Serum Ceftazidime Concentrations in Hemodialysis Patients

Background In case of severe infections, the first 24h of treatment are critical to limit the severity and the mortality (1). Thus, the antibiotic treatment must be initiated as soon as possible, be efficacious against the causative organism, and reach concentrations in the infected tissues larger than the minimal inhibitory concentration (MIC). In case of sepsis, pharmacokinetic parameters are modified (e.g., increase of the volume of distribution \[Vd\]), increase \[initially\] and decrease \[later\] of the renal clearance), which impacts on the actual concentration and activity of the antibiotic. For most antibiotics, the adaptation of the dosages is performed in a very crude fashion in case of renal failure (5). A first study has demonstrated that the antibiotic doses prescribed in case of sepsis were, in most of the cases, insufficient to obtain a concentration of 4 times or even 1 time the MIC of the causative organism during the desired fractions of the time between two successive administration (70% for ceftazidime) (6). A second study has shown that the doses of antibiotic prescribed in case of sepsis and of renal failure requiring extracorporeal hemofiltration are insufficient (7). Both studies showed also that there is a large interpatient variability in the antibiotic serum concentrations. An additional factor, rarely taken into account, is that only the free fraction of the antibiotic is active (8). Yet, the free fraction is rarely measured. The use of ceftazidime in first intention probabilistic antibiotherapy in patients undergoing chronic hemodialysis is supported by its broad spectrum of activity against the most frequently observed organisms in this patient population, including Pseudomonas aeruginosa. The low incidence of adverse effects reported for this antibiotic makes it a better choice than ciprofloxacin. Moreover, its rather simple and controllable mode of administration (one intravenous dose administered at the end of the dialysis period) guarantees a better compliance compared to a daily oral antibiotherapy. Ceftazidime shows a low protein binding, and its activity is therefore not expected to be grossly affected in patients compared to what can be measured in vitro in reference media (9). However, mots reported data on ceftazidime protein binding have been obtained on healthy volunteers and the extent of binding of beta-lactams to serum proteins may considerably vary according to the type of subjects (10). Presently, the recommendations for the dosages of ceftazidime in patients undergoing chronic hemodialysis are based on rather old pharmacokinetic studies assuming a plasma half-life of ceftazidime reduced to 3.3 h during hemodialysis vs. 33.6 h in anuric patients and recommending to administer half of the total daily dose at the end of the hemodialysis (11). But these recommendations do not take into account the progresses made since then in hemodialysis techniques. The international recommendations and those of the US Food and Drug Administration (FDA) recommend to administer 1 g of ceftazidime after each hemodialysis session (12). Different dialysis parameters modulating the level of removal of ceftazidime in modern dialyzers have been studied (13). This showed that the ceftazidime clearance increases with the rate of dialysis fluid flow, the ultrafiltration volume, and the urea clearance. There is a positive correlation between the increase in blood flow and the ceftazidime clearance that, however, is observed only in patients for whom an AN69 membrane was used (14). Another study has examined the impact of the use of membranes with high permeability on the pharmacodynamics of ceftazidime using mathematical modeling. This study led to a recommendation that ceftazidime should be administered at a daily dose of 1 g to obtain a serum trough concentration higher than the MIC of the causative (or suspected) organism during at least 70% of the dosing interval. A dose of 2 g of ceftazidime after each dialysis session was effective for isolates for which the MIC of ceftazidime was not superior to 16 mg/L (14). This study, however, is based on data collected in 1983 and included only 6 patients. Thus, there has been no in vivo study looking as to whether a dose of 1 or 2 g of ceftazidime administered after each dialysis session is sufficient when using contemporary dialysis equipment. Moreover, no measure of the free fraction has been made. Objectives The study will measure the ceftazidime serum concentrations in patients undergoing hemodialysis (before the hemodialysis session and at the end of the hemodialysis session,) and receiving standard doses of ceftazidime in order to determine whether these concentrations are sufficient based on accepted literature-based criteria concerning suspected or documented infections (15). The data will be confronted with the evolution of the patients' clinical conditions in order to draw possible correlations between these two aspects of the treatment. Primary objective Assess whether the standard doses of ceftazidime (1 or 2 g) administered at the end of the dialysis session (intermittent dialysis) allow to obtain the following serum concentrations: * If the causative organism is not identified: a trough level equal or superior to 8 mg/L (1 time the MIC of the least susceptible isolates of Pseudomonas aeruginosa based on the interpretative criteria of EUCAST (European Committee on Antimicrobial Susceptibility Testing; resistance set at \> 8mg/L). * If the causative organism is identified and MIC determined and lower or equal to 8 mg/L: a trough level equal or superior to 1 time this MIC Secondary objectives * Assess whether the standard doses of ceftazidime (1 or 2 g) administered at the end of the dialysis session (intermittent dialysis) allow to obtain a trough concentration at 32 mg/L if the causative organism is not identified and 4 times its MIC if it is identified and its MIC determined and lower or equal to 8 mg/L * Assess whether the criteria mentioned in the primary objectives also apply to the free fractions of ceftazidime. * Assess whether reaching the desired free and total trough concentrations of 8 mg/L or 1 time the MIC impacts the clinical outcome of the patient (as judged by the attending physician). * Assess whether reaching the desired free and total trough concentrations of 32 mg/L or 4 time the MIC impacts the clinical outcome of the patient (as judged by the attending physician). * Assess the impact of main hemodialysis parameters (type of membrane, flow rate of the dialysis fluid, length of the dialysis session; volume of ultrafiltration) on the ceftazidime total and free serum concentrations. * Assess the impact of patient's residual renal function on the ceftazidime serum free and total concentrations. * Assess the impact of potential drug-drug interactions (based on the record of all drugs administered to the patient) impact on ceftazidime serum free and total concentrations. * Asses how the MIC of the causative organism (if known) affects the expected effectiveness of ceftazidime, considering both the total and the free serum concentrations. Materials and Methods * Prospective monocentric study in the hemodialysis center of the Centre hospitalier-universitaire (CHU) of Charleroi, Belgium. * Drug assay: all assays will be made in duplicate in two laboratories: * total serum concentrations in the laboratory of Biological Chemistry of the hospital where the study will be conducted (CHU) using a validated High Performance Liquid Chromatography (HPLC) separation method and an UV photometric detection * total and free concentrations in the laboratory of Cellular and Molecular Pharmacology of the Université catholique de Louvain (Drug Research Institute) using an HPLC separation method and a tandem mass spectrometry (MS-MS) detection method (to mitigate the risk of interferences by other drugs or metabolites in abnormal quantities). Assay of the free fraction will be made after separation by membrane sieving. * Expected number of enrolled patients: 20. This number has been chosen arbitrarily but is compatible with (i) previous studies indicating that it is sufficient to provide useful information; (ii) the present number of patients admitted to the hospital for hemodialysis per year (from 100 to 120) with a proportion of about 20% needing antibiotic treatment for an infection for which ceftazidime is indicated. * Sampling: Total: 6 per patient * after the first administration of ceftazidime (loading dose): sampling #1: at trough before 1st dialysis; sampling #2: after the end of the dialysis session * after the second administration of ceftazidime (maintenance dose): sampling #3: at trough before 2d dialysis session * after the third administration of ceftazidime (maintenance dose): sampling #4: at trough before 3d dialysis session * after the fourth third administration of ceftazidime (maintenance dose): sampling #5: at trough before 4th dialysis session; sampling #6: after the end of the dialysis session \- Ceftazidime doses * loading dose (1st administration): 2 g * maintenance dose (all subsequent administrations): 1 g (the dose may be modified by the clinician in charge if deemed necessary and recorded accordingly) * Patient's clinical and biological data collection * collection patient's data: * indication(s) of ceftazidime administration * dose and time of ceftazidime administration * documentation of the infection and treatment outcomes: (fever (Celsius degree); suspected site (confirmed or suspected); inflammatory syndrome (biological data) (C reactive protein (mg/l), leucocytes count (/µl)); bacteriological data: clinical outcome: cure - improved - failure - relapse (by the same organism), infection-related adverse effects: septic metastasis, death; administration-related adverse effects: catheter-clogging and infection; treatment-related adverse effects: any that could be related to the drug * parameters of the hemodialysis: type of dialysis membrane and membrane area (Polysulfone. 2 m2, 2,2m2), blood flow (ml/min (350ml/min)), dialysate flow rate (500 ml/min (fixed)), KT/V ratio (at 2d and 5th sessions) (1,2 to 1,5) where K = dialyzer clearance of urea; T= dialysis time; V = volume of distribution of urea \[approximately equal to patient's total body water\], access route: catheter (double lumen) or direct arterio-venous connection, length of dialysis session (minutes), added solutes (bicarbonate and other filling fluids) (mmol/liter). * nutritional status: albumins and total proteins (g/l) * associated medications: other antibiotics (e.g., vancomycin in case of suspicion of infection by a methicillin-resistant Gram-positive organism \[e.g., S. aureus or S. epidermidis\]), antivitamin K, non-steroidal anti-inflammatory agent(s). Modeling and population pharmacokinetics Serum concentration data will be analyzed using appropriate softwares to determine the pertinent pharmacokinetic parameters for the population of patients studied. The analysis will focus on * median and mean ceftazidime clearances and half-lives out of and during the dialysis periods * the extraction rate of ceftazidime during the hemodialysis sessions After establishing the best model and based on Monte-Carlo simulations, the results will be used to determine whether the threshold for efficacy defined above holds true when considering the population of patients routinely undergoing hemodialysis. The general methodology and calculations approaches will be similar to those described in previous publications with adaptation if necessary (16, 17).