Serum standards for the bioassay of aminoglycosides in cerebrospinal fluid.

Abstract

SUMMARY Four diluents were compared as reference standards for the assay of gentamicin in cerebrospinal fluid (CSF): human CSF, human serum, distilled water, and 150 mmol NaCl/4 5 mmol CaCl2. Standards prepared in pooled human serum were the best alternative to CSF for the assay of gentamicin and were also useful for the assay of tobramycin, netilmicin, amikacin, and sisomicin. The pH (6.0-9 8) of CSF did not alter the results of the assay.The accurate determination of microbiologically active gentamicin in cerebrospinal fluid (CSF) is often critical in the management of patients with Gramnegative bacterial meningitis. Several clinical studies have reported CSF drug concentrations; however, standards for the assay were not specified (Kaiser and McGee, 1971;McCracken, 1972;McCracken and Mize, 1976). Although CSF would seem to be the best diluent for standards, it is often not available. A recent report described an alternative solution for CSF standards (Deacon, 1976a). We compared this artificial diluent (150 mmol NaCI/4 5 mmol CaCl2) with distilled water, human CSF, and pooled human serum. Since cerebrospinal fluid pH may vary markedly, the effect of pH was also examined. Unlike Deacon's report, our for nutrient agar as the former produced clearer zones of inhibition. A 30-mm square agar plate was flooded with a 1/100 dilution of an overnight growth of the test organism in nutrient broth. The excess inoculum was removed by pipette and the plate was dried by incubation for approximately 45 minutes at 370C with the lid ajar. Wells (2 mm diameter) were cut in the seeded plate and numbered so that samples were distributed randomly on the plate. Gentamicin standards were prepared in concentrations of 20, 10, 5, 2-5, and 125 ,ug/ml in the various test fluids. Four replications of each sample were performed for each of three tests, and zones of inhibition (mm) were recorded and averaged after 18-20 hours' incubation. Standard curves were plotted on semilog paper. The same method was employed when four other aminoglycosides were assayed to compare CSF and serum as diluents.CSF was obtained from hospital patients without leptomeningitis or ventriculitis. CSF 1 was from a hydrocephalic patient; CSF 2, 3, and 4 were CSF pools obtained from diagnostic lumbar puncture in febrile infants without meningitis. Human serum was collected from adult volunteers. Glucose, protein, and cell count of the CSF were determined by standard methods. CSF 1, 2, and 3 and different pools of human serum were used for each of three experiments involving gentamicin (Table 1). The NaCl/CaCI2 solution was prepared as recommended by Deacon (1976a).The effect of pH was examined with CSF 4. Using 2% HCO and 4%Y. NaOH the samples were adjusted to pH 6, 7, 8-2, 9, and 9-8. Gentamicin was added to each sample to yield a final concentration of 5 ,ug/ml.