Simultaneous determination of 8 beta-lactams and linezolid by an ultra-performance liquid chromatography method with UV detection and cross-validation with a commercial immunoassay for the quantification of linezolid.

Linezolid and beta-lactams are anti-infective drugs frequently used in intensive care unit patients. Critical illness could induce alterations of pharmacokinetic parameters due to changes in the distribution, the metabolism and the elimination process. Therapeutic drug monitoring (TDM) is therefore recommended to prevent mainly under-dosing of beta-lactams or hematological and neurological toxicities of linezolid. In Multi-or Extensively-Drugs Resistant-Tuberculosis Bacteria, the regimen could include linezolid with meropenem and amoxicillin/clavulanate justifying the development of a method allowing their simultaneous quantification. The aim of this work was to develop an in-house ultra-performance liquid chromatography method with UV detection (UHPLC-PDA) allowing the simultaneous determination of 8 beta-lactams (amoxicillin, aztreonam, cefepime, ceftazidime, ceftriaxone, cefuroxime, meropenem and piperacillin) and linezolid and to cross-validate the linezolid quantification with a new commercial immunoassay (ARK kit) tested on a Cobas analyzer. The main advantages of the immunoassay are a 24/24 h random access assay which is fully automated and results provided within 2 h. The interference due to potential co-administrated drugs was evaluated on both methods. The preanalytical factors (type of matrix, stability) for linezolid were also investigated. The influence of hemolysis, icteria or lipemia on the spectroscopic detection of the immunoassay was assessed. The analytical performances were evaluated using the accuracy profiles approach with acceptance limits fixed at ±30%. Seventy patient samples were measured using both methods. No cross-reaction with the tested anti-infective drugs as well as no influence of hemolysis, lipemia, icteria were observed. The linezolid concentration could be measured on heparinized plasma or serum without a significant difference and remained stable for at least 72h at 4°C.The UHPLC-PDA method performed well in the analytical range investigated (0.25-50 mg/L for meropenem, 0.75-50 mg/L for linezolid and 1-200 mg/L for other beta-lactams) with an intermediate precision and a relative bias below 7.6 and 7.7%, respectively. The analytical range of the immunoassay was narrower, from 0.85 to 18.5 mg/L. The precision and relative bias were lower than 8.1% and 4.2%, respectively. Results obtained on clinical samples showed an acceptable difference between methods with a mean bias of -1.8% [95% confidence interval: -5.2% - 1.6%]. To conclude, both methods showed acceptable performance to perform TDM of linezolid considering the therapeutic through target of 2-8 mg/L. The choice of the method should be made according to the degree of emergency of the response required and the field of application justifying or not the simultaneous quantification of beta-lactams and linezolid.

Investigation of unbound colistin A and B in clinical samples using a mass spectrometry method.

Colistin, used as a last-resort drug, has a narrow therapeutic range that justifies therapeutic drug monitoring. Few data are available in the literature regarding the in vivo unbound fraction of colistin. The objectives of this study were to develop a method to isolate unbound colistin in clinical samples by ultrafiltration and to quantify it. The association between unbound colistin and biological parameters (total protein, albumin, alpha-1-acid glycoprotein and creatinine) was investigated. The measured ranges were 0.036-7.160 mg/L for colistin A and 0.064-9.630 mg/L for colistin B. The process of isolation and determination of unbound colistin was applied to clinical samples (n = 30) within 40 min and no non-specific binding was observed during the ultracentrifugation step. The median unbound fractions of colistin measured were 34.3% (12.8-51.0%) and 53.4% (27.0-77.8%) for colistin A and B, respectively. High interindividual biological variation of binding was observed for colistin A and B that was not explained by the biochemical parameters studied. The method developed could be useful to improve outcomes for patients.