The study site, Instituto Nacional de Cardiologia, is a public tertiary care medical centre offering high-complexity cardiology care and cardiac surgery, including heart transplantation and, more recently, lung transplantation, in the city of Rio de Janeiro, Brazil. The institute has 165 beds, 45 of which are adult intensive care beds, and it is a teaching hospital, training cardiology and cardiac surgery residents. Annually, it performs approximately 1200 surgeries and conducts 50,000 medical outpatient consultations. The Department of Heart Valve Diseases provided an average of 5989 outpatient consultations annually between 2019 and 2023, with an average of 207 valve surgeries per year during this time span.

This was a prospective observational cohort study of consecutive adult patients with IE hospitalized at institution. The study included only adult patients meeting the definite criteria for IE according to the modified Duke criteria, [6] spanning from 1 January 2006 to 30 June 2023. Study variables were collected using standardized International Collaboration on Endocarditis case report forms as previously described [2] Data analysis was conducted post hoc.

Data included patient demographics, epidemiological and clinical data, underlying cardiac conditions, comorbidities, and predisposing factors for IE. Clinical manifestations, complications, causative microorganisms, echocardiographic findings, surgeries, and outcomes were assessed. Laboratory parameters such as erythrocyte sedimentation rate, C-Reactive Protein (CRP) levels, rheumatoid factor, and urinalysis results were obtained.

A literature search was conducted using the virtual library databases PubMed and Virtual Health Library (Biblioteca Virtual em Saúde, BVS). The following descriptors were used: ‘endocarditis’, ‘epidemiology’, ‘cohort studies’, and their respective terms in Portuguese and Spanish. The inclusion criteria were epidemiological studies from the last 10-years (2014–2024), with a minimum sample size of 100 participants, most of whom fulfilled definite IE according to the modified Duke criteria. The exclusion criteria were multicentre studies, studies published more than 10-years ago, and studies with a sample size of less than 100 participants. These data were used to build a Supplementary Table used in the discussion. However, large multicenter series were also read and used in the discussion.

The definitions of IE were based on the modified Duke criteria, as well as guidelines from the European Society of Cardiology (ESC) in 2015 and 2023 for the management of IE [1,5,6]

Community-acquired and healthcare-associated IE were defined according to the ESC 2009 guidelines [7]. Individuals with healthcare-associated IE were categorized as those manifesting IE 48 h after hospital admission or those who acquired IE through an invasive procedure performed within the past 8-weeks before symptom development [1,6] Early PVE was defined as that occurring within 1-year of heart valve surgery. Late PVE was defined as occurring 12 or more months after valve implantation. Pacemaker (PM)-and Implantable Cardioverter Defibrillator (ICD)-related IE were considered healthcare-associated if they occurred within 1-year of device insertion [1,5,6]

Minor clinical criteria included fever >38.0°C, predisposition to IE, and vascular or immunologic phenomena. Major criteria included typical microorganisms isolated from two separate blood cultures and evidence of endocardial involvement on echocardiography [1,5] The diagnosis of definite IE requires two major criteria, or one major and three minor criteria. St. Thomas’ minor criteria were also employed to improve diagnostic yield [8]

IE was classified as acute, if it evolved with sudden onset and rapid progression, presenting within 1-month; subacute, with symptoms evolving between 1- and 6-months after onset; and chronic, with symptoms persisting for more than 6-months [1,5,6]

Embolic events were diagnosed clinically and/or radiologically. All patients with definite IE had systematic imaging techniques performed at the study site, usually contrast-enhanced computed tomography scans, for evaluation of emboli to the central nervous system, abdomen (in left-sided endocarditis), and thorax (in device-related and right-sided endocarditis) [1,5,6]

Comorbidities were diagnosed based on medical records. Acute kidney injury was defined as a creatinine clearance rate <60 mL/min/1.73 m2 if not previously present[1,5,6]

Surgical indications were based on the ESC 2005 guidelines (and its subsequent versions) for the management of IE, including heart failure, uncontrolled infection, vegetation >10 mm, or evidence of embolic events [1]

Prior antibiotic use referred to the administration of antibiotics at any time before IE diagnosis [1,5,6]

Transthoracic Echocardiography (TTE) was routinely performed in all patients. Transesophageal Echocardiography (TEE) was performed to detect cases with negative TTE results, to complement TTE data, and in those with prosthetic or device-related IE. Both TTE and TEE were performed on site. The variables included in the standardized case report form were the presence of vegetation, regurgitation, abscess, leaflet perforation, pseudoaneurysm, fistulae, left ventricular ejection fraction and systolic pulmonary artery pressure.

Peripheral venous blood cultures were collected using sterile techniques, with a minimum of two sets, as per institutional protocol, and incubated for 5-days in the BACTEC 9240 system (BACTEC/ALERT®, BioMérieux, Durham, NC, USA). Bacterial identification and antimicrobial susceptibility testing were performed using an automated VITEK 2 system (BioMérieux). The minimum inhibitory concentrations of vancomycin and daptomycin were determined using the E-test for methicillin-resistant Staphylococcus aureus.

A subset of patients had surgically excised valves studied by 16SrRNA sequencing [9] Serological and molecular analyses of Coxiella burnetii and Bartonella spp. were performed at an external reference laboratory in cases of culture-negative endocarditis with strong epidemiological suspicion and no reported history of prior antimicrobial use [9]

Qualitative data were expressed as absolute and relative frequencies. Quantitative data were presented as frequencies, mean ± standard deviation, median, and interquartile range. Statistical analyses were conducted using Jamovi®, version 1.2.2. Statistical significance was set at p<0.05. Categorical data were described as frequencies and percentages. The normality of the numerical variables was assessed using the Kolmogorov-Smirnov test. Proportions were compared using the Chi-Squared test, or the Fisher's exact test, where appropriate. The Student's t-test was used to compare means.

This study was approved by the institution's Ethics Committee. Informed consent was obtained from each patient or their legal representative, and the study protocol adhered to the ethical guidelines of the Helsinki Declaration of 1975 and its modifications.

Between 1 January 2006 and 30 June 2023, our database identified 502 episodes of definite IE in 481 patients (19 patients had two episodes and one patient had three episodes during the study period).

The mean age of the patients in the cohort was 48±17.2 years, with a ratio of approximately two males for each female (327/502, 65.1 % vs. 175/502, 34.9 %). Individuals over 60-years of age accounted for 149/502 (29.7 %) of the sample. A total of 260/501 (51.9 %) patients were referred from other centers.

Native and prosthetic valve IE occurred in 68.5% and 31.5% of cases, respectively. Early prosthetic valve IE occurred in 59/502 ( 11.8%) cases and late prosthetic valve IE in 99/502 ( 19.7%) cases. Community-acquired IE occurred in 324 ( 64.7%) patients, healthcare-associated nosocomial IE in 128 ( 25.5%), and healthcare-associated non-nosocomial IE in 49 (9.8%).

Half of the patients (50.3 %) had systemic arterial hypertension, 196 (39.0 %) had Congestive Heart Failure (CHF), 107 (21.4 %) had Chronic Kidney Disease (CKD), and 40.6 % had undergone previous cardiac surgery, mostly valve replacement.

The main predispositions for IE were valve prostheses (158/502; 31.5%), RVD in 151/492 ( 30.7%), congenital heart disease in 70/502 ( 13.9%), and previous IE in 63 ( 12.6%) patients (Table 1).

Most endocarditis cases presented acutely (52 %), followed by subacute (41.3 %) and chronic progression (32/496 cases, 6.4 %).

The most common clinical-laboratory manifestation of IE was fever in 454/501(9 0.6%), followed by elevated C-reactive protein (CRP) in 329/455 (7 2.3%) and presence of new regurgitant murmurs (5 0.7%). Approximately one-third of the patients had an elevated erythrocyte sedimentation rate 104/328 (3 1.7%), nearly half had an embolic vascular event (45%), and splenomegaly (1 9.2%) was also a frequent clinical manifestation. Osler's nodes, Janeway lesions, and conjunctival hemorrhage were observed in less tha n 5% of the cases. The clinical manifestations are shown in Fig. 1.

Fig. 1.

Clinical-laboratory manifestations of IE in adult patients with definite IE, 2006‒2023.

(0.09MB).

Blood cultures were drawn from 495/502 (98.6 %) patients. They were positive in 339/495 (68.5 %) patients. Serology was performed in less than 5 % of the cases (23/502; 4.6 %). Culture-negative IE cases accounted for 33.1 % of all blood cultures collected. Among positive blood cultures, oral streptococci and enterococci were the most frequent, followed by S. aureus and coagulase-negative staphylococci. Typical microorganisms such as streptococci of the bovis group and the HACEK group (Haemophilus species, Aggregatibacter species, Cardiobacterium hominis, Eikenella corrodens and Kingella species) were infrequent and found in less than 5 % of the cases. Non-HACEK gram-negative bacteria were more frequent than the HACEK group. The etiology of IE is shown in Fig. 2.

Fig. 2.

Causative agents in 502 episodes of definite IE in adult patients, 2006‒2023.

(0.08MB).

Bovis, Bovis group streptococci; viridans group, viridans group streptococci; CNS, Coagulase-Negative Staphylococci; non-HACEK GNB, non-HACEK Gram-Negative bacteria; HACEK, Haemophilus species, Aggregatibacter species, Cardiobacterium hominis, Eikenella corrodens and Kingella species; Other GP: Abiotrophia defectiva (2); Corynebacterium diphtheriae (2); Corynebacterium amycolatum (1); Cutibacterium acnes (1); Gemella morbillorum (3); Gemella bergeri (1); Granulicatella spp (4); Kocuria kristinae (1); Listeria monocytogenes (1); Streptococcus agalactiae (3); Streptococcus pneumoniae (2); Streptococcus pyogenes (1). Others: HACEK (5); Trichosporon sp (2), Coxiella burnetii (2); Bartonella henselae (2).

TTE was performed in 425/500 (85 %) cases, whereas TEE was performed in 390/500 (78 %) cases. Regardless of the involvement of native or prosthetic valves, the most frequent location of vegetations was the Mitral Valve (MV) (221/498, 46.9 %) and the aortic valve (189/499, 37.9 %). PM or ICD infections were reported in 37/501 (7.4 %) cases. The main echocardiographic findings were vegetation (82.9 %), MV vegetations 235/501 (46.9 %), MV regurgitation (221/498; 44.4 %), and aortic valve regurgitation (207/501; 37.9 %). Complications of IE observed on echocardiography included abscesses (15.2 %), perforation of the valve apparatus (17.2 %), and prosthetic valve dehiscence (4.5 %). When comparing early and late prosthetic valve IE and native IE, fistulae were found in 2/59 (3.4 %) of early Prosthetic Valve Endocarditis (PVE), 2/98 (2 %) of late PVE, and 19/345 (5.5 %) of native valve IE patients, respectively, whereas abscesses were found in 11/59 (18.6 %) of early PVE, 24/98 (24.5 %) of late PVE, and 41/345 (11.9 %) of native valve IE patients, respectively. Dehiscence occurred in Early PVE in 5 of 59 (8.5 %) episodes, and in 10 of 95 (10.5 %) episodes of late PVE. The mean pulmonary arterial systolic pressure was 45.9 mmHg. Echocardiographic findings are summarized in Table 2.

Among the complications of IE episodes (Table 3), acute heart failure was the most frequent, occurring in more than half of the patients (291/502, 58 %), followed by acute renal failure (156/476, 32.8 %), central neurological events (131/501, 26.1 %), and myocardial or paravalvular abscesses (108/501, 21.6 %). Embolization was also frequently seen, with splenic emboli being the most common with 176/500 cases (35.2 %), followed by emboli to the central nervous system in 131/501 cases (26.1 %). The lungs were affected in 10/35 cases (28.5 %) of device-related endocarditis.

The overall mortality rate was 125/ 500 (25%) (Table 5). Among all the patients, 415 (82.6 %) had surgical indications. Of these patients, 347 (83.6 %) underwent surgery. Among patients who underwent surgery, the mortality rate was 21.3 % (74/347). Among the 68 patients who had a surgical indication but did not operate, 39/68 died (mortality rate: 57.4 %). This difference was statistically significant (p < 0.001, Chi-Squared test). Only 53 patients did not have surgical indications; among them, 10 died, representing a mortality rate of 18.9 %.

Selected variables were compared between two time periods of the cohort study: 2006‒2014, and 2015‒2023. These are presented in Table 4. The number of episodes of IE were 220 and 282 respectively, a 28 % increase in absolute numbers. There were significantly more episodes involving patients older than 60-years, patients on hemodialysis, patients with implantable cardioverter defibrillator, those who had previous cardiac surgery and patients with late prosthetic valve IE. Regarding predisposition, less patients with RVD were seen in the more recent period, and more patients with congenital heart disease. There were more patients with high degree AV block and coagulase negative staphylococcal IE. The frequency of surgical indication, the rate of surgery and the mortality did not change over the years.