Population pharmacokinetics of Ceftriaxone in intensive care unit (ICU) adult patients
C Le Guellec (1), N Simon (2), D.Garot (3), R. Respaud (1), P Lanotte (4), H. Blasco (1), PF Dequin (3).
(1) Department of clinical pharmacology, University hospital, Tours, France. (2) Department of clinical pharmacology, Université de la méditerranée, APHM, Marseilles, France. (3) Intensive acre unit, University hospital, Tours, France. (4) Laboratory of microbiology, University hospital, Tours, France
Objectives: Pharmacokinetics may be altered in severely ill patients. We describe population pharmacokinetics of ceftriaxone in a large group of critically ill patients suffering from sepsis, severe sepsis or septic shock. The influence of several clinical and biological covariates was analyzed. The model was used for dose simulation in reference to MIC of common ICU pathogens.
Methods: The dose of ceftriaxone was 1 g or 2 g once a day. A full or semi-rich pharmacokinetic (PK) profile was obtained on two occasions for each patient. The first PK was drawn on the second day of ceftriaxone therapy; the second PK occurred following the resolution of sepsis. Population pharmacokinetic analysis was performed using NONMEM VI. Fourteen potential covariates were evaluated, including demographic, biological and clinical characteristics, and co-administered drugs. Between Occasion Variability (BOV) was analyzed from PK1 to PK2. Simulations were made at various doses for hypothetical patients having different covariates values.
Results: We included 54 patients: 19 suffered from sepsis, 9 from severe sepsis and 26 from septic shock. Eleven patients were hemofiltrated or haemodyalized. For the other patients creatinine clearance (CLcr) ranged 5.5 to 214 ml/min. Renal function improved from PK1 to PK2. A 2-compartments model best fitted the concentration data and the covariate selection analysis showed that V1 increased with more severe type of sepsis (V1 = 8.21 ± 3.55 L, 9.77 ± 3.83 L and 11.5 ± 4.16 L for sepsis, severe sepsis and septic shock, respectively) but its inclusion in the model was not significant. Ceftriaxone CL was independent of CLcr for values below 60 ml/min and then increased linearly with GFR [CL=THETA(1)+THETA(2)*(CLcr/4.26);THETA (1)=0.56; rse=19.6; THETA(2)= 0.32; rse=37.5]. The model did not supported BOV on any parameter. The other PK parameters were V2 (7.35 L; rse=10.2%; CV=65%) and intercompartment clearance (5.28 h-1; fixed). The residual variability modelled as proportional was 24%. Simulations performed for 2 different ceftriaxone doses and 4 different renal functions indicated that most patients will achieve effective concentrations, even with a dose of 1 g.
Conclusions: We found a wide inter-patient but weak intrapatient variability of ceftriaxone PK. The only parameter that influenced ceftriaxone pharmacokinetics was CLcr, but only when CLcr was > 60 ml/min, suggesting participation of non-renal phenomenon. Simulations indicated that the risk of being under 4 MIC for the entire dosing interval is very low and exists only in patient with high glomerular filtration rate (CLcr > 120 ml/min).
