Blood culture-negative endocarditis: Epidemiology, microbiology, and diagnosis

INTRODUCTION — 

Infective endocarditis (IE) refers to infection of the endocardial surface of the heart; it usually refers to infection of one or more heart valves.

The primary means of IE diagnosis involves microbiologic testing in the form of blood cultures and imaging with echocardiography; other diagnostic tools can be used to provide supplemental information. The most widely used criteria for the diagnosis of IE are the 2023 Duke-ISCVID Criteria [1]. Blood culture results are fundamental to defining an appropriate treatment course; however, they are not always diagnostic.

Issues related to the causes and clinical approach to diagnostic evaluation of patients with blood culture-negative endocarditis (BCNE) due to infectious etiologies will be reviewed here.

Issues related to clinical manifestations and diagnosis of native valve endocarditis are discussed separately. (See "Clinical manifestations and evaluation of adults with suspected left-sided native valve endocarditis".)

DEFINITION AND CATEGORIES

●Definition – Blood culture-negative endocarditis (BCNE) refers to endocarditis with no definitive microbiologic etiology following inoculation of at least three independently obtained blood samples in a standard blood-culture system, with negative cultures after five days of incubation.

The incidence of BCNE ranges from 2 to 69 percent [2-4]. Factors influencing the likelihood of BCNE include antibiotic administration prior to diagnostic evaluation, availability of diagnostic tools, and local epidemiology (including prevalence of zoonotic infections).

●Categories – There are three main categories of BCNE:

•Infective endocarditis (IE) due to typical bacterial causes of IE whose growth in culture is inhibited by antecedent antibiotic administration – This is the most common cause of BCNE [2,5]. This category includes typical bacterial pathogens such as staphylococci, streptococci, and enterococci. (See 'Organisms capable of growth in standard blood cultures' below.)

•IE due to organisms with fastidious growth characteristics in vitro – Fastidious organisms include fungi, atypical mycobacteria, Brucella spp, Mycoplasma spp, Legionella spp, and Cutibacterium acnes. (See 'Fastidious organisms' below.)

•IE due to organisms that cannot be readily cultured from blood using standard microbiologic methods – These include Coxiella burnetii, Bartonella spp, Tropheryma whipplei, and Chlamydia spp. (See 'Intracellular organisms' below.)

CAUSES — 

Causes of blood culture-negative endocarditis (BCNE) include infectious and noninfectious etiologies (table 1).

Infectious etiologies

Organisms capable of growth in standard blood cultures — A five-day incubation period is usually sufficient for detecting the pathogens below using modern instrument-based continuous-monitoring blood culture systems.

Typical bacterial pathogens — Typical bacterial infective endocarditis (IE) pathogens (such as staphylococci, streptococci, and enterococci) are the most common cause of BCNE, given findings from studies of supplementary diagnostic tools (such as polymerase chain reaction [PCR] of valvular tissue) [2,5,6]. Most culturable microbes can be recovered by modern automated continuously monitored blood culture systems; therefore, in such cases it is presumed growth in blood culture was inhibited by antecedent antibiotic administration. (See "Native valve endocarditis: Epidemiology, risk factors, and microbiology", section on 'Microbiology' and "Prosthetic valve endocarditis: Epidemiology, clinical manifestations, and diagnosis", section on 'Microbiology'.)

In one series including 283 patients with BCNE who underwent a multimodal diagnostic strategy including systematized testing of blood, and when available, valvular biopsy specimens using serological, broad range molecular, and histopathological assays, the proportion of cases due to typical IE pathogens was as follows [6]:

●Staphylococcus: 10.9 percent

●Streptococcus: 8.5 percent

●Enterococcus: 5.3 percent

Nutritionally variant streptococci — Nutritionally variant streptococci, Abiotrophia spp and Granulicatella spp, account for a very small proportion of streptococcal IE cases. In one center study in which 16S deoxyribonucleic acid (DNA) analysis was performed on heart valves of 279 patients with IE between 2012 and 2021, Abiotrophia spp was identified in two samples and Granulicatella spp was identified in one sample [7].

HACEK organisms — HACEK group are gram-negative bacteria that colonize the oropharynx:

●Haemophilus species (including H. parainfluenzae, the most likely cause of endocarditis among HACEK organisms)

●Aggregatibacter species (including Aggregatibacter actinomycetemcomitans [previously called Actinobacillus actinomycetemcomitans], Aggregatibacter aphrophilus [previously called Haemophilus aphrophilus], and Aggregatibacter paraphrophilus [previously called Haemophilus aphrophilus])

●Cardiobacterium species (including Cardiobacterium hominis)

●Eikenella corrodens

●Kingella species (including Kingella kingae)

Traditionally, these organisms were considered BCNE pathogens; however, with contemporary blood culture technologies, these organisms should grow in standard blood cultures within five days [8].

Since HACEK organisms typically grow in blood cultures, they are uncommon causes of BCNE; in one series including 283 patients with BCNE, a HACEK organism was identified in one case [6].

Fastidious organisms — There are several fastidious organisms to consider as a potential cause of BCNE, depending on individual patient risk factors (table 1).

An extended incubation period may be required for recovery of certain fastidious organisms but is not routinely necessary. Extended blood cultures may be reasonable for patients with a recent history of Legionella pneumonia or in the setting of epidemiologic concern for Brucella spp. (See "Detection of bacteremia: Blood cultures and other diagnostic tests", section on 'Duration of incubation'.)

If infection due to an atypical mycobacteria is suspected (see 'Atypical mycobacteria' below), mycobacterial blood cultures should be obtained; the typical incubation period is six weeks.

Fungi — Fungal causes of IE include Candida spp and Aspergillus spp. Rarer fungal causes of IE include Histoplasma capsulatum, Cryptococcus neoformans, Coccidioides spp, as well as other yeasts and molds [9-11]. The sensitivity of blood cultures for fungi is low. (See "Candida endocarditis and suppurative thrombophlebitis" and "Epidemiology and clinical manifestations of invasive aspergillosis", section on 'Endocarditis' and "Pathogenesis and clinical manifestations of disseminated histoplasmosis", section on 'Endocarditis'.)

Fungal IE is rare [6,12-15]. In one series including 283 patients with BCNE in France whose diagnostic evaluation included use of PCR on blood and valve samples, the frequency of fungal infection was 0.3 percent [6].

Risk factors associated with Candida IE include intravascular catheters, prosthetic heart valves, injection drug use, and immunosuppression [9,12]. Risk factors associated with Aspergillus IE include immunosuppression, exposure to a healthcare setting, and presence of a prosthetic valve [15,16].

In a review of fungal IE from 1965 to 1995, the sensitivity of diagnostic tools for the detection of fungal IE was as follows: blood culture – 54 percent, cardiac vegetation culture – 73 percent, histologic examination of valve vegetation – 95 percent, and histologic examination of peripheral embolus – 63 percent [17]. Aspergillus is rarely isolated from blood cultures; therefore, the diagnosis of Aspergillus IE is usually made by histopathology/culture of a resected heart valve or peripheral embolus [9,15].

In the above review, Candida was the most common fungal cause of IE (52 percent) followed by Aspergillus (24 percent) and Histoplasma (6 percent) [17]. Among Candida IE cases, C. albicans is the most commonly isolated, followed by C. parapsilosis [9,17].

Atypical mycobacteria — Atypical mycobacteria are a rare cause of IE.

Mycobacterium chimaera, a slow-growing mycobacterium, has been a cause of IE after cardiac surgery associated with contaminated heater-cooler units serving heart lung machines. Patients with M. chimaera IE have a delayed presentation, with a median time from cardiac surgery to IE presentation of 20 to 30 months. Symptoms include fatigue and weight loss with clinical findings of hepatitis, cytopenias, splenomegaly, and chorioretinitis [18-20]. These infections are noteworthy for their multisystem involvement and poor outcomes, even with intense medical and surgical treatment.

Rapid growing mycobacteria such as M. abscessus, M. chelonae, and M. fortuitum are rare causes of prosthetic valve IE. Studies examining valve origin, histology and cultures of patients with IE due to rapidly growing mycobacteria have implicated manufacturer contamination of bioprosthetic heart valves [21-23]. The time from initial valve implantation to IE ranges from 2 to 48 months. Patients with IE due to these organisms tend to present with dyspnea and echocardiographic signs of valve dysfunction, with or without fever [21,23]. Ziehl-Neelsen staining of valvular tissue was the key to diagnosis in most cases [21,23].

Atypical mycobacteria are discussed further separately. (See "Overview of nontuberculous mycobacterial infections".)

Brucella spp — Brucella spp are gram-negative coccobacilli transmitted to humans from infected animals (cattle, sheep, goats, camels, pigs, or other animals), by ingestion of food products, via contact with tissue or fluids or via inhalation of aerosols. The majority of reported Brucella IE cases are from Turkey, followed by Spain; cases have also been reported in travelers [24]. Most patients with Brucella IE have a history of valvular disease or a prosthetic valve [24].

In a review of 308 cases of IE due to Brucella, most cases (56 percent) were due to B. melitensis or B. abortus. Epidemiologic risk factors for infection included contact with infected animals or consumption of unpasteurized dairy (50 percent) and occupational exposure such as shepherding/farming (25 percent). The diagnosis was made by blood culture (positive in 60 percent of cases) and/or serology.

Issues related to brucellosis are discussed further separately. (See "Brucellosis: Epidemiology, microbiology, clinical manifestations, and diagnosis" and "Brucellosis: Treatment and prevention".)

Mycoplasma spp — Mycoplasma spp are intracellular organisms with specific nutritional requirements that grow poorly in routine blood cultures. They lack a cell wall and therefore are not detectable on gram stain. Mycoplasmas are colonizers of the human genital tract and also can be found in the respiratory tract. Mycoplasma spp are a rare cause of IE; most cases are due to M. hominis [25-27] rather than M. pneumoniae [28].

Almost all cases of M. hominis IE have occurred in patients with a prosthetic valve or after surgical repair for congenital heart disease. The median interval between cardiac surgery and IE presentation is six months (range 1 to 12 months) [26]. This epidemiology suggests a nosocomial source of infection, with potential inoculation around the time of surgery.

Most patients presented with fever and heart failure. Cases have been identified by PCR of valvular tissue [25-27].

Legionella spp — Legionella spp, fastidious gram-negative bacilli, are a rare of cause of IE. IE due to L. pneumophila, L. micdadei, L. longbeachae, L. anisa have been reported. Risk factors for Legionella IE include a prior history of Legionella pneumonia, immunosuppression, and presence of a prosthetic valve [29]. (See "Microbiology, epidemiology, and pathogenesis of Legionella infection".)

The first case of Legionella IE was reported in 1984 and occurred on a bioprosthetic aortic valve. Since then, less than two dozen other cases have been reported. Diagnoses were made mostly by PCR of valve tissue and serologies. Culture techniques are discussed separately. (See "Clinical manifestations and diagnosis of Legionella infection".)

Cutibacterium acnes — Cutibacterium acnes is a slow-growing facultatively anaerobic gram-positive bacillus. Since C. acnes is skin commensal, culture results must be interpreted carefully. (See "Invasive Cutibacterium (formerly Propionibacterium) infections", section on 'Endovascular infection'.)

In one study including 18 patients with negative blood cultures whose heart valves were submitted for 16S ribosomal RNA (rRNA) gene PCR sequencing, C. acnes was detected by PCR in five cases; C. acnes was also recovered from valve culture in two of these cases [30]. In another series including 31 cases of BCNE, two cases of C. acnes were identified (one by valve PCR and valve culture; the other by thrombus PCR) [2].

Intracellular organisms — Intracellular bacteria that are not culturable from blood or valves include Bartonella spp, Coxiella burnetii, Chlamydia spp, and Tropheryma whipplei. IE due to these organisms may be diagnosed via serologic testing or PCR of valves or other tissues.

In the series described above including 283 patients with BCNE in France whose diagnostic evaluation included use of PCR on blood and valve samples, the frequency of Bartonella spp, C. burnetii, and T. whipplei was 6.7 percent, 8.1 percent, and 1.1 percent, respectively [6]. IE due to Chlamydia, Mycoplasma, and Legionella is limited to case reports.

Coxiella burnetii — Coxiella burnetii are gram-negative intracellular bacteria that are transmitted via the respiratory or gastrointestinal route. C. burnetii is the causative agent of Q fever, an illness whose clinical manifestations may include fever, pneumonia, and hepatitis [31,32]. The main animal reservoirs are cattle, sheep, and goats; cats and dogs can also transmit infection [33]. (See "Q fever: Epidemiology, microbiology, and diagnostic tests" and "Chronic Q fever, including endocarditis".)

IE is a chronic form of Coxiella infection that can develop months to years after initial infection [33]. In one series including 1569 patients with acute Q fever, 7.6 percent developed IE [31].

In a review of 185 cases of Coxiella IE between 1950 and 2019, most patients had a history of animal exposure (72 percent) and either underlying heart valve disease or prosthetic cardiac material (89 percent) [34].

Coxiella IE is usually diagnosed by serology (anti-phase I IgG antibody titer >1:800), or by culture or PCR of valvular tissue. Valvular pathology of Coxiella IE is characterized by more fibrosis and calcification and less inflammation than IE due to typical bacterial causes [35].

Bartonella spp — Bartonella spp are fastidious, slow-growing, gram-negative bacteria that require specific laboratory conditions for optimal growth. Most cases of IE caused by Bartonella spp are due to B. quintana or B. henselae [36]. Risk factors for B. quintana IE include homelessness, alcoholism, and infestation with body lice; risk factors for B. henselae IE include contact with cats and previous valvular disease [36-38].

Serological testing and PCR from valve tissue have high sensitivity for diagnosis of Bartonella IE (sensitivities: immunofluorescence assay – 58 percent, Western blot – 100 percent, PCR valve – 92 percent, PCR blood – 33 percent, PCR serum – 36 percent) [36]. Valvular pathology of Bartonella IE is characterized by more fibrosis and calcification and less inflammation than IE due to typical bacterial causes [39]. (See "Microbiologic diagnosis of Bartonella infections".)

Issues related to IE due to Bartonella are discussed further separately. (See "Endocarditis caused by Bartonella".)

Tropheryma whipplei — T. whipplei is a gram-positive bacillus found in the environment [40]. Whipple's disease is a multisystemic process that can present over time. Heart valve involvement can occur in the absence of other manifestations of Whipple's disease (eg, joint and gastrointestinal manifestations) [41]. IE due to T. whipplei often has a subacute presentation with no or low grade fever [40]. Risk factors for IE due to T. whipplei include immunosuppression and pre-existing heart valve disease [40]. (See "Whipple's disease".)

In one study including 1135 patients in Germany who underwent valve explantation and analysis for bacterial infection with conventional culture and PCR amplification of the bacterial 16S rRNA gene followed by sequencing, bacterial IE was diagnosed in 255 cases. T. whipplei was observed in 6.3 percent of cases; it was the fourth most frequent pathogen (after streptococci, staphylococci, and enterococci) [41]. Other studies have reported a frequency of 1 to 3 percent [2,6].

Chlamydia spp — Chlamydia pneumoniae, C. psittaci, and C. trachomatis are obligate intracellular organisms that are rare causes of culture-negative IE [42,43]. C. pneumoniae causes respiratory tract infection; C. psittaci causes respiratory and systemic symptoms and is transmitted to humans predominantly from birds. C. trachomatis is a sexually transmitted infection. (See "Pneumonia caused by Chlamydia pneumoniae in adults" and "Psittacosis" and "Clinical manifestations and diagnosis of Chlamydia trachomatis infections in adults and adolescents".)

Most cases have of Chlamydia IE have been diagnosed by serology; a few cases have been diagnosed by PCR of valvular tissue [43,44],

Given cross-reactivity between the serologic tests for Chlamydia pneumoniae and Bartonella quintana, it is believed that many of the reported cases of IE attributed to Chlamydia [42] were actually due to Bartonella [45,46]. Furthermore, use of serologies to diagnose IE is complicated by the high prevalence of C. pneumoniae antibody titers in healthy adults [47]. For these reasons, we do not pursue serologic testing for Chlamydia spp [45].

Noninfectious etiologies — Patients with valvular vegetation(s) or arterial emboli in the absence of bacteremia may have sterile vegetation(s) due to a noninfectious etiology.

●Nonbacterial thrombotic endocarditis – Nonbacterial thrombotic endocarditis (NBTE) is a form of noninfectious endocarditis characterized by deposition of thrombi on heart valves (mostly aortic and mitral). NBTE is associated with a number of conditions; advanced malignancy (marantic endocarditis) and systemic lupus erythematosus (Libman-Sacks endocarditis) are the most common. (See "Nonbacterial thrombotic endocarditis".)

●Antiphospholipid syndrome – Antiphospholipid syndrome is a condition characterized by presence of antiphospholipid antibodies and related complications which may include venous thrombosis, arterial thrombosis, and nonthrombotic manifestations (eg, heart valve thickening or vegetation). (See "Antiphospholipid syndrome: Diagnosis".)

●Behcet disease – Behcet syndrome is a rare condition characterized by recurrent oral aphthae and any of several systemic manifestations including genital aphthae, ocular disease, skin lesions, gastrointestinal disease, neurologic disease, vascular disease, and arthritis. Cardiac involvement is uncommon but may include endocarditis. (See "Clinical manifestations and diagnosis of Behçet syndrome".)

●Pork allergy – An allergic reaction to pork was implicated as a cause of BCNE in a case report of a patient with history of porcine prosthetic valve replacement [48]. Proposed diagnostic tools include serum IgE, peripheral eosinophilia, and histopathology with eosinophilic infiltrate [49].

CLINICAL APPROACH

Clinical suspicion — The diagnosis of blood culture-negative endocarditis (BCNE) should be suspected in patients with clinical manifestations and cardiac imaging suggestive of infective endocarditis (IE), in the setting of negative blood cultures. Fever is the most common symptom of IE (up to 90 percent of patients); it is often associated with chills, anorexia, and weight loss. (See "Clinical manifestations and evaluation of adults with suspected left-sided native valve endocarditis".)

Clinical manifestations should be considered in the context of relevant risk factors and predisposing conditions – including prior IE, history of valvular disease, history of congenital heart disease, presence of cardiac prosthetic material, presence of indwelling intravenous lines, immunosuppression, injection drug use, or a recent dental or surgical procedure.

It is useful to categorize patients as having community-associated or health care-associated IE [50].

●For patients with community-associated IE, oral bacteria are common causes of community-associated IE that can be difficult to recover in blood cultures among patients who have received antecedent antibiotics. In addition, consider epidemiologic risks such as exposure to specific animals (ruminants, cats, birds) or regions typical for endemic fungi.

●For patients with health care-associated IE, a history of recent cardiac surgery should raise concern for atypical mycobacteria, mycoplasma infection, or microbes that have been epidemiologically associated with postoperative infections specific to an institution (eg, Serratia spp).

For patients who develop IE while on broad-spectrum antibiotics, pathogens associated with typical nosocomial infections that tend to be masked by the concurrent use of antibiotics should be considered (eg, staphylococci, enterococci, enteric gram-negative bacilli).

History and physical examination

●History – Clinical history should include the following details:

•Signs and symptoms – Signs and symptoms of IE are discussed in detail separately. (See "Clinical manifestations and evaluation of adults with suspected left-sided native valve endocarditis", section on 'Clinical manifestations'.)

•Acuity (eg, duration of symptoms).

•Recent surgery (including cardiac surgery) or dental procedure.

•History of valvular disease or congenital heart disease.

•Presence of cardiac prosthetic material and/or intravascular devices.

•Immunosuppression.

•Prior antibiotic use.

•Geographic location and travel history (risk for endemic fungi, Brucella).

•Exposure to ruminant/farm animals, cats, or birds (risk for Coxiella, Brucella, Bartonella, and C. psittaci).

•Injection drug use (risk for staphylococci, oral flora, and fungi).

•Social history including homelessness and alcohol use (risk for B. quintana).

●Physical examination – Physical examination should include cardiac auscultation, evaluation for stigmata of IE, and evaluation for extra-cardiac manifestations. (See "Clinical manifestations and evaluation of adults with suspected left-sided native valve endocarditis", section on 'Clinical manifestations' and "Complications and outcome of infective endocarditis".)

Diagnostic evaluation — Diagnostic evaluation for IE includes laboratory evaluation (microbiologic and/or histopathologic evaluation) in conjunction with cardiac imaging and/or surgical findings (table 2 and table 3). Issues related to microbiologic evaluation are discussed below; other components of diagnostic evaluation are discussed separately. (See "Clinical manifestations and evaluation of adults with suspected left-sided native valve endocarditis", section on 'Establishing a diagnosis of IE'.)

Obtaining blood cultures — Blood cultures remain the mainstay for diagnosis of IE; these should be optimized with respect to timing, volume, and number.

At least two sets of blood cultures should be obtained from separate venipuncture sites prior to initiation of antibiotic therapy. Three or more blood culture sets are ideal for distinguishing IE from non-IE bloodstream infection in the setting of microbes that are common skin colonizers or are unusual causes of IE. A blood culture set consists of an aerobic bottle and an anaerobic bottle; 10 mL of blood should be inoculated into each bottle.

For those who have already had blood cultures already drawn at the time of evaluation, we suggest obtaining additional blood cultures until at least two sets are obtained. For patients who were started on antibiotics empirically, we do not discontinue antibiotics prior to obtaining more blood cultures if there is a reasonable suspicion for IE.

To maximize diagnostic sensitivity, ideally blood for molecular testing should be obtained at the time of initial blood culture collection (with instructions for processing, freezing, and holding for subsequent molecular testing if needed). (See 'Further evaluation' below.)

Issues related to the timing, volume, and number of blood culture sets are discussed in detail separately. (See "Detection of bacteremia: Blood cultures and other diagnostic tests".)

Further evaluation — For patients with suspected IE and negative blood cultures, an approach to diagnostic evaluation is outlined in the algorithm (algorithm 1).

●Culture from alternative site(s) – Patients should be evaluated for potential sites of metastatic infection. Imaging (such as computed tomography [CT]) should be performed to evaluate focal signs or symptoms. In some cases, cultures obtained from such sites (ie, deep organ or tissue sites such as psoas abscess) can provide important diagnostic information.

●Laboratory testing (serology and polymerase chain reaction [PCR] from serum and/or plasma) – In the absence of culture data from blood or other sites, patients should be evaluated for causes of IE that are not recoverable in blood cultures (table 1) [51]. (see 'Causes' above):

•Serologic testing – Serologic testing for C. burnetii and Bartonella spp should be pursued. Diagnostic consideration may be guided by assessing risk factors like epidemiological exposure, keeping in mind that not all risk factors may be apparent or explicit. In one study evaluating use of serologic testing in conjunction with blood culture, serologic testing yielded etiologic identification in an additional 8 percent of cases [52].

Serum is used for most assays; it is usually collected in a clot activator tube with gel separator. Immunofluorescence antibody (IFA) is the most widely used serologic tool; testing is available at reference laboratories with a typical turnaround time of two to three days.

A positive serologic test should be reflexed to a titer with immunoglobulin differentiation (IgM versus IgG). In general, IgM reflects recent infection and IgG reflects recent or past infection; however, interpretation of antibody kinetics can be challenging [53].

-C. burnetii – During acute C. burnetii infection, phase II antigens circulate; during chronic infection, phase I antigens circulate. Anti-phase I IgG antibody titer >1:800 is diagnostic for C. burnetii IE based on the Duke Criteria [1]; an even higher cutoff titer has been suggested to avoid false positive results [36].

-Bartonella spp – Bartonella spp serologic tests are available for specific species or in a panel including B. henselae and B. quintana. A single IgG titer >1:800 is diagnostic for Bartonella spp IE [1,54].

•PCR – PCR (from serum and/or plasma) for Bartonella spp, C. burnetii, and T. whipplei should be obtained, if feasible.

•Additional testing – Further diagnostic evaluation should be guided by epidemiologic risk factors (table 1).

●Surgical specimen – If laboratory testing outlined above is negative and a surgical specimen is available (such as excised valve or embolic specimen), further evaluation is suggested, as outlined below. In practice, depending on the timing of surgery, these diagnostics are often done in parallel.

•Microbiology and histopathology – Tissue should be sent for Gram stain, bacterial culture (aerobic and anaerobic), fungal culture, and mycobacterial culture. Histopathology should include general and specific stains guided by epidemiologic risk factors (table 1).

•Targeted or broad-range PCR – If the above evaluation has not yielded a microbiologic diagnosis, we send tissue for targeted or broad-range PCR (with reflex to sequencing). We generally prefer targeted PCR due to its higher sensitivity. In one study including 123 cases, use of targeted PCR was more likely to establish a diagnosis than broad-range PCR (62 versus 2 percent, respectively) [51,55].

-Targeted PCR – Targeted PCR involves use of species-specific primer/probe pairs that bind to targeted sites on the pathogen DNA (if present) and amplify until the fluorescent dye on the probe is detected by the PCR instrument. This approach requires clinical decision-making regarding the suspected organisms to target.

-Broad-range PCR – Broad-range PCR involves use of primers against conserved regions to amplify various segments of the ribosomal genes; these amplicons are then sequenced and matched onto a sequence database to obtain species level identification. This approach does not require selection of suspected organisms to target.

•Specimen type – The diagnostic sensitivity for tissue is greater than for blood or serum; in one study including 106 patients with Bartonella IE, cardiac valve PCR had greater sensitivity than blood or serum PCR (92, 33, and 36 percent, respectively) [36].

Use of fresh tissue is preferable over formalin-fixed and paraffin-embedded (FFPE) blocks; FFPE followed by subsequent de-paraffin processing causes specimen damage including DNA fragmentation [56,57]. In one study including 1276 fresh tissue samples and 997 FFPE blocks, Bartonella spp were identified more frequently from fresh tissue samples (16 versus 10 percent) [58]. Emerging techniques may compensate for some limitations associated with FFPE [59].

●Metagenomic sequencing – If initial laboratory testing and surgical specimen evaluation is unrevealing, we send plasma for cell-free metagenomic sequencing. The approach consists of whole blood collection in ethylenediaminetetraacetic acid (EDTA) tubes, followed by prompt processing to stabilize the sample.

Metagenomics utilizes next-generation sequencing (NGS), in which all nucleic acid in a sample (including host DNA) is sequenced. Use of various post-sequencing bioinformatics applications allows identification of pathogens, if present [60,61].

This tool may have important diagnostic utility for settings in which antibiotic suppression is a major limitation in establishing a diagnosis. In one study including 23 patients with definite IE who underwent diagnostic evaluation with blood cultures and metagenomic sequencing, at 11 days after initiation of antibiotic therapy, metagenomic sequencing had greater diagnostic sensitivity (87 versus 20 percent) [60].

There are two important drawbacks of metagenomic sequencing for diagnosis. First, results require careful interpretation within the clinical context to distinguish between infection and contamination. Second, metagenomic sequencing does not provide susceptibility data to guide therapy; since most causes of IE can be cultured, every effort should be made to recover isolates for susceptibility testing.

●Fluorescence in situ hybridization and PCR sequencing (FISHseq) – FISHseq is a diagnostic approach that combines specialized valve histologic examination with 16S rRNA gene sequencing; thus far, it is not widely available [62]. In this technique, fluorescently labeled probes hybridize with ribosomes within target microorganisms, enabling visual identification as well as differentiation between active and inactive microorganisms. PCR sequencing permits identification of the microbial species.

In a retrospective study including 124 patients who underwent heart valve surgery due to suspected IE, standard diagnostic methods (ie, blood culture, valve culture, histopathology, and PCR sequencing) yielded inconclusive results in 46 percent of cases. FISHseq added diagnostic information in 63 percent of cases and reduced the number of inconclusive cases by 86 percent (from 57 to 8). Use of FISHseq had the potential to impact therapy in 11 patients (in whom only FISHseq resulted in identification of the causative organism), and in 40 patients (in whom treatment de-escalation could be considered).

FISHseq also improved diagnosis of prosthetic valve IE in a study including 113 prosthetic valves (from 105 patients) [63]. Conventional blood cultures were positive in 44 percent of cases, while valve cultures were positive in 28 percent of cases. FISHseq detected bacteria in 77 percent of cases; of these, the causative pathogen was identified in 45 percent of cases.

Establishing a diagnosis

Definite IE — Definite IE is established if pathologic criteria (either microbiologic or histopathologic criteria) or clinical criteria are met (table 2) [1]. Most patients with suspected IE do not meet pathologic criteria; the diagnosis often relies on clinical features.

●Pathologic criteria (either microbiologic or histopathologic criteria):

•Microbiologic criteria – Microorganisms identified in the context of clinical signs of active endocarditis:

-In a vegetation

-From cardiac tissue

-From an explanted prosthetic valve or sewing ring

-From an ascending aortic graft (with concomitant evidence of valve involvement)

-From an endovascular cardiac implantable electronic device (CIED)

-From an embolus

•Histopathologic criteria – Histopathologic findings of active endocarditis identified (may be acute or subacute/chronic):

-In or on a vegetation

-From cardiac tissue

-From an explanted prosthetic valve or sewing ring

-From an ascending aortic graft (with concomitant evidence of valve involvement)

-From a CIED

-From an embolus

Tools for microorganism identification include culture, staining, immunologic techniques, PCR or other nucleic acid-based tests. Molecular techniques and tissue staining should be interpreted cautiously, particularly in patients with a prior episode of IE; such tests can remain positive for extended periods following successful treatment.

Test specificity is influenced by several factors, and false positives can occur. Test results should be interpreted in the context of clinical and histological evidence of active endocarditis. A single finding of a skin bacterium by PCR on a valve without additional clinical or microbiological supporting evidence should be regarded as minor criterion and not definite IE.

●Clinical criteria – Clinical criteria for diagnosis of IE (major and minor) include:

•Two major clinical criteria

•One major and three minor clinical criteria

•Five minor clinical criteria

Major criteria for diagnosis of IE include (table 3):

•Microbiologic criteria (for patients with negative blood cultures, this includes one of the following laboratory tests):

-Positive PCR or other nucleic acid-based technique from blood for Coxiella burnetii, Bartonella spp, or Tropheryma whipplei

-Coxiella burnetii antiphase I IgG antibody titer >1:800 or Coxiella burnetii isolated from a single blood culture

-Indirect immunofluorescence assays for detection of IgM and IgG antibodies to Bartonella henselae or Bartonella quintana, with IgG titer >1:800

•Imaging criteria

-Echocardiography and/or cardiac CT imaging

-[¹⁸F]-fluorodeoxyglucose positron emission tomography (FDG-PET)/CT Imaging

•Surgical major criterion – Evidence of IE observed by direct inspection during cardiac surgery, in the absence of major microbiologic or imaging criteria, and in the absence of pathologic (microbiologic or histologic) criteria.

Minor criteria for diagnosis of IE are outlined in the table (table 3).

Possible or rejected IE — Definitions for possible or rejected IE are outlined in the table (table 2).

MANAGEMENT — 

Issues related to selection of empiric therapy for native valve and prosthetic valve infective endocarditis (IE) are discussed separately. (See "Antimicrobial therapy of left-sided native valve endocarditis", section on 'Empiric therapy' and "Antimicrobial therapy of prosthetic valve endocarditis", section on 'Empiric therapy'.)

For cases in which a definitive diagnosis is established, treatment should be tailored accordingly:

●(See "Candida endocarditis and suppurative thrombophlebitis".)

●(See "Treatment and prevention of invasive aspergillosis".)

●(See "Diagnosis and treatment of disseminated histoplasmosis in patients without HIV".)

●(See "Treatment of Mycobacterium avium complex pulmonary infection in adults".)

●(See "Rapidly growing mycobacterial infections: Mycobacteria abscessus, chelonae, and fortuitum".)

●(See "Mycoplasma pneumoniae infection in adults".)

●(See "Brucellosis: Treatment and prevention", section on 'Endocarditis'.)

●(See "Treatment and prevention of Legionella infection".)

●(See "Invasive Cutibacterium (formerly Propionibacterium) infections", section on 'Endocarditis'.)

●(See "Endocarditis caused by Bartonella".)

●(See "Whipple's disease".)

For cases in which no definitive diagnosis is established, the approach to empiric treatment is discussed separately. (See "Antimicrobial therapy of left-sided native valve endocarditis", section on 'Culture-negative endocarditis' and "Antimicrobial therapy of prosthetic valve endocarditis", section on 'Culture negative PVE'.)

SUMMARY AND RECOMMENDATIONS

●Definition – Blood culture negative endocarditis (BCNE) refers to endocarditis with no definitive microbiologic etiology following inoculation of at least three independently obtained blood samples in a standard blood-culture system, with negative cultures after five days of incubation. (See 'Definition and categories' above.)

●Categories and causes – There are three main categories of BCNE (see 'Definition and categories' above):

•Infective endocarditis (IE) due to typical bacterial causes of IE whose growth in culture is inhibited by antecedent antibiotic administration – This is the most common cause of BCNE. This category includes typical bacterial pathogens such as staphylococci, streptococci, and enterococci.

•IE due to organisms with fastidious growth characteristics in vitro – Fastidious organisms include fungi, atypical mycobacteria, Brucella spp, Mycoplasma spp, Legionella spp, and Cutibacterium acnes.

•IE due to organisms that cannot be readily cultured from blood using standard microbiologic methods – These include Coxiella burnetii, Bartonella spp, Tropheryma whipplei, and Chlamydia spp.

Infectious causes of BCNE are discussed above and summarized in the table (table 1). (See 'Infectious etiologies' above.)

Noninfectious causes of BCNE are discussed above. (See 'Noninfectious etiologies' above.)

●Clinical suspicion (see 'Clinical suspicion' above):

•The diagnosis of BCNE should be suspected in patients with clinical manifestations and cardiac imaging suggestive of IE, in the setting of negative blood cultures. Fever is the most common symptom of IE (up to 90 percent of patients); it is often associated with chills, anorexia, and weight loss.

•These findings should be considered in the context of relevant risk factors and predisposing conditions – including prior IE, history of valvular disease, history of congenital heart disease, presence of cardiac prosthetic material, presence of indwelling intravenous lines, immunosuppression, injection drug use, or a recent dental or surgical procedure.

●Diagnostic evaluation

•Obtaining blood cultures – Blood cultures remain the mainstay for diagnosis of IE; these should be optimized with respect to timing, volume, and number. (See 'Obtaining blood cultures' above.)

•Further evaluation – An approach to diagnostic evaluation for patients with suspected BCNE is outlined in the algorithm (algorithm 1). The approach includes the following tools (see 'Further evaluation' above):

-Laboratory testing for C. burnetii (serology and/or polymerase chain reaction [PCR]) and Bartonella spp (serology and/or PCR), Tropheryma whipplei (PCR), and other causes, guided by epidemiologic risk factors (table 1).

-Surgical specimen (such as excised valve tissue or embolic specimen) evaluation including microbiology studies, histopathology, and PCR.

-Use of cell-free metagenomic sequencing, if initial laboratory testing and surgical specimen evaluation is unrevealing.

●Establishing a diagnosis – Definite IE is established if pathologic criteria (either microbiologic or histopathologic criteria) or clinical criteria are met. (See 'Establishing a diagnosis' above.)

●Management

•For cases in which a definitive diagnosis is established, treatment should be tailored accordingly.

•For cases in which no definitive diagnosis is established, the approach to empiric treatment is discussed separately. (See "Antimicrobial therapy of left-sided native valve endocarditis", section on 'Culture-negative endocarditis' and "Antimicrobial therapy of prosthetic valve endocarditis", section on 'Culture negative PVE'.)