Comparison of penicillin and vancomycin, individually and in combination with gentamicin and amikacin, in the treatment of experimental endocarditis induced by nutritionally variant streptococci.

Six different antibiotic treatment regimens were compared for efficacy in rabbits with endocarditis induced by inoculation with a nutritionally variant strain of streptococcus. Seven untreated animals, sacrificed at day 11, had vegetations containing 8.89 +/- 1.35 log10 CFU/g, none of which was sterile. The vegetations from the rabbits in all treated groups had bacterial titers significantly lower than those of the controls (P less than 0.001). Vegetations from penicillin-treated animals averaged 5.14 +/- 1.00 log CFU/g, and no vegetations were sterile. Treatment with penicillin plus gentamicin or amikacin was more effective than treatment with penicillin alone, resulting in 3.99 +/- 0.94 log CFU/g of vegetation and sterile lesions in 5 of 12 animals. Treatment with vancomycin alone was as least as efficient as that with penicillin plus an aminoglycoside, resulting in an average of 3.33 +/- 0.96 log CFU/g of vegetation and sterile lesions in five of eight animals. Treatment with vancomycin plus an aminoglycoside was not superior to treatment with vancomycin alone, resulting in an average of 3.68 +/- 1.37 log CFU/g of vegetation and sterile lesions in 8 of 13 animals. These in vivo results correlated poorly with the in vitro susceptibility of the strain to the various antibiotics, as measured by the time-kill method. These results support the current practice of using vancomycin as alternative therapy when a penicillin-aminoglycoside combination is ineffective or contraindicated in patients with endocarditis caused by nutritionally variant streptococci.