Continuous vs. Intermittent Administration of Antimicrobial Agents: Tissue Penetration and Efficacy in Vivo

Abstract

Two experimental models, fibrin clot and pyelonephritis, were used to study the influence of continuous and intermittent administration of antimicrobial agents on penetration of drug into tissues and efficacy in vivo. In the fibrin-clot model, clots infected with Streptococcus faecalis were inserted subcutaneously into rabbits; in the pyelonephritis model, Escherichia coli was injected into the left kidneys of rats. In the experiments with rabbits, the bactericidal activity of the serum against S. faecalis was two to eight times greater after bolus injections of a combination of penicillin G and gentamicin than after continuous infusion of a similar dose of either drug alone. Synergism in vivo, as determined by counts of live bacteria in the clots, was demonstrated after bolus injections of these two drugs, but constant infusion of the combination was not synergistic. In the experiments with rats, two types of pyelonephritis were produced: a severe infection induced by direct inoculation of E. coli into the left kidney and a less severe infection caused by reflux of infected urine into the right ureter and right kidney. Treatment with gentamicin or ampicillin alone or in combination, or with the combination trimethoprim-sulfamethoxazole, was started 24 hr after inoculation. Injections of gentamicin (10 mg/kg) every 12 hr for seven days resulted in continuous levels of drug in the medulla that persisted above the minimal inhibitory concentration (MIC) for the infecting strain of E. coli throughout a period > 32 days. Eighty percent of the left and right kidneys and urine were found to be sterile on the 25th day after cessation of therapy. Ampicillin (100 mg/kg), which was administered every 6 hr for seven days, was found in high concentrations in the urine and papilla but had a very short half-life in the medulla. Ampicillin sterilized only 22o7 of the left kidneys, but the cure rate for the right kidneys was 75Wo. Even though seven days of treatment with the combination of trimethoprim (10 mg/kg every 12 hr) and sulfamethoxazole (50 mg/kg every 12 hr) induced peak drug concentrations in the medulla and papilla that were up to 500 times higher than the MIC, none of the left kidneys and only 33% of the right kidneys were sterile 25 days after cessation of therapy. Gentamicin (10 mg/kg every 8 hr) given for three days in association with ampicillin (100 mg/kg every 6 hr) for seven days was the most effective treatment. The greater effectiveness of this combination might have been the result of what we call anatomic synergy. Concentrations of drug in serum and urine were poor predictors of both the intrarenal distribution of drugs and the outcome of pyelonephritis. On the other hand, the level of drug in the renal tissue was not always a reliable indicator of efficacy in vivo.