Measurement of antimicrobial agents in cerebrospinal fluid using the Abbott TDx analyzer.

Occasionally, requests are made by our physicians for the measurement of gentamicin, tobramycin, or vancomycin in cerebrospinal fluid (CSF) specimens during the course of treating patients for bacterial meningitis. We evaluated CSF as a specimen type for the measurement of amikacin, gentamicin, tobramycin, and vancomycin on the Abbott TDx analyzer. Coefficients of variation for CSF spiked with these antimicrobial agents ranged from 0.8% to 6.5% for intra-assay values and from 2.1% to 2.3% for inter-assay values. Serum and CSF specimens were spiked at various levels with equal amounts of the antibiotics. Correlation coefficients for serum vs. CSF for these agents were 0.999. Recoveries ranged from 86% to 134%. Sensitivity for these assays is about fourfold better for CSF than for serum. CSF appears to be an acceptable specimen type for the measurement of these antibiotics using the Abbott TDx analyzer.

Concentrations of ceftazidime, tobramycin and ampicillin in the cerebrospinal fluid of newborn infants.

Thirty-five neonates with suspected septicaemia were randomized to treatment with tobramycin or ceftazidime, both in combination with ampicillin. Concentrations of antibiotics in the CSF were measured 1 h after the third, fourth or fifth injection. In 13 of 17 neonates tobramycin CSF concentrations were below 0.5 mg/l. Ceftazidime CSF concentrations ranged from 2.5 to 17 mg/l, which should be sufficient for treatment of infections with group B streptococci and most aerobic gram-negative bacilli but not all strains of Staphylococcus aureus. Ampicillin CSF concentrations ranged from 1 to 80 mg/l, which should be sufficient for treatment of meningitis caused by enterococci and Listeria monocytogenes, the most important neonatal pathogens not covered by ceftazidime.

Effect of hyperbaric oxygen on the blood: cerebrospinal fluid transfer of tobramycin.

The blood:cerebrospinal fluid (CSF) transfer of the aminoglycoside antibiotic, tobramycin (TOB) was assessed in rabbits. The CSF:blood ratio of TOB, 90 min after a subcutaneous injection, was approximately 1:100 as measured by an agar disc diffusion bioassay. Hyperbaric oxygen therapy at a pressure of 3 ATA caused a slight, non-significant, increase in this ratio, while high dose carbon dioxide, a treatment known to damage the blood:brain barrier, more than doubled this ratio. Thus, hyperbaric oxygen has no significant effect on CSF concentration of TOB in rabbits with intact meninges.

Effect of mannitol infusions into the internal carotid artery on entry of two antibiotics into the cerebrospinal fluid and brains of normal rabbits.

The rapid infusion of hyperosmolar solutions into the internal carotid artery transiently disrupts the integrity of the blood-brain barrier, permitting entry of substances normally excluded from the nervous system. This study examined the effect of such an infusion on the penetration of vancomycin and tobramycin into the cerebrospinal fluid and brains of normal rabbits. Groups of five to seven animals were assigned either to control groups, receiving vancomycin or tobramycin alone, or to experimental groups, receiving these antibiotics with 40-s intracarotid infusions of 8.0 ml of 25% mannitol. Serum, cerebrospinal fluid, and brain specimens were obtained for assay 135 min after dosage. Concentrations of these antibiotics in serum were comparable in control and experimental groups; concentrations of the antibiotics in cerebrospinal fluid were significantly higher in the experimental groups. Mean (+/- standard deviation) concentrations of the antibiotics in cerebrospinal fluid were, respectively, less than 0.5 and 4.1 +/- 4.6 micrograms of vancomycin per ml and 0.6 +/- 0.5 and 3.4 +/- 2.4 micrograms of tobramycin per ml in the control and experimental groups. The concentrations of these antibiotics in brains were also higher in the experimental groups. This trend achieved statistical significance in the tobramycin studies, where control animals uniformly had tobramycin concentrations in brains of less than 0.1 microgram/g and experimental group animals had a mean concentration of 0.4 microgram/g. Additional studies with tobramycin indicated that intracarotid antibiotic administration immediately after the mannitol infusion further augmented antibiotic concentrations in brains.

Tobramycin levels in cerebrospinal fluid of patients with slightly and severely impaired blood-cerebrospinal barrier.

Concentrations of tobramycin in cerebrospinal fluid (CSF) and serum of neurosurgical and of meningitis patients were determined following systemic application of 80-mg doses. Tobramycin was detectable only in 8 tests out of 17 in CSF of 11 neurosurgical patients. The highest drug amount in CSF in these tests was 0.47 microgram/ml. In CSF of 7 meningitis patients with CSF cell counts ranging from 173/3 to 20,500/3 cells, tobramycin was detectable in all 11 determinations. However, the tobramycin level was above 1.0 microgram/ml in only one CSF sample. These results also confirm the poor penetrability of aminoglycosides through the blood-CSF barrier for tobramycin. In proven gram-negative bacillary meningitis and tobramycin treatment, additional intraventricular aminoglycoside application is necessary to obtain effective levels in CSF. An exception may be treatment of neonatal meningitis.

Comparison of four aminoglycoside antibiotics in the therapy of experimental E. coli meningitis.

Tobramycin, sisomicin, and amikacin were compared with gentamicin in the therapy of experimental E. coli meningitis in rabbits. Meningitis was produced in 40 animals by intracisternal injection of 10(5) E. coli. Three dosages of each antibiotic were administered intravenously over 8 hours. Serum and CSF samples were obtained at 0,2,4,6, and 8 hours for determination of aminoglycoside concentrations and CSF bacteria counts. The four aminoglycosides demonstrated comparable penetration into the CSF. The mean percent penetration (CSF conc./serum conc. X 100%) with the three dosages was 10 to 50% for gentamicin, 8 to 23% for tobramycin, 6 to 16% for sisomicin, and 11 to 23% for amikacin. This variation in penetration reflected individual differences in each dosage group and the increase in percent penetration that was observed during therapy. Sisomicin and gentamicin were consistently bactericidal in vivo. Mean CSF bacterial titers gentamicin were consistently bactericidal in vivo. Mean CSF bacterial titers decreased 3.07 logs in rabbits treated with sisomicin and 2.44 logs in animals treated with gentamicin. Even through CSF concentrations were comparable, the group treated with tobramycin demonstrated only a 0.64 log decrease and the amikacin group had a 0.45 log increase in mean CSF titers. The bactericidal effect of sisomicin appeared to be more rapid than that of gentamicin. During the first 2 hours of therapy CSF titers declined 1.02 logs in animals treated with sisomicin as compared to 0.37 log in animals receiving gentamicin even though gentamicin concentrations were higher (mean 7.4 vs. 4.1 gm./ml.). Sisomicin also demonstrated greater bactericidal activity than the other aminoglycosides in normal CSF in vitro. The results of this study suggest that sisomicin may be of value in the therapy of E. coli meningitis.

Aminoglycoside therapy of gram-negative bacillary meningitis.

The distribution of aminoglycosides in the cerebrospinal fluid (CSF) space was examined after intralumbar, intraventricular, and systemic administration during seven episodes of gram-negative bacillary meningitis. Six episodes were associated with culture proved ventriculitis. Parenteral therapy with gentamicin or tobramycin produced low concentrations of aminoglycoside (less than 1.0 mug/ml) in the lumbar, ventricular, and cisternal CSF. Administration of 5 to 10 mg of aminoglycoside into the lumbar intrathecal space resulted in 27-81 mug/ml in the lumbar CSF, but 0-2.1 mug/ml in the ventricular CSF. In contrast, aminoglycoside administered into the cerebral ventricles produced concentrations in the lumbar CSF of 11.5-27.5 mug/ml and ventricular CSF of 12.8-40 mug/ml. All six episodes treated via the ventricular route resulted in a bacteriologic cure. Intraventricular administration of aminoglycosides offers a reliable means of achieving high aminoglycoside concentrations throughout the subarachnoid space.