Infective endocarditis in patients after percutaneous pulmonary valve implantation with the stent-mounted bovine jugular vein valve: Clinical experience and evaluation of the modified Duke criteria.

Percutaneous pulmonary valve implantation (PPVI) has proven good hemodynamic results. As infective endocarditis (IE) remains a potential complication with limited available clinical data, we reviewed our patient records to improve future strategies of IE prevention, diagnosis and treatment. Medical records of all patients diagnosed with Melody® valve IE according to the modified Duke criteria were retrospectively analyzed in three Belgian tertiary centers. 23 IE episodes in 22 out of 240 patients were identified (incidence 2.4% / patient year) with a clear male predominance (86%). Median age at IE was 17.9 years (range 8.2-45.9 years) and median time from PPVI to IE was 2.4 years (range 0.7-8 years). Streptococcal species caused 10 infections (43%), followed by Staphylococcus aureus (n = 5, 22%). In 13/23 IE episodes a possible entry-point was identified (57%). IE was classified as definite in 15 (65%) and as possible in 8 (35%) cases due to limitations of imaging. Echocardiography visualized vegetations in only 10 patients. PET-CT showed positive FDG signals in 5/7 patients (71%) and intracardiac echocardiography a vegetation in 1/1 patient (100%). Eleven cases (48%) had a hemodynamically relevant pulmonary stenosis at IE presentation. Nine early and 6 late percutaneous or surgical re-interventions were performed. No IE related deaths occurred. IE after Melody® valve PPVI is associated with a relevant need of re-interventions. Communication to patients and physicians about risk factors is essential in prevention. The modified Duke criteria underperformed in diagnosing definite IE, but inclusion of new imaging modalities might improve diagnostic performance.

Copyright © 2020 Elsevier B.V. All rights reserved.

Conflict of interest statement Marc Gewillig is proctor for Numed, Medtronic and Edwards. All other authors and study collaborators have no conflict of interest to declare.