**3. Risk**

cumulative risk of 3% at 5 years and 5% at 10 years [1, 2]. The incidence of PVE in the aortic position is significantly higher than in the mitral position [3]. In comparison, the mitral valve is more commonly affected than the aortic valve in native valve endocarditis [4]. Patients undergoing simultaneous aortic and mitral valve replacement have an even greater risk of

In 1885, Osler observed an association between perioperative bacteremia and endocarditis [7]. In 1935, Okell and Elliott noted that 11% of patients with poor oral hygiene had positive blood cultures for *Streptococcus viridans*, and that 60% of patients had bacteremia associated with dental extraction [8]. Not long after, initial reports of valve replacements by Starr and Harken, the first reports of PVE appeared in the literature [9, 10]. Before the routine use of prophylactic antibiotics, Geraci and Stein reported incidences of early PVE of 10 and 12%, respectively [11, 12]. The use of routine prophylactic antibiotics was noted to reduce the incidence of early PVE to 0.2% [11]. From the outset, the surgical management of PVE has been a formidable challenge. In the 1960s and 1970s, surgery for PVE was associated with an extremely high mortality rate. Discouraged by early operative experience, cardiac surgeons avoided intervention in cases of PVE despite recognition that antibiotic therapy alone was ineffective and often fatal. Surgery for PVE was reserved for high risk cases, and the surgical outcomes were predictably poor. Hence, a vicious cycle developed in which surgery was avoided for fear of poor surgical outcomes, and poor surgical outcomes achieved in high risk cases

In 1972, Ross successfully performed an aortic root replacement for PVE using an aortic homograft [13]. His report stressed surgical principles still true today: complete surgical debridement of all infected tissue, the use of homograft for reconstruction, and the minimal use of foreign material in the infected area [13]. In 1977, Olinger and Maloney reported replacement of an infected aortic prosthesis and external felt buttressing for correction of aortoventricular discontinuity [14]. In 1980, Frantz reported successful repair of an aortoventricular discontinuity from endocarditis and abscess formation by aortic root replacement using a synthetic valved conduit [15]. In 1981, Reitz successfully applied this technique to the treatment of prosthetic aortic valve endocarditis [16]. In 1982, Symbas combined aortic valve replacement with patch repair of a periannular abscess cavity [17]. In 1987, David and Feindel described techniques to reconstruct the mitral annulus with pericardium after debridement for PVE [18]. Surgical treatment of PVE remains a significant challenge, but outcomes improved in the

**1.** widespread use of transthoracic and trans-esophageal echocardiography in making an

**2.** an appreciation that, like surgical infections elsewhere, surgery for PVE requires radical

1990s. Factors that contributed to improved outcomes included:

debridement of infected and devitalized tissue,

PVE than with a single valve replacement [5, 6].

**2. Historical note**

116 Advanced Concepts in Endocarditis

reinforced this fear.

early, accurate diagnosis,

The risk of PVE to the patient is lifelong. However, as assessed by hazard function analysis, the risk of infection is greatest during the first 3 weeks after valve implantation [19]. Most deaths occur within 3 months of PVE development [18]. By clinical convention, PVE is classified as early or late [20]. Early PVE is acquired perioperatively and accounts for approximately one-third of all cases [20, 21]. Although traditionally defined as occurring within 60 days of initial valve replacement, the contemporary literature variably defines early PVE as occurring within 2, 6, or 12 months of initial valve replacement [20–22]. Late PVE results from infection unrelated to the initial valve operation and accounts for the remaining two-thirds [20]. The prognosis for early PVE is significantly worse than that of late PVE and often requires surgical intervention [20].

The distinction between early and late PVE provides insight into the acquisition of infection, expected clinical course, and appropriate management. Early PVE arises from the contamination of the valve during the perioperative period of valve implantation [20]. However, a patient with a prosthetic valve placed more than 12 months prior remains at risk for PVE commonly related to a healthcare-associated infection [20]. In a prospective, multicenter study of 171 patients with prosthetic heart valves by Fang et al., 43% developed endocarditis [23]. At the time, bacteremia was discovered, 33% had prosthetic valve endocarditis. These cases were described as having endocarditis at the outset. In comparison, 11% developed endocarditis at a mean of 45 days after the bacteremia was discovered. These cases were described as having new endocarditis. All cases of new endocarditis were health care associated with 33% developing bacteremia from intravascular devices [23].

Patients with central venous catheters are at particular risk for bacteremia. In the United States, approximately 80,000 central venous catheter-related cases of bacteremia are reported annually [24]. Urinary catheters are another source of bacteremia [25]. Catheter-associated bacteriuria develops at a rate of 8% per day in the first week of catheterization [25]. After the tenth day of catheterization, over 50% of patients are bacteriuric. Bacteremia develops in 0.4– 4% of patients with catheter-associated bacteriuria [25]. Bacteremia per se does not invariably cause PVE. In a 10-year review of 890 patients by Parker et al., 3.6% undergoing cardiac valve replacement developed bacteremia in the early postoperative period. Only 6% of bacteremic patients developed PVE, though uniformly fatal [26]. Other authors have suggested that the risk of PVE may be significantly higher in cases of bacteremia; Murray reported an infective endocarditis rate of up to 25% in cases of *Staphylococcus aureus* bacteremia [27].

Although PVE secondary to candidemia is rare, accounting for 5–10% of all cases, it carries a high mortality rate [28]. In a retrospective study of 44 cases of nosocomial fungemia in patients with prosthetic heart valves by Nasser et al., 9% developed fungal endocarditis at a mean of 232 days after documented candidemia [29]. Hence, patients with candidemia must be treated aggressively in the acute setting and be provided close long-term follow-up.

the study period, from 10 to 68.4% [36]. The prognosis of endocarditis in hemodialysis patients

Prosthetic Valve Endocarditis

119

http://dx.doi.org/10.5772/intechopen.79758

Healthcare-associated infections are a significant source of PVE, accounting for 10–34% of all cases [38]. The majority of cases of healthcare-associated PVE develop more than 72 h following hospital admission [38]. The source of healthcare-associated PVE is frequently an intravascular device, such as a pacemaker or implantable cardioverter defibrillator. PVE is

The incidence of PVE in mechanical and bioprosthetic valves is comparable [39]. Patients with mechanical prostheses have a higher risk of PVE in the first 3 months following valve replacement than those with bioprostheses [19]. The reason for higher risk of PVE in the early postoperative period with mechanical prostheses is unclear. Allografts lack prosthetic material and have a very low incidence of PVE in the early postoperative period. This suggests that mechanical prostheses have a tendency to develop early PVE, attributed to surface contami-

PVE in mechanical and bioprosthetic valves differs in anatomic involvement [40]. Infection of mechanical valves involves the junction between the sewing ring and annulus. This leads to the development of perivalvular abscesses, valve dehiscence, pseudoaneurysms, and fistulas. In comparison, infection of bioprosthetic valves is localized to the leaflets, leading to vegetations, cusp rupture, and perforation [40]. Endocarditis after mitral valve repair is rare. In a study of 30 patients, Gillinov et al. reported only 3% of cases of failed mitral valve repair as being caused by endocarditis [41]. In a study of 1275 mitral valve repairs over a 9-year period, Karavas et al. reported a 0.7% incidence of mitral valve endocarditis requiring surgical intervention [42]. The reason for this low incidence is likely related to less prosthetic material for

Aortic PVE is associated with substantial early morbidity and mortality. Regardless of the type of infected valve, mechanical or bioprosthetic, extensive tissue destruction may complicate aortic PVE. In a 20-year study of surgical treatment of aortic PVE by Perrotta et al., perivalvular abscess was reported in 83% of patients [43]. Comparably, Sabik et al. reported a 78% abscess rate in 103 patients with aortic PVE [44]. Abscess formation may be complicated by pseudoaneurysm and fistulisation [40]. Complete aortoventricular discontinuity has been reported in 40% of patients with aortic PVE [44]. Medical therapy alone has been associated with mortality rates as high as 70%, improved to 4–20% with surgical intervention. Significant risk factors for mortality include older age, higher preoperative creatinine, shorter interval from initial valve operation to reoperation for PVE, and fistula development. Mortality results

is poor, with in-hospital death rates of 25–45% and 1-year death rates of 46–75% [37].

classified as healthcare-associated if it occurs within 1 year of device insertion [38].

**4. Type of prosthesis**

nation at the time of surgery [19].

potential infection with mitral valve repair than replacement.

**4.1. Aortic valve prosthetic valve endocarditis**

from sepsis and multiple organ failure [44].

Implantation of a prosthetic valve in the setting of native valve contamination without known active infection may increase the risk of PVE. For this reason, many surgeons routinely culture excised valve leaflets to ensure that the new valve is not contaminated at the time of implantation. In a study of 222 patients by Campbell et al., 14.4% who underwent elective valve replacement had positive valve cultures [30]. Coagulase-negative *Staphylococcus* was the most common bacterial isolate [30]. None of these patients had clinical evidence of infection. Only 3% of patients with positive valve cultures developed PVE. Most positive native valve cultures were thought to be false positives. Campbell concluded that positive cultures did not predict PVE and recommended against routinely obtaining native valve cultures [30].

Nonetheless, the potential morbidity and mortality of PVE may justify the practice of culturing excised valve tissue and treating patients with positive cultures. Intraoperative contamination at the time of valve implantation may occur from a variety of sources. Cardiac surgical procedures are complex and entail numerous intravascular monitoring devices as well as the circuit of the cardiopulmonary bypass machine. This complexity may contribute to the incidence of positive intraoperative blood cultures. In 1969, Ankeney and Parker reported positive intraoperative blood cultures in 19% of patients undergoing open cardiac surgery [31]. In a 1974 study of 66 patients undergoing open cardiac surgery, Kluge et al. reported a 71% incidence of positive intraoperative cultures from at least one site and a 20% incidence from two or more sites [32]. Several decades later, the issue remains unresolved.

In a 2004 study of 64 patients who underwent cardiovascular surgery, Shindo et al. reported positive intraoperative blood cultures in 16% of patients who underwent cardiopulmonary bypass [33]. Intraoperative blood salvage is routinely used in cardiac surgical procedures to avoid homologous blood transfusion. Autotransfusion is associated with lower risk of hypersensitivity reactions and infections compared to transfusion of homologous blood [33]. However, intraoperative blood salvage is associated with a high incidence of positive cultures. Shindo et al. reported positive blood cultures in 67% of cases using intraoperative blood salvage, excluding cardiopulmonary bypass [33]. In a 1992 study of 31 patients, Bland et al. reported positive cultures in 97% of cases using intraoperative blood salvage [34]. In a 1999 study of 10 patients by Reents et al., 90% of cases using a cell-saving device had bacterial contamination [35].

Hemodialysis has also been associated with endocarditis, particularly with the increasing prevalence of dialysis dependence. In a study of 329 patients with endocarditis by Cabell et al., 20.4% were hemodialysis dependent [36]. Hemodialysis was independently associated with the development of *Staphylococcus aureus* endocarditis. The frequency of hemodialysis dependence also significantly increased during the 7-year study period, from 6.7 to 21% [36]. There was a corresponding significant increase in *Staphylococcus aureus* endocarditis during the study period, from 10 to 68.4% [36]. The prognosis of endocarditis in hemodialysis patients is poor, with in-hospital death rates of 25–45% and 1-year death rates of 46–75% [37].

Healthcare-associated infections are a significant source of PVE, accounting for 10–34% of all cases [38]. The majority of cases of healthcare-associated PVE develop more than 72 h following hospital admission [38]. The source of healthcare-associated PVE is frequently an intravascular device, such as a pacemaker or implantable cardioverter defibrillator. PVE is classified as healthcare-associated if it occurs within 1 year of device insertion [38].
