Oral step-down antibiotic treatment in patients with prosthetic valve endocarditis; a POETry substudy

Abstract Background Oral step-down antibiotic treatment (PO) was included in the 2023 ESC guidelines for treatment of infective endocarditis. In the nationwide POETry study(1), the use PO in patients with left-sided infective endocarditis (IE) was examined in Denmark, showing that almost half of possible candidates with left-sided IE received PO, with no signs of increased treatment failure. Purpose The aim of this study was to examine the prevalence and outcomes in patients with left-sided prosthetic valve IE (PVE) who recieved PO, compared to patients who recieved conventional intravenous antibiotic treatment only (IV). Method Patients with a left-sided PVE, diagnosed between May 2019 and December 2020 in Denmark were included retrospectively. Data was collected by medical chart review. Patients were followed for 6 months. Patients were considered possible candidates for PO if they had left-sided PVE with one of the following bacterial pathogens: Staphylococcus aureus, Enterococcus faecalis, any streptococci or coagulase-negative staphylococci. Patients were excluded if he/she died or were discharged as terminally ill prior to end of antibiotic treatment. The composite outcome consisted of I) embolism, II) non-planned heart valve surgery, III) relapse of bacteremia or IV) all-cause mortality, from end of antibiotic treatment until end of follow-up. Comparisons of primary outcome and all-cause mortality was analyzed with log-rank test. Results In total 158 patients (median age 77 years (IQR: 69-81, 42 female (23%)) with PVE were considered possible candidates for PO, of which 70 (45%) were switched to PO. Patients who received PO more often had diabetes (PO: 27% vs IV: 14%, p=0.03) and were more often non-surgically managed (PO: 86% vs IV: 71%, p =0.02). Median length of stay was reduced by 22 days (PO: 26 days IQR: 17-44 vs IV: 48 days, IQR: 43-61, p<0,001). After 6 months follow-up primary outcome had occurred in 7 (10%) patients who received PO and in 13 (15%) who received IV (p=0.34, Figure, panel A). Prevalence of all-cause mortality in patients who received PO was 4% (n=3) compared to 13% (n=11) for patients who received IV (p=0.07, Figure, panel B). There was no difference in the prevalence of primary outcome according to bacterial pathogen. Conclusion In patients with left-sided PVE who were potential candidates for oral step-down treatment, about half were switched to PO, with a relevant reduction of in-hospital length-of-stay of 22 days. There was no difference in the prevalence of primary outcome or all-cause mortality after 6 months between the two groups. These results support the safety of PO treatment for patients with PVE after stabilized infection.