic earlier in the course of the illness. The persistence of this man's symptoms for weeks, with fever and subsequent complications, limits the scope of the diagnostic possibilities. Overall, this complex clinical course is most consistent with a bacterial cause of diarrhea. Salmonella, campylobacter, and yersinia would be at the top of my list of possible causes.1 In the United States, salmonella and campylobacter are much more common food-borne pathogens than yersinia.2
This case begins to distinguish itself from routine infectious diarrhea in its subsequent course. The weight loss and pressure in the right upper quadrant that the patient had after the diarrheal illness are unusual and suggest other problems that may have been initiated by the original gastrointestinal infection.
Persistent Symptoms with Abnormal Results on Liver-Function Tests
The increased serum level of alanine aminotransferase and a borderline low serum level of albumin, along with the weight loss and pressure in the right upper quadrant, suggest the evolution of a new process involving the liver or perhaps the presence of underlying liver disease that predisposed the patient to more severe infection. The incubation period was too short for hepatitis to have been acquired during the patient's recent contact with apparently contaminated food. The workup for hepatitis B and EBV-related hepatitis was negative. A workup for hepatitis A was not done.
Liver disease due to iron overload or hemochromatosis would predispose a patient to yersiniosis. Yersiniae exploit host iron and use it to enhance their virulence. The detection by CT of nonspecific heterogeneous liver attenuation suggests a diffuse hepatic process. Heterogeneous attenuation, as compared with homogeneous attenuation, is less consistent with infiltrative disease due to iron deposition and more consistent with a vascular process. The patient's serum iron levels and total iron-binding capacity make hemochromatosis an unlikely diagnosis, and no aspect of his history suggests iron overload.
Deterioration of Liver Function
Over the next few weeks, the results of tests of liver function became increasingly abnormal. Alkaline phosphatase rose to a very high level, and the level of lactate dehydrogenase was high, with ongoing elevation of hepatocellular enzymes. The erythrocyte sedimentation rate continued to rise, suggesting an ongoing inflammatory process. A process that connects the diarrheal disease to the abnormal results on liver-function tests must be considered. Although pyogenic liver abscesses would have been high on my diagnostic list, repeated CT two weeks after the initial CT did not support this diagnosis. It did show ongoing inflammatory change around the cecum and possibly the terminal ileum, suggesting that the initial infection may not have resolved. The liver continued to show low-level heterogeneous attenuation, with no abscesses or biliary obstruction.
A liver biopsy showed a nonspecific increase in inflammatory cells. Without definitive pathological findings in the liver, vascular disruption causing diffuse heterogeneous liver changes must be strongly considered. The development of liver disease after infectious diarrhea, the subsequent radiologic studies, and the continuing fevers all suggest the portal venous system as a site where infection might arise, where organisms might disseminate from the bowel to the systemic circulation, and where vascular occlusion may develop, possibly as a result of portal-vein thrombosis.
Ileocecitis and Periappendicitis Associated with Food-Borne Pathogens
The upper gastrointestinal series with small-bowel follow-through showed a lack of distention in the terminal ileum, as well as mucosal irregularities. At this point, a possible link between the diarrheal illness and the continuing mucosal irregularities in the small bowel becomes more apparent. Yersinia, salmonella, and campylobacter have been associated with ileocolitis, periappendiceal inflammation, and appendicitis following acute infection.3,4,5,6,7,8,9,10,11,12 Clinically, these gastrointestinal infections may be indistinguishable once appendicitis develops.5 Of these three organisms, yersinia is most often associated around the world with subsequent inflammation of the appendix, periappendiceal areas of the ileum, or mesenteric lymph nodes.4,13 In the United States, Yersinia enterocolitica has reportedly been found in up to 9 percent of appendixes removed because of symptoms of appendicitis.4 It is not clear whether Y. enterocolitica is the causative agent of appendicitis in these cases. Y. pseudotuberculosis is a pathogen that is found primarily in animals and rarely in humans.8 Y. enterocolitica is usually transmitted to humans from domestic and wild animal reservoirs by way of food consumption. Although self-limited enterocolitis is much more commonly caused by Y. enterocolitica than by Y. pseudotuberculosis, the latter is more often associated with mesenteric lymphadenitis and terminal ileitis.8
Salmonella, which can cause pseudoappendicitis or even an illness mimicking inflammatory bowel disease,10 may have initiated the disease process in this patient. In the United States, several studies indicate that diarrheal disease due to campylobacter is more common than that due to salmonella and shigella combined.6 Campylobacter jejuni is the principal diarrheal pathogen belonging to this genus; it causes tissue injury in the jejunum, ileum, and colon and has also been associated with pseudoappendicitis.6
During an exploratory laparotomy, an inflamed retrocecal appendix was removed. Periappendiceal and pericolonic inflammation was reported. The disease process in the ileal and appendiceal areas was probably associated with the original infectious cause of diarrhea. In spite of an initial postoperative improvement, fever and night sweats recurred.
A workup at this hospital included what turned out to be an unremarkable medical history and an essentially normal physical examination. However, MRI of the liver with gadolinium contrast material yielded findings of interest. May we review these studies?
Dr. Dushyant Sahani: Contrast-enhanced CT of the abdomen, performed in arterial and portal venous phases of liver enhancement on the day before admission, showed heterogeneous enhancement of the liver parenchyma in the arterial phase (Figure 1A). However, in the portal venous phase, the liver enhanced homogeneously. An occlusive thrombus was identified in the left main branch of the portal vein and a nonocclusive thrombus in the superior mesenteric vein (Figure 1B). In addition, there was stranding in the fat, which indicated inflammation in the right lower quadrant, along the medial aspect of the cecum (Figure 1C). There was no ascites.
Figure 1. Contrast-Enhanced CT Images of the Liver Obtained One Day before Admission.
The arterial-phase image (Panel A) shows heterogeneous enhancement of the liver parenchyma at the level of the porta hepatis (arrows). In the subsequent image from the portal venous phase (Panel B), a filling defect, or thrombus, is evident in the main portal vein and the left branch (arrow). In an image obtained at the level of the right lower quadrant (Panel C), there is stranding in the pericecal fat (arrows), indicating inflammation.
One day after admission, a contrast-enhanced MRI of the liver confirmed the finding of a nonenhancing, occlusive thrombus in the portal vein. When this scan was compared with the CT scan, progression of the thrombus in the main portal vein, intrahepatic branches, and the superior mesenteric vein (Figure 2) was evident.
Figure 2. A Dynamic Contrast-Enhanced MRI Scan of the Liver Obtained on the Second Hospital Day.
An occlusive thrombus is present in the superior mesenteric vein (arrow).
Pylephlebitis Following Appendicitis
Dr. Kasper: The imaging tests and the case presentation are consistent with pylephlebitis (septic thrombophlebitis of the portal veins). This condition appears to have developed as a result of the appendicitis and periappendicitis, which were probably due to one of the diarrheal pathogens I have already mentioned. It is less likely, but possible, that these food-borne organisms were responsible for the pylephlebitis. Only rarely do case reports describe pylephlebitis as having been caused by salmonella14; yersinia and campylobacter are also unlikely suspects. Appendicitis itself is a well-known cause of pylephlebitis.15 The venous drainage from the appendiceal area and terminal ileum flows directly into the portal system (Figure 3). Bacteroides species are components of the normal ileal and cecal flora.16 In appendicitis, both vascular flow to the appendix and vascular leakage of bacteria are increased. In one series of appendicitis cases in Croatia,17 portal-vein bacteremia was found in 19 of 50 cases, but patients had systemic bacteremia in only 3 of the cases. Bacteroides and Escherichia coli predominated among the isolates.
Figure 3. The Pathogenesis of Pylephlebitis and Liver Abscess after Acute Appendicitis.
Infection in the appendix can result in bacterial leakage into the portal venous system. Bacteroides species, particularly B. fragilis, have unique virulence factors that contribute to thrombosis.
In a 1995 review of 19 cases of pylephlebitis reported in the literature after 1979,15 appendicitis was second only to diverticulitis as a predisposing condition for pylephlebitis. The organisms most frequently cultured from the blood were bacteroides (especially Bacteroides fragilis) and E. coli. Pylephlebitis often leads to inflammation and, not uncommonly, to abscess of the liver.
Bacteroides and Thrombotic Disease
Anaerobic infections are strongly associated with thrombotic diseases. Thrombophlebitis is found in as many as 5 to 12 percent of cases of anaerobic bacteremia.18 The responsible mechanism is not fully understood, but several factors may contribute. Bacteroides species elaborate enzymes that degrade heparin.19 Bacteroides can accelerate coagulation in mice, as shown by decreased clotting time; this acceleration is secondary to a cell-wall component.20 Surface components of bacteroides substantially enhance fibrin clotting, probably in tandem with a capsular component that activates macrophages and initiates the clotting cascade.21 B. fragilis expresses capsular polysaccharides that have structural characteristics that make them fibrinogenic and thus capable of facilitating abscess formation22,23; the underlying mechanism is probably quite similar to that which enables the organism to cause thrombophlebitis.21 The unique immunologic properties found in the capsular zwitterionic polysaccharides of B. fragilis promote this procoagulant activity.23 Although I would usually be inclined to favor a unifying diagnosis, in this case I think the probability of two more common associations — food-borne infection causing appendicitis and B. fragilis causing pylephlebitis — is greater. Therefore, I believe that the most definitive diagnostic test in this case was the blood culture, which I suspect grew B. fragilis (although it may have grown yersinia, salmonella, or campylobacter).
Dr. Nancy Lee Harris (Pathology): Dr. Ivers, you were the infectious-diseases fellow who cared for this patient; can you give us your clinical impressions?
Dr. Louise Ivers (Infectious Diseases): We reviewed the patient's radiology images from previous admissions with radiology staff members on the day of admission. On the basis of the finding of portal-vein thrombosis in combination with the history of subacute appendicitis and prolonged fever, our leading differential diagnosis was septic portal-vein thrombosis. Since the patient was clinically stable, we recommended that antibiotics be withheld for the first 24 hours after admission while blood cultures were obtained.
Dr. Harris: Would stool cultures have been of any use in determining what caused the patient's original diarrhea?
Dr. Kasper: The patient still had signs of active ileocecitis, as seen in the CT scan. I would like to know whether stool cultures were done.
Clinical Diagnosis
Pylephlebitis.
Dr. Dennis L. Kasper's Diagnosis
Ileocecitis with appendicitis in association with infection by yersinia, salmonella, or campylobacter, leading to pylephlebitis with B. fragilis bacteremia.
Pathological Discussion
Dr. Joseph Misdraji: The specimen obtained during liver biopsy is perhaps most remarkable for the absence of pathological findings in the portal vein, given the imaging studies that showed complete occlusion of this vessel. The portal tracts are edematous and the bile ducts appear reactive and injured. Inflammatory cells are sparse, but a few neutrophils infiltrate the bile duct epithelium (Figure 4A). Although these changes are not well developed, they raise the possibility of some degree of bile-duct obstruction. In the setting of portal-vein thrombosis, this may be due to a condition known as portal biliopathy, in which extrahepatic biliary obstruction results from an engorgement of collateral veins around the bile duct.24
Figure 4. Biopsy Specimens from the Liver and the Appendix (Hematoxylin and Eosin).
The specimen from the liver biopsy (Panel A) contains an edematous portal tract. The bile-duct epithelium shows reactive changes with nuclear enlargement, irregular nuclear placement, and cytoplasmic eosinophilia. A very few inflammatory cells are seen infiltrating the duct epithelium. The specimen from the appendix (Panel B) shows patchy mucosal ulceration consistent with appendicitis. A vein in the periappendiceal fat (Panel C) reveals a vein occluded by loose fibrous tissue, indicating an organized thrombus.
The appendix showed areas of mucosal ulceration (Figure 4B), associated with hemosiderin deposition, and periappendiceal fat necrosis, indicating an acute and chronic inflammatory process. In the periappendiceal fat, there was a thrombosed vein (Figure 4C), which is of interest in the context of the patient's portal-vein thrombosis. As Dr. Kasper predicted, a blood culture yielded B. fragilis, establishing the diagnosis of pylephlebitis as a result of infection with B. fragilis.
Pylephlebitis begins with thrombophlebitis of the small veins draining the infected site, which in this case was the appendix. The involvement, by extension, of larger veins leads to septic thrombophlebitis of the portal vein, and further extension to the large mesenteric veins may result in bowel ischemia — a rare complication that carries a high risk of death. Embolization of an infected thrombus results in liver abscesses.25 The organisms that are most often cultured from the blood are E. coli and B. fragilis. Although the role of these organisms in precipitating appendicitis is controversial, their role in abdominal infectious complications of appendicitis has been well established; these two organisms are those most frequently recovered from inflamed, gangrenous appendixes or the peritoneal fluid in cases of appendicitis.26,27,28,29,30,31,32,33 Clinical trials have shown that the use of antibiotics with activity against B. fragilis in the treatment of complicated appendicitis can reduce the rate of postoperative infectious complications.34,35
Discussion of Management
Dr. Harris: Dr. Kasper, could you discuss your recommendations for the management of bacteroides pylephlebitis?
Dr. Kasper: In a case such as this one, until culture results are known, metronidazole and a quinolone agent should be administered to treat bacteroides and food-borne infection, respectively. If the blood yielded only a food-borne pathogen, metronidazole would not be needed. Anticoagulant therapy in this clinical setting has not been proved essential but has been associated with improvement in the condition of some patients, with a decrease in septic embolization to the liver from infected portal thrombi and consequent prevention of liver abscesses. Many patients respond to antibiotics alone, whether or not they receive heparin.15
Dr. Harris: Dr. Ivers, would you tell us how you treated this patient's condition and how he is doing?
Dr. Ivers: Anticoagulation with intravenous heparin was begun. When we received the positive results of the blood cultures, we started treatment with cefepime and metronidazole and switched to levofloxacin and metronidazole after the identification of B. fragilis. The patient's fever immediately resolved. Subsequent blood cultures were negative. Although the infectious-diseases consultants had requested stool cultures, these were not done. A transthoracic echocardiogram did not suggest the presence of valvular vegetations. Esophagogastroduodenoscopy revealed grade 1 esophageal varices. The heparin was ultimately discontinued and oral warfarin was begun. A workup to evaluate the presence of a hypercoagulable state revealed a mildly elevated titer of anticardiolipin IgM, which was to be followed up after discharge. The patient was discharged to his home. He was well and preferred to follow up with his physician in his home city, who reported that the patient was doing well. The testing for hypercoagulability has not been repeated.
Dr. Sahani: Follow-up CT scans performed at another hospital two and a half months after the patient was discharged showed a persistent portal-vein thrombus, with the development of venous collaterals or cavernous transformation of the portal vein, suggestive of portal hypertension, as well as evidence of chronic liver disease, with enlargement of the caudate lobe (Figure 5).
Figure 5. A Contrast-Enhanced CT Scan Obtained Two and a Half Months after the Patient's Discharge.
There are signs of chronic portal hypertension, including the development of venous collaterals around the thrombosed portal vein (arrow) and hypertrophy of the caudate lobe of the liver (arrowhead).
Dr. Kenneth McIntosh (Pediatrics, Children's Hospital): Had blood cultures been done before the patient's admission to this hospital?
Dr. Ivers: No cultures were done before his admission here. It appeared from the record that the primary concern of the physicians caring for him before the laparoscopy was that he had an occult malignant condition — in particular, lymphoma.
Dr. Nesli Basgoz (Infectious Diseases): Should people with a thrombotic complication of infection be investigated for a hypercoagulable state? Is there any reason to expect that such patients are uniquely susceptible to thrombotic conditions because of preexisting hypercoagulable states?
Dr. Kasper: Not to my knowledge. It is the infection with B. fragilis that predisposes the patient to hypercoagulation. I know of no association of the organism with a preexisting hypercoagulable state or the subsequent development of thrombotic disease.
Anatomical Diagnosis
Portal-vein thrombosis due to B. fragilis infection.
Dr. Kasper reports having received consulting fees from Chiron Vaccines and reports holding patents in several areas of group B streptococcal vaccines and in zwitterionic polysaccharides as immunomodulators. Dr. Sahani reports having received grant support from Bracco Diagnostics.
Source Information
From the Department of Medicine, Brigham and Women's Hospital (D.L.K.); the Departments of Radiology (D.S.) and Pathology (J.M.), Massachusetts General Hospital; and the Departments of Medicine (D.L.K.), Radiology (D.S.), and Pathology (J.M.), Harvard Medical School — all in Boston.
References
Butterton JR, Calderwood SB. Acute infectious diarrheal diseases and bacterial food poisoning. In: Kasper DL, Fauci AS, Longo DL, Braunwald E, Hauser SL, Jameson LJ, eds. Harrison's principles of internal medicine. 16th ed. New York: McGraw-Hill, 2005:754-9.
Preliminary FoodNet data on the incidence of infection with pathogens transmitted commonly through food -- selected sites, United States, 2003. MMWR Morb Mortal Wkly Rep 2004;53:338-343.
Arda IS, Ergin F, Varan B, Demirhan B, Aslan H, Ozyaylali I. Acute abdomen caused by Salmonella typhimurium infection in children. J Pediatr Surg 2001;36:1849-1852.
Bennion RS, Thompson JE Jr, Gil J, Schmit PJ. The role of Yersinia enterocolitica in appendicitis in the southwestern United States. Am Surg 1991;57:766-768.
Blakelock RT, Beasley SW. Infection and the gut. Semin Pediatr Surg 2003;12:265-274.
Blaser MJ. Infections due to Campylobacter and related species. In: Kasper DL, Fauci AS, Longo DL, Braunwald E, Hauser SL, Jameson LJ, eds. Harrison's principles of internal medicine. 16th ed. New York: McGraw-Hill, 2005:907-9.
Chandler ND, Parisi MT. Yersinia enterocolitica masquerading as appendicitis. Arch Pediatr Adolesc Med 1994;148:527-528.
Dennis DT, Campbell GL. Plague and other Yersinia infections. In: Kasper DL, Fauci AS, Longo DL, Braunwald E, Hauser SL, Jameson LJ, eds. Harrison's principles of internal medicine. 16th ed. New York: McGraw-Hill, 2005:921-9.
Deutsch A, Wasserman D, Ruchelli E, Johnson J, Broussard DL. An uncommon presentation of Salmonella. Pediatr Emerg Care 1996;12:285-287.
Lesser CF, Miller SI. Salmonellosis. In: Kasper DL, Fauci AS, Longo DL, Braunwald E, Hauser SL, Jameson LJ, eds. Harrison's principles of internal medicine. 16th ed. New York: McGraw-Hill, 2005:897-902.
Naktin J, Beavis KG. Yersinia enterocolitica and Yersinia pseudotuberculosis. Clin Lab Med 1999;19:523-536.
Puylaert JB, Van der Zant FM, Mutsaers JA. Infectious ileocecitis caused by Yersinia, Campylobacter, and Salmonella: clinical, radiological, and US findings. Eur Radiol 1997;7:3-9.
Dionisio D, Belli A, Dionisio A, et al. Appendicitis: microbial interactions and new pathogens. Recenti Prog Med 1992;83:330-336.
Knobel B, Sommer I, Menashes Z. A rare presentation of systemic salmonellosis. Infection 1985;13:70-72.
Plemmons RM, Dooley DP, Longfield RN. Septic thrombophlebitis of the portal vein (pylephlebitis): diagnosis and management in the modern era. Clin Infect Dis 1995;21:1114-1120.
Hentges DJ. Anaerobes as normal flora. In: Finegold SM, George WL, eds. Anaerobic infections in humans. San Diego, Calif.: Academic Press, 1989:37-53.
Juric I, Primorac D, Zagar Z, et al. Frequency of portal and systemic bacteremia in acute appendicitis. Pediatr Int 2001;43:152-156.
Gorbach SL, Bartlett JG. Anaerobic infections. N Engl J Med 1974;290:1289-1294.
Gesner BM, Jenkin CR. Production of heparinase by Bacteroides. J Bacteriol 1961;81:595-604.
Bjornson HS, Hill EO. Bacteroidaceae in thromboembolic disease: effects of cell wall components on blood coagulation in vivo and in vitro. Infect Immun 1973;8:911-918.
Rosenthal GA, Levy G, Rotstein OD. Induction of macrophage procoagulant activity by Bacteroides fragilis. Infect Immun 1989;57:338-343.
Tzianabos AO, Onderdonk AB, Rosner B, Cisneros RL, Kasper DL. Structural features of polysaccharides that induce intra-abdominal abscesses. Science 1993;262:416-419.
Choi Y-H, Roehrl MH, Kasper DL, Wang JY. A unique structural pattern shared by T-cell-activating and abscess-regulating zwitterionic polysaccharides. Biochemistry 2002;41:15144-15151.
Perego P, Cozzi G, Bertolini A. Portal biliopathy. Surg Endosc 2003;17:351-352.
Slovis TL, Haller JO, Cohen HL, Berdon WE, Watts FB Jr. Complicated appendiceal inflammatory disease in children: pylephlebitis and liver abscess. Radiology 1989;171:823-825.
Martirosian G, Bulanda M, Wojcik-Stojek B, et al. Acute appendicitis: the role of enterotoxigenic strains of Bacteroides fragilis and Clostridium difficile. Med Sci Monit 2001;7:382-386.
Mosdell DM, Morris DM, Fry DE. Peritoneal cultures and antibiotic therapy in pediatric perforated appendicitis. Am J Surg 1994;167:313-316.
Rautio M, Saxen H, Siitonen A, Nikku R, Jousimies-Somer H. Bacteriology of histopathologically defined appendicitis in children. Pediatr Infect Dis J 2000;19:1078-1083.
Heseltine PN, Yellin AE, Appleman MD, et al. Perforated and gangrenous appendicitis: an analysis of antibiotic failures. J Infect Dis 1983;148:322-329.
Bennion RS, Baron EJ, Thompson JE Jr, et al. The bacteriology of gangrenous and perforated appendicitis -- revisited. Ann Surg 1990;211:165-171.
Pieper R, Kager L, Lindberg AA, Nord CE. Acute appendicitis and Bacteroides fragilis. Scand J Infect Dis Suppl 1979;19:92-97.
Pieper R, Kager L, Weintraub A, Lindberg AA, Nord CE. The role of Bacteroides fragilis in the pathogenesis of acute appendicitis. Acta Chir Scand 1982;148:39-44.
Elhag KM, Alwan MH, Al-Adnani MS, Sherif RA. Bacteroides fragilis is a silent pathogen in acute appendicitis. J Med Microbiol 1986;21:245-249.
Wilson SE, Boswick JA Jr, Duma RJ, et al. Cephalosporin therapy in intraabdominal infections: a multicenter randomized, comparative study of cefotetan, moxalactam, and cefoxitin. Am J Surg 1988;155:61-66.
Bennion RS, Thompson JE, Baron EJ, Finegold SM. Gangrenous and perforated appendicitis with peritonitis: treatment and bacteriology. Clin Ther 1990;12:Suppl C:31-44.(Dennis L. Kasper, M.D., D)