Transcatheter aortic valve implantation (TAVI) has become an accepted alternative to surgical aortic valve replacement (SAVR). TAVI has been carefully evaluated in head-to-head comparisons in the PARTNER studies, with the most recent PARTNER 3 study (1,000 patients) showing that TAVI was superior to SAVR in preventing death, stroke and repeat hospitalisation at one-year follow-up1. Similarly, the EVOLUT Low Risk trial (1,403 patients) showed no difference in two-year outcomes (mortality, stroke) between SAVR and TAVI2. More importantly, in both trials, the one-year/two-year event rates were extremely low.
In addition to these “traditional events”, other events include development of valve thrombosis and endocarditis. Recent studies have highlighted an increased TAVI (and, to a lesser extent, SAVR) thrombosis rate, ranging from 10% to 15% when computed tomography (CT) was used3. The definition of thrombosis was either thickened leaflets or reduced mobility of the leaflets. Valve thrombosis on CT, however, appears a subclinical finding, since these patients rarely show an increased gradient over the valve3.
The development of infective endocarditis after SAVR or TAVI is rare, with an estimated incidence varying between 0.1% and 2.3%, not different between SAVR and TAVI. Three larger studies have specifically reported on prosthetic valve endocarditis (PVE); the details of these studies are summarised in Supplementary Table 1,4,5,6,7.
Amat-Santos and colleagues4 set up a large multicentre registry, including 21 centres in North America, South America, and Europe over a time period between 2007 and 2014. The diagnosis of PVE was based on the modified Duke criteria, or the findings during surgery. Out of 7,981 patients undergoing TAVI, 53 developed PVE over a mean follow-up of 1.1±1.2 years, yielding an incidence of 0.67%, with 0.50% in the first year after TAVI. The authors reported that the most frequent causal microorganisms were coagulase-negative staphylococci (24%), Staphylococcus aureus (21%) and enterococci (21%). In 77% of the patients with PVE, vegetations were noted on the transcatheter aortic valve – on the TAVI leaflets (39%), on the stent frame (17%) and on the mitral valve (21%). Complications of the PVE occurred in 87% of patients - heart failure (67.9%), acute kidney injury (54.7%), septic shock (20.8%), stroke (7.5%), systemic embolism (9.4%) and persistent infection (28.3%). Valve intervention was rare (11%), and the in-hospital mortality was 47.2%, with 66% one-year mortality, related to heart failure and septic shock.
Regueiro and colleagues7 reported on the TAVI international registry, which included 20,006 patients undergoing TAVI from 47 centres in Europe, North America, and South America during the period between 2005 and 2015. PVE occurred in 250 patients, with an incidence of 1.1% per person-years. The median time between TAVI implantation and symptoms was 5.3 months. Early PVE was diagnosed in 178 patients (71.2%), with an incidence of 0.9% (and included 72 patients [28.8%] diagnosed within two months of TAVI). Risk factors for developing PVE were younger age, male gender, diabetes and newly diagnosed moderate to severe aortic regurgitation. The authors reported that the most frequent causal microorganisms were enterococci species (24.6%) and Staphylococcus aureus (23.3%). The in-hospital mortality rate was 36%, and surgery was performed in 14.8%. The in-hospital mortality was related to higher logistic EuroSCORE, heart failure and acute kidney injury. The two-year mortality rate was 67%. In this study, the specific findings on echocardiography are highlighted - vegetations in 67%, either anchored to the stent frame or the leaflets of the transcatheter valve. Newly diagnosed mitral or aortic regurgitation was noted in 13.9% and 9.8%, respectively. Periannular complications included abscess, fistula or pseudoaneurysms in 18%.
Butt and co-workers5 used the Danish nationwide registries and included all patients undergoing TAVI (2,632 patients) and isolated SAVR (3,777 patients) during the period between 2008 and 2016, without previous infective endocarditis, and alive at discharge. During a mean follow-up of 3.6 years, 115 patients (4.4%) with TAVI and 186 patients (4.9%) with SAVR were admitted with PVE. The median time from TAVI/SAVR to hospitalisation with PVE was 352 days for TAVI (with a reported 25th to 75th percentile of 133 to 778 days) versus 625 days for SAVR (with a reported 25th to 75th percentile of 209 to 1,385 days). The authors reported an incidence of PVE of 1.6 per 100 person-years for TAVI, and 1.2 per 100 person-years for SAVR. The factors associated with greater risk of PVE in TAVI were male gender and chronic kidney disease, versus male gender and diabetes in SAVR. The cumulative one-year risk of PVE was not different in TAVI (2.3%) versus SAVR (1.8%); the cumulative five-year risks were also similar for TAVI (5.8%) and SAVR (5.1%).
In the current issue of EuroIntervention, Moriyama and co-workers6 report on a comparison between SAVR and TAVI for the development of PVE.
The source of the data is the FinnValve registry, a nationwide registry which includes retrospective data from consecutive and unselected patients who underwent TAVI or SAVR (bioprosthesis only) between January 2008 and October 2017 in any of the five university hospitals in Finland. In total, 6,463 consecutive patients were included, with 4,333 patients undergoing SAVR and 2,130 patients undergoing TAVI. The diagnosis of PVE was based on the modified Duke criteria. The incidence of PVE was 2.9/1,000 person-years in SAVR, and 3.4/1,000 person-years in TAVI, which was not significantly different. In total, 68 patients developed PVE, most frequently related to staphylococci (38.2%) as the causal microorganisms. On multivariate analysis, patients with PVE more often were male, had deep sternal wound infection, or active malignancy. On multivariate analysis, male gender and deep sternal wound infection remained significant. In-hospital death occurred in 19 patients (27.9%). Surgical treatment for PVE was the only independent predictor of in-hospital death. The mortality rate after PVE was 37.7% at one month, and 52.5% at 12 months.
All four of these studies confirm the low prevalence of PVE after TAVI4,5,6,7, which is comparable to SAVR. However, if PVE occurs, the mortality is high. Enterococci and staphylococci are the predominant causal organisms. Treatment with surgery or antibiotics is poor, with high mortality rates secondary to heart failure and acute kidney injury. It remains difficult to diagnose the patients, since the presentation is often aspecific and the echocardiographic findings are diverse. A relatively new diagnostic tool could be the use of fusion imaging with positron emission tomography-computed tomography (PET-CT): this technique permits anatomical assessment with CT (and anatomical co-registration) and particularly the assessment of inflammation with PET and F18-fluorodeoxyglucose8. The inclusion of PET-CT for diagnosis of endocarditis after SAVR is already included in recent European Society of Cardiology guidelines9, and may be expanded to use in patients with suspected PVE after TAVI.
Conflict of interest statement
J. Bax has received speaker fees from Abbott Vascular. V. Delgado has received speaker fees from Abbott Vascular. The Department of Cardiology of the Leiden University Medical Center has received unrestricted research grants from Biotronik, Medtronic, Boston Scientific and Edwards Lifesciences.
