QJM Advance Access originally published online on October 3, 2005
QJM 2005 98(11):821-827; doi:10.1093/qjmed/hci126
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Risk factors for bacteraemia and endovascular infection due to non-typhoid salmonella: a reappraisal
From the Department of Surgery, National Taiwan University Hospital, National Taiwan University College of Medicine, Taipei, Taiwan, ROC
Address correspondence to Dr F.-Y. Lin, National Taiwan University Hospital, No.7, Chung-Shan S. Rd, Taipei, Taiwan 100, ROC. email: fylin1{at}ha.mc.ntu.edu.tw
Received 17 March 2005 and in revised form 13 August 2005
 |
Summary |
|---|
 Top Summary IntroductionMethodsResultsDiscussionReferences |
|---|
Background: Endovascular infections are rare complications of non-typhoid salmonellosis. The diagnosis is frequently not established until the infection is advanced. It is important to identify high-risk patients and treat them as early as possible.
Aim: To identify risk factors for bacteraemia and endovascular infection in patients with non-typhoid salmonellosis.
Design: Retrospective study.
Setting: A single tertiary-care hospital in Taiwan.
Methods: Data were collected by retrospective chart review. Log-logistic regression modelling was used to identify independent risk factors for bacteraemia and endovascular infection. We analysed the characteristics of patients with gastroenteritis vs. those with bacteraemia, and of bacteraemic patients with vs. without endovascular infection.
Results: Between 1984 and 2004, there were 373 adult cases of non-typhoid salmonellosis. There were 76 intestinal Salmonella infections, 290 bloodstream infections (including 47 endovascular infections), and 7 extra-intestinal non-bacteraemic infections. The independent positive predictors of bacteraemia were systemic lupus erythematosus, liver cirrhosis, HIV infection, and solid organ cancers. The only independent positive predictor of endovascular infection was atherosclerosis. The independent negative predictors of endovascular infection were solid organ cancers and immunodeficiency.
Discussion: Risk factors for atherosclerosis predisposed our patients with bacteraemia to endovascular infection. Although immunodeficiency predisposed patients to bacteraemia, it was associated with a low incidence of endovascular infection.
|
Introduction |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Infections due to non-typhoid Salmonella are common, and their occurrence appears to be increasing in many countries.1–3 Non-typhoid Salmonella are important food-borne pathogens that cause gastroenteritis, bacteraemia, and subsequent focal metastatic infection.1,4 Endovascular infection is an important focal complication associated with a significant mortality and morbidity.5–18 Although endovascular infections are rare complications of salmonellosis, the diagnosis is frequently not established until the patient has an advanced stage of the disease. Because of its insidious onset and poor prognosis, it is important to treat patients with non-typhoid salmonellosis at an early stage, before the development of endovascular infection.9 Imaging studies with computed tomography and magnetic resonance imaging are the only sensitive diagnostic methods, but the cost is high. Therefore, it is important to identify high-risk patients and treat them as early as possible. Traditionally, old age, atherosclerosis and immunodeficiency were considered risk factors for endovascular infection. However, because of small case numbers, the risk factors for endovascular infection in patients with non-typhoid salmonellosis have not been evaluated in a detailed way.13 In a previous study, we reported that immunodeficiency predisposed patients with non-typhoid salmonellosis to acquire bacteraemia, but did not predispose bacteraemic patients to acquire endovascular infection.14 Among the traditional risk factors for atherosclerosis, only old age has been shown to be associated with increased incidence of endovascular infection.13 To ascertain whether patients with risk factors for atherosclerosis and immunodeficiency are more or less likely to have endovascular infection, we retrospectively analysed the medical records of all adult patients with non-typhoid salmonellosis in a longer study interval. This is an update, adding 213 new patients to our previously 160 reported patients,14 and also a reappraisal of patient risk factors for bacteraemia and endovascular infection in a larger patient cohort.
|
Methods |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Setting
The National Taiwan University Hospital (Taipei, Taiwan) is a 2000-bed tertiary care hospital. It serves an urban population of 2 million people as both first-line and tertiary facilities. It also serves as a referral centre for other hospitals in the country (population 22 million).
Patients
Patients were recruited from a retrospective analysis of the clinical and bacteriological data for adult patients (
18 years old) with non-typhoid salmonellosis from September 1984 through September 2004. Patients were included if they had at least one culture positive for non-typhoid Salmonella. Cases of S. typhi and S. paratyphi were excluded from the study. Where recurrence of Salmonella infection with the same pathogen occurred in the same patient, only the first episode was included in analysis. During the study interval, the diagnosis and measurements were consistent over time.
Data collection
The medical records of patients with non-typhoid salmonellosis were reviewed, and data were recorded for demographics (age and gender), underlying medical diseases (diabetes mellitus, hypertension, peripheral arterial occlusive disease, previous ischaemic stroke, coronary artery disease, liver cirrhosis, uraemia, systemic lupus erythematosus, human immunodeficiency virus infection, haematological malignancy and solid organ malignancy), bacteriology (serogroups A, B, C, D, E, G, and H) and hospital outcome (death). Based on the assumed pathogenic mechanisms, we divided those patients with non-typhoid salmonellosis into three groups: (i) intestinal salmonellosis or gastroenteritis, those with diarrhoea and a positive stool culture of Salmonella; (ii) bacteraemia, those with a positive blood culture of Salmonella; and (iii) endovascular infection, those with infective endocarditis, infected aortic aneurysm, or infected intravascular implants, which was documented by a imaging study or a tissue culture.
We defined immunodeficiency as having acquired human immunodeficiency virus (HIV) infection, use of corticosteroid and chemotherapeutic agents for malignancy or autoimmune diseases, or use of immunosuppressive agents for organ transplantation.14 Atherosclerosis was defined as having traditional risk factors for atherosclerosis (diabetes mellitus, and hypertension), or certain atherosclerotic diseases (peripheral arterial occlusive disease, previous ischaemic stroke, and coronary artery disease).
Bacteriology
Isolates of Salmonella were initially identified by biochemical tests. Serogrouping was determined by agglutination testing with use of antisera specific to O antigen. Serotypes were not routinely determined.
Statistical analysis
Analysis used the 
2 test, Fisher's exact test and Mann-Whitney U test. Log-logistic regression modelling, as a generalized linear model with binomial error structure and log link, was used to identify independent risk factors for bacteraemia and endovascular infection. First, we analysed patient characteristics that distinguished between patients with gastroenteritis and patients with bacteraemia. Second, we analysed patient characteristics that distinguished between bacteraemic patients with endovascular infection and those without. We used our clinical knowledge to choose the important variables in univariate analysis for a log-logistic regression model.
|
Results |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Patients
Between 1984 and 2004, 373 adult cases of non-typhoid salmonellosis were admitted to our hospital. Among them, there were 76 intestinal Salmonella infections, 290 blood stream infections, and seven extra-intestinal non-bacteraemic infections. Among the 290 bacteraemic patients, 47 patients had endovascular infections. Among the 47 endovascular infections, there were five infective endocarditis, 40 infected aortic aneurysms, one infected aneurysm of superior mesenteric artery and one infected vascular graft of arteriovenous fistula.
Of the 366 patients with either bacteraemia or intestinal infection, 26 (7%) had diabetes mellitus, 40 (11%) had hypertension, 17 (5%) had coronary artery disease, 3 (1%) had peripheral arterial occlusive disease, 12 (3%) had previous ischaemic stroke, 13 (4%) had uraemia, 32 (9%) had liver cirrhosis, 40 (11%) had systemic lupus erythematosus, 43 (12%) had haematological malignancies, 66 (18%) had solid organ cancers, and 175 (48%) had immunodeficiency.
Salmonella serogroups
Three (0.8%) isolates were Group A, 140 (38%) isolates were Group B, 85 (23%) isolates were Group C, 111 (30%) isolates were Group D, 5 (1%) isolates were Group E, 1 (0.3%) isolate was Group G, 1 (0.3%) isolate was Group H and 27 (7%) isolates were unidentified. The distribution of serogroups did not change during the study period.
Risk factor analysis
Table 1 shows the clinical data in patients who had bacteraemia and in those who had gastroenteritis alone. Patients with bacteraemia were older, and had more underlying diseases of liver cirrhosis, systemic lupus erythematosus, immunodeficiency and solid organ cancers. Table 2 shows the clinical data in bacteraemic patients who had endovascular infections and in those who did not. Bacteraemic patients with endovascular infection were older, had more underlying disease of diabetes mellitus, hypertension, peripheral arterial occlusive disease, and previous ischaemic stroke, but less liver cirrhosis, immunodeficiency and solid organ cancers. Table 3 shows the independent patient risk factors for bacteraemia by log-logistic regression. The independent positive predictors of bacteraemia were systemic lupus erythematosus, liver cirrhosis, HIV infection and solid organ cancers. Table 4 shows the independent patient risk factors for endovascular infection in non-typhoid Salmonella bacteraemia by log-logistic regression. The only independent positive predictor of endovascular infection was atherosclerosis. The independent negative predictors of endovascular infection were solid organ cancers and immunodeficiency.
|
|
|
|
|
Discussion |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
In our patients with non-typhoid salmonellosis, the major patient risk factors for bacteraemia were certain immunocompromised conditions. Endovascular infection is a feared complication of non-typhoid Salmonella bacteraemia, and the prognosis is poor.4–9 Although the prognosis was poor in previous reports,5 we have reported that the prognosis of Salmonella-infected aortic aneurysms in Taiwan is not invariably poor; timely surgical management and prolonged postoperative antibiotic treatment can provide an improved outcome.17 Early identification of high-risk patients and aggressive treatment is most important in achieving a good outcome.
In patients presenting with non-typhoid Salmonella bacteraemia in the absence of gastroenteritis, underlying immunodeficiency should be excluded first.10 These findings point to the importance of the immune system in the defence against Salmonella infection. Salmonella invasion requires the coordinated expression of complex arrays of virulence factors that allow the bacterium to evade the host's immune system. All Salmonella serotypes share the ability to invade the host by inducing their own uptake into cells of the intestinal epithelium and establishing systemic infection through their ability to survive and replicate in mononuclear phagocytes.19–21 The pathogenesis of endovascular infection can thus be divided into three steps. The first step is transmission to a susceptible host. Transmission of Salmonella to humans is usually by consumption of contaminated food, most commonly beef, poultry or eggs.21 Significant advances in the understanding of how Salmonella interacts with the intestinal mucosa to cause disease have been made during the past decade, and attachment of the bacterium to the host cell surface may be an essential step in pathogenesis. This interaction involves surface structures on the bacteria that are capable of binding membrane components on the host cell surface and extracellular matrix proteins such as laminin and collagens. Fimbriae of the Salmonella may play a critical role in virulence by allowing bacteria to interact with host cells.21 The second step is establishment of systemic infection. The mechanisms of systemic Salmonella infections are characterized by survival and proliferation of bacteria inside macrophages.19 The final step is establishment of endovascular infection. However, little is known about the mechanisms or patient risk factors involving in disease progression from bacteraemia to endovascular infection.
Immunodeficiency
In previous reports, HIV infection and systemic lupus erythematosus were prominent risk factors for non-typhoid salmonellosis and bacteraemia.22–26 Traditionally, patients with immunodeficiency are thought to be susceptible to Salmonella endovascular infection. We believe we are the first group to question whether immunodeficiency is a risk factor for endovascular infection.14 In this study, although patients with bacteraemia had more underlying diseases with certain immunodeficient states, solid organ cancers and immunodeficiency were negative predictors of endovascular infection.
Immunodeficiency has a contradictory role in the pathogenesis of endovascular infection. Immunodeficiency predisposed patients to acquire non-typhoid Salmonella bacteraemia, but the incidence of endovascular infection was low in immunodeficient patients. This concept is supported by two important clinical observations. First, despite the prominence of bacteraemic infection with non-typhoid Salmonella species in patients with acquired HIV infection or systemic lupus erythematosus, only a few series have described endovascular infection in these patients,27–29 and none of our HIV-infected patients had endovascular infection (Table 2). This suggests that the immune system may play a critical role in the process of inflammation and destruction of aorta and heart valves. Second, pathological examination of surgical specimens from patients with infected aortic aneurysms showed acute inflammation was present in 62% and chronic inflammation in another 27%.30,31 But bacterial culture from the aneurysm wall was positive in less than half of surgical cases.31 This also suggests that the aortic wall destruction commonly seen in infected aortic aneurysm results from an immune-mediated mechanism rather than direct destruction by the bacteria. Further understanding of these mechanisms may yield new approaches for the prevention and treatment of life-threatening endovascular infections caused by non-typhoid Salmonella or other pathogens.
Atherosclerosis
Traditionally, atherosclerosis is considered a risk factor for endovascular infection. In previous reports, patients with Salmonella-infected aortic aneurysms have usually had heavy calcification of the aortic wall.31–33 Aortic atherosclerosis was present in 75% of surgical specimens obtained from patients with infected aortic aneurysm.32 Traditional risk factors for atherosclerosis included old age, diabetes mellitus, and hypertension.34 However, previous studies have not adequately quantified the risk of atherosclerosis by a statistical comparison of risk factors for atherosclerosis.14,15 In this study, patients with atherosclerotic disease, such as coronary artery disease, peripheral arterial occlusive disease and previous ischaemic stroke, had a high risk of endovascular infection (Table 2). Risk factors for atherosclerosis predisposed patients with blood stream infection to acquire endovascular infection (Table 4).
Liver cirrhosis
Bacterial infection is a well-described complication of liver cirrhosis and is a major cause of death. Bacteraemia was reported to occur in approximately 4–9% of hospitalized patients with cirrhosis. The incidence of bacteraemia was significantly higher in patients with decompensated liver disease than in compensated liver disease. Escherichia coli and Klebsiella pneumoniae were the two micro-organisms most frequently isolated. Non-typhoid salmonella was an infrequent cause of bacteraemia in cirrhotic patients.35 In this study, liver cirrhosis was associated with high incidence of non-typhoid Salmonella bloodstream infection. Reticuloendothelial blockade predisposed patients to non-typhoid Salmonella infection.1 Multiple immunological abnormalities have been described in patients with liver cirrhosis. Patients with liver cirrhosis have defects in both humoural and cell-mediated immunity. The reticuloendothelial cells represent a potent filtering mechanism for portal blood-borne pathogens. In patients with liver cirrhosis, filtering is impaired.36
Study limitations
The major limitations of our study were that it was retrospective, and that patients with severe disease were referred to our hospital. Because our study was retrospective, other traditional risk factors for atherosclerosis, such as serum levels of cholesterol and triglycerides were not evaluated. Although some of our results have been reported in previous studies,14,18 there are 373 patients and 47 endovascular infections in this current series, and only 160 patients and 28 endovascular infections in our previous report.14 Although 35 patients with endovascular infection were reported in our another previous report,18 the main objectives of that paper was risk factors for death in infected aortic aneurysm, not risk factors for endovascular infection.
Conclusions
In this study, systemic lupus erythematosus, liver cirrhosis, solid organ cancers and immunodeficiency predisposed patients with non-typhoid salmonellosis to acquire blood stream infections, and risk factors for atherosclerosis predisposed patients with blood stream infections to acquire endovascular infection. Immunodeficiency and solid organ cancers were associated with a lower incidence of endovascular infection. We recommend that imaging studies be done in high-risk patients with non-typhoid Salmonella bacteraemia. Risk factors include old age, diabetes mellitus, absence of certain immunodeficient states and the presence of certain atherosclerotic diseases.
|
References |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
1. Hohmann EL. Nontyphoidal salmonellosis. Clin Infect Dis 2001; 32:263–9.[CrossRef][Web of Science][Medline]
2. Centers for Disease Control. Outbreaks of Salmonella serotype enteritidis infection associated with eating shell eggs-United States, 1999–2001. MMWR Morb Mortal Wkly Rep 2003; 51:1149–52.[Medline]
3. Tirado C, Schmidt K. WHO surveillance programme for control of foodborne infections and intoxications: preliminary results and trends across greater Europe. World Health Organization. J Infect 2001; 43:80–4.[Web of Science][Medline]
4. Cohen JI, Bartlett JA, Corey GR. Extra-intestinal manifestations of Salmonella infections. Medicine 1987; 66:349–88.[Medline]
5. Soravia-Dunand VA, Loo VG, Salit IE. Aortitis due to Salmonella: report of 10 cases and comprehensive review of the literature. Clin Infect Dis 1999; 29:862–8.[Web of Science][Medline]
6. Veraldi GF, de Manzoni G, Laterza E, et al. Extraintestinal infection by group C Salmonella: a case report and review of the literature. Hepatogastroenterology 2001; 48:471–4.[Medline]
7. Katz SG, Andros G, Kohl RD. Salmonella infections of the abdominal aorta. Surg Gynecol Obstet 1992; 175:102–6.[Web of Science][Medline]
8. Meerkin D, Yinnon AM, Munter RG, Shemesh O, Hiller N, Abraham AS. Salmonella mycotic aneurysm of the aortic arch: case report and review. Clin Infect Dis 1995; 21:523–8.[Web of Science][Medline]
9. Oskoui R, Davis WA, Gomes MN. Salmonella aortitis. A report of a successfully treated case with a comprehensive review of the literature. Arch Intern Med 1993; 153:517–25.
10. Brown M, Eykyn SJ. Non-typhoidal Salmonella bacteraemia without gastroenteritis: a marker of underlying immunosuppression. Review of cases at St. Thomas' Hospital 1970–1999. J Infect 2000; 41:256–9.[CrossRef][Web of Science][Medline]
11. Shimoni Z, Pitlik S, Leibovici L, Samra Z, Konigsberger H, Drucker M, Agmon V, Ashkenazi S, Weinberger M. Nontyphoid Salmonella bacteremia: age-related differences in clinical presentation, bacteriology, and outcome. Clin Infect Dis 1999; 28:822–7.[Web of Science][Medline]
12. Ramos JM, Garcia-Corbeira P, Aguado JM, Arjona R, Ales JM, Soriano F. Clinical significance of primary vs. secondary bacteremia due to nontyphoid Salmonella in patients without AIDS. Clin Infect Dis 1994; 19:777–80.[Web of Science][Medline]
13. Cohen PS, O'Brien TF, Schoenbaum SC, Medeiros AA. The risk of endothelial infection in adults with salmonella bacteremia. Ann Intern Med 1978; 89:931–2.[CrossRef][Web of Science][Medline]
14. Hsu RB, Tsay YG, Chen RJ, Chu SH. Risk factors for primary bacteremia and endovascular infection in patients without acquired immunodeficiency syndrome who have nontyphoid salmonellosis. Clin Infect Dis 2003; 36:829–34.[CrossRef][Web of Science][Medline]
15. Benenson S, Raveh D, Schlesinger Y, Alberton J, Rudensky B, Hadas-Halpern I, Yinnon AM. The risk of vascular infection in adult patients with nontyphi Salmonella bacteremia. Am J Med 2001; 110:60–3.[CrossRef][Web of Science][Medline]
16. Hsu RB, Tsay YG, Wang SS, Chu SH. Surgical treatment for primary infected aneurysm of descending thoracic aorta, abdominal aorta and iliac arteries. J Vasc Surg 2002; 36:746–50.[Web of Science][Medline]
17. Hsu RB, Tsay YG, Wang SS, Chu SH. Management of aortic aneurysm infected with Salmonella. Br J Surg 2003; 90:1080–4.[CrossRef][Web of Science][Medline]
18. Hsu RB, Chen RJ, Wang SS, Chu SH. Infected aortic aneurysms: clinical outcome and risk factor analysis. J Vasc Surg 2004; 40:30–5.[CrossRef][Web of Science][Medline]
19. Ohl ME, Miller SI. Salmonella: a model for bacterial pathogenesis. Annu Rev Med 2001; 52:259–74.[CrossRef][Web of Science][Medline]
20. Fierer J, Guiney DG. Diverse virulence traits underlying different clinical outcomes of Salmonella infection. J Clin Invest 2001; 107:775–80.[Web of Science][Medline]
21. Darwin KH, Miller VL. Molecular basis of the interaction of Salmonella with the intestinal mucosa. Clin Microbiol Rev 1999; 12:405–28.
22. Lim E, Koh WH, Loh SF, Lam MS, Howe HS. Non-thyphoidal salmonellosis in patients with systemic lupus erythematosus. A study of fifty patients and a review of the literature. Lupus 2001; 10:87–92.
23. Tsao CH, Chen CY, Ou LS, Huang JL. Risk factors of mortality for salmonella infection in systemic lupus erythematosus. J Rheumatol 2002; 29:1214–18.
24. Arthur G, Nduba VN, Kariuki SM, Kimari J, Bhatt SM, Gilks CF. Trends in bloodstream infections among human immunodeficiency virus-infected adults admitted to a hospital in Nairobi, Kenya, during the last decade. Clin Infect Dis 2001; 33:248–56.[CrossRef][Web of Science][Medline]
25. Angulo FJ, Swerdlow DL. Bacterial enteric infections in persons infected with human immunodeficiency virus. Clin Infect Dis 1995; 21:S84–93.
26. Blaser MJ, Feldman RA. Salmonella bacteremia: reports to the Centers for Disease Control, 1968–1979. J Infect Dis 1981; 143:743–6.[Web of Science][Medline]
27. Fernandez Guerrero ML, Ramos JM, Nunez A, Nunez A, de Gorgolas M. Focal infections due to non-typhi Salmonella in patients with AIDS: report of 10 cases and review. Clin Infect Dis 1997; 25:690–7.[Web of Science][Medline]
28. Fernandez Guerrero ML, Torres Perea R, Gomez Rodrigo J, Nunez Garcia A, Jusdado JJ, Ramos Rincon JM. Infectious endocarditis due to non-typhi Salmonella in patients infected with human immunodeficiency virus: report of two cases and review. Clin Infect Dis 1996; 22:853–5.[Web of Science][Medline]
29. Taylor G, Goodfield MJ. Fatal salmonellosis in systemic lupus erythematosus. Clin Exp Dermatol 1995; 20:255–7.[CrossRef][Web of Science][Medline]
30. Salzberger LA, Cavuoti D, Barnard J. Fatal salmonella aortitis with mycotic aneurysm rupture. Am J Forensic Med Pathol 2002; 23:382–5.[CrossRef][Web of Science][Medline]
31. Miller DV, Oderich GS, Aubry MC, Panneton JM, Edwards WD. Surgical pathology of infected aneurysms of the descending thoracic and abdominal aorta: clinicopathologic correlations in 29 cases (1976 to 1999). Hum Pathol 2004; 35:1112–20.[CrossRef][Web of Science][Medline]
32. Lee MH, Chan P, Chiou HJ, Cheung WK. Diagnostic imaging of Salmonella-related mycotic aneurysm of aorta by CT. Clin Imaging 1996; 20:26–30.[CrossRef][Web of Science][Medline]
33. Macedo TA, Stanson AW, Oderich GS, Johnson CM, Panneton JM, Tie ML. Infected aortic aneurysms: imaging findings. Radiology 2004; 231:250–7.
34. Fruchart JC, Nierman MC, Stroes ES, Kastelein JJ, Duriez P. New risk factors for atherosclerosis and patient risk assessment. Circulation 2004; 109:III15–19.
35. Brann OS. Infectious complications of cirrhosis. Curr Gastroenterol Rep 2001; 3:285–92.[Medline]
36. Johnson DH, Cunha BA. Infections in cirrhosis. Infect Dis Clin North Am 2001; 15:363–71.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||