Infective Endocarditis

Introduction

Although relatively uncommon, infective endocarditis ( IE ) is of great importance for several reasons:

• It has a high mortality rate, and almost always kills the patient if left untreated;
• It is often preventable, with appropriate prophylaxis; conversely failure to administer prophylaxis can result in clinical and medicolegal disasters;
• Early diagnosis and correct management are vital if the wellbeing and life of the patient is to be preserved;

Diagnosis of Infective Endocarditis

Thorough clinical assessment is vital. Apart from the above two sentinel findings, one should particularly look for:

• Skin petechiae (also sometimes on conjunctiva or other mucosal membranes), as well as subungual splinter haemorrhages;
• Pallor (anaemia is present in 70+ percent of cases);
• Haematuria (usually microscopic, in up to one half of all patients!);
• Evidence of emboli: these may be minor (to peripheral vessels), or major, with organ infarction. They occur in one to two fifths of all patients.

"Classical" features of IE such as clubbing, Osler's nodes (tender lesions on the pulps of the fingers), Roth's spots (fundal haemorrhages with whitish centers), and splenomegaly are relatively infrequent findings, and their absence means little. Subungual splinter haemorrhages are unfortunately rather common in normal individuals, where they are usually related to trauma - those seen with IE are said to be more proximally located and less obviously "traumatic". Vague musculoskeletal complaints are common - these include arthralgias in about a sixth of patients, and myalgias. As a consequence of severe valve involvement patients are often in heart failure. Another much-neglected finding is central nervous system disease, with mycotic aneurysms of the cerebral circulation a fairly frequent occurrence (often presenting when they leak), as well as embolic stroke.

The cornerstone of diagnosis is growing the offending organism on blood culture. This is vital. Adequate volumes of blood must be obtained (for example 3 x 20ml). There is little evidence that taking these cultures at different times or different sites is better than taking them at the same time from one site, but pickup is probably improved by taking samples from the patient while his/her temperature is on the rise! DO NOT start antibiotics before adequate cultures have been taken! One can even make a case for waiting for cuture results and then reculturing if no organism is found, provided the clinical condition of the patient justifies this!? Cultures should be taken by direct needle stab of a peripheral vein, and not from venous or arterial lines.

If cultures are negative:

• Some twit gave antibiotics;
• Insufficient or inadequate specimens were sent;
• Your laboratory is not up to scratch; or
• The organisms are slow-growing or fastidious (Brucella, Chlamydia, Rickettsia, fungi, etc, as well as the HACEK group of organisms).
• It's not infective endocarditis.

Echocardiography is invaluable in detecting the presence of vegetations on the valves, as well as providing evidence about the nature and severity of the valve lesions. As important as defining the valvular vegetations is examination for perivalvular infection. If there is a strong clinical suspicion of infective endocarditis, trans-thoracic echo cannot rule out IE: a trans-oesophageal echocardiogram should be obtained.

The Duke criteria have supplanted previous diagnostic criteria:

• Definite IE: Two major criteria, 1 major + 3 minor, or five minor criteria;
  OR if pathology or microbiology of the vegetations (or emboli, or intracardiac abscess) shows infective endocarditis.
• Possible IE: "Not definite but not rejected";
• Rejected IE:
  • Firm alternate diagnosis explaining clinical findings; OR
  • Resolution of manifestations with antibiotic therapy for 4 days or less; OR
  • no evidence of IE at surgery or autopsy.

Major Criteria are:

• positive blood culture for IE:
  typical IE micro-organisms from 2 separate cultures; OR
  one of the following organisms: S viridans / S bovis / HACEK / S aureus / enterococci, WITH NO primary focus and WITHOUT a "persistently positive blood culture";
• Evidence of endocardial involvement on echocardiogram, including an oscillating mass on or near the valve, abscess, new partial dehiscence of a prosthetic valve, or new valve regurgitation. Note that a "changing murmur" is NOT a criterion.

Minor criteria include:

• Predisposition (heart condition or IV drug use);
• Fever over 38.0 ^o C;
• Vascular phenomena: (major arterial emboli, septic pulmonary infarcts, mycotic aneurysm, intracranial haemorrhage, conjunctival haemorrhage, Janeway lesions);
• Immune phenomena: (glomerulonephritis, Osler's nodes, Roth spots, rheumatoid factor +ve);
• Microbiological evidence not meeting major criterion, unless these organisms "never" cause IE;
• Serological evidence of active infection with an organism consistent with IE;
• Echocardiogram, but not meeting major criteria;

Causative Organisms

These vary, depending on:

• The underlying valve problem: for example, patients who have recently undergone valve replacement present a different spectrum of infection;
• The patient: for example, IV drug abusers tend to get right-sided endocarditis with S aureus a common pathogen;
• The situation: ICU patients with indwelling lines will also get right-sided lesions; fungi will be more common with long ICU stays and multiple prior antibiotic use;

Common organisms include:

• Streptococci esp. "S. viridans";
• Staphylococcus aureus;
• Staphylococci: coagulase negative;
• Enterococci;

Prophylaxis of Infective Endocarditis

This is still a controversial topic, with many conflicting sets of recommendations. We can however extract a few general rules:

1. Although most infective endocarditis is NOT related to procedures performed by doctors/dentists, prophylaxis is said by some over 50% effective when properly administered. (But see Note 1). If not given to patients at substantial risk of endocarditis, or incorrectly administered, endocarditis may occur, and this may be difficult do defend medico-legally;
2. Dental procedures may be the most important risk factor. Bacteraemia occurs with periodontal work, scaling etc. (Even when you bite hard on your teeth). Maintenance of good dental hygiene in patients at risk is thus very important.
3. Administration of prophylactic antibiotics is not benign: the risk of adverse reactions and cost should be weighed up against the risk of endocarditis. Risk varies widely, depending on the patient.

• prosthetic valves;
• congenital cyanotic heart disease;
• surgical shunts (systemic to pulmonary) for correction of congenital heart disease;
• patients with a past history of IE.

Patients with valvular dysfunction, other congenital cardiac malformations, mitral valve prolapse WITH mitral regurgitation, and possibly those with hypertrophic cardiomyopathy present a lower risk, but should still generally receive prophylaxis.

Prophylaxis should be administered for:

• oral and upper respiratory procedures, including scaling, teeth cleaning, periodontal surgery, tonsillectomy, rigid bronchoscopy;
• gastrointestinal procedures, such as biliary surgery, oesophageal dilatation, colonoscopy;
• genitourinary procedures, especially prostatic surgery, and where the urine is infected. Use before vaginal delivery is unsubstantiated, but many would argue that it is needed in high-risk cases!

Antibiotics can be given either orally or parenterally. As with all prophylaxis, the antibiotic should be in the blood when the bug hits the blood. We therefore generally give these agents:

• Orally: one hour before the procedure;
• Parenterally: thirty minutes before the procedure;

Table VII in the article by Stamboulian & Carbone neatly summarises American Heart Association and European recommendations. Briefly:

Procedure	Usual Rx	Penicillin Allergy
Dental, Upper respiratory tract	amoxicillin 2g PO, OR ampicillin IV / IM	Replace ampicillin with ONE of: clindamycin 600mg PO / IV clarithromycin 500mg PO azithromycin 500mg PO ? roxithromycin 300mg PO
Genitourinary	ampicillin 2g IV / IM AND gentamicin 1.5mg/kg (maximum 120mg)	Replace ampicillin with: Vancomycin 1g IV 1 hour before procedure, OR teicoplanin 400mg IV. STILL give the gentamicin.
	In addition, some authorities recommend a a follow-up dose of 1g amoxicillin or ampicillin, SIX HOURS after the genitourinary procedure!

Paediatric doses are: ampicillin 50mg/kg, azithromycin/clarithromycin 15mg/kg, clindamycin 20mg/kg and teicoplanin 6mg/kg.

In other words, for dental and URT procedures, give ampicillin or amoxicillin, unless penicillin allergic. For genitourinary procedures, give ampicillin parenterally and add aminoglycoside. With penicillin allergy, give clindamycin or one of a variety of other agents for the dental work, or replace the penicillin with vancomycin for genitourinary work. The possible second dose of ampi/amoxicillin seems peculiar to us.

Note that the above are guidelines . Therapy may need to be tailored to the special patient. Physicians performing lower gastrointestinal procedures should probably follow the guidelines for genitourinary procedures, and many would argue that this holds good for upper GIT procedures as well.

Little appears to have been written about prophylaxis in patients who have already been in hospital for some length of time. This is a pity, as such patients will most likely be colonised by the local (often multiresistant) flora, so more aggressive prophylaxis might be indicated.

Why not use the URT regimen with upper GIT endoscopy, unless there is severe disease with probable colonisation by more aggressive organisms? When in doubt, best stick to the "rules". Several surveys have shown low compliance by doctors with established guidelines. Do NOT trust your patient to take the antibiotics 1 hour before the procedure if you prescribe an oral regimen. Make sure they take it! Patients should also ideally carry an "endocarditis risk" card! There is little/no evidence to support prophylaxis for insertion of tympanostomy tubes, endotracheal tube insertion, flexible bronchoscopy, cardiac catheterisation, urethral catheterisation in the absence of infection, therapeutic abortion, laparoscopy and sterilisation.

Further notes:

• Mani et al [Endoscopy 1997 29(2) 114-9pp] suggest that one adds IV metronidazole if performing endoscopy on neutropaenic patients. They use oral ciprofloxacin or parenteral gentamicin, or parenteral quinolone, or cephalosporin or ureidopenicillin for ERCPs.
• A worrying study from J Antimicrob Chemother [1996 37(4) 783-95] showed that blood cultures still showed S viridans in four fifths of patients who received correct doses of clindamycin or erythromycin prior to invasive oral procedures.
• Quinolones have borderline activity against many clinically important Gram +ve cocci, including S. viridans & S. aureus. See [Giamarellou H: Drugs 1995 49S2 58-66pp].
• An interesting debunking of "myths" associated with IE is found in [Wahl M: Arch Intern Med Jan 1994 154(2) 137-44pp].

References

Note 1

• www.med.umich.edu/pediatrics/ebm/cats/endocarditis.htm
• www.rcseng.ac.uk/fds/Documents/Patient%20Information%20Sheet.doc
• www.upenn.edu/ldi/issuebrief6_6.pdf

Older references

1. Drugs Nov 1997 54(5) 730-44pp
   Stamboulian D & Carbone E
   A fairly good recent article on recognition, management and prophylaxis. Much of the above is based on this article.
   
   
2. Br J Hosp Med Oct 1995 54(7) 341-7pp
   Prasad A & Fraser A
   Notes on British recommendations.
   
   
3. JAMA 1997 277(22) 1794-801pp
   Dajani A et al
   Contains the American Heart Association recommendations.
   
   
4. Eur Heart J 1995 16SB 126-31pp
   Leport C et al
   This is a European Consensus report from the International Society for Chemotherapy.
   
   
5. N Engl J Med Jan 1995 332(1) 38-44pp
   Durack D
   Yet another review on the prevention of IE, from the UK.
   
   
6. Lancet 1990 335 88-9
   Simmons N
   These are the older British recommendations. Fairly straightforward, easy to use, and things haven't really changed that much (although now Erythromycin is less popular).