Summary
The epidemiology of infective endocarditis is changing rapidly due to the emergence of resistant microorganisms, the indiscriminate use of antibiotics, and an increase in the implantation of cardiovascular devices including percutaneous valves. Percutaneous pulmonary valve implantation has achieved standard of care for the management of certain patients with right ventricular outflow tract dysfunction. With its expanding use, several cases of early and delayed infective endocarditis with higher morbidity and mortality rates have been reported. This review summarizes the trends in percutaneous pulmonary valve infective endocarditis, postulates proposed mechanisms, and elaborates on the prevention and management of this unique and potentially fatal complication.
Résumé
L’épidémiologie des endocardites infectieuses change rapidement en raison de l’émergence de micro-organismes résistants, à l’utilisation large des antibiotiques, et à l’augmentation des implants cardiovasculaires. L’implantation de valves percutanées fait désormais partie de la prise en charge conventionnelle des patients avec dysfonction de la voie d’éjection droite. Avec son expansion rapide, plusieurs cas d’endocardites infectieuses précoces et tardives ont été rapportés avec un risque accru de morbi-mortalité. Cette revue résume les connaissances actuelles sur le sujet, propose des mécanismes physiopathologiques et élabore des conduites à tenir pour prévenir et traiter cette complication unique et potentiellement fatale.
Introduction
Percutaneous pulmonary valve implantation (PPVI) (Melody ® valve; Medtronic, Minneapolis, MN, USA) has become established as a valuable treatment option for patients with right ventricular outflow tract (RVOT) dysfunction since its clinical introduction in 2000 . Excellent early, mid-term and even long-term success rates, with improvements in functional status, peak exercise capacity and ventricular function, have been reported since then . Unfortunately, because of factors that are not clearly understood, several cases of early and delayed percutaneous pulmonary valve infective endocarditis (PPV IE) have been reported. IE is a burden in the congenital heart disease (CHD) population, particularly in patients with prosthetic valves in whom diagnosis is more difficult, prognosis is worse and the need for surgery is more frequent compared with CHD patients with native valves. Around half of the patients with CHD-associated IE develop severe episode-related complications and the mortality rate for surgery for IE is very high (40–50%) . In this review article, we summarize the trends in PPV IE, postulate proposed mechanisms, and elaborate on the prevention and management of this unique and potentially fatal complication.
Clinical presentation and diagnosis
IE is a clinical diagnosis and requires a high index of suspicion, especially in the CHD population, as patients may present with non-specific symptoms. The Duke criteria for the diagnosis and management of IE were initially drafted in 1994; they were later modified in 2002 to include echocardiography criteria, and again in 2007 for optimal clinical use. The two major clinical criteria are abnormal blood cultures and evidence of endocardial involvement. The five minor clinical criteria are a predisposition to IE, a fever of > 38 °C, vascular phenomena, immunological phenomena or microbiological evidence of IE not meeting major criteria. Requirements for a clinical diagnosis of IE are: two major clinical criteria; one major and three minor clinical criteria; or five minor clinical criteria . Although these criteria are still used universally for IE, the patient characteristics, timing and presentation of PPV IE are quite complex, and one may have to keep a high index of suspicion in such patients. It is important to note that the clinical presentation with atypical organisms, such as Coxiella burnetii endovascular infection, is usually insidious, lacks the typical features of bacterial endocarditis and often results in delayed diagnosis. Patients are often afebrile and vegetations are usually absent or small .
Definitive IE is diagnosed based on modified Duke criteria, although in patients with prosthetic or Melody valves, true valve involvement may be difficult to determine with a high degree of certainty, because of the acoustic shadowing artefacts from the prosthesis and the unusual anatomy of the RVOT. Transoesophageal echocardiography might be needed. However, the anterior position of the RVOT makes its visualization very difficult and the result is not always inconclusive. An increase in RVOT gradient from postprocedure echocardiogram to hospital admission is present in most patients with IE. The Duke criteria should be modified for this particular substrate (i.e. PPVI) because, in our opinion, any degree of increase of RVOT gradient (unexplained by a structural complication, such as stent fracture) demonstrates the valvular involvement and should be considered as a major criterion similar to new onset of pulmonary regurgitation, unless proved otherwise. Based on these features, blood stream infection with a rise in RVOT gradient may be reclassified as definitive endocarditis. This criterion alone may very well explain the difference noted in the incidence of PPV IE between reported studies. Some other diagnostic tools may be helpful, such as intracardiac echocardiography, three-dimensional echocardiography or positron emission tomography. Positron emission tomography-computed tomography fusion imaging is a good diagnostic tool in case of difficulties in assessing hot spots and slow mouldering cardiac involvement.
Incidence
Several types of presentations and risk factors have been reported. The first series in 2008 by Lurz et al. reported five cases of PPV IE out of 155 patients with an age range of 7–71 years. The possible risk factors reported included dental treatment ( n = 1), septic wound after arm trauma ( n = 1), reactivation of previously treated fungal infection ( n = 1) and previous history of medically treated IE before PPVI ( n = 2) . Subsequently, there have been several anecdotal reports of PPV IE. Our group published a case series on four PPVI-related IE cases involving Streptococcus sanguis ( n = 2), Streptococcus mitis and Staphylococcus epidermidis . Strikingly, three of these had a history of abrupt aspirin discontinuation. Two of the four patients died: one patient (with S. epidermidis ) died on the day of presentation, with heart failure and ventricular fibrillation; and the other (with S. sanguis ) died of multiorgan failure following emergency surgery . Subsequently, Buber et al. reported on 14 subjects who developed bloodstream infection, four of them with definite PPV IE. There were two deaths with methicillin-resistant Staphylococcus aureus bacteraemia without valve involvement. All of these patients had recent onset of obstructive RVOT dysfunction before PPV IE. Male sex, previous IE, high number of stents and altered RVOT anatomy were identified as important risk factors. Only one of the four patients with documented valve involvement underwent surgical valve replacement. A more recent report combining the results of three manufacturer-sponsored prospective North American and European studies showed that 16 of 311 patients were diagnosed with definite or presumed IE, including six who met their criteria for PPV IE. One patient with valve-related IE died after bioprosthesis explantation and multiple complications. Another patient died with severe haemoptysis, 3 weeks after a second PPVI for progressive pulmonary stenosis. Only four patients required Melody valve replacement and two had a second PPVI. Two subjects presented with recurrent IE, 1.6 and 2.7 years after the initial diagnosis. One of the two patients had a second PPVI for stent fracture-related pulmonary stenosis after antibiotic treatment for the initial IE episode, and the other had vegetation at initial diagnosis but received non-surgical treatment . Cheung et al. reported six cases of PPV IE, and demonstrated the unique utility of intracardiac echocardiography for identifying small and subtle vegetations. The most recent report by Villafane et al. described four paediatric patients with PPV IE: two had documented vegetations; three showed worsening pulmonary stenosis at the time of presentation; three underwent surgical valve replacement with no deaths.
It is interesting to note that there is no specific microbiology pattern of Melody valve involvement. Most organisms, from cutaneous commensals to pathogens commonly encountered in IE, have been reported. The annualized rate of first episode of valve-related IE is 0.88% per patient-year. None of the patients seen in our institution or reported in the literature showed any septic emboli to the pulmonary circulation. Table 1 shows the various studies and reports in the literature with Melody valve IE.
| Reference | Year of publication | PPV IE cases/total cases | Patient age (years) | PPVI to IE interval (months) | Pathogen | Proposed major risk factors/trigger | Direct valve involvement | Mode of treatment | Outcomes of IE |
|---|---|---|---|---|---|---|---|---|---|
| Lurz et al. | 2008 | 5/155 | 21 (7–71) | 1.9–23.2 | Staph aureus; Strep aurelius ; Candida albicans | Unprotected dental treatment ( n = 1); reactivation of previously-treated fungal infection ( n = 1); septic wound after arm trauma ( n = 1); prior IE ( n = 2) | ND | Surgery ( n = 3); medical ( n = 2) | No deaths |
| Eicken et al. | 2011 | 1/102 | 21.5 (16–30) | 6 | Staph aureus | ND | 1 | Surgery ( n = 1) | No deaths |
| Atamanyuk et al. | 2011 | 1 | 15 | 65 | Bartonella henselae | Paravalvular leak | 1 | Surgery ( n = 1) | No deaths |
| Roberts et al. | 2011 | 1/15 a | 9 | 2 | ND | ND | 1 | Surgery | No deaths |
| Patel et al. | 2012 | 4 | (11–26) | 3–28 | Strep sanguis ( n = 2); Strep mitis ; Staph epidermidis | Abrupt discontinuation of aspirin; inherent to bovine tissue; residual gradient | 4 | Surgery ( n = 3, including one with emergency RVOT stent) | Two deaths |
| Alsoufi et al. | 2012 | 1 | 16 | 2.5 | Aspergillus | Thrombus; stenosis | 1 | Surgery ( n = 1) | No deaths |
| Bhat et al. | 2012 | 1 | 19 | 4 | Staph aureus | ND | 1 | Surgery ( n = 1) | No deaths |
| Gillespie et al. | 2012 | 3 (confirmed in 2)/104 | 26 (3–63) | 13, 18 | ND | 2 | Surgery ( n = 2) | No deaths | |
| Butera et al. | 2013 | 2/63 | 24 (11–65) | 2–3 | Staph aureus | Disregarded IE prophylaxis | 2 | Surgery ( n = 2) | No deaths |
| Buber et al. | 2013 | 14 BSI (4 PPV IE)/147 | 19 (3–59) | 1–56 | Strep (50%); Staph (43%); Haemophilus (7%) | High risk of systemic bacterial infections with male sex; previous endocarditis events; high number of stents at the RVOT; altered RVOT anatomy | 4 | Surgery ( n = 1); medical ( n = 3) | No deaths in confirmed IE |
| McElhinney et al. | 2013 | 16 (6 PPV IE)/311 | 21 (11–41) | 1.7–56 | Strep ; Staph ; Moraxella catarrhalis | Prior IE; invasive procedure; bleeding oral ulcer; IVDU; high pre- and postimplantation gradients | 4 | Surgery ( n = 4) | Two deaths |
| Cheung et al. | 2013 | 6/43 | 25 (6–67) | Strep ( n = 2); Staph aureus ( n = 2); Staph epidermidis ; Staph gordonii | ND | 2 | Surgery ( n = 2) | One death | |
| Villafane et al. | 2014 | 4/143 | 11–20 | 1.7–16 | Strep viridans ; Strep sanguis ; Staph aureus ; Corynebacterium | Tattoo | 3 | Surgery ( n = 3) | No deaths |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
