Validation of a HPLC-MS/MS assay for the determination of total and unbound concentration of temocillin in human serum

Temocillin is a β-lactam antibiotic active against gram-negative bacteria, including β-lactamase producers. It is currently indicated for the treatment of septicemia, urinary tract infections, and lower respiratory tract infections. In the recent years the interest on this drug has been greatly enhanced, due to the increased incidence of infections caused by microorganisms producing extended-spectrum β-lactamases. This makes temocillin a strategic and interesting alternativeto carbapenems, especially with the current driving force to spare the use of these drugs.

Pharmacodynamic studies predict that therapeutic success with β-lactams depends on the time during which their unbound concentration in serum remains above the MIC of the offending organism. Dosage should thus be adapted to take into account not only the bacterial susceptibility, but also the rate of elimination of the drug. There is evidence of a wide interindividual variability in the temocillin serum levels so that standard dosing regimens may not be adequate in special patient populations such as intensive-care and end-stage renal disease patients. These patients may benefit from individualized dose regimens based on therapeutic drug monitoring.

Two validated assays (HPLC-UV) have been described in the literature to quantify temocillin. The first one was used for controlling purity and detecting related substances in temocillin batches, and the other one, for quantifying the drug in serum. The later method involves a laborious sample preparation with solid-phase extraction to eliminate possibly interfering substances andwas validated for the total concentration only. The present work aims at validating a new, rapid, sensitive and selective HPLC-MS/MS method for determination of both total and unbound concentrations of temocillin in serum.


The correlation coefficient (R2) for each calibration curve was >0.99, with a mean value of 0.993 and 0.995 for the total (n = 15) and unbound (n = 15) temocillin concentration, respectively. LOQ and LOD were found to be 1 and 0.10 mg/L (total concentration) and 0.50 and 0.05 mg/L (unbound concentration). As shown in this paper, the retention times (RT) were 3.14 and 3.18 min for temocillin and IS, respectively. For the total concentrations, inter-assay (3 days; n = 15) precision and accuracy ranged from6.48 to 8.88% and from2.26 to 11.64% respectively, while the intra-assay (3 days; n = 5) precision and accuracy ranged from 7.14 to 9.18% and from 6.86 to 13.91% respectively. For the unbound concentrations, inter-assay (3 days; n = 15) precision and accuracy ranged from 3.4 to 4.8% and from 3.3 to 7.6%, with an intra-assay (3 days; n = 5) precision and accuracy ranging from 1.1 to 6.7% and from 8.4 to 12.5%, respectively. Temocillin extraction recovery ranged from 85.8 to 99.4% with a precision ranging from 7.5 to 11.9%, and appeared concentration independent. No interfering peak greater than 20% of the response at the

LOQ was detected at the m/z transitions and RT of temocillin and IS. Typical ion chromatograms of extracts from blank serum sample, serum spiked with temocillin (1 mg/L) and ticarcillin (40 mg/L) are shown in this paper.

At the three QC levels, matrix effect was acceptable, ranging from −36.2 to −17.3% ion suppression for both temocillin and IS. This effect was significantly corrected using ticarcillin as IS. Indeed, the function slope between calibration curves made in HPLC solvent A and different serum samples was comparable (mean = 0.571; CV = 4.8%; n = 7).The results of these experiments suggested that matrix effects have minimal influence on the results.

This method was then applied to determine the concentration–time profile of total and unbound temocillin for temocillin in the serum of an end-stage renal disease patient, having received 2 g of temocillin by IV route. Total concentration matched those determined by a previously validated HPLC-UV method. Unbound concentrations oscillated between 24% at 0.5 h and 13% at 44 h. Unbound AUC represented 14% of the total value. These observations are in linewithwhat is stated in the summary of product characteristics, namely that the unbound fraction depends on the total concentration due to saturation of binding sites on serum proteins and that the protein binding is about 85% for total concentrations around 16 mg/L.





Perrin Ngougni Pokem, Ana C. Miranda Bastos, PaulM.Tulkens, PierreWallemacq, Françoise Van Bambeke, Arnaud Capron, Clinical Biochemistry 48 (2015) 542–545