Comparative assessment of in vitro inactivation of gentamicin in the presence of carbenicillin by three different gentamicin assay methods.

Inactivation of gentamicin (G) is known to occur secondarily to the formation of complexes with certain beta-lactam antibiotics. However, aminoglycosides in the presence of aminoglycoside-beta-lactam complexes may not be recognized uniformly by all assay methods. We tested this hypothesis by using mixtures of G plus carbenicillin (C), with and without the addition of penicillinase, in pooled sera under several in vitro conditions: at 25 and 35 degrees C and at low and high C concentrations. Samples were assayed for G with the EMIT and TDx systems, and a microbiological assay was performed with a strain of Klebsiella pneumoniae resistant to C. In the presence of C (500 micrograms/ml) at 35 degrees C, the initial G concentration of 5 micrograms/ml decreased markedly over 48 h as assessed by all three assay methods. However, significantly greater degradation was noted when samples were measured by microbiological assay and TDx than by EMIT. Differences between assays were less marked when mixtures were studied at a lower temperature and with a lower G to C ratio (5 micrograms of G plus 100 micrograms of C per ml). The addition of penicillinase to the antibiotic mixtures prevented the degradation of G over time when measured by all three assay systems. We concluded (i) that EMIT measures higher serum concentrations of G than do TDx or microbiological assays when complexes of G and C are present and (ii) that the addition of penicillinase to serum samples containing C and G would be effective in preventing G degradation during prolonged (greater than 24-h) periods between the time of sampling and assay.