No abstract

Pathologic and bacteriologic studies were carried out in all 24 dogs subjected to arteriovenous anastomoses and bacteremia. The single etiologic agent, Streptococcus mitis, was stained and cultured from 10 of the 11 hearts with active endocarditis (one animal had a healed tricuspid valvulitis) and from 16 of 17 animals with patent fistulas and active caval vegetations. The latter lesions occurred shortly after the initiation of injections and served as an important source of bacteria to the blood stream and heart valves. The endocarditis resembled that found in human cases with destructive and reparative phenomena coexisting. Focal myocarditic lesions and cardiac failure were noted only in animals with endocarditis. The other secondary (embolic) lesions were similar to those seen in man with the exception that in the dog the lesions had a suppurative tendency. Acute diffuse proliferative glomerulonephritis was found in animals with endocarditis and in those with infected fistulas and bacteremia.

It is important in order to advance the increasingly successful attack on subacute bacterial endocarditis to be able to produce the disease experimentally and to treat it with new and improved therapeutic regimens. Since endocarditis caused by Streptococcus mitis has not been consistently produced previously in the dog, a study was undertaken in which cardiovascular stress was created by the establishment of aortico-inferior vena caval and bilateral iliofemoral arteriovenous shunts. After a period of antibacterial treatment and of stabilization, daily intravenous injections of broth cultures of S. mitis were begun. In 50 per cent of 24 animals bacterial endocarditis developed, simulating that found in clinical cases at necropsy. These animals had widely patent fistulas and 11 showed evidences of congestive heart failure. Of the remaining 12 animals, five with patent fistulas died of the effects of multiple infected pulmonary emboli. Seven others whose shunts had closed were killed and found to have no significant lesions other than healed caval infections.T HE evaluation of new treatment regimens for bacterial endocarditis would be greatly facilitated by the development of a simple method of producing the disease in animals. In the past most of the methods employed in the production of experimental endocarditis involved one major approach; that is, the intravenous injection of organisms into the animal, either with or without preceding injury to one or several of the heart valves. More recently, attempts have been made to increase the cardiac work load by various procedures followed by the injection of organisms. Although positive results were sometimes obtained by each of these procedures, Streptococcus mitis endocarditis in the dog has not been consistently produced.