Invasion by organisms of the endothelium can involve the valve leaflets, the chordae tendineae, the endocardium, the valve rings, and the great vessels in aortic coarctation or patent ductus arteriosus. The clinical signs and symptoms are variable and often non-specific, and the vegetations are not always readily apparent on imaging. The most widely accepted diagnostic criteria were those of von Reyn. etaL (.1981.), which included definite, probable, possible, and rejected categories
(Table 1). 0.yr§ck...eL§L (1994) observed that some patients with a clinical diagnosis of endocarditis failed to satisfy these criteria, particularly when the diagnosis was made by echocardiography or acute endocarditis occurred in the absence of previous cardiac disease. A revised set of criteria were produced following evaluation of 405 episodes of endocarditis (Table...?). In 69 cases of endocarditis which were proved pathologically, the sensitivity of the new criteria (excluding pathology) was 80 per cent compared with 51 per cent for the von Reyn criteria.
Table 1 Criteria for diagnosis of infective endocarditis
Table 2 Revised criteria for diagnosis of infective endocarditis
Acute endocarditis is caused by Staphylococcus aureus, Streptococcus pyogenes, and Streptococcus pneumoniae. It affects normal heart valves and follows a short aggressive course with high fever, leukocytosis, and death in a few weeks if untreated. 'Subacute' disease develops on already damaged valves, follows a prolonged indolent course with low-grade fever and night sweats, and is caused by a-hemolytic streptococci. Although responsible for a third of all cases of endocarditis, infection of valvular prostheses often fails to demonstrate the diagnostic clinical features. Division into early (operative) and late cases is arbitrary, and may be taken at 60 days, 4 months, or 1 year.
Blood cultures must be collected before starting antibiotic treatment; only three sets are necessary, taken 1 h apart preferably from different sites. Bacteremia is low grade but continuous, and blood cultures will be positive in 95 per cent of cases. In streptococcal disease, 96 per cent of the first blood cultures are positive, as are 98 per cent of one of the first two. For other causes, 86 per cent of first cultures are positive, as are almost all of one of the first two. If cultures are negative, repeated specimens are rarely helpful unless the patient was given antibiotic therapy in the previous 2 weeks (only 65 per cent of first cultures are positive). At least half of cases of fungal endocarditis are culture negative. A normochromic normocytic anemia is usual, but a leukocytosis is common only in acute endocarditis. Both the erythrocyte sedimentation rate and the C-reactive protein are markedly elevated, and rheumatoid factor is detectable in half the cases of subacute disease. Urinalysis will often reveal microscopic hematuria and proteinuria.
Echocardiography is important in confirming the presence of vegetations, underlying cardiac lesions, and any complications, but can only confirm and not exclude the diagnosis. Planar vegetations on a prosthetic or myxomatous valve are easily missed and the skill of the operator is a major factor in its success ( Fig 2).
Transthoracic echocardiography may fail to demonstrate ring abscesses. Although prolongation of the PR interval on the electrocardiogram can suggest the presence of an abscess in the conducting system, transesophageal echocardiography is often the most reliable means of diagnosis. Vegetations are demonstrable by transthoracic imaging in 70 per cent of clinically suspected cases compared with 80 to 100 per cent when the transesophageal approach is used. An increase in diagnostic accuracy from 46 to 92 per cent is obtained for abscesses. Diagnosis of vegetations on prosthetic valves is difficult by conventional echocardiography (20-30 per cent) but rises to 90 per cent accuracy when a transesophageal method is used. Obesity and emphysema can reduce the quality of a transthoracic image but not of a transesophageal image. Echocardiography can be used to monitor the size of vegetations and hence the outcome of treatment. Therefore it is important in establishing the need for surgery.
Fig. 2 A xenograft showing small vegetations that could easily be missed on echocardiography.
Was this article helpful?