Author + information
- Received July 27, 2015
- Revision received October 5, 2015
- Accepted October 13, 2015
- Published online January 19, 2016.
- François Delahaye, MD, PhD∗∗ (, )
- Ali M’Hammedi, MD†,
- Brice Guerpillon, MD‡,
- Guy de Gevigney, MD†,
- André Boibieux, MD‡,
- Olivier Dauwalder, PharmD, PhD§,
- Coralie Bouchiat, PharmD, PhD§ and
- François Vandenesch, MD, PhD§
- ∗Department of Cardiology, Hôpital Louis Pradel, Hospices Civils de Lyon, Université Claude Bernard, Lyon, France
- †Department of Cardiology, Hôpital Louis Pradel, Hospices Civils de Lyon, Lyon, France
- ‡Department of Infectious Diseases, Hôpital de la Croix-Rousse, Hospices Civils de Lyon, Lyon, France
- §Centre National de Référence des Staphylocoques, Hospices Civils de Lyon, Lyon, France
- ↵∗Reprint requests and correspondence:
Prof. François Delahaye, Hôpital Louis Pradel, 28, Avenue du Doyen Lépine, 69677 Bron Cedex, France.
Background Looking for and treating the portal of entry (POE) of infective endocarditis (IE) is important, but published research on this topic is nonexistent.
Objectives The goal of this study was to systematically search for the POEs of present and potentially new episodes of IEs.
Methods Patients were systematically seen by a stomatologist, an ear, nose, and throat specialist, and a urologist; women were systematically seen by a gynecologist; patients were seen by a dermatologist when there were cutaneous and/or mucous lesions. Colonoscopy and gastroscopy were performed if the microorganism came from the gastrointestinal tract in patients ≥50 years of age and in those with familial histories of colonic polyposis. Treatment of the POE was systematically considered.
Results The POEs of the present IE episodes were identified in 74% of the 318 included patients. The most frequent POE was cutaneous (40% of identified POEs). It was mainly (62% of cutaneous POEs) associated with health care and with intravenous drug use. The second most frequent POE was oral or dental (29%). A dental infectious focus was more often involved (59% of oral or dental POEs) than a dental procedure (12%). POEs were gastrointestinal in 23% of patients. Colonic polyps were found in one-half of the patients and colorectal adenocarcinomas in 14%. Performance was good regarding the search for an oral or dental or a colonic potential POE, which were found in 53% and 40% of patients, respectively.
Conclusions Our search for the POEs of present IEs was often successful, as was searching for an oral or dental or a gastrointestinal POE of a new IE episode. We advise the systematic performance of stomatologic examinations in patients with IE and performance of colonoscopy in patients ≥50 years of age or at high risk for colorectal cancer.
Infective endocarditis (IE) is a severe disease, with an in-hospital mortality rate of about 20% (1). Five percent to 10% of patients will have additional episodes of IE (2). Thus, looking forand treating the portal of entry (POE) of IE is particularly important. The POE of the present episode must be identified in order to treat it. The potential POE of a new episode must be searched for in order to eradicate it and thus lower the risk for a new IE episode. Yet published research on this topic is nonexistent. The search for and treatment of the POE are not even mentioned in the most recent guidelines on IE (3,4). We thus undertook a study of the performance of a systematic search for the POE of the present episode of IE and of a potential new episode of IE.
Since January 2005, we have been prospectively enrolling all patients hospitalized at our tertiary hospital for definite IE according to the Duke-Li criteria (5). Since then, we have been systematically looking for the POE of the present IE episode and for the potential POE of a new IE episode (e.g., a patient’s present IE is due to Streptococcus gallolyticus, the POE of the present IE episode is a colorectal adenocarcinoma, systematic stomatologic examination identifies several dental infectious foci, which are considered potential POEs for a new IE episode). Patients were informed of the study but did not have to provide individual consent, in accordance with French ethics laws.
Patients were systematically seen by a stomatologist (who performed an orthopantomogram), an ear, nose, and throat (ENT) specialist, and a urologist; women were systematically seen by a gynecologist. When there were cutaneous or periorificial mucous lesions on the initial examination, patients were seen by a dermatologist. Cerebral and thoracoabdominopelvic scans were systematically performed. Colonoscopy and gastroscopy were performed if the microorganism came from the gastrointestinal tract, in patients ≥50 years of age, and in those with familial histories of colonic polyposis. Because our center is a tertiary center with cardiac surgery facilities, most patients who are hospitalized for IE at our hospital are transferred from other hospitals. Either the whole antibiotic course and all investigations for the search for the POE were performed during the patient’s stay in our hospital, or the patient was transferred to the hospital of origin before the end of the antibiotic course, and we requested that these investigations be performed there.
For each microorganism, the most probable POE was inferred from its natural habitat or site of colonization in humans on the basis of a search of published research (Table 1). Treatment, if any, of the POE was systematically considered. It was either performed during the patient’s stay in our hospital or prescribed.
Health care–associated IE was defined as either IE developing in a patient hospitalized for more than 48 h before the onset of signs or symptoms consistent with IE or IE diagnosed within 48 h of admission in an outpatient with extensive health care contact (received intravenous therapy, wound care, or specialized nursing care at home within 30 days; underwent hemodialysis; received intravenous chemotherapy; resided in a nursing home or long-term care facility). Community-acquired IE was defined as IE diagnosed at the time of admission (or within 48 h of admission) in a patient not fulfilling the criteria for health care–associated infection (6).
Among 444 patients hospitalized at our institution between 2005 and 2011, 318 (320 episodes) were included in the present study (we excluded 82 patients who died during hospitalization; 44 medical charts were unavailable for technical reasons).
The median age of the patients was 61 ± 2 years; 75% were men; 29% had native valve disease, 22% had ≥1 valvular prosthesis, and 49% did not have previously known heart disease; 11% had cardiac implantable electronic devices (pacemakers or defibrillators). Microorganisms were streptococci in 41%, staphylococci in 31%, and enterococci in 8%.
POE for the present IE episode
The POEs for the present IE episodes were identified in 238 patients (74%). Among identified POEs, 40% were cutaneous, 29% were oral or dental, and 23% were gastrointestinal (Table 2).
POEs were cutaneous in 96 patients. Cutaneous POEs were health care associated in 41% of these patients, community acquired in 34%, related to intravenous drug use in 22%, and related to inoculation diseases in 3% (louse bite, Bartonella quintana, n = 1; tick bite, Coxiella burnetii, n = 1; cat scratch disease, Bartonella henselae, n = 1).
Vascular access was the main health care–associated cutaneous POE (44%), followed by infection of a cardiac implantable electronic device (28%) and infection of the operative site (28%) (Table 3). Wounds, nonsuppurative skin and soft-tissue infections, and diabetic foot ulcers were the most frequent community-acquired cutaneous POEs.
Staphylococci were responsible for 87% of the 39 cases of IE with health care–associated cutaneous POEs (Staphylococcus aureus, 38%; coagulase-negative staphylococci, 49%) (Table 4). S. aureus was responsible for 82% of 33 cases of IE with community-acquired cutaneous POEs and for 52% of cases of IE in intravenous drug users.
Oral or dental POE
Overall, a stomatologist saw 62% of patients during their stays in our hospital. Oral or dental POEs were identified in 68 patients. The distribution of lesions is detailed in Table 5, and the distribution of microorganisms is presented in Table 6. Oral streptococci were responsible for 69% of the cases of IE with oral or dental POEs.
Sixty-five of the 68 patients with oral or dental POEs (96%) saw a stomatologist during their stay in our hospital. For organizational reasons, the other 3 patients with oral or dental POEs did not see a stomatologist during their hospital stays but had seen their dentists within the previous 3 months.
Dental procedures to treat POEs were undertaken during 24 patients’ stays in our hospital. All other patients were given instructions on dental procedures to be performed.
Gastrointestinal POEs were identified in 56 patients. Colonic polyps were present in 46% of these patients (Table 7). Colorectal adenocarcinoma was diagnosed in 14% of the patients. Streptococcus bovis group and Enterococcus faecalis were responsible for 50% and 29% of cases of IE with gastrointestinal POEs, respectively (Table 8).
Urinary POEs were acute pyelonephritis (n = 4), benign prostatic hypertrophy with acute urine retention (n = 1), transurethral resection of the prostate (n = 1), prostate needle biopsy (n = 1), transurethral resection of bladder cancer (n = 1), and urinary self-probing because of chronic urethral stenosis (S. bovis group, n = 2; Enterococcus, n = 1; Streptococcus agalactiae, n = 1; Escherichia coli, n = 1).
One female patient had an infection (S. agalactiae) of an aseptic necrobiosis of a uterine fibroma. The POE was ENT in 5 patients: pansinusitis with cerebral abscesses (Streptococcus pneumoniae, n = 1; Haemophilus parainfluenzae, n = 1), tonsil phlegmon (S. pneumoniae), recurrence of a laryngeal epidermoid carcinoma (S. agalactiae), and repetitive epistaxis with iterative plugging and cauterization of a nasal polyp (S. aureus). Three patients had pneumonia, and their blood cultures grew S. pneumoniae.
Among 82 episodes with nonidentified POEs, the microorganism habitat was cutaneous in 49%, oral or dental in 22%, and gastrointestinal in 22% (Table 9).
Potential POE of a new IE
Potential POEs for future IE episodes were as follows:
• continuation of intravenous drug use in 21 patients;
• cutaneous disease in 2 patients: Klippel-Trenaunay syndrome with varicose ulcer and psoriasis with scratching lesions;
• oral or dental infective foci in 66 of 125 patients (53%) who underwent stomatologic examinations: dental infectious focus in 41, radiological dental infectious focus (cyst, granuloma) without clinical lesion in 9, endodontal and periodontal disease in 11, and periodontal disease in 5;
• colonic lesions (polyps, diverticulosis, adenocarcinoma) in 32 of 80 patients (40%) who underwent colonoscopy because they were ≥50 years of age or had familial histories of colonic polyposis: polyps in 13 patients, sigmoid diverticulosis in 15 patients, sigmoid diverticulosis with polyps in 2 patients, diffuse angiodysplasia in 1 patient, and colorectal adenocarcinoma in 1 patient;
• urinary lesions in 11 of 52 patients (21%) who underwent urinary examinations: prostate cancer in 3 patients, benign prostatic hypertrophy with urine retention in 2 patients, urethral stenosis in 2 patients, pyelonephritis in 1 patient, cystinuria with repetitive renal lithiasis in 1 patient, post-radiotherapy bladder in 1 patient, and extrinsic urethral compression by colon cancer in 1 patients (no gynecologic lesions were found in the 16 women >79 years of age who underwent gynecologic examinations); and
• ENT lesions (sinusitis, otomastoiditis, and so on) in 6 of 180 examinations.
It seems obvious that the POE in a patient with IE should be searched for and eradicated, ideally during the initial stay, while the patient is receiving antibiotics. Many physicians probably look for and treat the POEs in their patients with IE. Yet there is no recommendation about the POE in recent guidelines on IE (3,4), and there is almost never information on the POE in reports of large series of IE.
At our institution, where the POE of IE is systematically searched for, the POEs of the current IE episodes were found in as many as three-quarters of patients. We consider this very good performance and an a posteriori justification of the systematic search for IE POE.
However, one might argue that the proportion of patients (up to a quarter) for whom POEs were not found is too high. The microorganisms were known in almost all (81 of 82) episodes of IE for which the POEs were not found (Table 9), and this may help uncover a possible POE, even if it is not indubitably identified.
The most frequent POE was cutaneous (40% of identified POEs). It was mainly (62%) associated with health care and with intravenous drug use. The most frequent microorganisms were staphylococci, which were identified in 78% of episodes of IE with cutaneous POEs, as expected from their ecology (Table 1), S. aureus in 55%, and coagulase-negative staphylococci in 23%.
The second most frequent POE was oral or dental (29%). Among oral or dental POEs, a dental infectious focus was much more often involved (59% of oral or dental POEs) than dental procedures (12%). Periodontal disease was involved in 28%. The most frequent microorganisms were oral (viridans) streptococci (69%), then HACCEK bacteria (Haemophilus spp., Aggregatibacter [Actinobacillus] actinomycetemcomitans, Capnocytophaga spp., Cardiobacterium hominis, Eikenella corrodens, and Kingella kingae) (10%). The habitat of viridans streptococci is dental plaque, oral mucosa, and the oropharynx. Their POEs are dental and periodontal disease (Table 1) (7). The oropharynx is the habitat of HACCEK organisms. Their POEs are buccal and dental infections and dental procedures (Table 1) (8).
The third most frequent POE was gastrointestinal (23%). Colonic polyps were found in almost one-half of the patients and colorectal adenocarcinoma in 14%. As may be expected, the most frequent responsible microorganisms were S. bovis group (S. gallolyticus) (50%) and E. faecalis (29%). The habitat of the S. bovis group is the gastrointestinal tract, and its POEs are colorectal adenoma and adenocarcinoma. The habitat of enterococci is the gastrointestinal and genitourinary tracts, and its POEs are the biliary tree and gastrointestinal or urinary tract infections (9) (Table 1).
Concordance between POE and microorganisms (i.e., intravenous drug use and S. aureus, dental infection and viridans streptococci, colonic polyps and S. bovis group) was excellent. But the POE should not be only presumed because of the microorganism. It should be looked for and treated, if needed.
Potential POEs for additional IE episodes were obvious in drug users continuing drug use and in some patients with chronic cutaneous lesions. The performance of a systematic search for potential POE was low for the ENT region and the genitourinary tract. Performance was good regarding the search for oral or dental or colonic potential POEs, which were found in 53% and 40% of patients, respectively. We limited systematic colonoscopy to patients who had familial histories of colonic polyposis or were ≥50 years of age, because the incidence of colorectal cancer increases in patients aged ≥50 years (10).
Our study showed that with a systematic approach to source identification, the POE can often be identified, and in a substantial proportion of these patients, risk modification can be attempted. This topic is of clinical importance, as it relates to our understanding of the sources of infection in patients with IE and also influences management of patients (e.g., ordering colonoscopy in a patient with S. bovis group IE, recommending better maintenance of oral hygiene).
The present study was performed at a single center; thus, the results may not be applicable to other areas of the world. A POE can be established with certainty only when the microorganism responsible for IE is also identified at the site of the POE and is genetically the same. Thus, the POEs in our study were presumed, not definite.
A systematic search for the POEs of IE was successful in as many as 74% of patients. Systematically searching for potential oral or dental, gastrointestinal, or genitourinary POEs of new IE episodes was also successful in many patients.
We would advise the systematic performance of a stomatologic examination in patients with IE and performance of colonoscopy in patients ≥50 years of age or at high risk for colorectal cancer. A flowchart for the identification and treatment of POEs is shown in the Central Illustration.
COMPETENCY IN PATIENT CARE AND PROCEDURAL SKILLS: A systematic search can identify the source of bacteremia in three-quarters of patients with IE, and in more than one-third of cases, additional potential POEs can be found that pose a risk for future infections.
TRANSLATIONAL OUTLOOK: Additional studies could be conducted to confirm these findings and assess the efficacy of eradicating potential portals of bacterial entry for the prevention of recurrent endocarditis.
For a reference list of the studies cited in Table 1, please see the online version of this article.
This study was supported financially by a research grant from Fédération Française de Cardiologie. The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- ear, nose, and throat
- infective endocarditis
- portal of entry
- Received July 27, 2015.
- Revision received October 5, 2015.
- Accepted October 13, 2015.
- American College of Cardiology Foundation
- Selton-Suty C.,
- Célard M.,
- Le Moing V.,
- et al.,
- for the AEPEI Study Group
- Habib G.,
- Hoen B.,
- Tornos P.,
- et al.
- Nishimura R.A.,
- Otto C.M.,
- Bonow R.O.,
- et al.
- Li J.S.,
- Sexton D.J.,
- Mick N.,
- et al.
- Sinner S.W.,
- Tunkel A.R.
- Steinberg J.P.,
- Burd E.M.
- Arias C.A.,
- Murray B.E.