Nosocomial endocarditis: still a challenging diagnosis.

Abstract

Infective endocarditis (IE) is a systemic illness whose hallmarks are cardiac disease and bloodstream infection. Although both of these conditions are exceedingly common among medical inpatients, and even more so in critically ill patients, IE is rarely found as the underlying cause of bacteremia with heart failure. Population-based studies [1] estimate that the current rate of community acquired, ‘spontaneous’ IE ranges from 3 to 10 cases every 100,000 person-years. Fewer data are available regarding hospital-acquired IE, even though risk factors for this condition are commonly encountered [2]. The diagnosis of IE has always been a clinical challenge. Historically, IE was first recognized in the sixteenth century at post-mortem examinations, but only in the mid1800s were signs and symptoms of live affected patients described. In 1885, in the Gulstonian lectures on ‘malignant endocarditis’, Sir William Osler presented a comprehensive review of the disease, delineating the evidence that was, and still remains, key to the diagnosis, including persistent bacteremia, heart valve involvement, embolic events and immunologic phenomena. Over subsequent years, it has become clear that there is a need to adopt strict case definitions, in order to improve diagnostic accuracy and to allow comparison of research data. In 1981, Von Reyn et al. introduced the Beth Israel criteria, to differentiate between IE and bacteremia from other foci in patients with heart disease. The concept of ‘definite’ IE was based on microbiologic (Gram’s stain and culture) or pathologic (vegetations or active valvulitis) evidence of the disease at either autopsy or surgically-explanted valves. In the absence of these ‘definite’ IE criteria, cases were categorized as ‘probable’, ‘possible’ or ‘rejected’. The subsequent definitions of IE diagnostic criteria resulted from the evolving epidemiology of the disease as well as the advances of bacteriology and imaging techniques. The advent of echocardiography, and in particular transesophageal echocardiography, paved the way for a revolution in the clinical diagnosis of IE, generating the still viable Duke University criteria, developed in 1994 by Durack et al. Indeed, when performed with adequate equipment and by experienced personnel, TEE is thought to possess a nearly absolute negative predictive value, showing either typical or atypical findings in essentially all IE patients. These criteria were further revised in 2001 to include serologic and molecular diagnosis of difficult-to-grow causative agents. Thus, the Duke criteria were a significant step forward toward a ‘clinically definite’ diagnosis of IE, incorporating the major criteria relating to microbiologic and echocardiographic finding. Still, their application depends on clinical probability, and several issues remain unanswered regarding their validity in special patient subgroups, such as the elderly, implantable electronic device carriers and prosthetic valve recipients, in whom clinical and echocardiographic findings may be more problematic and elusive. For these reasons, the adjunctive diagnostic significance of newer techniques, such as 3D echocardiography, multislice cardiac CT scan and 18FDGpositron emission tomography scan, is currently being evaluated. Thus, the greater the use of increasingly accurate diagnostic modalities, the lesser the role for pathologic criteria. At least theoretically, this would seem to be true. E. Durante-Mangoni (&) R. Utili Cattedra di Medicina Interna, U.O.C. Medicina Infettivologica e dei Trapianti, Ospedale Monaldi, II Universita di Napoli, Via Bianchi, 80131 Naples, Italy e-mail: emanuele.durante@unina2.it