Population pharmacokinetics of cefazolin in maternal and umbilical cord plasma, and exposure in term neonates
Omar Elkayal (1), Karel Allegaert (1,2,3), Isabel Spriet (1,4), Anne Smits (2,5), Seghaye Marie-Christine (6), Corinne Charlier (7), Erwin Dreesen (1,8).
(1) Department of Pharmaceutical and Pharmacological Sciences, KU Leuven, Leuven, Belgium, (2) Department of Development and Regeneration, KU Leuven, Leuven, Belgium, (3) Department of Clinical Pharmacy, Erasmus MC Rotterdam, Rotterdam, the Netherlands, (4) Pharmacy Department, University Hospitals Leuven, Leuven, Belgium, (5) Neonatal Intensive Care Unit, University Hospitals Leuven, Leuven, Belgium, (6) Pediatrics NDB, Centre Hospitalier Universitaire de Liège, Liège, Belgium, (7) Department of Toxicology, Centre Hospitalier Universitaire de Liège, Liège, Belgium, (8) Department of Pharmacy, Uppsala University, Uppsala, Sweden.
Introduction: Intrapartum administration of cefazolin, a first-generation cephalosporin, is used to prevent vertical transmission of Group B Streptococcus (GBS) in mothers allergic to penicillin without a history of anaphylaxis. Early-onset sepsis in the newborn may be prevented when exposure to unbound cefazolin in the mother and the neonate serum exceeds 1 mg/L (worst-case clinical breakpoint).
Objectives: The objectives of our work were (i) to build a population pharmacokinetic (popPK) model of cefazolin in healthy women undergoing vaginal delivery, and (ii) to investigate pharmacokinetic/pharmacodynamic (PK/PD) target attainment (time above 4x the epidemiological cutoff (ECOFF) of 0.25 mg/L) in maternal and neonatal blood during labor, and (iii) to investigate the early postpartum cefazolin exposure in the newborn.
Methods: Data were obtained from a prospective study in which 24 healthy, GBS colonized, pregnant women (20-41 years), giving vaginal delivery (gestational age ≥37 weeks) were recruited. During labor, all women received a 2g cefazolin intravenous infusion. Eight hours later, eight women received another 1g infusion since delivery did not occur yet. Next to maternal total plasma concentrations (up to ten per dosing interval until delivery), venous and arterial umbilical cord total concentrations were determined at delivery. Based on previous literature, protein-binding was assumed 75% in maternal plasma and 60% in neonatal plasma [1,2,3]. A maternal popPK model of cefazolin was built (NONMEM 7.4).
Subsequent neonatal exposure was predicted using Monte Carlo simulations (n=1,000) with the arterial umbilical cord concentration at delivery – predicted from the maternal model – and a previously published neonatal cefazolin popPK model (assuming elimination only) [1]. Arterial umbilical cord plasma concentrations were considered to represent fetal plasma concentrations since arterial cord blood originates directly from the fetal circulation. Missing neonatal serum albumin was multiply imputed (n=1,000) with values sampled from a uniform distribution within the range reported by Smits et al. (28.2 g/L to 43.7 g/L) [4].
PK simulations were performed of standard predelivery maternal cefazolin dosing to predict the probability of target attainment (PTA) in maternal and neonatal plasma at delivery up to 12 hours after the first infusion. The therapeutic target was set at 1 mg/L for 60% of the dosing interval (unbound cefazolin).
Results: At delivery, maternal blood and arterial umbilical cord unbound cefazolin was >1 mg/L in 23/24 (95.8%) and 11/12 (91.7%), respectively. The popPK of cefazolin in pregnant women was described by a two-compartment model with first-order elimination. Two additional compartments were added to resemble the venous and arterial umbilical cord (neonate). Interindividual variability was estimated on the maternal clearance (43% coefficient of variation [CV]), central volume of distribution (67% CV), and peripheral volume of distribution (68% CV). A combined additive and proportional error model was selected to describe the residual variability. None of the available variables (body weight and gestational age) were retained in a stepwise covariate analysis (αforward=0.010, αbackward=0.001). The neonatal PK simulations showed adequate antimicrobial exposure for protection against GBS in neonates at birth in our study cohort (median [interquartile range; IQR] 3.6 [2.5-4.8] mg/L). The maternal and arterial umbilical cord PK simulations showed that the median maternal unbound cefazolin plasma concentration is above the therapeutic target of 1 mg/L for the entire 12-hour simulation period.
Conclusion: Our study showed that the standard predelivery maternal cefazolin dosing provided adequate maternal and neonatal antimicrobial protection against GBS. In addition, we developed a popPK model to describe the cefazolin PK in women during labor with subsequent vaginal delivery. PopPK model simulations showed that standard predelivery cefazolin dosing provides adequate unbound cefazolin concentrations during the entire 12-hour time window after the first infusion.
References:
[1] De Cock RFW, Smits A, Allegaert K, et al. J Antimicrob Chemother 2014; 69: 1330–8.
[2] Allegaert K, Van Mieghem T, Verbesselt R, et al Clin Pharmacol 2009; 31: 25–8.
[3] Van Hasselt, et al. BioMed Res Int 2014; 2014: 1–9.
[4] Smits A, Kulo A, Verbesselt R, et al. Eur J Clin Microbiol Infect Dis 2012; 31: 3359–65.