Population Pharmacokinetics and Pharmacodynamics of Beta-lactams of Interest in Adult Patients From Intensive Care Units

1. Background, Literature Survey and Justification of The Study 1.1. Introduction β-lactams efficacy depends primarily from the time interval during which the plasma concentration remains above the minimal inhibitory concentration (MIC) of the antibiotic against the target organism(s) (Craig, 1998). It is generally accepted that the free concentration of the antibiotic must remain above the MIC for at least 40 to 70% of the interval between two successive administrations, and should even reach 100% for severe infections in patients hospitalized in Intensive Care Units (MacGowan, 2011). The free concentration must reach a value of 4 x the MIC for 40 to 70% (Mohd Hafiz et al., 2012) or even 100 % (Tam et al., 2005) of the dosing interval in order to prevent the emergence of resistance. Due to the large inter- and intraindividual variations between patients, it is difficult to reach the desired concentrations if relying only on usual dosage recommendations and/or using standard dosing regimens. Moreover, Intensive Care patients are difficult patients in this context (Roberts et al., 2014) due to gross perturbations related to underlying diseases and abnormalities (arterial hypertension, cardiac rhythm alterations, renal and/or hepatic insufficiency) and the necessary interventions (artificial ventilation, surgery, artificial feeding, and so on…). They also show important variations in the level of plasma proteins and rapid and unpredictable fluctuations of their renal function (Beumier et al., 2015; Goncalves-Pereira and Povoa, 2011; Roberts and Lipman, 2009), all of which are known to modulate the pharmacokinetics of β-lactams (Goncalves-Pereira and Povoa, 2011; Hayashi et al., 2013; Sime et al., 2012; Udy et al., 2012; Wong et al., 2013). The concentration of the free fraction will be especially modified for those β-lactams with large protein binding such as temocillin or ceftriaxone (Schleibinger et al., 2015; Ulldemolins et al., 2011; Van Dalen et al., 1987; Wong et al., 2013), but may also be altered for β-lactams that are mainly excreted via the renal route (temocillin, ceftriaxone, meropenem) (Carlier et al., 2013; Simon et al., 2006; Vandecasteele et al., 2015). 1.2. Clinical Interest of temocillin, ceftriaxone and meropenem and State of the Art Concerning their dosing Temocillin is a carboxypenicillin with useful activity against Gram-negative bacteria (excluding P. aeruginosa) and with a large stability towards most β-lactamases, including ESBL), AmpC cephalosporinases, and some carbapenemases (Livermore et al., 2006; Zykov et al., 2016). Temocillin may stand as an alternative to carbapenems (Balakrishnan et al., 2011; Livermore and Tulkens, 2009). About 85% of temocillin in plasma is protein-bound and about 80% of the administered dose is eliminated in 24 h under an intact form by glomerular filtration and tubular secretion (Temocillin Summary of Product Characteristics [SmPC], 2015) Ceftriaxone shows a moderately enlarged spectrum of activity and is stable towards certain β-lactamases but not to ESBLs, AmpC cephalosporinases and certain carbapenemases (Suankratay et al., 2008). It is an alternative to carbapenems when dealing with an infection with susceptible organisms (Paradis et al., 1992). Its protein binding is about 95%, and its elimination is mainly via the renal route (50 to 60% under an unchanged form) with the remaining eliminated via the bile to form microbiologically inactive metabolites (Ceftriaxone SmPC, 2015). Meropenem shows a very large spectrum and is used in empiric therapy when fearing the presence of ESBL-producing organisms to which other antibiotics are resistant (Zykov et al., 2016). Meropenem has an unpredictable pharmacokinetic profile in patients with renal insufficiency or under hemodialysis (Carlier et al., 2013; Goncalves-Pereira et al., 2014). Meropenem is mainly excreted via the renal route (50 - 75 % under an unchanged form; SmPC meropenem, 2014). Antibiotic are often prescribed empirically with doses based on what has been found appropriate for the general population, with some adaptation for weight and renal and/or hepatic function. As for any drug, however, there is increasing evidence that the concentrations observed after administration of a standard dose are actually highly variable and often different from the expected ones, leading to risks of sub-therapeutic or toxic effects. Recent studies have shown that an intravenous administration of 6 g of temocillin by continuous infusion (Laterre et al., 2015), of 4 g of ceftriaxone in two administrations at 12 h interval (Roberts et al., 2007; Salvador et al., 1983), or of 6 g of meropenem in 3 administrations by prolonged infusion (3 h) at 8 h interval (Dulhunty et al., 2013; Frippiat et al., 2015; Jamal et al., 2015), allow to reach free plasma concentrations of 4 x the MIC against susceptible organisms during 40-70%, or even 100% of the dosing interval, with, however, large inter-individual variations, especially for molecules with high protein biding (temocillin, ceftriaxone) due to variations in their renal elimination. Very little information is available about their tissue levels but it is suspected that large inter-individual variations are also frequent.
2. Study Objectives The goal is to measure the total and free concentrations of the antibiotics in plasma, accessible body fluids and, if possible, tissues after intravenous administration of:

• temocillin: 6 g by continuous infusion over 24 h;
• ceftriaxone: bolus administration of 2 g in a 30 min infusion twice daily;
• meropenem: prolonged infusion (3 h) of 2 g three times daily. These doses will be adjusted in patients based on their renal function. Primary objective: To calculate and assess the values of key pharmacokinetic parameters (total clearance, volume of distribution, constants of elimination, plasma and tissue total exposure, and maximal and minimal plasma and body fluid concentrations. Secondary objectives:
• the correlation between the plasma protein profile and the actual free antibiotic concentrations;
• the impact of the alterations of the renal function on the free and total plasma concentrations of the antibiotics;
• the impact of the level and nature of circulating proteins on the free fraction of the antibiotics;
• the extent of the tissular penetration of the antibiotics (in accessible samples) and of their penetration in pertinent body fluids (bronchoalveolar lavage, drainage and ascites fluids.
• to model the population pharmacokinetics of the antibiotics in the whole set of patients included in the study;
• to investigate and assess the influence of co-variates (using biometric, biochemical and clinical data) on the variability of the individual pharmacokinetic profiles.

1. Outcome measures

• Primary Outcome Measure: Impact of renal function on total plasma concentrations (Measurement of total plasma antibiotic concentrations)
• Secondary Outcome Measures:
• Impact of the plasma protein concentration and of their nature on the free concentration of antibiotics
• Tissular and fluid penetration of antibiotics (lung tissue, bronchoalveolar lavage, drainage fluids)
• Pharmacokinetic modeling
• Co-variates analysis

1. Conduct of the Study 4.1. Eligible patients Patients hospitalized in Intensive Care Units and treated for pulmonary or abdominal infection, septicemia, or any other infection calling for the prescription of one of the three antibiotics mentioned above. 4.2. Study groups Patients will be divided in two groups: :

• Group 1: patients with a glomerular filtration rate (GFR) ≥ 30 mL/min
• Group 2: patients with renal insufficiency or under hemodialysis 4.3. Safety considerations The three β-lactams have each a long record of safe use in patients hospitalized in Intensive Care Units but may cause an alteration of the commensal flora, allergic reactions, neurotoxicity (at high doses). Ceftriaxone may case hemolytic anemia. 4.4. Exclusion criteria
• Patients of <18 years
• Allergy to β-lactams
• Hypersensitivity to penicillin (IgE-mediated)
• Any biological abnormality considered by the attending physician as susceptible to interfere in a significant manner on the interpretation of the data
• Absence of consensus
• Therapeutic limitation 4.5. Treatment duration: 7 days except for deep, non-controlled foci (extended to 10-14 days). 4.6. Follow up: First visit (visit #1) to determine eligibility criteria. Additional visits: each day during the treatment period.

1. Calculation of the number of patients As this is a descriptive pharmacokinetic study without formal predefined hypothesis, no calculation of the size of the population has been made. Based on literature data and the experience of the investigators, a total of 20 patients in each arm should be sufficient to draw meaningful conclusions.
2. Sampling and processing of samples

• Sampling of serum and body fluids: typically at equilibrium for all three antibiotics, and at fixed times after administration of the bolus (ceftriaxone) or the prolonged infusion (meropenem), and performed by a research nurse according to a predefined schedule.
• Sampling of tissues: by Medical personnel when justified for diagnostic or treatment reasons.
• All samples will transferred to the laboratory where they will be treated using predefined and validated protocols. Antibiotic assay: validated liquid chromatography - mass spectrometry methods (protocols and performance of the assay methods available upon request). The free fraction of each antibiotic will be measured after separation of the bound fraction by molecular sieving (Ngougni Pokem et al., 2015).

1. Statistical analysis and data analysis Pharmacokinetic analyses will be performed using either NONMEM (NONlinear Mixed Effect Modeling) (http://www.iconplc.com/innovation/nonmem/ ) or PMETRICS (http://www.lapk.org/pmetrics.php) software. Mono-, bi-, and tri-compartmental models will be tested using plasma, tissular and body fluids antibiotic free and total antibiotic concentrations. The First-Order Conditional Estimation with Interaction (FOCE-I) method will be used to assess the objective functions (Jaruratanasirikul et al., 2015) in order to select the most appropriate model for the calculation of the pharmacokinetic parameters (Roberts et al. 2009).
2. Confidentiality and Rights of patients. The identity and the personal data of the patients will remain confidential according to the applicable Belgian Law Before enrollment, each patient (or his/her guardian) will provide a written informed consent. Each enrolled patients (or his/her guardian) will be allowed to withdraw from the study at any time without impact on his/her treatment.
3. Contacts

All questions concerning the study can be addressed to

• the responsible investigator: Professor Pr Pierre-François Laterre (phone: 00-32-2-764- 2733 (Intensive Care Unit) or 764-2735 (direct) at the Cliniques universitaire St Luc, Brussels, Belgium
• the associated investigators: Professor Françoise Van Bambeke (phone: 00-32-2-764-7378) and Pharm. Perrin Ngougni Pokem (phone 00-32-2-764-7225) at the Université catholique de Louvain (Louvain Drug Research Institute), Brussels, Belgium.