Annals
Established in 1927 by the American College of Physicians
:
Advanced search
 
box Article
 arrow  Table of Contents                
space
 arrow  Abstract of this article Free
space
 arrow  Figures/Tables List
space
 arrow  Articles citing this article
space
box Services
 arrow  Send comment/rapid response letter
space
 arrow  Notify a friend about this article
space
 arrow  Alert me when this article is cited
space
 arrow  Add to Personal Archive
space
 arrow  Download to Citation Manager
space
 arrow  ACP Search                        
space
 arrow  Get Permissions
space
box Google Scholar
 arrow  Search for Related Content
space
box PubMed
Articles in PubMed by Author:
  arrow  Kearney, R. A.
space
  arrow  Wolf, J. E.
space
 arrow  Related Articles in PubMed
space
 arrow  PubMed Citation
space
 arrow  PubMed
space

REVIEW

Nonvalvular Infections of the Cardiovascular System

right arrow Rosemary A. Kearney; Howard J. Eisen; and Judith E. Wolf

1 August 1994 | Volume 121 Issue 3 | Pages 219-230

Purpose: To review the risk factors, clinical presentation, diagnosis, and treatment of nonvalvular infections of the cardiovascular system.

Data Sources: A MEDLINE search of English-language articles from 1966 to 1993 relating to myocardial abscesses, mural endocarditis, infective endarteritis and mycotic aneurysms, infection of pacemakers and implantable defibrillators, prosthetic vascular graft infections, and infected atrial myxomas; manual review of article bibliographies.

Study Selection and Data Extraction: Case series and single reports, prospective and retrospective clinical studies, autopsy studies, and reviews were selected if they contained sufficient information about the prevalence, clinical manifestations, microbiologic features, management, and outcome of nonvalvular cardiovascular infections.

Results: Nonvalvular infections of the cardiovascular system most commonly occur on previously damaged endocardium or vascular intima and are usually associated with intravascular devices such as graft material or pacemakers. Rarely, they can involve primary cardiac tumors such as myxomas. Most patients affected are beyond the fifth decade of life. Risk factors include cardiovascular disease, diabetes mellitus, and malignancy. The sexes are affected equally. The clinical presentations of nonvalvular endovascular infections are subtle and diagnoses are difficult to make, often requiring not only a high index of suspicion but also the use of sophisticated radiologic techniques. Hence, for many of these infections, the diagnosis is made late in the course of the infection, and survival rates are poor. Complications include peripheral embolization, cardiac rupture, vascular aneurysm rupture, and pericarditis. Therapy frequently involves surgical intervention in addition to the use of antibiotics.

Conclusions: Although uncommon, nonvalvular infections of the cardiovascular system will increase in frequency as the use of implantable devices and prosthetic materials increases in the elderly. Studies are needed to determine the most appropriate diagnostic methods, treatment regimens, and methods for prevention of these infections.


Most reviews of cardiovascular infections focus on valvular endocarditis and pericarditis. However, infections of abnormal or previously damaged nonvalvular endocardium or vascular intima, often associated with intravascular devices and prosthetic material, have been increasingly recognized. The most common nonvalvular endovascular infections are primary bloodstream infections and septic thrombophlebitis related to temporary intravascular catheters, used in more than 20 million patients in the United States each year [1]. Although they account for 4% to 14% of nosocomial bacteremias [2], we do not review catheter-related infections. Instead, we present an overview of pyogenic infections of myocardium, mural endocardium, cardiac tumors, implantable devices, arterial walls, and prosthetic vascular grafts.


Myocardial Abscesses
space

Myocardial abscesses are rare. We identified 202 patients in four autopsy studies and several case reports. The frequency of myocardial abscess in routine adult patient autopsies ranged from 0.18% to 1.52% [3-5].

Risk Factors

Myocardial abscesses develop either after bacteremia or fungemia or from direct extension from valvular or mural endocarditis [3, 4]. Patients at greatest risk from hematogenous seeding include those with prolonged hospitalizations, indwelling venous catheters, and extended courses of antibiotic therapy. Several areas of myocardium are often involved. For example, the results of a study of 63 patients with abscesses documented at autopsy by Kim and colleagues [5] found all to have multiple myocardial abscesses and 81% to have abscesses in other organs. In contrast, contiguous extension from local disease occurs more commonly in persons with valvular endocarditis. Abscesses involving the paravalvular region are most often associated with prosthetic valve endocarditis, reportedly occurring in 15% to 45% of cases [6]. Myocardial abscesses have also been reported to develop in areas of previous myocardial infarction [7-9]. This may involve seeding via collateral circulation compounded by reduced arterial flow in the infarction region [7].

Microbiological Characteristics

As might be expected, Staphylococcus aureus is the most frequently reported bacterial isolate in patients with myocardial abscesses [10]. Isolated case reports have described infections secondary to group F streptococci [7], Bacteroides spp [8]., Escherichia coli [5], Salmonella spp [11]., Clostridium perfringens [9], as well as other bacteria. Fungal myocardial abscesses with Candida and Aspergillus spp. may also occur, usually in patients with prolonged hospitalizations, in those having surgical procedures, and in immunosuppressed patients [5, 12-14]. The combination of prolonged courses of antibiotics that alter the normal bacterial flora and cytotoxic immunosuppressive agents, underlying malignancy, and poor nutritional status frequently promotes fungal superinfection in these patients. In an autopsy series described by Ihde and colleagues [13], myocardial abscesses were found in 17 of 85 (20%) patients with cancer and disseminated candidiasis. In contrast, none of the patients who presented with localized candidiasis was found to have myocardial involvement. In immunocompromised patients, fungal myocarditis appears to be more common than fungal valvular endocarditis [14].

Clinical Presentation

With the exception of paravalvular involvement, the presentation of myocardial abscesses is usually subtle and clinically silent such that the diagnosis is only made at autopsy. Patients may have nonspecific findings such as low-grade fevers, chills, and leukocytosis. Conduction system abnormalities and nonspecific electrocardiographic changes may also occur. Many patients have no history of underlying cardiac disease. For example, Palank and colleagues [15] described a 19-year-old patient who presented with fever, chest pain, and an acute anterior myocardial infarction pattern on electrocardiogram 12 days after treatment for a dentoalveolar abscess. Histologically, bacteria originating in the oral cavity were present in multiple patchy areas of inflammation and necrosis.

Complications typically relate to the size and location of the abscess or abscesses. Cases of ruptured myocardial abscesses resulting in tamponade, hemopericardium, and purulent pericarditis have been described [8, 9, 16]. Other complications include the development of mural endocarditis with embolization [9], diffuse myocarditis [15], intracardiac fistulae [12], cardiac dysrhythmias [17], and generalized sepsis [5, 16].

Diagnosis

Delays in diagnosis occur because of a low index of clinical suspicion or because other more profound clinical signs associated with bacteremia or fungemia mask the subtle signs of myocardial abscess. Of the 44 patients with bacterial myocardial abscesses for whom blood culture results were reported, the bacterial isolation rate was 75%. In those patients with fungal abscesses, however, isolation of fungal pathogens from routine blood cultures rarely exceeded 50%, even in patients with disseminated disease. In the autopsy study by Ihde and colleagues [13] described above, 12 of the 17 patients had hematologic malignancies and evidence of extensive hematogenous dissemination but only 7 had positive cultures before death. In their autopsy series, Kim and colleagues found a similar disparity [5]; candidal myocardial abscesses were identified histologically in 37% of patients, yet fungal pathogens were recovered in less than 20% of blood cultures. This low yield can be improved by following several guidelines. Because fungi are generally present in lower concentrations than bacteria, a sufficient volume of blood (preferably 20 to 30 mL) should be cultured to ensure optimal recovery [18]. Similarly, the yield from routine blood cultures can be enhanced by culturing multiple samples of blood, venting the bottles, subculturing onto media promoting fungal growth, and incubating cultures for at least 4 weeks [18, 19]. Finally, when the suspicion for fungemia or disseminated fungal infection is high, the optimal blood culture method for recovery of fungi, lysis centrifugation (Isolator, Dupont; Wilmington, Delaware), should be used. With this technique, cellular components of the blood are lysed and organisms subsequently concentrated via centrifugation, allowing for more rapid isolation and enhanced fungal recovery [20].

Serial electrocardiograms may reveal conduction abnormalities such as first-degree atrioventricular block and complete heart block [17] resulting from myocardial abscesses in patients with an otherwise unexplained febrile illness. Since transesophageal echocardiography has been shown to improve the diagnosis of abscesses associated with valvular endocarditis [21], this technique should also be applied to patients with suspected myocardial abscess from any source. Although the usefulness of magnetic resonance imaging (MRI) is unknown, Akins and colleagues [22] found that MRI complemented echocardiography in detecting paravalvular abscesses and pseudoaneurysm formation complicating valvular endocarditis in five patients, especially in those infections involving prosthetic valves.

Clinical Course

The mortality rate in the 202 patients we identified as having myocardial abscesses varied with the location and pathogenesis of the infection. Of the 58 patients with paravalvular abscesses, more than 50% had associated staphylococcal aortic valve endocarditis and 75% survived [21, 23, 24]. In contrast, the mortality rate for the remaining 144 nonparavalvular infections was 100%, in part because of the patients' underlying diseases but also because of delays in diagnosing the abscess or abscesses. None of the patients described had surgery for nonparavalvular abscesses and, to date, no reports have been published that compare differences in outcome between patients treated with antibiotics alone and those treated with antibiotics and surgery. Likewise, no formal recommendations have been made regarding the duration of antibiotic therapy for this type of infection. Nevertheless, the experience with paravalvular abscesses suggests that outcome is improved by early surgical intervention and by administering a minimum of 4 weeks of parenteral antibiotics postoperatively, especially in patients with persistence or extension of infection despite adequate antibiotic therapy, heart block, or congestive heart failure [21, 23, 24]. Therefore, in myocardial abscesses unassociated with valvular endocarditis, early surgical intervention with resection of the affected area (that is, aneurysm or loculation) may improve survival.


Mural Endocarditis
space

Mural endocarditis is inflammation and disruption of the nonvalvular endocardial surface of the cardiac chambers. We found 52 patients with the condition in three autopsy studies and five case reports. Its presentation is remarkably similar to that of infective valvular endocarditis.

Risk Factors

Mural endocarditis typically results from seeding of an abnormal area of endocardium during bacteremia or fungemia or as an extension of infection from underlying myocardial abscesses (Table 1). Reports have also been published of infectious thrombi extending from the pulmonary veins onto left atrial endocardium [25]. Other cardiac abnormalities associated with mural endocarditis include ventricular aneurysms or pseudoaneurysms, mural thrombi, chordal friction lesions (reactive plaques incorporating chordae and mural endocardium) [26], pacemaker lead insertion sites, idiopathic hypertrophic subaortic stenosis, jet lesions from ventriculoseptal defects, and other congenital defects [25, 27, 28].


View this table:
[in this window]
[in a new window]
 
Table 1. Clinical Features of Myocardial Abscesses and Mural Endocarditis

 

Certain groups of patients seem to be predisposed to fungal infection of the mural endocardium. In two series describing 21 patients with fungal mural endocarditis, 70% received prolonged courses of antibiotics and 86% were immunocompromised patients treated with cytotoxic chemotherapy, steroids, or other immunosuppressive agents [25, 28]. Moreover, in these patients, there appeared to be a strong association with the presence of myocardial abscesses, possibly reflecting the duration and severity of infection. For example, myocardial abscesses were found at autopsy in 4 of 6 cases of fungal mural endocarditis reported by Buchbinder and colleagues [25] and in all 14 autopsy cases described by Lang and colleagues [28]. Blood cultures were infrequently positive for fungi (16% and 7% in these series, respectively) despite the fact that all patients had evidence of extracardiac systemic infection.

Microbiological Characteristics

The organisms associated with mural endocarditis include staphylococci, viridans streptococci, Enterococcus spp [26]., Salmonella spp. [29], Klebsiella spp. [27], Bacteroides fragilis [30], Candida spp. [25], and Aspergillus spp. [12, 28]. Endocardial and myocardial involvement with cytomegalovirus has been described in a pediatric patient in whom the viral inclusions were discovered on autopsy [31].

Clinical Presentation

Fever and chills have been reported in virtually all cases [26]. In two reported cases of patients with Salmonella [29] and B. fragilis [30] mural endocarditis, symptoms of nonspecific abdominal pain and diarrhea of at least 5 days duration before admission were present. The gastrointestinal tract was presumed to be the source of infection in both patients. In another report, a 38-year-old woman presented with overwhelming bloodstream infection and hypotension secondary to acute S. aureus endocarditis involving an endocardial friction lesion [26]. Blood cultures, reported in 8 of the 19 patients with bacterial mural endocarditis, were positive in 7 patients [26, 27, 29-31].

Diagnosis

It is difficult to confirm the diagnosis of mural endocarditis. Although evidence of peripheral embolization and splenomegaly may be found on physical examination, routine laboratory studies usually reveal only a mild to moderate leukocytosis and an elevated erythrocyte sedimentation rate. When present, electrocardiographic changes are nonspecific. The size and location of the lesions have hampered echocardiographic visualization.

Clinical Course

Peripheral embolization is the complication most frequently associated with mural endocarditis, although fistulous tracts and cardiac rupture have also been described. The mortality rate is virtually 100% in patients with either bacterial or fungal infection. Not surprisingly, the cause of death in these patients is overwhelming systemic infection and not mural endocarditis per se [25].

Early diagnosis with subsequent surgical intervention when possible may prevent the development of complications such as embolization and myocardial rupture with tamponade in patients with large mural or fungal vegetations or both, infected thrombi, aneurysms, or pseudoaneurysms [28, 29]. Appropriate long-term antibacterial or antifungal therapy should be administered in a manner similar to the management of infective valvular endocarditis.


Myxomas
space

Cardiac myxomas are the most common primary cardiac tumors. Most are solitary lesions located in the left atrium attached to the fossa ovalis, but they may also be found in the right atrium and ventricles. Most occur sporadically; fewer than 7% are familial (autosomal dominant) or part of a myxoma syndrome complex.

The clinical presentation of cardiac myxomas can be quite varied, depending on the location of the tumor. Right-sided myxomas may be associated with pulmonary emboli and may mimic tricuspid valve disease. Left-sided lesions may present with systemic embolization or with signs and symptoms similar to mitral stenosis or regurgitation. More obscure presentations include fever of unknown origin, myocardial infarction secondary to coronary artery embolization, pericarditis, ventricular arrhythmias including sudden death, atrial fibrillation or flutter, right bundle-branch block, and cerebrovascular accidents from systemic embolization [32]. Often, signs and symptoms suggestive of infective endocarditis or vasculitis may be present, including an elevated sedimentation rate, anemia, low-grade fevers, arthralgias, and rash.

Risk Factors

Infected cardiac myxomas have been reported in 13 patients [33-35]. The age at presentation has ranged from 16 to 69 years, with both sexes affected equally. No consistent predisposing factors have been found, although dental work and a minor gynecologic procedure without antibiotic prophylaxis were implicated in one case [36]. Eleven of the 13 patients had left-sided lesions.

Microbiological Characteristics

Although rare, the diagnosis of infected cardiac myxoma should be considered when patients with presumed infective valvular endocarditis continue to have symptoms despite appropriate antibiotic therapy. The organisms isolated in the reported infections include S. aureus, S. mutans, E. faecalis, Veilonella spp., Candida parapsilosis [37], and Histoplasma capsulatum [38].

Clinical Presentation

Illness associated with an infected myxoma typically is insidious. Most patients described in the literature report weeks to months of temperatures greater than 38 °C, fatigue, and malaise, sometimes accompanied by arthralgias. One patient presented with paroxysmal atrial fibrillation [36]. Another patient, with a remote history of embolic stroke, developed fever and a new left cerebellar stroke while receiving anticoagulation [39]. Eleven patients with infected cardiac myxomas had evidence of embolization at admission or during treatment, including pulmonary embolism [39] or cerebrovascular events such as transient ischemic attacks or strokes [35, 36, 38]. Peripheral embolization appears to be more common with cardiac myxomas, especially infected myxomas, than with infective endocarditis. Bough and colleagues [40] reported embolization in 88% of patients with infected myxomas compared with 40% in noninfected myxomas and 33% in infective endocarditis. One particularly dramatic case report described spontaneous aortic saddle embolization that, by echocardiography, left no remaining evidence of the left atrial tumor [36].

Diagnosis

Only five patients have been described with physical signs typically seen in infective endocarditis, such as splinter hemorrhages, Osler nodes, petechiae, or Roth spots [34, 35]. Findings on cardiac examination have included various murmurs, a prominent first heart sound, and an early diastolic sound consistent with a "tumor plop" [41]. Hepatosplenomegaly has been described in four patients, weight loss in three, and glomerulonephritis in one. Laboratory values, when reported, have been notable for elevated leukocyte counts in 70% of cases, elevated erythrocyte sedimentation rates in 66%, positive blood cultures in 85%, and hematocrits ranging from 26% to 36% in all cases. Advances in echocardiography, especially the advent of transesophageal echocardiography, allow the diagnosis of myxoma to be made much earlier.

Clinical Course

Because the infection has been known to persist despite antibiotic therapy [33], definitive treatment usually requires surgical removal of the infected myxoma. In a series reported by Jadimar and colleagues [39], for example, only those patients who had surgical resection survived. Surgical intervention should not be delayed so as to enhance rapid clearance of the source of bacteremia and to prevent catastrophic embolization. Although the optimal duration of antibiotic therapy is unknown, as with other endovascular infections, the use of several weeks of postoperative parenteral antibiotics seems appropriate.


Pacemakers
space

Infectious complications of transvenous pacemakers have declined recently with improvements in surgical technique [42]. However, the risk for infection during the first 3 years after pacemaker insertion is still estimated to be 1% to 6% [42, 43]. We identified 14 case reports, 5 retrospective analyses, 2 prospective studies, and 2 clinical reviews with detailed information about the localization, cause, microbiologic findings, and therapy of infected cardiac pacemakers, involving a total of 206 patients.


Risk Factors
space

Factors that predispose patients to the development of pacemaker infections include chronic underlying conditions such as diabetes, malignancy, skin disorders, malnutrition, and the use of anticoagulants, steroids, or other immunosuppressive agents [42, 44, 45]. Postinsertion hematomas or seromas, recurrent surgical manipulation at the pacemaker site, and poor skin and wound healing also increase the risk for infection [46]. Although the practice is still controversial [43, 47], it is common to use antistaphylococcal antibiotics before insertion and to continue them for 24 hours afterward.

Pacemaker lead infections tend to develop at least 1 month after insertion. Most occur via contiguous extension from an infected generator site or from contamination after the wire has eroded through the skin. However, metastatic seeding of the endovascular leads during transient bacteremia unrelated to the pacing system itself has also been reported [48-51]. The reasons for such a low incidence of pacemaker lead infections is unclear but may involve endothelialization of the pacemaker leads as well as an inert coating that diminishes the likelihood of bacterial adherence [48, 52]. Complications of infected endovascular leads include valvular endocarditis, infected mural thrombi, localized abscesses, and late electrode perforation [48, 53, 54]. One unusual case described seeding of a prosthetic patch via an epicardial pacemaker after a tetralogy of Fallot repair [55].

Microbiological Characteristics

The microbiologic findings of pacemaker infections are similar to those of prosthetic devices in general [42, 44]. Early infections, defined as those developing within 2 to 4 weeks of insertion, are likely to be related to wound contamination at the time of surgery. More than three quarters of early infections are caused by S. aureus [45] (Table 2). In later infections, those occurring at least 1 month after surgery, S. epidermidis is more prevalent [42, 44]. This finding may be explained by the organism's relatively low virulence yet strong propensity for causing foreign body infections [56]. Other bacteria less commonly reported in pacemaker infections include Enterococcus spp., Proteus spp., E. coli, Klebsiella spp., Pseudomonas spp., Peptostreptococcus spp., and micrococcus [42]. There has been one reported case of Mycobacterium avium-intracellulare infection in a young, malnourished woman without evidence of human immunodeficiency virus infection [57]. Although rare, fungal infections involving Candida spp [58]. and Aspergillus spp [53]. may develop in debilitated patients, particularly those receiving a prolonged course of antibiotics.


View this table:
[in this window]
[in a new window]
 
Table 2. Microbiologic Findings and Presentation of Pacemaker Infections

 

Clinical Presentation

Pacemaker infections are generally classified according to the site of infection, clinical presentation, and the presence or absence of associated bacteremia. Anatomically, infections can involve the generator pocket, the subcutaneous pacemaker leads, the endovascular pacemaker leads, or a combination. Localized infections of the generator pocket and subcutaneous tract are most common [42]. In one series of 75 patients described by Lewis and colleagues [44], 76% presented with pocket infections, and erosion of part of the pacing system through the overlying skin occurred in 24%. Bacteremia occurred in 23% of patients in this particular series. In contrast, Morgan and colleagues [59] studied 1235 consecutive patients who had permanent pacemaker placement between 1964 and 1977; the overall incidence of bacteremia was only 1%. Moreover, in the past 15 years, only 80 cases of documented pacemaker-related bacteremia have been reported [44, 45, 51, 59-61].

Diagnosis

Patients with infected generator pockets typically present with local erythema, tenderness, occasional purulent discharge, and wound breakdown [42]. Cultures of fluid or drainage surrounding the generator site can be easily obtained. Infections involving the endovascular portion of the pacemaker leads are manifested by fevers, chills, and a sustained bacteremia without other apparent sources [62]. Echocardiography is of limited value unless associated with tricuspid valve endocarditis.

Clinical Course

The management of pacemaker infections depends on the site of the infection and whether there is an associated bacteremia. With localized infection, treatment with parenteral antibiotics alone is often sufficient to eradicate the offending microorganisms [52, 63, 64]. However, in some cases, clearance of the infection requires removal of the entire system [44, 60]. This may necessitate cardiotomy if entrapment of the pacemaker leads occurs. In cases of pacemaker infection with associated bacteremia, most investigators tend to favor removal of all pacemaker components to eradicate the infection [46, 49, 50, 59, 65]. Nevertheless, there have been occasional reports of patients with infected transvenous pacemakers and associated bacteremia who have been cured with systemic antibiotics alone [54, 66] (Table 3). In a recent review of sustained bacteremia in 26 patients with permanent endocardial pacemakers by Camus and colleagues [51], all patients with staphylococcal bacteremia, whether related to the pacing system or not, required removal of the entire device. However, five of six patients with nonstaphylococcal bacteremia were cured with antibiotics alone. Bluhm and colleagues [49] recommend a combined approach using systemic antibiotics and removal of as much foreign material as possible. Patients with relapse of infection after this initial treatment should be treated with cardiotomy to remove residual parts as well as a 6-week course of antibiotics. Although some investigators prefer antibiotic prophylaxis for patients with pacemakers who are having dental, gastrointestinal, or genitourinary procedures [46], this procedure is not recommended by the American Heart Association [67].


View this table:
[in this window]
[in a new window]
 
Table 3. Clinical Presentation and Outcome in Patients with Infected Transvenous Pacemakers

 


Implantable Defibrillators
space

The automatic implantable cardioverter-defibrillator has been used in the management of life-threatening ventricular arrhythmias refractory to medical therapy for almost 10 years. Defibrillators in current use are extravascular devices consisting of a pulse generator unit implanted in the abdomen, connected by wires to epicardial rate-sensing electrodes and two mesh pericardial defibrillator patches (Figure 1). They are inserted via a median sternotomy, lateral thoracic or subdiaphragmatic approach, sometimes concurrently with other cardiac surgery such as bypass or valve replacement. We reviewed one retrospective and three prospective studies, three clinical reviews, and three case reports in which a total of 84 patients with infectious complications of implantable defibrillators were described.



View larger version (74K):
[in this window]
[in a new window]
 
Figure 1. Schematic drawings of the automatic implantable cardioverter defibrillator. Top. The epicardial electrodes, mesh patches, and pulse generator of the defibrillator in current use are shown. Bottom. In the transvenous defibrillator, the tripolar lead is inserted intravenously and implanted into the superior vena cava, right atrium, and right ventricle.

 

Risk Factors

The current infection rate within 1 year after implantable defibrillator placement ranges from 1% to 6% [68]. Contamination at the time of insertion is thought to be the primary source of infection. This theory is supported by the increased risk for infection in patients having prolonged operative procedures and reoperation or generator replacement [69, 70]. Although skin lesions over the generator pocket can serve as a portal of entry for organisms, hematogenous seeding, a well-documented mechanism of pacemaker infections [51, 71], has also been described. Other factors associated with an increased risk for infections include concomitant catheter-related sepsis, sternal wound infections, and diabetes mellitus [69, 72].

Microbiology

Staphylococci have been isolated in over half of all reported cases of infected implanted defibrillators [69, 73, 74]. Other organisms such as Serratia spp. [73], Candida spp. [75], Torulopsis glabrata [70], Proprionibacteria spp., and Peptostreptococcus spp. [76] have occasionally been reported.

Clinical Presentation

More than 50% of patients showed local signs of inflammation and infection surrounding the generator site within the first 3 months of insertion [69]. Involvement of the electrodes and patches, however, presented more insidiously with nonspecific findings such as low-grade fever, leukocytosis, and an elevated erythrocyte sedimentation rate. Blood cultures, reported in only 17 cases, were positive in 5 [69, 74, 76].

Diagnosis

Radiographic studies have occasionally proved helpful in defining the site and extent of infection. In one report, for example, distortion of the epicardial patches detected on a computed tomographic (CT) scan led to the diagnosis of pericardial infection [70]. In another study, gallium scanning was helpful in guiding an invasive diagnostic procedure that confirmed the presence of a suspected infection [76].

Clinical Course

Treatment of defibrillator-related infections depends on the site involved [77]. As with pacemakers, explantation of the entire device via median sternotomy or left thoracotomy may be required for complete eradication of infection [70, 74]. Management must be individualized.

New investigational transvenous defibrillators that obviate the need for sternotomy or thoracotomy may soon become available for general use. As the design of the implantable defibrillator changes to include an endovascular component Figure 1, the presentation of associated infections may also change.


Infective Endarteritis and Mycotic Aneurysms
space

Endarteritis is defined as inflammation of an arterial wall. The term, mycotic aneurysm, coined by Osler [78], is used to describe any nonsyphilitic infectious aneurysm that develops from endarteritis. In the pre-antibiotic era, over 85% of mycotic aneurysms developed as a consequence of bacterial valvular endocarditis [79] and were associated with septic embolization and invasion of arterial walls or microembolization to the vaso vasorum of the peripheral vasculature or both. With the introduction of antibiotics, however, the cause and location of endarteritis and associated aneurysm formation changed. We have therefore confined our review of this area to the last 45 years and have selected 4 clinical reviews and 22 case reports in which recent trends in predisposing factors, infecting organisms, diagnostic and treatment modalities, and mortality statistics are described.

Risk Factors

The pathogenesis of infective endarteritis and mycotic aneurysms involves several different mechanisms, one or more of which may cause infection in a given patient. Although isolated cases of septic embolization secondary to bacterial endocarditis still occur, most infected aneurysms result from hematogenous seeding of arteries during bacteremia [80, 81]. The affected vessels usually have underlying anatomic abnormalities such as atherosclerotic plaques or preexisting aneurysms. Certain congenital abnormalities, such as patent ductus arteriosus and aortic coarctation, are also predisposed to the development of mycotic aneurysms [82-84], most likely because of intimal damage caused by turbulent blood flow. Depressed host immunity secondary to diabetes, cirrhosis, collagen vascular disease, and corticosteroid therapy are contributing factors in 24% of cases [80]. Other pathogenetic mechanisms include contiguous extension of infection from an adjacent process and inoculation of bacteria at the time of local arterial trauma from illicit intravascular drug injection, gunshot wounds, intra-aortic balloon pumps, arterial catheters, hemodialysis shunts, and vascular surgery [81]. In one review of 220 patients with infected aneurysms, for example, arterial trauma accounted for 29% of cases [80]. Moreover, the risk for infection in patients with indwelling arterial catheters has been shown to be increased by longer durations of cannulation, local inflammation, and placement via surgical cutdown [85].

Infected aneurysms account for approximately 2.6% of all aneurysms [86]. Preferred sites include areas of bifurcation or narrowing. Men outnumber women by a ratio of 3:1, with an average age of 65 years [87, 88]. This age and sex preponderance is not surprising given that 70% of cases involve the aorta, the vessel most commonly and severely damaged by atherosclerosis [86]. In contrast to the pre-antibiotic era, infection of the peripheral and cerebral vasculature now most often occurs in intravenous drug abusers [89].

Microbiological Characteristics

Unlike infective valvular endocarditis in which gram-positive organisms predominate, endarteritis and mycotic aneurysms are associated with a broader range of infecting organisms. For example, Salmonella spp., an uncommon cause of valvular endocarditis, have been isolated in up to 50% of cases with infections of the aorta [87, 90]. It is not known why Salmonella should be such a successful pathogen in aneurysmal infections. However, certain species, especially Salmonella typhimurium and Salmonella choleraesuis, appear to have a particular predilection for seeding abnormal tissues including diseased arterial walls [91]. Staphylococcus aureus, most often associated with extra-abdominal mycotic aneurysms [89, 92], has been reported to account for approximately 30% of cases overall [81]. Other organisms such as E. coli, Pseudomonas spp., Enterobacter spp., Proteus spp., Yersinia spp., Campylobacter fetus, group B streptococci, B. fragilis, Candida spp., and other fungi have also been reported [92-94], implicating the gastrointestinal tract as the primary source of infection.

Clinical Presentation

The clinical manifestations of endarteritis and mycotic aneurysms are influenced by the anatomic site affected. More than 70% of patients present with fevers and leukocytosis [87]. Localized pain, erythema, sinus tract formation, ischemia distal to the affected area, or a combination are characteristic of infected peripheral vasculature. Patients with aortic involvement have a palpable mass in approximately 50% of cases and may report abdominal and back pain [95]. Although these symptoms are not diagnostic, the presence of fever strongly suggests an infected aneurysm. Underlying thoracic or lumbar osteomyelitis, present in up to one third of patients with aortic endarteritis, may account for some of this pain [96]. One possible explanation for the association between vertebral osteomyelitis and aortic aneurysms is occlusion of the lumbar arteries by the enlarging infected aneurysm, altering blood flow to the lumbar vertebrae and thus creating a "locus minoris resistentiae" with increased susceptibility to infection [88]. However, in many cases, it is unclear whether the osteomyelitis precedes the development of the mycotic aneurysm or develops as a result of contiguous spread of infection. Signs of peripheral embolization from infected aneurysms are indistinguishable from those associated with valvular bacterial endocarditis. A new or changing bruit may suggest pseudoaneurysm formation. Catastrophic presentations such as cerebrovascular hemorrhage or exsanguinating aortic hemorrhage can occur with aneurysm rupture.

Diagnosis

Because this type of infection often presents insidiously, the diagnosis of infective endarteritis and mycotic aneurysms can be difficult. With the exception of peripheral vascular involvement, which represents a more superficial and therefore more easily detectable area of infection, most cases of infective endarteritis without aneurysm formation are not diagnosed until autopsy. Pathologically, endarteritis and mycotic aneurysms may exhibit signs of acute or chronic inflammation, necrosis, hemorrhage, or abscess formation, often with an intact intimal lining. Therefore, blood cultures may be negative in up to 50% of cases [87, 90, 92]. When positive, however, the bacteremia is characteristically persistent.

Vascular ultrasonography and CT scanning are helpful in showing the presence of an aneurysm or pseudoaneurysm and perivascular fluid collections [97]. Computed tomography scanning with contrast enhancement has been shown to be more sensitive than arteriography in the early stages of disease [98]. Findings suggestive of a mycotic aneurysm include irregular peripheral enhancement of the arterial wall, unusual luminal shape, loss or disruption of intimal calcification, rapid expansion of the aneurysm, and adjacent vertebral osteomyelitis or fluid collections [97, 98]. Although these studies may not be able to reliably distinguish among sterile fluid, abscesses, and areas of hemorrhage, they may be useful in guiding percutaneous needle aspiration of the affected area. Extraluminal gas in the periaortic region on a CT scan should alert the physician to the possible presence of an aortoenteric fistula [99]. Magnetic resonance imaging may be especially useful when the patient has a contraindication to the use of contrast material and may also obviate the need for more invasive procedures such as angiography, which has traditionally been used to define the exact extent of aneurysmal involvement preoperatively [100].

Imaging studies such as gallium and indium-labeled leukocyte scans can be used to localize inflammatory lesions and complement CT, MRI, and angiographic findings. However, because information comes from individual cases and small series, the sensitivity and specificity of these studies is unknown. Gallium-67 has been shown to accumulate in vessels with focal vasculitis even before structural changes occur [101]. Indium-111 leukocyte imaging has been reported to add specificity for the presence of infection in aneurysms detected by other anatomic imaging methods [102]. In one retrospective analysis of seven patients, five of whom had infected aneurysms, leukocyte scintigraphy was positive in four patients with infected aneurysms, equivocal in one, and negative in the two patients with noninfected vessels [92].

Clinical Course

The diagnosis of infected endarteritis is often considered only after antibacterial treatment has already begun. Continued bacteremia despite appropriate antibiotic therapy in a patient without signs of infective endocarditis should lead to the consideration of infective endarteritis or mycotic aneurysm. In persistent bacteremia, antibiotic therapy alone is insufficient, and early surgical intervention is of utmost importance because the risk for rupture of the infected vascular segment is approximately 75% [92]. The risk for rupture is greater in the first 2 weeks after blood cultures become positive for gram-negative compared with gram-positive organisms. In one study, 1 of 10 aneurysms with gram-positive organisms ruptured compared with 5 of 6 aneurysms with gram-negative organisms [95].

Despite antibiotic therapy and advances in diagnostic and surgical technology, morbidity and mortality remain quite high. Vascular infections of the lower extremities result in amputation in as many as 21% of cases [89]. Aneurysm rupture of the aorta can be rapidly fatal. The mortality rate for patients in whom infected atherosclerotic aneurysms are detected late in their clinical course exceeds 70% [92]. More recently, however, earlier detection and aggressive surgical intervention has led to improved survival, with rates as high as 75% to 100% [81]. Optimally, complete resection with extensive local debridement and long-term antibiotic therapy provides the greatest chance for cure. If possible, bypass procedures or placement of prosthetic material through the involved area should be avoided as they pose a high risk for reinfection. Antibiotic therapy, guided by blood and intraoperative cultures, should be continued for at least 4 weeks after surgery, with careful long-term follow-up [83, 87, 90, 103].


Prosthetic Vascular Graft Infections
space

The diagnosis and management of prosthetic vascular graft infections are similar to those of infective endarteritis and mycotic aneurysms (Table 4). From our review of the medical literature, we found the most detailed descriptions of these infections in six clinical reviews and in five prospective and two retrospective studies.


View this table:
[in this window]
[in a new window]
 
Table 4. Comparison of Infected Aneurysms and Vascular Graft Infections

 

Risk Factors

Of the patients at risk, 75% are men with significant atherosclerotic disease and a mean age of 65 years. The incidence of infection ranges from 0.77% [104] to more than 5% [105], depending on the graft's anatomic location. For example, infections occur in fewer than 1% to 1.5% of aortoiliac grafts compared with 2% to 7% of femoral-popliteal arterial grafts [106]. In one series of more than 2400 vascular graft implantations, graft infections occurred in 62 patients, all of whom had incisions of the groin [104]. Other factors that increase the risk for graft infection include adjacent soft-tissue infection, superficial graft location, and underlying diabetes mellitus, as well as surgical revision and emergency aortic repair.

More than 95% of infections are thought to arise from contamination at the time of graft insertion. Hematogenous seeding of the graft, described anecdotally [107, 108], occurs within the first few months of implantation before endothelialization and development of a fibrin and connective tissue pseudointima. Incomplete formation of the pseudointima has also been proposed as a factor in the development of late infections, occurring several years after graft insertion [109]. Contiguous spread of infection from anastomotic endarteritis, infected lymphatics, and inflamed bowel compose the remainder of vascular graft infections.

Microbiological Characteristics

Staphylococci have been implicated in more than 40% of prosthetic vascular graft infections. Early postoperative infections are usually caused by S. aureus, whereas late infections are more frequently associated with coagulase-negative staphylococci [110]. Gram-negative organisms such as E. coli, Pseudomonas spp., and Proteus spp. have been reported in fewer than 15% of both early and late infections [108]. Only a few reports of fungi and anaerobic bacteria have been published. Fungal infections tend to occur late in the course of the patient's illness and are associated with prolonged use of antibiotics [111] and a delay in diagnosis. Polymicrobial infections can also be seen, usually with vascular graft infections of the groin and abdomen, such as aortoenteric fistulae.

Clinical Presentation

Manifestations of prosthetic vascular graft infections depend on their anatomic location. Graft infections of the groin and extremities usually occur within a few months of insertion and are manifested by signs of localized infection such as overlying erythema, skin breakdown, or purulent drainage [112]. In one study of 128 cases of infected grafts, for example, 80% of groin infections became apparent within 5 weeks of surgery [113]. Abscess formation, graft thrombosis, draining sinus tracts, and false aneurysms with associated bruits are easily recognizable signs of graft infection in these locations [6]. In contrast, intra-abdominal vascular graft infections typically present more insidiously during the first year after insertion. As with mycotic aneurysms occurring in this anatomic site, fever, vague abdominal or back pain, or a combination may be the only symptoms. Complications such as septic emboli, retroperitoneal or gastrointestinal hemorrhage, development of a false aneurysm with a bruit or mass, or fistula formation have been reported [109].

In general, patients with vascular graft infections do not present with sepsis. Reasons for this include the low virulence or small inoculum of the organisms involved and the fact that infection may be limited to the perigraft tissues or the vascular wall. Leukocytosis and positive blood cultures are therefore not routinely present. The most consistent laboratory finding, an elevated erythrocyte sedimentation rate, is unfortunately nonspecific [112].

Diagnosis

Certain medical imaging studies, useful in the diagnosis of endarteritis and mycotic aneurysms, also facilitate the diagnosis of prosthetic vascular graft infections. Ultrasonography, often the first and least invasive study, may show thrombosis, perigraft fluid, or pseudoaneurysm formation. However, a negative examination does not exclude infection [112]. Computed tomography scanning is currently the preferred method for visualization of the graft during the early postoperative period. Although air or fluid collections surrounding the graft may be normal findings during the first 2 to 3 months after graft insertion, one of the most reliable indicators of infection on CT scans beyond this period is the presence of perigraft gas [114]. Computed tomography has been shown to detect infection in the presence or absence of fistula formation between the graft and bowel with 94% sensitivity and 85% specificity [115]. As with mycotic aneurysms, CT scanning or ultrasonography can be used to guide percutaneous needle aspiration of a suspicious collection.

Magnetic resonance imaging is also helpful in evaluating infected grafts and perigraft areas [116]. In two studies evaluating the accuracy of MRI in the diagnosis of prosthetic aortic graft infections in 75 patients, it helped to establish the correct diagnosis of retroperitoneal infections in 86% [117, 118]. In contrast to CT scanning, one limitation of MRI is its inability to diagnose accurately aorto-enteric fistulae, because both gas and atherosclerotic plaques appear black on MRI scans [118].

Indium-111-labeled leukocyte scans can be used to confirm the presence of graft infection and estimate the total area involved. In a study of 32 patients with aortic vascular graft implantations, 23 of whom had suspected graft infections, indium-111-labeled leukocyte scanning diagnosed infection with 83% accuracy [119]. However, although other studies have reported sensitivities of 100%, specificities have ranged between 50% and 87% [120, 121]. False-positive scans, caused by localization of indium-labeled leukocytes to areas of noninfectious inflammation such as hematomas, account for the low specificity. Conversely, however, in patients with initially negative studies in whom clinical suspicion of graft infection remains high, serial scans can be helpful in ultimately confirming or excluding the presence of infection [119]. A new technique that uses indium-labeled human IgG to detect vascular graft infections has also recently been introduced [122]. Although the accuracy of this technique must be assessed, its potential advantages over leukocyte scans include easier preparation, elimination of blood product handling, and improved resolution.

Clinical Course

Once the vascular graft infection is diagnosed, the treatment of choice is removal of the graft with debridement of surrounding tissues. Angiography can be used to define the anatomy preoperatively as well as to visualize the formation of thrombus, rupture, or aneurysm [109]. After surgery, a minimum of 4 weeks of antibiotics should be administered. Documentation of positive operative cultures from the proximal end of the vascular graft predicts an increased incidence of recurrent infection and possible suture rupture [123]. Therefore, a postoperative antibiotic course of at least 6 weeks is recommended. As in the case of endarteritis and mycotic aneurysms, bypass of the infected area rather than replacement of prosthetic material through the affected site is preferred [109]. When surgery is not an option, a prolonged course of parental antibiotics followed by long-term suppressive therapy is recommended. Several publications have reported successful treatment of infected but patent peripheral grafts using local debridement and antibiotics [124, 125].

The outcome of vascular graft infections can be devastating, resulting in loss of limb or even death despite appropriate therapy. Morbidity and mortality seem to correlate most strongly with the site of the vascular graft infection. Although mortality rates are low in patients with graft infections of the extremities, morbidity can be quite high. Several studies have reported amputation rates of 20% to 50% [104, 113]. Because of the serious nature of prosthetic graft infections, prophylactic antibiotic therapy beginning before surgery and continuing for 24 to 36 hours postoperatively is recommended for all vascular graft insertions [109, 126]. The value of prophylaxis, as is traditionally used with patients at high risk for valvular endocarditis before specific genitourinary, dental, or gastrointestinal procedures [67], has not been established at this time. However, experimental animal studies have shown that vascular grafts are most susceptible to bacteremic seeding during the first 4 months after insertion [107]. On the basis of these findings, Karchmer [6] has suggested that the use of antibiotic prophylaxis during this period may be of benefit.


Conclusions
space

Nonvalvular infections of the cardiovascular system are associated with anatomic abnormalities, intravascular devices, prosthetic material, and the immunocompromised state. These infections are often insidious in presentation, posing a special diagnostic challenge for the clinician. Though their prognosis is generally worse than that of infective valvular endocarditis, outcome varies considerably among the different types of nonvalvular infections. The mortality rates associated with myocardial abscesses and mural endocarditis are high, more often reflecting the degree of patient debility and systemic disease than the severity of cardiac infection per se. The prognosis of patients with infected aneurysms, prosthetic vascular grafts, and pacemakers, on the other hand, has improved substantially in the last 10 years because of earlier detection and aggressive surgical management. More studies are needed to establish the most appropriate guidelines for the diagnosis, treatment, and prevention of these infections.


Author and Article Information
space
up arrowTop
dotAuthor & Article Info
down arrowReferences

From Jefferson Medical College of Thomas Jefferson University, and the Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania.
Requests for Reprints: Judith E. Wolf, MD, Division of Infectious Diseases, Jefferson Medical College, Thomas Jefferson University, 1020 Locust Street, #329, Philadelphia, PA 19107.
Acknowledgment: The authors thank Dr. John Miller for help in preparing the illustrations in this manuscript.


References
space
up arrowTop
up arrowAuthor & Article Info
dotReferences

1. Maki D. Infection due to infusion therapy. In: Bennett J, Brachman P, eds. Hospital Infections. Boston: Little, Brown; 1986:561-80.

2. Raad II, Bodey GP. Infectious complications of indwelling vascular catheters. Clin Infect Dis. 1992; 15:197-208.

3. Samson J, Slodki S, Gruhn J. Myocardial abscesses. Am Heart J. 1963; 66:301-8.

4. Ryon DS, Pastor BH, Myerson RM. Abscess of the myocardium. Am J Med Sci. 1966; 251:698-705.

5. Kim HS, Weilbaecher DG, Lie JT, Titus JL. Myocardial abscesses. Am J Clin Pathol. 1978; 70:18-23.

6. Karchmer AW, Bisno AL. Infections of prosthetic heart valves and vascular grafts. In: Bisno AL, Waldvogel FA, eds. Infections Associated with Indwelling Medical Devices. Washington, D.C.: American Society for Microbiology; 1989:129-59.

7. Gopalakrishna KV, Kwon KH, Shah A. Metastatic myocardial abscess due to group F streptococci. Am J Med Sci. 1977; 274:329-32.

8. Case records of the Massachusetts General Hospital. Weekly clinicopathological exercises. Case 27-1970. N Engl J Med. 1970; 28:1477-85.

9. Guneratne F. Gas gangrene (abscess) of heart. N Y State J Med. 1975; 75:1766-9.

10. Sheagren JN.Staphylococcus aureus. The persistent pathogen. N Engl J Med. 1984; 310:1368-73.

11. Kortleve JW, Duren DR, Becker AE. Cardiac aneurysm complicated by Salmonella abscess. A clinicopathologic correlation in two patients. Am J Med. 1980; 68:395-400.

12. Walsh TJ, Hutchins GM.Aspergillus mural endocarditis. Am J Clin Pathol. 1979; 71:640-4.

13. Ihde D, Roberts WC, Marr KC, Brereton HD, McGuire WP, Levine AS, et al. Cardiac candidiasis in cancer patients. Cancer. 1978; 41:2364-71.

14. Parker JC Jr. The potentially lethal problem of cardiac candidosis. Am J Clin Pathol. 1980; 73:356-61.

15. Palank EA, Janardhana ML, Utell M. Fatal acute bacterial myocarditis after dentoalveolar abscess. Am J Cardiol. 1979; 43:1238-41.

16. Winek RR, Schwenk NM, Edwards WD. Fatal hemopericardium due to ruptured solitary myocardial abscess unassociated with infective endocarditis. Am J Cardiovasc Pathol. 1988; 1:255-7.

17. DiNubile MJ, Calderwood SB, Steinhaus DM, Karchmer AW. Cardiac conduction abnormalities complicating native valve active infective endocarditis. Am J Cardiol. 1986; 58:1213-7.

18. Merz W, Roberts G. Detection and recovery of fungi from clinical specimens. In: Balows A, Hausler WJ, eds. Manual of Clinical Microbiology. 5th ed. Washington, D.C.: American Society for Microbiology; 1991:588-600.

19. Crislip M, Edwards J. Candidiasis. Infect Dis Clin of North Am. 1989; 3:103-33.

20. Gray L, Roberts G. Laboratory diagnosis of systemic fungal diseases. Infect Dis Clin North Am. 1988; 2:779-803.

21. Daniel WG, Mugge A, Martin RP, Lindert O, Hausmann D, Nonnast-Daniel B, et al. Improvement in the diagnosis of abscesses associated with endocarditis by transesophageal echocardiography. N Engl J Med. 1991; 324:795-800.

22. Akins EW, Slone RM, Wiechmann BN, Browning M, Martin TD, Mayfield WR. Perivalvular pseudoaneurysm complicating bacterial endocarditis: MR detection in five cases. AJR Am J Radiol. 1991; 156:1155-8.

23. Mildvan D, Goldberg E, Berger M, Altcheck MR, Lukban SB. Diagnosis and successful management of septal myocardial abscess: a complication of bacterial endocarditis. Am J Med Sci. 1977; 274:311-6.

24. Croft CH, Woodward W, Elliott A, Commerford PJ, Barnard CN, Beck W. Analysis of surgical versus medical therapy in active complicated native valve infective endocarditis. Am J Cardiol. 1983; 51:1650-5.

25. Buchbinder NA, Roberts WC. Active infective endocarditis confined to mural endocardium. A study of six necroscopy patients. Arch Pathol. 1972; 93:435-40.

26. Clinicopathologic conference. A thirty-eight year old woman with overwhelming sepsis. Am J Med. 1972; 53:233-41.

27. Herzog CA, Carson P, Michaud L, Asinger R. Two-dimensional echocardiographic imaging of left ventricular mural vegetations. Am Heart J. 1988; 115:684-6.

28. Lang DM, Leisen JC, Elliott JP, Lewis JW Jr, Wendt DJ, Quinn EL. Echocardiographically silent Aspergillus mural endocarditis. West J Med. 1988; 149:334-8.

29. Catherwood E, Mintz GS, Kotler M, Kimbiris D, Lemmon W, Parry WR. Pseudoaneurysm of the left ventricle complicated by Salmonella typhimurium infection. Recognition by two-dimensional echocardiography. Am J Med. 1980; 68:782-5.

30. Esteban A, Wilson WR, Ruiz-Santana S, Fernandez-Segoviano P, Martinez-Cabruja R. Endocarditis caused by B. fragilis. Chest. 1983; 84:104-7.

31. Persaud V. Two unusual cases of mural endocarditis, with a review of the literature. Am J Clin Pathol. 1970; 53:832-8.

32. McAllister H, Fenoglio J. Tumors of the cardiovascular system. In: Atlas of Tumor Pathology. 2d ed. Washington, D.C.: Armed Forces Institute of Pathology; 1978:15.

33. Frandsen TM, Prichard JG, Storz SO.Streptococcus viridans bacteremia associated with atrial myxoma. Am Heart J. 1985; 110:180-1.

34. Graham HV, vonHartitzsch B, Medina JR. Infected atrial myxoma. Am J Cardiol. 1976; 38:658-61.

35. Rajpal RS, Leibsohn JA, Liekweg WG, Gross CM, Olinger GN, Rose HD, et al. Infected left atrial myxoma with bacteremia simulating infective endocarditis. Arch Intern Med. 1979; 139:1176-8.

36. Schweiger MJ, Hafer JG Jr, Brown R, Gianelly RE. Spontaneous cure of infected left atrial myxoma following embolization. Am Heart J. 1980; 99:630-4.

37. Dick H, Mullin E. Myxoma of the heart complicated by bloodstream infection by Staphylococcus aureus and Candida parapsilosis. N Y State J Med. 1956; 56:856-9.

38. Rogers EW, Weyman AE, Noble RJ, Bruins SC. Left atrial myxoma infected with Histoplasma capsulatum. Am J Med. 1978; 64:683-90.

39. Jadimar HA, Webb SW, Adgey AA. Survival of a patient with an infected right atrial myxoma following surgery. Eur Heart J. 1988; 9:110-3.

40. Bough EW, Johnson EE, Zacks SI, Boden WE, Mandel A, Medeiros AA, et al. Echocardiographic diagnosis of an infected myxoma in an atypical location. Am Heart J. 1987; 113:1031-2.

41. Mills P, Craige E. Echophonocardiography. Prog Cardiovasc Dis. 1978; 20:337-58.

42. Beeler BA. Infections of permanent transvenous and epicardial pacemakers in adults. Heart Lung. 1982; 11:152-6.

43. Bluhm G, Nordlander R, Ransjo U. Antibiotic prophylaxis in pacemaker surgery: a prospective double blind trial with systemic administration of antibiotic versus placebo at implantation of cardiac pacemakers. PACE Pacing Clin Electrophysiol. 1986; 9:720-6.

44. Lewis AB, Hayes DL, Holmes DR Jr, Vlietstra RE, Pluth JR, Osborn MJ. Update on infections involving permanent pacemakers. Characterization and management. J Thorac Cardiovasc Surg. 1985; 89:758-63.

45. Bluhm G. Pacemaker infections. A clinical study with special reference to prophylactic use of some isoxazolyl penicillins. Acta Med Scand. 1985; 699:1-62.

46. Eichhorn EJ, Winters WL Jr, Crawford ES, Musher DM, Middleton JW, Hettig RA. Bacterial endocarditis and right atrial vegetation: detection by two-dimensional echocardiography. JAMA. 1981; 246:2724-5.

47. Bluhm G, Jacobson B, Julander I, Levander-Lindgren M, Olin C. Antibiotic prophylaxis in pacemaker surgery—a prospective study. Scand J Thor Cardiovasc Surg. 1984; 18:227-34.

48. Bryan CS, Sutton JP, Saunders DE Jr, Longaker DW, Smith CW. Endocarditis related to transvenous pacemakers. Syndromes and surgical implications. J Thorac Cardiovasc Surg. 1978; 75:758-62.

49. Bluhm G, Julander I, Levander-Lindgren M, Olin C. Septicaemia and endocarditis—uncommon but serious complications in connection with permanent cardiac pacing. Scand J Thorac Cardiovasc Surg. 1982; 16:65-70.

50. Marrie TJ, Nelligan J, Costerton JW. A scanning and transmission electron microscopic study of an infected endocardial pacemaker lead. Circulation. 1982; 66:1339-41.

51. Camus C, Leport C, Raffi F, Michelet C, Cartier F, Vilde JL. Sustained bacteremia in 26 patients with a permanent endocardial pacemaker: assessment of wire removal. Clin Infect Dis. 1993; 17:46-55.

52. Har-Shai Y, Amikam S, Ramon Y, Kahir G, Hirshowitz B. The management of exposed cardiac pacemaker pulse generator and electrode using restricted local surgical interventions; subcapsular relocation and vertical-to-horizontal bow transposition techniques. Br J Plast Surg. 1990; 43:307-11.

53. Kramer L, Rojas-Corona RR, Sheff D, Eisenberg ES. Disseminated aspergillosis and pacemaker endocarditis. PACE Pacing Clin Electrophysiol. 1985; 8:225-9.

54. Zeller NH, Lusk RH, Palmer DL. Nonsurgical cure of endocarditis associated with a pacemaker. Arch Intern Med. 1975; 135:580-1.

55. Yener A, Ayrancioglu K, Aslamaci S, Ekici E, Ikizler C. An unusual late complication following surgical repair of tetralogy of Fallot. J Cardiovasc Surg (Torino). 1990; 31:587-8.

56. Lowy FD, Hammer SM.Staphylococcus epidermidis infections. Ann Intern Med. 1983; 99:834-9.

57. Amin M, Gross J, Andrews C, Furman S. Pacemaker infection with Mycobacterium avium complex. PACE Pacing Clin Electrophysiol. 1991; 14:152-4.

58. Cohen T, Pons V, Schwartz J, Griffin J.Candida albicans pacemaker site infection. PACE Pacing Clin Electrophysiol. 1991; 14:146-8.

59. Morgan G, Ginks W, Siddons H, Leatham A. Septicemia in patients with an endocardial pacemaker. Am J Cardiol. 1979; 44:221-4.

60. Harjula A, Jarvinen A, Virtanen KS. Pacemaker infections—treatment with total or partial pacemaker system removal. Thorac Cardiovasc Surg. 1985; 33:218-20.

61. Smyth EG, Pallister D. Septicaemia in patients with temporary and permanent endocardial pacemakers. J R Soc Med. 1989; 82:396-8.

62. Corman LC, Levison ME. Sustained bacteremia and transvenous cardiac pacemakers. JAMA. 1975; 233:264-6.

63. Furman RW, Hiller AJ, Playforth RH, Bryant LR, Trinkle JK. Infected permanent cardiac pacemaker. Management without removal. Ann Thorac Surg. 1972; 14:54-8.

64. Hurst LN, Evans HB, Windle B, Klein GJ. The salvage of infected cardiac pacemaker pockets using a closed irrigation system. PACE Pacing Clin Electrophysiol. 1986; 9:785-92.

65. Hong-Barco P, O'Toole J, Gerber ML, Domat I, Moquin M, Jackson SC. Endocarditis due to six entrapped pacemaker leads and concomitant recurrent coronary arteriosclerosis. Ann Thorac Surg. 1988; 46:97-9.

66. Tylman T, Seaworth B. Pacemaker-associated sepsis: successful medical management. South Med J. 1985; 78:1140-1.

67. Dajani AS, Bisno AL, Chung KJ, Durack DT, Freed M, Gerber MA, et al. Prevention of bacterial endocarditis. Recommendations of the American Heart Association. JAMA. 1990; 264:2919-22.

68. El-Sherif N, Samet P. The automatic implantable cardioverter defibrillator (AICD): follow-up and complications. In: El-Sherif N, Samet P, eds. Cardiac Pacing and Electrophysiology. 3d ed. Philadelphia: W.B. Saunders; 1991:750-1.

69. Spratt KA, Blumberg EA, Wood CA, Kutalek SP, Reboli AC. Infections of implantable cardioverter defibrillators: approach to management. Clin Infect Dis. 1993; 17:679-85.

70. Goodman LR, Almassi GH, Troup PJ, Gurney JW, Veseth-Rogers J, Chapman PD, et al. Complications of automatic implantable cardioverter defibrillators: radiographic, CT, and echocardiographic evaluation. Radiology. 1989; 170:447-52.

71. Shepard RB. Cardiac pacemaker experience. 1961-1970. Am Surg. 1971; 37:691-9.

72. Borbola J, Denes P, Ezri MD, Hauser RG, Serry C, Goldin MD. The automatic implantable cardioverter-defibrillator. Clinical experience, complications, and follow-up in 25 patients. Arch Intern Med. 1988; 148:70-6.

73. Winkle RA, Mead RH, Ruder MA, Gaudiani VA, Smith NA, Buch WS, et al. Long-term outcome with the automatic implantable cardioverter-defibrillator. J Am Coll Cardiol. 1989; 13:1353-61.

74. Wunderly D, Maloney J, Edel T, McHenry M, McCarthy PM. Infections in implantable cardioverter defibrillator patients. PACE Pacing Clin Electrophysiol. 1990; 13:1360-4.

75. Almassi GH, Olinger GN, Troup PJ, Chapman PD, Goodman LR. Delayed infection of the automatic implantable cardioverter-defibrillator. Current recognition and management. J Thorac Cardiovasc Surg. 1988; 95:908-11.

76. Kelly PA, Wallace S, Tucker B, Hurvitz RJ, Ilvento J, Mirabel GS, et al. Postoperative infection with the automatic implantable cardioverter defibrillator: clinical presentation and use of the gallium scan in diagnosis. PACE Pacing Clin Electrophysiol. 1988; 11:1220-5.

77. Taylor RL, Cohen DJ, Widman LE, Chilton RJ, O'Rourke RA. Infection of an implantable cardioverter defibrillator: management without removal of the device in selected cases. PACE Pacing Clin Electrophysiol. 1990; 13:1352-5.

78. Osler W. The Gustonian lectures on malignant endocarditis. Br Med J. 1885; 1:467-70.

79. Stengel A, Wolferth C. Mycotic (bacterial) aneurysms of intravascular origin. Arch Intern Med. 1923; 123:527-35.

80. Johansen K, Devin J. Mycotic aortic aneurysms. A reappraisal. Arch Surg. 1983; 118:583-8.

81. Gomes MN, Choyke PL, Wallace RB. Infected aortic aneurysms. A changing entity. Ann Surg. 1992; 215:435-42.

82. Gorodezky M, Maroto JM, Mota J, Cardenas M, Contreras R. Mycotic aneurysms of the pulmonary artery. Chest. 1974; 66:214-6.

83. Schneider JA, Rheuban KS, Crosby IK. Rupture of postcoarctation mycotic aneurysms of the aorta. Ann Thorac Surg. 1979; 27:185-90.

84. Jaffe RB, Condon VR. Mycotic aneurysms of the pulmonary artery and aorta. Radiology. 1975; 116:291-8.

85. Grantham RN, Munnell ER, Kanaly PJ. Femoral artery infection complicating intraaortic balloon pumping. Am J Surg. 1983; 146:811-4.

86. Parkhurst G, Decker J. Bacterial aortitis and mycotic aneurysm of the aorta: a report of twelve cases. Am J Pathol. 1955; 31:821-35.

87. Mundth ED, Darling RC, Alvarado RH, Buckley MJ, Linton RR, Austen WG. Surgical management of mycotic aneurysms and the complications of infection in vascular reconstructive surgery. Am J Surg. 1969; 117:460-70.

88. Reichle FA, Tyson RR, Soloff LA, Lautsch EV, Rosemond GP. Salmonellosis and aneurysm of the distal abdominal aorta: case report with a review. Ann Surg. 1970; 171:219-28.

89. Johnson JR, Ledgerwood AM, Lucas CE. Mycotic aneurysm. New concepts in therapy. Arch Surg. 1983; 118:577-82.

90. Davies OG Jr, Thorburn JD, Powell P. Cryptic mycotic abdominal aortic aneurysms: diagnosis and management. Am J Surg. 1978; 136:96-106.[Medline]

91. Finseth F, Abbott WM. One-stage operative therapy for salmonella mycotic abdominal aneurysms. Ann Surg. 1974; 179:8-11.

92. Ben-Haim S, Seabold JE, Hawes DR, Rooholamini SA. Leukocyte scintigraphy in the diagnosis of mycotic aneurysm. J Nucl Med. 1992; 33:1486-93.

93. Plotkin GR, O'Rourke JN Jr. Mycotic aneurysm due to Yersinia enterocolitica. Am J Med Sci. 1981; 281:35-42.

94. Rutherford EJ, Eakins JW, Maxwell JG, Tackett AD. Abdominal aortic aneurysm infected with Campylobacter fetus subspecies fetus. J Vasc Surg. 1989; 10:193-7.

95. Jarrett F, Darling RC, Mundth ED, Austen WG. Experience with infected aneurysms of the abdominal aorta. Arch Surg. 1975; 110:1281-6.

96. Bennett DE, Cherry JK. Bacterial infection of aortic aneurysms. A clinicopathologic study. Am J Surg. 1967; 113:321-6.

97. Vogelzang RL, Sohaey R. Infected aortic aneurysms: CT appearance. J Comput Assist Tomogr. 1988; 12:109-12.

98. Gomes MN, Choyke PL. Infected aortic aneurysms: CT diagnosis. J Cardiovasc Surg (Torino). 1992; 33:684-9.

99. Moriarty JA, Edelman RR, Tumeh SS. CT and MRI of mycotic aneurysms of the abdominal aorta. J Comput Assist Tomogr. 1992; 16:941-3.

100. Ewart JM, Burke ML, Bunt TJ. Spontaneous abdominal aortic infections. Essentials of diagnosis and management. Ann Surg. 1983; 49:37-50.

101. Rivera JN, Blanco G, Perez M, Ruiz W. Gallium-67 localization in a mycotic aneurysm of the thoracic aorta. Clin Nucl Med. 1985; 10:814-6.

102. Seabold JE, Binet EF, Schaefer RF. Mycotic aortic aneurysm diagnosed by In-111 leukocyte scintigraphy and computed tomography. Clin Nucl Med. 1983; 8:486-7.

103. Wilson SE, Gordon HE, Van Wagenen PB.Salmonella arteritis: a precursor of aortic rupture and pseudoaneurysm formation. Arch Surg. 1978; 113:1163-6.

104. Lorentzen JE, Nielson OM, Arendrup H, Kimose HH, Bille S, Andersen J, et al. Vascular graft infection: an analysis of sixty-two graft infections in 2411 consecutively implanted synthetic vascular grafts. Surgery. 1985; 98:81-6.

105. Goldstone J, Moore WS. Infection in vascular prostheses. Clinical manifestations and surgical management. Am J Surg. 1974; 128:225-33.

106. Karchmer A. Approach to the patient with infection in a prosthetic device. In: Gorbach SL, Bartlett JG, Blacklow NR, eds. Infectious Diseases. Philadelphia: W.B. Saunders; 1992:1374-84.

107. Malone JM, Moore WS, Campagna G, Bean B. Bacteremic infectability of vascular grafts: the influence of pseudointimal integrity and duration of graft function. Surgery. 1975; 78:211-6.

108. Bunt TJ. Synthetic vascular graft infections. I. Graft infections. Surgery. 1983; 93:733-46.

109. Freischlag JA, Moore WS. Infection in prosthetic vascular grafts. In: Rutherford RB, ed. Vascular Surgery. v 1. 3d ed. Philadelphia: W.B. Saunders; 1989:510-21.

110. Kaebnick HW, Bandyk DF, Bergamini TW, Towne JB. The microbiology of explanted vascular prostheses. Surgery. 1987; 102:756-62.

111. Bandyk DF. Diagnosis and treatment of biomaterial-associated infections. Infect Dis Clin North Am. 1992; 6:719-29.

112. Golan JF. Vascular graft infection. Infect Dis Clin North Am. 1989; 3:247-58.

113. Liekweg WG Jr, Greenfield LJ. Vascular prosthetic infections: collected experience and results of treatment. Surgery. 1977; 81:335-42.[Medline]

114. O'Hara PJ, Borkowski GP, Hertzer NR, O'Donovan PB, Brigham SL, Bevan EG. Natural history of periprosthetic air on examination of the abdomen following abdominal aortic aneurysm repair. J Vasc Surg. 1984; 1:429-33.

115. Low RN, Wall SD. Aortoenteric fistula and perigraft infection: evaluation with CT. Radiology. 1990; 175:157-62.

116. Justich E, Amparo EG, Hricak H, Higgins CB. Infected aortoiliofemoral grafts: magnetic resonance imaging. Radiology. 1985; 154:133-6.

117. Olofsson PA, Auffermann W, Higgins CB, Robahie GN, Tavares N, Stoney RJ. Diagnosis of prosthetic aortic graft infection by magnetic resonance imaging. J Vasc Surg. 1988; 8:99-105.

118. Auffermann W, Olofsson PA, Rabahie GN, Tavares NJ, Stoney RJ, Higgins CB. Incorporation versus infection of retroperitoneal aortic