Outcomes With Percutaneous Debulking of Tricuspid Valve Endocarditis

Abstract

Background: In patients with tricuspid valve infective endocarditis, percutaneous debulking is a treatment option. However, the outcomes of this approach are less well known.

Methods: We performed a retrospective analysis of all patients who underwent percutaneous vegetation debulking for tricuspid valve infective endocarditis from August 2020 to November 2022 at a large academic tertiary care public hospital. The primary efficacy outcome was procedural success defined by clearance of blood cultures. The primary safety outcome was any procedural complication. For the composite outcome of in-hospital mortality or heart block, outcomes were compared (sequential noninferiority and superiority) with published surgical outcomes data.

Results: Of the 29 patients with tricuspid valve infective endocarditis who underwent percutaneous debulking, the average age was 41.3±10.1 years, all patients had septic pulmonary emboli with 27 (93.1%) patients having cavitary lung lesions before the procedure. For the efficacy outcomes, 28 patients (96.6%) had clearance of cultures after their procedure, mean white blood cell count significantly decreased from 16.8±1.4×10³ to 12.6±1.0×10³ per µL (P<0.01), and mean body temperature significantly decreased from 99.8F ±0.30 to 98.3F ±0.20 (P<0.001) post-procedure. For safety outcomes, there were no procedural complications (0%). Two patients (6.9%) died during the follow-up period, both during the index hospitalization due to severe necrotizing pneumonia. When compared with published data on surgical outcomes, percutaneous debulking was noninferior and superior for the composite of in-hospital death or heart block (noninferiority, P<0.001; superiority, P=0.016).

Conclusions: Percutaneous debulking is feasible, effective, and safe in treating patients with tricuspid valve infective endocarditis refractory to medical therapy.

Keywords: body temperature; endocarditis; follow-up studies; hospital mortality; lung.