Prosthetic Valve Endocarditis Diagnosis and Management— New Paradigm Shift Narratives Free

Abstract

Prosthetic valve endocarditis (PVE) represents a heterogeneous group of syndromes, clinically and microbiologically, influenced by valve type and location. The general topic of PVE has been recently reviewed in-depth by Palraj et al. [1], and these global clinical and microbiologic features of PVE will not be further addressed in this narrative. Moreover, “variants” of PVE (eg, intracardiac pacing-related or ventricular assist device-related infections) have also been reviewed elsewhere [2, 3].

We will focus on 3 topics pertaining to the diagnosis and management of PVE that often confront multidisciplinary infective endocarditis (IE) “teams” [4, 5]: i) Mycobacterium chimaera PVE associated with contaminated surgical theater equipment;, ii) PVE complicating transcatheter aortic valve replacement (TAVR-IE); and iii) use of positron emission tomography-computerized tomography (PET-CT) to diagnose PVE and its complications. It should be emphasized that these represent recent and often controversial aspects of PVE for which data from definitive, well-designed and controlled randomized trials are not available.

Mycobacterium chimaera PVE Associated With Contaminated Cardiac Surgical Theater Equipment

Mycobacterium. chimaera is a slow-growing, nontuberculous mycobacterial sequevar of M. avium and M. intracellulare within the MAC complex [6]. Prior to 2013, M. chimaera was cited mainly as an environmental colonizer, occasionally reported to cause pulmonary and/or disseminated infection in immunocompromised patients, or those with underlying structural lung disease [7]. In 2013, Achermann et al. in Zurich [8] reported the first 2 well-described, fatal cases of M. chimaera bacteremia, one in a patient with clinical and echocardiographic evidence of PVE, and the other representing multi-organ dissemination of infection, without pathologic evidence of PVE at autopsy. In the PVE patient, there was: i) a 2-year delay between dual aortic–mitral valve prosthetic valve placements in 2008 and the onset of a “fever of unknown origin” syndrome; and ii) a 3-year delay, until the diagnosis of M. chimaera PVE was ultimately made. The resected valve demonstrated what has now become a classical histopathologic “signature” of invasive M. chimaera infections (ie, acute and chronic necrotizing granulomatous involvement, with numerous acid-fast bacilli). Similarly, in the bacteremic patient without PVE, chorioretinitis was observed (a now well-chronicled complication of invasive M. chimaera infections [9, 10]), as well as granulomatous involvement of multiple visceral organs (spleen, liver, and kidneys). The authors suspected a nosocomial source of M. chimaera infection, including direct contamination of the operating field during surgery or the operating room water supply. However, no specific epidemiologic cause was identified.

In 2 studies published in 2015, the Swiss team, plus Dutch and German collaborators, identified the “missing epidemiologic link” underlying such outbreaks of M. chimaera PVE following open heart surgery. Sax et al. [11] studied 6 adult men who presented clinically ~1.5-to->3 years post-open heart surgery. They cultured M. chimaera from both the water circuits of heater-cooler units (HCUs) connected to cardiopulmonary bypass machines, as well as in air samples from the operating theater when these units were in active use. They were then able to match molecular genotypes from these environmental samples to patient isolates from clinical infectious materials. These studies underscored the paradigm that the likely route of transmission was direct contamination of the operative field with M. chimaera aerosols during cardio-pulmonary bypass.

A companion clinical analysis reviewed 10 cases (9 adults and 1 neonate, representing the 6 patients reported above, plus 4 additional cases [12]). These patients featured: placement of a variety of prosthetic valve and graft materials in the aortic or mitral valve positions, or in the aorta itself; and a mean delay of nearly 2 years from the time of initial surgery to presentation of infection. Clinical analyses emphasized: i) nonspecific initial syndromes (fever, shortness of breath, fatigue, and weight loss); ii) splenomegaly; iii) long mean diagnostic delays (~18 months post-open heart surgery); iv) misdiagnoses of “sarcoidosis”; and v) the ultimate diagnosis generally made at cardiac surgery with positive cultures and/or PCR from operative samples (although blood cultures were also positive in 4 patients). Clinical outcomes were suboptimal despite multi-drug anti-M. chimaera therapy (based on in vitro antimicrobial susceptibilities). Five patients died (mainly of uncontrollable infection), 2 patients relapsed, while the remaining 3 patients were in a post-therapy monitoring period at the time of the report.

In 2019, the CDC’s Morbidity and Mortality Weekly Report (MMWR) reported on M. chimaera HCU-related, postcardiac surgical PVE in patients from Los Angeles County between 2012–2016 [13]. They highlighted that the HCUs in question were from a single manufacturer (Sorin Stockert 3T; LiveNova PLC, London, UK). A health alert was released in October 2016 to notify relevant medical care-givers and their Los Angeles County patients who had undergone open heart surgery in hospitals utilizing that particular HCU system (~4000 patients), concerning cardinal clinical symptoms. By May 2017, 20 additional confirmed M. chimaera cases were identified. Fifteen were diagnosed during their hospitalization workups and/or subsequent surgical treatments. Of note, 5 other patients were confirmed following the patient notification letters above, with prospective evaluations by their physicians.

The actual cumulative world-wide number of M. chimaera postcardiac surgical cases to-date is not known, in part, because of the long interval between surgery and clinical presentations (up to 5 years post-surgery) [14]. Given the long delay between epidemiologic exposure and onset of the clinical syndrome, it is likely that such cases will continue to occur for at least the next several years, despite manufacturing efforts to modify HCU units within infection control guidelines.

The diagnostic approach to M. chimaera PVE and related infections involves a multidisciplinary approach of: i) clinical suspicion (patients postcardiac bypass surgery with fever of unknown origin syndromes, chorioretinitis, or granulomatous infections of undefined etiology; ii) mycobacterial blood cultures); iii) PCR identification of blood culture isolates or from any removed cardiac tissues for M. chimaera; and iv) transesophageal echocardiography and PET-CT scanning of prior cardiac surgical sites [15, 16]. In addition, cell-free plasma DNA assays may be useful for the rapid and quantitative detection of M. chimaera in blood samples [17, 18].

Two interesting clinical nuances of M. chimaera-associated PVE have been recently published. Among 7 such patients in Italy, 3 had preceding staphylococcal or enterococcal PVE [19]; the authors postulated that underlying undiagnosed M. chimaera PVE had caused ongoing granulomatous damage to perivalvular tissue, providing an anatomic foundation for these cases of bacterial PVE. Also, investigators from Alberta, Canada detailed 3 fatal cases of progressive M. chimaera granulomatous encephalitis following valve replacement surgeries (4–15 months), associated with profound neuro-cognitive declines [20].

In 2017, the International Society of Cardiovascular Infectious Diseases (ISCVID) organized a global, multi-specialty taskforce to promulgate guidance in this syndrome. In 2020, international guidelines for M. chimaera cardiovascular infections were published [16]. This document provides a comprehensive analysis of all aspects of this infection. It should be consulted and serve as a reference for details not elucidated in the current report, especially related to specific clinical hallmarks, diagnostic workups, antimicrobial therapy, surgical indications, and infection control issues.

PVE Complicating Transcatheter Aortic Valve Replacement (TAVR) and Pulmonic Valve Replacement (TPVR)

Physiologically significant calcific aortic stenosis is a common geriatric problem, as part of age-related degenerative disease. Traditional aortic valve replacement surgery is problematic in this population due to significant comorbid diseases (eg, diabetes, renal disease, and/or pulmonary compromise), making operative mortalities daunting. In 2006, TAVR methodologies were introduced in which a bioprosthetic aortic valve with metal stent frame is introduced transvascularly, and was associated with high placement success rates (>90%), and low postprocedural mortality rates (<10%) [21]. The most common post-TAVR complications involved perivalvular leaks or conduction system abnormalities [22].

In 2015, Amat-Santos et al. [23], described 32 TAVR-IE patients following a systematic review of 28 literature publications between 2000–2013 (mainly case reports or small series). Two-thirds of the patients were men, with a mean overall age of ~80 years and a high mean logistic Euroscore (>30). Most cases had balloon-expandable (58%) vs self-expandable (42%) TAVRs. The transfemoral artery approach was most frequently employed (66% of cases). The main presenting features were fever (80%), heart failure (22%), and/or stroke syndromes (12%). Enterococci were the predominant blood culture pathogen (34%), followed by coagulase-negative staphylococci (12.5%) and S. aureus (6.2%). The enterococcal predominance presumably reflected the preponderance of trans-femoral approaches, implicating perineal colonization as a likely seeding source. The in-hospital mortality of this TAVR-IE cohort was substantial, with ~one-third dying after valve explantation and one-third dying during medical therapy.

Twenty eight pediatric cases of post-TPVR IE were assessed, mainly following Tetrology of Fallot corrective surgery (~50%). The relative rates of IE at 1-year post-procedure were modestly higher in TPVR vs TAVR (2.2 vs 1.3%). In contrast to TAVR-IE cases, S. aureus was the most common cause of post-TPVR IE (~one-third of cases). Of interest, post-TPVR-IE reports have emphasized a predominance of skin flora and oral flora pathogens (potentially related to acne and orthodontic adjustments, respectively, in such target populations) [23, 24].

Amat-Santos et al. [23] also emphasized the reduced sensitivity of echocardiography, including transesophegeal echocardiogram (TEE), in identifying vegetations on valves or periannular structures in TAVR-IE (~50%), ascribed in part to metal stents creating hyper-reflectance and shadowing artifacts [25, 26]. Vegetations were most frequently found on valve leaflets.

The findings above were amplified by Reguiero et al. [27] in a retrospective analysis of a TAVR registry of more than 20 000 patients from 47 countries between 2005–2015. The incidence of TAVR-IE (250 cases) was 1.1% per patient-year, with most occurring within 6 months of TAVR. Three major independent risk factors for TAVR-IE were: male sex, diabetes mellitus, and significant residual aortic valve regurgitation post-TAVR. Approximately one-half of the cases were health care-associated, with a relatively equal frequency in patients with self-expandable vs balloon-expandable TAVRs. As above [23], the transfemoral TAVR route was predominant (> 80% of cases), and enterococci (~25%), S. aureus (~24%) and coagulase-negative staphylococci (~17%) were the 3 most frequent causative pathogens.

Echocardiography (mainly TEE) was more sensitive at identifying vegetations than in the prior publication above [23], with ~two-thirds of patients having positive studies; ~50% demonstrated valve leaflet vegetations vs ~18% having stent-associated vegetations. Moreover, 18% of patients also had periannular extension of TAVR-IE. Of interest, stent frame vegetations were 3-fold more likely to occur in those with self-expandable TAVRs, presumably related to their larger stent frame (vs balloon-expandable TAVRs).

Greater than 80% of these 250 patients with TAVR-IE had 1 or more classical indications for valve explantation surgery; however, such surgeries were only done in 18% of this cohort, while the remainder underwent appropriate medical therapy. This disparity appeared to reflect the high logistic Euroscores in this population (predictive of high surgical mortality), reflecting older age plus significant comorbidities. In-hospital mortality was 36%, and disturbingly, the 2-year mortality was 67%. These TAVR-IE mortality figures are significantly higher than those with TAVR without IE in other large trials [28, 29].

The preponderance of enterococcal and staphylococcal TAVR-IE again raised the spectre of skin (especially perineal) sources of infection in patients with transfemoral TAVR approaches. In the context of β-lactams being the most commonly used peri-procedural prophylaxis agents, the authors urged a re-examination of prophylaxis strategies in TAVR, with perhaps a more targeted approach against the above 3 predominant TAVR-IE pathogens. Of note, Pericas et al. [30] have recommended that glycopeptides plus an aminoglycoside might be a more reasonable approach to peri-TAVR antibiotic prophylaxis, in light of the microbiologic spectrum observed in post-TAVR-IE.

In 2020, Khan et al. [31] published a large systematic review of TAVR-IE. Using detailed literature-review metrics in cases reported up to October 2018, they found only 11/137 articles screened were appropriate for inclusion (including the 2 publications summarized above) [23, 27]. The 11 published articles ultimately employed in this analysis represented ~35 000 TAVR patients. This review emphasized many of the same findings outlined above, including: incidence of TAVR-IE, microbiologic pathogen frequencies, antibiotic prophylaxis issues, and high mortality outcomes (both in-hospital and at longer-term follow-up).

Role of PET-CT in Diagnosis of PVE

Given the relatively high rate of negative TEEs in PVE (especially TAVR-IE), ¹⁸fluorodeoxygluose (FDG)-PET-CT has been validated as a useful adjunctive diagnostic interrogation in a number of such patients [32–36]. This technique relies on overexpression of major glucose uptake transporters (eg, GLUT1) within activated inflammatory cells (eg, polymorphonuclear leukocytes and macrophages) which are prevalent and metabolically-active at infection sites, and accumulate FDG in high concentrations [37–39]. Its potential for enhancing identification of bona fide cases of PVE led The European Society of Cardiology to recommend “routine” FDG-PET-CT in patients highly suspected of PVE [40].

A recently reported large cohort of patients with either PVE or infections of ascending aortic prostheses provided good clarity on this issue [41]. Abegao de Camargo et al. examined 188 such patients (including 151 with PVE, of which 75% had bioprosthetic valves) using PET-CT FDG imaging. FDG uptake profiles were classified as either: “negative or diffuse” vs “focal,” with the latter suggestive of localized valvular/perivalvular infection. In addition, FDG uptake intensity was quantified by standardized maximum uptake values. Further, histopathological features of postsurgical explanted valves were catalogued; the histopathologic parameters of “active inflammation/infection” were then compared head-to-head to PET-CT imaging findings in a subset of available patients. As a parallel control group, PET-CT was performed in a group of 115 patients with NVE.

The sensitivity, specificity, positive and negative predictive values, and accuracy, in 151 patients with PVE, were uniformly >90% (Figure 1). Further, PET-CT outcomes re-classified a large number of suspected PVE patients into the “clinically definite PVE” vs “possible PVE” by modified Duke Criteria [42]. Of note, imaging results were not significantly impacted by the length of antimicrobial therapy prior to interrogations. In sharp contrast, PET-CT did not enhance diagnosis of NVE, as confirmed in other recent series [43].

Figure 1.

Different degrees of focal FDG uptake around the aortic prosthetic valves in IE patients. Axial ¹⁸F-FDG-PET/CT, CT, and PET images of 4 patients (A–D) with aortic prosthetic valve endocarditis and different degrees of focal FDG update around the valve rings. The maximum standardized uptake value was 2.7, 5.1, 7.6, and 9.7, in A, B, C, and D, respectively. Abbreviations: ¹⁸F-FDG-PET/CT, ¹⁸F-fluorodeoxyglucose positron emission tomography/computed tomography; IE, infective endocarditis. Reproduced [permission pending] from: Abegao de Camargo R, Bitencourt MS, Meneghetti JC, et al. The role of 18F-fluorodexoyglucose positron emission tomography/computed tomography in the diagnosis of left-sided endocarditis: native vs prosthetic valves endocarditis. Clin Infect Dis 2020; 70:583–94.

Open in new tabDownload slide

In 20 PVE patients in which histopathological outcome parameters could be compared directly to PET-CT imaging metrics, the following results emerged in patients with positive vs negative PET-CTs: i) more fibrin deposition; ii) greater degree of inflammatory infiltrates, especially polymorphonuclear; iii) less fibrosis; and iv) quantitatively higher degrees of FDG uptake intensities. It was postulated that the enhanced polymorphonuclear influxes in PVE (vs NVE) reflects the presumed distinct pathogenesis of these two infections, in which PVE (e.g., TAVR-IE) likely begins by seeding at the stent site with active biofilm formation and inflammation, rather than being primarily “valvular” and relatively acellular (as in NVE) [44, 45].

PET-CT in suspected cases of TAVR-IE appears to provide important complementary diagnostic information, especially in cases where TEE is technically inadequate to define valvular-stent vegetations and/or periannular complications. Also, some articles have underscored the potential of PET-CT to disclose unsuspected metastatic foci of infection in IE patients, particularly bone-joint emboli [46, 47]. The addition of CT-angiography to FDG-PET scanning has been proposed as a better imaging strategy than standard FDG-PET-CT to diagnose PVE with peri-annular complications (e.g., myocardial abscess and between-chamber fistulae), as well as for concomitant pre-operative coronary artery angiography.

There are several limitations to PET-CT, including the lack of wide-spread availability of this technique, and the rather rigorous dietary preparations for interrogation in order to reduce physiologic myocardial FDG uptake.

The potential for false-positive PET-CT results during the early postoperative period has been raised. This relates to significant functional glucose uptake in the perivalvular tissue following surgical manipulations [48]. Also, the use of pharmacologic “gluing agents” to seal aortic graft anastomoses can elicit substantial FDG uptake [49]. Thus, FDG uptake patterns in the early postoperative period may not be accurate for diagnosing PVE. Current recommendations are to wait at least three months postcardiac surgery before fully relying on FDG-PET-CT techniques as a diagnostic tool for PVE.

Nonstandard Abbreviations

Nonstandard Abbreviations
 
• FDG

  fluorodeoxyglucose

 
• HCV

  heater-cooler unit

 
• IE

  infective endocarditis

 
• ISCVID

  International Society of Cardiovascular Infectious Diseases

 
• PET-CT

  positron emission tomography-computerized tomography

 
• PVE

  prosthetic valve endocarditis

 
• TAVR

  transcatheter aortic valve replacement

 
• TAVR-IE

  transvascular-placed PV

 
• TEE

  transesophageal echocardiogram

 
• TPVR

  transcatheter pulmonic valve replacement.

Notes

Author contributions. Both authors confirmed they have contributed to the intellectual content of this paper and have met the following 4 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; (c) final approval of the published article; and (d) agreement to be accountable for all aspects of the article thus ensuring that questions related to the accuracy or integrity of any part of the article are appropriately investigated and resolved.

Potential conflicts of interest. A. B. reports grants from Lysovant Sciences, Inc, ContraFect Corporation, and Akagera Medicines, outside the submitted work. H.C. reports no potential conflicts. Both authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

References