Impact of non-conventional rescue therapies on quantification of cardiac agents following severe poisoning: Analytical considerations

Abstract

To assess whether a TurboFlow-LC-MS/MS (TF-LC-MS/MS) method that quantifies calcium channel blockers (CCB) and beta-blockers (BB) in plasma is suitable for monitoring severely-poisoned patients when non-conventional rescue therapies, i.e. lipid emulsions (LE) and Molecular Adsorbent Recirculating System (MARS ® ) are instituted. CCB amlodipine (A), nicardipine (Nic), nifedipine (Nif) and BB nebivolol (Neb) were quantified in plasma with or without lipid emulsion (PLE) and in dialysates (D) by TF-LC-MS/MS using a TLX-1 System interfaced with a TSQ Quantum Ultra ® (ThermoFisher). Impact of intravenous LE was studied with blank plasma spiked with Intralipid 20% (final concentrations between 5 to 30%). The albumin-containing dialysate that circulates through the MARS ® was replicated with an Ydralbum ® solution diluted in physiologic serum to obtain final concentrations of 0, 5, 10 and 15% of human albumin. All samples were analyzed by TF-LC-MS/MS after protein precipitation with acetonitrile containing deuterated internal standards (IS). Calibration curves were established by plotting the area of each analyte with dedicated deuterated IS in plasma samples. Impact of LE and MARS ® was studied with Quality Controls (QCs) spiked with analytes at low and high level concentrations (5 and 150 ng/mL) in PLE and in D, respectively. The method was finally applied to analyze samples obtained from two patients severely poisoned to amlodipine and treated with MARS ® and LE. Calibrations were linear from 2.5 to 200 ng/mL in plasma using 1/X 2 weighted least square regression. When plasma was spiked with LE, intra-day precisions were less than 4, 3, 4, and 8% (expressed as relative standard deviation) for A, Nic, Nif and Neb, respectively, irrespective of the LE final concentration. Intra-day accuracies were not statistically different in PLE compared to blank plasma. QCs prepared artificially with human albumin were confronted to the calibrations established in plasma. Precisions were less than 7% for all analytes, and accuracies did not exceed ±15% as requested by the European Medicine Agency (EMA). Intensities of absolute areas (AA) measured by TF-LC-MS/MS for each analyte and IS were similar in PLE, D and blank plasma. Concentrations of amlodipine were determined in plasma and dialysates sampled from 2 patients treated with MARS ® and ILE. Peak amlodipine concentrations were measured at 664 and 708 ng/mL (therapeutic range: 1–25), respectively. For patient 1, 30% of circulating amlodipine was trapped in LE, and for both patients, 2% of the toxicant was extracted by the MARS ® membrane. Time course levels of amlodipine were followed in plasma until patient's 1 death and patient's 2 full recovery. Massive overdoses with CCB and BB are often responsible for fatal refractory cardiovascular failure. In adjunction with symptomatic pharmacological agents, efficacy of extracorporeal life support like MARS ® and trapping strategy with IL is currently being assessed. Thus, concentrations of toxicants should be monitored in plasma but also in hyperlipemic plasma due to IV LE and in dialysates resulting from the extracorporeal circuitry of MARS ® . Hence, the analytical method must be validated for all the involved matrices. Irrespective of LE rates in plasma and albumin concentrations in physiologic serum, all QCs fulfilled acceptability criteria for precision and accuracy according to EMA. Furthermore, similar intensities of AA were found in blank plasma, PLE and D. Thus, combination of protein precipitation, TF-LC-MS/MS and use of deuterated IS contributes to a polyvalent and robust analytical method. In conclusion, our method initially validated to monitor A, Nic, Nif and/or Neb in plasma was also applicable to patients severely poisoned and treated with non-conventional therapies such as LE and MARS ® .