Differential response to benzylpenicillin in vivo of tolerant and non-tolerant variants of Streptococcus sanguis II.

A variant that was highly tolerant to benzylpenicillin was obtained from a non-tolerant clinical isolate of Streptococcus sanguis II by repeated exposure to penicillin. The rabbit model of endocarditis was used to investigate the efficacy of a high dose regimen of benzylpenicillin (250 mg/kg; peak serum concentration c. 25 mg/l) in the prophylaxis and treatment of endocarditis during challenge or infection with the non-tolerant parent strain or its tolerant variant. The two strains exhibited a similar capacity to initiate infection. A single dose of penicillin administered 0.5 h before bacterial challenge protected six of nine rabbits infected with the non-tolerant parent strain, but none of nine infected with the tolerant variant. Treatment of established infection with penicillin administered twice daily for four days cured eight of 13 (61%) rabbits infected with the non-tolerant parent strain, but only one of 14 (7%) rabbits infected with the tolerant variant. These results support the view that tolerance to penicillin has therapeutic implications.

Tolerance to penicillin in streptococci of viridans group.

Tolerance to penicillin was investigated in 40 isolates of optochin-resistant, alpha-haemolytic streptococci. Thirteen strains exhibited tolerance to penicillin (MBC:MIC ratio greater than or equal to 32) when stationary phase inocula were used, but only seven strains retained the tolerance phenotype in experiments with logarithmic phase inocula. There was a striking association between tolerance and Eagle's optimum dosage effect, particularly among strains that displayed tolerance in both the stationary and the logarithmic growth phases. Sequential viable counts on representative strains showed that reliance on the arbitrary criterion of bactericidal activity of 99.9% reduction of the original inoculum after 24 hours' exposure may occasionally lead to difficulties in the recognition of penicillin tolerance. In general, however, the 99.9% killing criterion provided a useful discriminator between strains that were rapidly killed by penicillin and those (tolerant strains) in which the bactericidal activity was much reduced.