Non-bacterial thrombotic endocarditis: manifestations and diagnosis in the age of echocardiography

Abstract

Sterile vegetations of the cardiac valves were first identified in 1888 by Zeigler, with case reports described eponymously in 1924 by Libman and Sacks. 2 These and subsequent early accounts identified an association between noninfectious endocarditis and a variety of diseases, specifically malignancy and rheumatological conditions such as systemic lupus erythematosus (SLE) and antiphospholipid antibody syndrome (AAS). Autopsy cases predominated the early literature; however, after the advent of echocardiography, many case series relied on noninvasive diagnosis (table 1). 4 These studies found a strong association with cancers, with a higher prevalence and greater mortality in patients with metastatic disease. Up to 31 per cent of cases are in fact culturenegative endocarditis rather than nonbacterial thrombotic endocarditis (NBTE). Therefore, a diagnosis of NBTE requires obtaining extended blood cultures to ensure infection is fully excluded. Currently, the diagnosis remains challenging as there are no pathognomonic echocardiographic or clinical features, and because the disease is associated with a variety of concomitant disorders. In Heart, QuinteroMartinez and colleagues provide an update on the epidemiology and conditions associated with NBTE. This singlecentre retrospective analysis included 48 patients defined by specialist consensus as having NBTE following an extensive echocardiographic and microbiological investigation. The study population was predominantly female (75%) with a median age of 60 years. Transoesophageal echocardiography (TOE) was the preferred initial imaging technique and was completed in 91.7% of the study population, with 54.2% of patients also undergoing transthoracic echocardiographic (TTE) imaging. Associated clinical conditions including connective tissue diseases (37.5%) and malignancies (52.1%) were defined. The authors observed many patients had echocardiographic evidence of a vegetation (85.4%) and/or valve thickening (89.6%) resulting in moderate to severe regurgitation in 54.2%. The mitral valve was most frequently affected (mitral vegetation in 54.2% and thickening in 70.8%), followed by the aortic valve (aortic vegetation in 41.7% and thickening in 56.3%). Twenty seven per cent of patients had evidence of multivalvular involvement. Based on their observation that TOE had superior diagnostic sensitivity when compared with TTE, the authors propose a diagnostic pathway which emphasises the role of TOE. The authors found that 79% of subjects suffered an embolic event (ischaemic stroke, peripheral ischaemic event, transient ischaemic attack or any combination thereof). This is in line with, if not increased, from similar postmortem reports. Anticoagulation was initiated in 91.7% of the study population, most commonly with low molecular weight heparin (45.8%) or warfarin (39.6%). Overall mortality was 33% at 1 year, which is unsurprising considering the strong association of NBTE with malignancy, particularly lung cancers and metastatic disease. Only four patients were treated with intervention including one mitral valve replacement, one aortic valve replacement, one surgical excision of vegetations and left atrial appendage closure, and one transcatheterbased procedure to excision mitral vegetations. QuinteroMartinez et al provide a detailed description of the TTE and TOE findings observed in this population emphasising the high prevalence of valve thickening, vegetations and regurgitation. Compared with prior reports, the sensitivity of TTE was adequate, but was out performed by TOE which demonstrated superior sensitivity particularly for the small (<5 mm) vegetations often found in NBTE. Unfortunately, the study does not address a direct comparison between TOE and TTE. The findings of the study provide an excellent update on the demographics, diagnosis and outcomes associated with this rare and poorly described disease. More work is needed to define the pathophysiological mechanism by which rheumatological diseases or malignancies trigger formation of sterile valvular plaques. Additionally, studies to address the role of anticoagulation and modern rheumatological and chemotherapeutic agents are lacking. The potential for superior treatment and survival with advances in cancer and rheumatological therapies underscores the importance of filling this knowledge gap. However, this study is an important step forward