Author + information
- Norman Mangner, MD∗ (, )
- Felix Woitek, MD,
- Stephan Haussig, MD,
- Florian Schlotter, MD,
- Georg Stachel, MD,
- Robert Höllriegel, MD,
- Johannes Wilde, MD,
- Anna Lindner, MD,
- David Holzhey, MD,
- Sergey Leontyev, MD,
- Friedrich W. Mohr, MD,
- Gerhard Schuler, MD and
- Axel Linke, MD
- ↵∗Department of Internal Medicine/Cardiology, University of Leipzig—Heart Center Leipzig, Struempellstrasse 39, D-04289 Leipzig, Germany
Studies of infective endocarditis (IE) in transcatheter aortic valve replacement patients are limited by univariate analysis of predictors (1), incomplete patient data (particularly non-IE cases ), or low number of endpoints (3). We aimed to investigate the incidence, treatment, predictors, and clinical outcomes of IE including individual data from a large cohort of well-characterized patients following transfemoral aortic valve replacement (TF-AVR) in a single center. Data about the occurrence of IE (modified Duke criteria) were collected retrospectively. All endpoint definitions were subject to the Valve Academic Research Consortium II definitions. The Ethics Committee, University of Leipzig approved the study (registration number: 167-10-12072010).
From February 24, 2006 to September 16, 2014, 55 of 1,820 patients (cumulative incidence 3.02%, incidence rate 1.82% per patient-year) developed IE following TF-AVR after a median follow-up of 366 days (interquartile range [IQR]: 161 to 1,033), with early IE (<12 months) in 74.5% after a median of 35 days (IQR: 13 to 115) post-procedure and late IE diagnosed in the remaining 25.5% after a median of 628 days (IQR: 437 to 1,038). Definitive or possible IE (modified Duke criteria) was diagnosed in 63.6% and 36.4%, respectively.
Patients with versus those without IE were younger (p = 0.012) and had a higher incidence of chronic obstructive pulmonary disease (p < 0.001), peripheral artery disease (p < 0.001), chronic kidney stage ≥3b (p = 0.006), and chronic hemodialysis (p < 0.001) (Table 1).
Implantations were performed under conscious sedation in >95% of cases with perioperative antibiotic prophylaxis, predominately cefazolin, with a self-expendable system used in approximately 75%. Procedural outcomes at 30 days differed between groups, with higher rates in patients developing IE of residual aortic regurgitation ≥grade 2 (p = 0.024), mean pressure gradient (p = 0.012), Valve Academic Research Consortium–defined strokes (p = 0.027; driven mainly by minor strokes, p = 0.003) and renal failure (p = 0.016; particularly renal failure stage 3, p = 0.019) (Table 1). New permanent pacemaker or implantable cardioverter defibrillator implantation (p = 0.197) were not different between groups.
In a multivariate binary logistic regression analysis, chronic hemodialysis (hazard ratio [HR]: 8.37; 95% confidence interval [CI]: 2.54 to 27.63; p < 0.001) and peripheral artery disease (HR: 3.77; 95% CI: 1.88 to 7.58; p < 0.001) remained the only independent predictors for developing IE.
A possible source of bacteremia was identified in 77.3% of cases, with 23 patients classified as nosocomial/health care–acquired IE. Most patients (94.5%) presented with fever >38.0°C, and 21.8% had an embolic event, with heart failure and sepsis frequent in IE patients on admission.
The 2 most common agents were coagulase-positive staphylococci (38.2%) including 3 cases by methicillin-resistant staphylococcus aureus and enterococci (30.9%). Coagulase-negative staphylococci (9.1%) and streptococci (3.6%) completed the list of “typical” microorganisms. Atypical organisms (18.2%) included extended spectrum beta lactamase E. coli, Proteus mirabilis, Propionibacterium acnes, Morganella morganii, Lactococcus garvieae, candida, and extended spectrum beta lactamase Klebsiella pneumoniae. Blood culture negative infective endocarditis was diagnosed in 2 patients (Table 1).
Transesophageal echocardiography was performed in 47 patients, but was not performed in the remaining 8 patients due to hemodynamic instability. There was no typical endocarditic vegetation in 15 patients. In 22 patients, prosthetic valve endocarditis without (n = 12) or with other affected locations (n = 10) was detected. Nine patients (19.1%) with prosthetic valve endocarditis had a paravalvular abscess. Infection of another valve than the prosthesis was diagnosed in 7 patients. Sole lead endocarditis was diagnosed in 3 patients (Table 1).
Antibiotic therapy was administered in 46 patients (83.6%) as the only therapy. Thirty-five patients (64.8%) had at least 1 formal indication for surgery including heart failure (37%), sepsis/septic shock (40.7%), large vegetation (18.9%), structural complications (19.1%), prosthetic valve endocarditis caused by methicillin-resistant staphylococcus aureus (5.5%), and systemic embolism (21.8%). In contrast, only 9 patients (16.4%) underwent surgery.
In-hospital mortality was 63.6% with a median survival 28 days (95% CI: 11 to 45 days) after diagnosis of IE. Cox regression analysis revealed chronic hemodialysis (HR: 8.60; 95% CI: 2.15 to 34.44; p = 0.002), heart failure (HR: 2.77; 95% CI: 1.20 to 6.39; p = 0.017), and sepsis/septic shock (HR: 5.50; 95% CI: 2.18 to 13.85; p < 0.001) as independent predictors of in-hospital mortality. Comparing early versus late endocarditis (HR: 0.83; 95% CI: 0.40 to 1.73; p = 0.615), definite versus possible IE (HR: 0.87; 95% CI: 0.44 to 1.73; p = 0.692), and echocardiographic evidence of IE versus negative imaging (HR: 0.77; 95% CI: 0.32 to 1.88; p = 0.576), no difference was detected between those groups. Six additional patients died 1 year after IE was diagnosed, resulting in a 1-year mortality rate of 74.5%.
This is the largest single-center study reporting independent predictors for the occurrence and outcome of patients developing IE after TF-AVR. The risk for developing IE increased in patients on chronic hemodialysis and with peripheral artery disease. There was a substantial proportion of patients with nosocomial/health care–associated IE, also reflected by the microbiological profile with coagulase-positive staphylococci and enterococci the most frequently isolated microorganisms. This highlights the need for preventive and aseptic techniques during health care procedures in these high-risk patients. Treatment of IE was a conservative approach in most patients, with a minority also undergoing surgery despite a high rate of potential indications for surgery. Prognosis of patients with IE after TF-AVR was poor, with nearly two-thirds dying in the hospital after diagnosis. IE-related complications such as heart failure and sepsis predicted in-hospital death, suggesting an early and aggressive therapy approach (e.g., surgery) may improve outcome. Patients requiring chronic hemodialysis remain the highest risk group for both development of and death by IE after TF-AVR.
Please note: Dr. Holzhev has received support from Symetis and Medtronic. Dr. Leontyev has received support from St. Jude Medical and Medtronic. Dr. Linke has received consulting fees and grant support from Medtronic; has received speaking honoraria from St. Jude Medical, Claret Medical, Boston Scientific, Bard, and Edwards Lifesciences, and Symetis; and owns equity in Claret Medical. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. The authors would like to thank Jennifer Adam, Kathrin Luderer, Carola Döhnert, and Jaqueline Föhlisch for their data management.
- American College of Cardiology Foundation
- Latib A.,
- Naim C.,
- De Bonis M.,
- et al.
- Amat-Santos I.J.,
- Messika-Zeitoun D.,
- Eltchaninoff H.,
- et al.
- Olsen N.T.,
- De Backer O.,
- Thyregod H.G.,
- et al.