Forty-four infants, children and adolescents with endocarditis were observed at the Paediatric Clinic of the Dfisseldorf University from 1970 to 1981. The mean follow-up was more than 5 years. There was a high incidence under the age of 2 years, but only two of the seven infants involved proved to have genuine bacterial endocarditis (BE) as was the case with all the older patients. One infant did present symptoms of typical BE; the others, however, had sepsis that led to death in all but one case. The constant finding at autopsy was a proliferative valve lesion. An endocardial ulceration was never observed, although there were always abscesses in different organs and once a small necrotic area in the myocardium of the papillary muscle. The only survivor with sepsis subsequently presented permanent mitral regurgitation.
In the group of 37 children and adolescents, 33 had an underlying congenital heart disease (CHD) confirmed by catheterisation, one a rheumatic involvement of the aortic and mitral valves. Half of the patients with CHD had cyanosis, eight times due to a Tetralogy of Fallot; among the acyanotic CHD coarctation was very common. Streptococcus viridans, still the most common pathogenic microorganism, was found in 24 cases (65%). The cultures remained negative in five instances (130/0). The course of the disease was almost always subacute. In one-third of the cases (n = 12) the BE caused valve lesions, four of which (three aortic and one mitral) later required surgical repair. Two children suffered a relapse of the infection immediately following the termination of antibiotic treatment. The BE was the ultimate cause of death eight times, i.e. for almost every fifth children.
In conclusion, BE in children still appears to be caused mainly by Streptococcus viridans, as was the case prior to the antibiotic era; these findings contrast with recent observations generally made in adults and occasionally in children. In infants, however, the septic aspect of the disease prevails. Moreover the endocardial lesions are sufficiently different from the ulcerative classical type of BE to be considered as a different entity, probably a secondary involvement, which does not necessarily influence the course of the disease.