Introduction

Endocarditis remains a lethal disease, with high morbidity and mortality. The incidence of endocarditis is increasing due to ongoing public health challenges. These include the opioid epidemic, an increasingly elderly and frail population, more patients on hemodialysis, and the increasing use of implanted cardiac valves and internal medical devices [1]. The microbial pattern of the disease continues to evolve as well, with many patients combating antibiotic-resistant infections [2]. Due to these factors, the disease continues to receive significant attention in the hope of improving diagnosis and therapy. The goal of this review is to summarize our current understanding of IE diagnosis and treatment, with a focus on published work from the last 5 years.

Clinical Presentation

Endocarditis is heterogeneous, with a wide range of clinical manifestations. Most patients present with non-specific signs and symptoms. These include fever, fatigue, generalized weakness, tachypnea, and tachycardia. Patients with valvular insufficiency are apt to present with additional cardiopulmonary symptoms such as dyspnea, orthopnea, and chest pain. The classical finding of a new murmur is present in less than 50% of cases [3]. Electrocardiogram findings may demonstrate conduction abnormalities, ranging from PR-interval prolongation to complete heart block [4]. IE often releases infectious emboli, and these can cause infections and vaso-occlusive phenomena in almost any organ. Patients can have peritonitis from mesenteric vasculature occlusion or neurologic sequelae from stroke when cerebral vasculature is affected [5]. Additionally, as part of initial imaging, CT scan can reveal a wide range of embolic phenomena. Patients can have cavitary lung lesions with pneumonia-like symptoms or splenic infarcts and present with left upper quadrant pain [3].

Characteristic dermatological findings include Janeway lesions and Osler nodes. Named after Dr. Edward Janeway and Dr. William Osler, and first described in the late 1800s, the Janeway lesion and Osler node correlate with the acute and subacute phases of the disease, respectively. Janeway lesions are non-tender, hemorrhagic macules, or papules, usually found on the hypothenar or thenar eminences lasting days to weeks. Osler nodes are tender, purple-to-pink nodules, usually found distally on the fingers and toes, and last hours to days [6]. Ultimately, the clinical presentation of IE can vary, with many patients displaying only non-specific signs and symptoms. Suspicion must remain high for those with predisposing risk factors [5].

Microbiology

Historically, the most prevalent causative organisms for IE have been Staphylococcus and Streptococcus species. This remains true in multiple recent cohort studies, in locations ranging from New York to Spain and to Hong Kong [7,8,9]. In these cohorts, Staphylococcus aureus remains the most common culprit. The prevalence of methicillin-resistant S. aureus varies between cohorts but can represent nearly half of cases with S. aureus infection [7]. The bacterial profile was similar in a large European infective endocarditis registry of 3116 patients from 40 countries. Staphylococci were the most identified organism (44%) in the registry, with streptococci second. This was divided between oral streptococci species (12%) and Streptococcus gallolyticus (7%)[1].

These cohorts demonstrate a significant percentage of patients with Enterococcus infection, ranging from 8 to 16%, and the percentage of patients with enterococci infection is increasing. It is possible this is due to the increasing age of endocarditis patients. Patients > 80 years old have much higher rates of infection with Enterococcus and Streptococcus gallolyticus and conversely lower rates of infection with S. aureus [10]. This is likely because more geriatric patients have the gastrointestinal and urinary tracts as their primary source of infection. The skin and oral cavity are more common sources for younger patients [10].

The microbiological profile is different for patients with specific risk factors, such as a bicuspid aortic valve, mitral valve prolapse, those on hemodialysis, or following transcatheter aortic valve implantation.

Data from the GAMES Registry in Spain (Grupo de Apoyo al Manejo de la Endocarditis infecciosa en Espana) showed that patients with either a bicuspid aortic valve or mitral valve prolapse are more likely to have infection with viridans group streptococci, when compared to endocarditis patients with other high-risk comorbidities [8]. Data from the Mayo Clinic Enterprise further identified the species of viridans group streptococci that affected patients with mitral valve prolapse. For these patients, Streptococcus mitis was the most identified viridans group streptococci, representing almost forty five percent of cases. Streptococcus anginosus and Streptococcus sanguinis were identified at lower rates [11].

Patients on hemodialysis (HD) have higher rates of infection with S. aureus, with rates of infection that can approach 50% [12]. HD patients are also more likely to have positive blood cultures when compared to other endocarditis patients (89% vs. 72%) [12].

A review of trials of endocarditis following transcatheter aortic valve implementation found that the most common isolate in these patients were Enterococcus species, ahead of S. aureus, by a significant margin (26% vs. 16%) [13]. Unfortunately, this group has a higher mortality rate of 34–40% [13].

Infection with gram-negative bacteria is divided into HACEK organisms (Haemophilus species, Aggregatibacter species, Cardiobacterium species, Eikenella species, and Kingella species) and non-HACEK organisms. Overall, gram-negative infection accounts for 1–10% of endocarditis cases, and the majority of these are due to HACEK organisms [14]. However, the proportion of non-HACEK infections has been increasing, again due to changing patient demographics. Infection with non-HACEK organisms is correlated with intravenous drug use and active malignancy and is more likely to be healthcare associated [15]. The distinction between HACEK and non-HACEK infection is clinically meaningful for two key reasons. First, the HACEK organisms are slow growing and not highly pathogenic. Infection with these organisms has a mortality rate of 2%, much lower than other causes of endocarditis. The non-HACEK organisms have a much higher associated mortality of 30% [14]. Second, non-HACEK organisms are more likely to have multi-drug resistance, which complicates therapy. Among non-HACEK cultured organisms, the predominant species are Escherichia coli, Klebsiella species, and Serratia marcescens [15].

Fungal endocarditis represents only 1–3% of cases, though the proportion is growing. This is due to increasing numbers of patients with ongoing critical care needs, invasive catheters, and immunosuppression. The mortality for fungal endocarditis is > 70%. Candida is by far the most identified fungal organism, and accounts for half of cases. Aspergillus and Histoplasma account for most of the remaining cases [16].

Diagnosis

Since their original publication in 1994, clinicians have used the Duke Criteria to assist with diagnosis. The Duke Criteria defines a list of major and minor clinical, laboratory, and imaging characteristics of IE. Patients are risk stratified into definitive IE; possible IE or the diagnosis is rejected based on the presence of these characteristics. Multiple modifications to the original criteria have been proposed since publication, to account for a changing understanding of microbiology (especially regarding S. aureus, Enterococcus faecalis, and Coxiella burnetii), as well as newer antibody and immunofluorescence detection techniques. Subsequent work has solidified the value of transesophageal echocardiography for diagnosis, and as well as CT imaging techniques [17]. These modifications culminated in the 2023 International Society for Cardiovascular Disease updates to the Modified Duke Criteria [18•].

Today’s patients with IE are increasingly ill and medically complex. They are more likely to be immunosuppressed, have prosthetic valves or other implanted cardiac devices, and are more likely to require intensive care. Despite these demographic changes, the Modified Duke criteria have been consistently validated to risk stratify patients with possible infectious endocarditis. The Modified Duke criteria are useful even when echocardiography findings are not considered, which is important clinically as echocardiography results are often unavailable during patient care [19]. When transesophageal echocardiography (TEE) findings are considered, the predictive value of the Modified Duke Criteria is exceedingly high. Improved echocardiography technology, with better image quality and 3-D image generation contributes to this diagnostic accuracy, especially for patients with prosthetic valves. While valvular vegetations are the classic findings, other abnormalities can also suggest the diagnosis. These include paravalvular abscesses, fistulas or pseudoaneurysms, and valve leaflet perforation and valve dehiscence. Two additional clinical factors, the presence of a heart murmur and the need for intensive care therapy, have been found to lead to a slight improvement in the performance of the Modified Duke Criteria and are especially important when ultrasound is not available [19].

As mentioned previously, there has been an increase in IE caused by Enterococcus species. This trend led to proposals to re-characterize how Enterococcus faecalis is characterized by the Duke Criteria. Previously, for Enterococcus culture to be considered a major diagnostic criterion, it must be community acquired, and without a primary focus. An analysis of patients with Enterococcus endocarditis found that if E. faecalis was instead classified as a “typical” organism (similar to S. aureus and viridans group streptococci), and then the sensitivity of the Duke Criteria for these patients improved dramatically, from 70 to 96%, as more patients were classified as having definitive, rather than possible, IE [20].

A persistent diagnostic challenge is the prevalence of culture negative endocarditis. In many studies, > 30% of patients will have negative blood cultures, complicating their care [21, 22]. This is often because patients receive antibiotics before cultures are obtained or due to infection with difficult to culture organisms. Thankfully, new tools are under investigation to help identify responsible pathogens and guide antimicrobial therapy. One technique is molecular analysis using amplification or ribosomal RNA genes found in bacteria and fungi. When combined with PCR arrays targeting antibiotic resistance genes, this approach is highly sensitive and specific for identifying the causative agent of endocarditis, as well as providing antibiotic susceptibility data. The technique can identify a pathogen in 92% of cases with culture negative endocarditis [21]. The major drawback of this technique is that it requires tissue. Thus, it is limited to patients undergoing cardiac valve replacement, where samples can be taken from the native valve or the perivalvular tissue of a prosthetic valve.

A newer technique, which demonstrates promise but is not yet widely clinically available, is targeting metagenomic sequencing on blood or plasma. This strategy employs PCR targeting of ribosomal RNA, but does not require tissue, thereby expanding its clinical utility to non-surgical cases of endocarditis. Pilot studies have shown promise, with pathogens identified in most culture negative cases [22].

The 2023 update to the Modified Duke Criteria was produced by a panel of 25 subject matter experts. They proposed several changes to the criteria, in both pathologic criteria, microbiologic criteria, and imaging criteria. A new major criterion—intraoperative surgical evidence of endocarditis—was also added. However, the overall structure of the criteria, with major and minor criteria used to risk stratify patients, remains unchanged.

New pathologic criteria for IE were expanded to include PCR or other nucleic-based techniques, as well as indirect immunofluorescence assays. When PCR techniques are used to identify Coxiella burnetti, Bartonella species, or Tropheryma whipplei, this now meets major criteria for diagnosis. If immunofluorescence assays detect IgM or IgG from Bartonella, major criteria are met as well [18•].

Regarding microbiologic criteria, the list of typical pathogens was expanded. As discussed previously, E. faecalis was included as a typical pathogen, due to increased sensitivity when included. Staphylococcus lugdunensis and the “streptococci-like” bacteria (Granulicatella, Abiotrophia, and Gemella species) were added as typical pathogens, due to the high risk for infective endocarditis in patients with bacteremia from these organisms. The list of typical pathogens in the presence of intracardiac devices was expanded as well to include eight new species, including Corynebacterium striatum, Serratia marcescens, Pseudomonas aeruginosa, and Candida. The expansion of the list of typical pathogens (both with and without associated intracardiac devices) will increase the number of patients with major microbiologic criteria for diagnosis and potentially identify more cases with today’s patient cohort [18•].

Emphasis on the importance of TTE for diagnosis has not changed in the 2023 guidelines. However, the additional imaging modalities of cardiac computed tomography (CT) and positron emission computed tomography (PET/CT) have been included as major criteria. While not as accurate for detecting vegetations, cardiac CT has better sensitivity for paravalvular lesions, such as a paravalvular abscess. PET/CT improves detection in patients with prosthetic material, such as those with multiple prosthetic valves, or prosthetic aortic valves combined with aortic grafts. Cardiac CT and PET/CT evidence of endocarditis have been included as major criteria in the proposed 2023 guidelines [18•].

Finally, the 2023 update includes intra-operative evidence of endocarditis as a major criterion. While this may not be applicable to most cases, it potentially can confirm the diagnosis in situations where other pathologic or microbiologic criteria have not been obtained, such as cases with negative cultures.

Treatment Recommendations

The current standard of care guidelines established by the American Heart Association (AHA) in 2015 underscore the importance of using blood and tissue cultures and sensitivities to guide antibiotic selection. Antibiotic initiation should be delayed in stable patients presenting with subacute endocarditis until cultures and sensitivity result [23]. Patients with acute presentations or who are unstable require empiric antibiotics. The choice of antibiotics depends on whether the patients have native heart valves or prosthetic heart valves [23, 24]. Empiric treatment is further determined by the likely source organism given the patient’s risk factors.

For patients with native valves, guidelines recommend vancomycin and either cefepime or ampicillin/sulbactam as empiric treatment, since the most likely organisms are S. aureus, Streptococcus, enterococci, and HACEK organisms [23]. While current guidelines recommend 2–6 weeks of IV antibiotics for gram-positive endocarditis, some studies have suggested it is possible to transition patients to oral regimens after initial intravenous treatment [23]. The subsequent use of dalbavancin, either 500 mg weekly or twice weekly, after initial treatment with parenteral antibiotics until achieving negative blood cultures has been shown to have a high success rate in treating gram-positive IE in one small study [25]. The use of dalbavancin to treat staphylococcal endocarditis is further proven in an ex vivo model that showed that the minimum inhibitory and eradication concentration of dalbavancin, which can break down the biofilm and eliminate the bacteria growing on valvular tissue, can be achieved with standard dosing of dalbavancin [26].

For patients with prosthetic valves, guidelines recommend vancomycin with either ceftriaxone if the valve is older than 1 year or the combination of rifampin, gentamicin, and cefepime if the valve is less than 1 year old as empiric treatment until cultures result [23].

The use of daptomycin has been suggested for gram-positive endocarditis either alone or in combination with either an aminoglycoside, cephalosporin, or beta-lactam. Data from a single hospital study found that daptomycin-containing regimens have reduced 30-day mortality and duration of antibiotics for patients with prosthetic valves. Further analysis found that high dose daptomycin (> 8 mg/kg/day) reduced mortality and treatment failure compared to standard dose daptomycin [27].

The continued use of rifampin for staphylococcal prosthetic valve endocarditis (SPVE) has come into question in recent studies. One large multicenter study found that rifampin had no effect on in-hospital or 1-year mortality for SPVE; interestingly, the group treated with rifampin did have longer hospital stays although the etiology of the prolonged hospital stay is unclear [28]. Similarly, while gentamicin is no longer recommended to treat native valve staphylococcal endocarditis, it is still part of the AHA recommendation for SPVE, though its use has been called into question [23]. Multiple recent studies found that the addition of gentamicin to antibiotic regimen containing rifampin for SPVE does not improve mortality or the rate of reoccurrence. Unfortunately, these studies did not directly compare regimens that used either only adjunctive rifampin or gentamicin versus those without any adjunctive rifampin or gentamicin [29, 30]. In short, the evidence from these articles is insufficient either to show or refute the utility of gentamicin and rifampin in IE [28,29,30].

Enterococcus faecalis infective endocarditis (EFIE) accounts for most enterococcus IE; guidelines recommend the use of either ampicillin/gentamycin (AG) or ampicillin/ceftriaxone (AC) regardless of the presence of prosthetic valves [23]. One meta-analysis study found that both treatment regimens have similar mortality and failure rates, but the AG regimen has higher side effects, including nephrotoxicity [31•]. As EFIE has been shown to have higher relapse rates compared to other IE, one large study based in France compared the rate of relapse of EFIE using an AC versus an AG antibiotic regimen and found comparable relapse rates. With a 1-year relapse rate of 8.2% for the AG group and 11.8% for the AC group, one must question whether current guidelines are sufficient [32].

Viridans group streptococci (VGS) and Streptococcus gallolyticus account for most of the endocarditis secondary to streptococcal organisms. Current treatment guidelines per the AHA recommend for native valve VGS/Streptococcus gallolyticus endocarditis either 4 weeks of penicillin or ampicillin or ceftriaxone if the organism is penicillin sensitive or 2 weeks of gentamicin and 4 weeks of either penicillin or ampicillin if the organism is penicillin resistant. For patients with prosthetic valves, the recommendations are similar, but the duration is 6 weeks [18•]. Data from one large multicenter study which compared the mortality rates of VGS/Streptococcus gallolyticus endocarditis with either the use of a beta-lactam antibiotic monotherapy regimen versus a beta-lactam and aminoglycoside regimen found that there was no difference in in-hospital and 1-year mortality for either regimen in patients with penicillin sensitive organisms, but patients with partial penicillin-resistant organisms had higher 1-year mortality with the beta-lactam monotherapy regimen [33].

Among the HACEK group, Haemophilus species are the leading cause of IE across all age groups [34]. However, culture, identification, and susceptibility testing of the HACEK organisms poses unique difficulties, often due to prolonged incubation periods and tailored growth mediums. Therefore, initial antibiotic recommendations for this group are either a broad-spectrum cephalosporin or fluoroquinolone (for patients with a β-lactam allergy) until susceptibility testing is obtained, with administration for 4 weeks in those with a native valves and 6 weeks in those with a prosthetic valve [35].

Antibiotic Prophylaxis

Guidelines regarding the use of prophylactic antibiotics to prevent IE have undergone multiple iterations. The latest AHA guidelines recommend prophylactic antibiotics be limited to patients who are at the highest risk of developing severe complications. These patients are defined as patients with previous IE, prosthetic or partial prosthetic valves, heart transplant patients with valvulopathy, and patients with a congenital heart disease that either have a cyanotic variant that’s unrepaired, have had prosthetic material within the past 6 months to repair the defect, or whose repairs have left tissue defects preventing endothelization [36]. One large observation study that tracked compliance with the new guidelines and the rates of IE found there was a significant decrease in the amount of prophylactic antibiotics prescribed before a procedure to patients deemed moderate risk of developing IE per AHA guidelines and an increase in the incidence of IE in those moderate risk patients. However, the incidence of IE in high-risk and moderate-risk patients was increasing before the 2007 change and continued to increase after the change without any notable jump in the moderate-risk group. This suggests the increased incidence of IE was not related to the guideline change [37].

Two large studies, which used ICD-10 codes from a medical insurance database, looked at the odds of developing IE after certain dental procedures. One of the studies found that in high-risk patients there was an increased risk of developing IE after any invasive dental procedure (IDP). Invasive dental procedures were defined as any procedure that had manipulation of the gingival tissue, periapical region of the teeth, endodontic procedures, and perforation of the mucosa [38]. The second study found that the odds of developing IE in high-risk patients were only limited to tooth extractions and scaling [39]. Both studies found that the risk of developing IE was highest 30 days after an IDP [38, 39]. Both studies found that the use of prophylactic antibiotics decreased the odds of developing IE in high-risk patients [38, 39].

The use of prophylactic antibiotics for dental procedures is not standardized around the world. These variations in care are likely due to conflicting evidence seen in observational studies and the lack of large experimental studies assessing the efficacy of prophylactic antibiotics in preventing IE [37,38,39,40]. For example, in 2012, health guidelines in Sweden recommended against the use of prophylactic antibiotics in any dental procedure even in high-risk individuals [40, 41•]. Two large Swedish studies then tracked the incidence of IE 5 years before and after the change. They found the incidence of both S. aureus and streptococcal viridians had been increasing over the 10-year period prior to the guideline update, but the rate of increase was not affected by the change. They also found there was no change in the adjusted hazard rate ratio of developing IE compared to low-risk patients after the guideline change [40, 41•].

Data regarding the association of non-dental invasive procedures with the development of endocarditis and whether prophylactic antibiotics would be effective in inhibiting IE from those invasive procedures is sparse. Two studies attempted to determine if there was an association between certain invasive procedures and IE. One study based in England used a crossover model to calculate the odds ratio of a procedure leading to IE and found that blood transfusions, upper endoscopies, and lower endoscopies were associated with IE [42]. The second study, based in Sweden, also a crossover study, found that colonoscopy, cystoscopies, bronchoscopies, and blood transfusion were associated with IE [43]. The absolute increase in risk associated with the invasive procedures listed above, while statistically significant, was minimal and likely clinically insignificant [42, 43]. This bolsters the AHA’s recommendation of not giving prophylactic antibiotics to patients undergoing routine respiratory, GI, and GU procedures unless there is an infection or disruption of the mucosa [36].

Conclusions

As the population ages, the use of invasive medical devices grows, and the opioid crisis continues, the incidence of endocarditis is increasing. The disease remains lethal and difficult to diagnose. Fortunately, our understanding of the disease continues to evolve. There are new immunoassay and imaging techniques to assist with diagnoses. The newly revised Modified Duke Criteria remain as relevant as ever. Antibiotic treatment recommendations continue to evolve, with newer antibiotics, especially dalbavancin, showing promise to improve therapy. It is important for all providers to be aware of the ongoing changes in our approach to infectious endocarditis.