A Model‐Based Approach to Assess Unstable Creatinine Clearance in Critically Ill Patients

Abstract

Creatinine clearance is an important tool to describe the renal elimination of drugs in pharmacokinetic (PK) evaluations and clinical practice. In critically ill patients, unstable kidney function invalidates the steady-state assumption underlying equations, such as Cockcroft-Gault. Although measured creatinine clearance (mCrCL) is often used in nonsteady-state situations, it assumes that observed data are error-free, neglecting frequently occurring errors in urine collection. In contrast, compartmental nonlinear mixed effects models of creatinine allow to describe dynamic changes in kidney function while explicitly accounting for a residual error associated with observations. Based on 530 serum and 373 urine creatinine observations from 138 critically ill patients, a onecompartment creatinine model with zero-order creatinine generation rate (CGR) and first-order CrCL was evaluated. An autoregressive approach for interoccasion variability provided a distinct model improvement compared to a classical approach (Δ Akaike information criterion (AIC) −49.0). Fat-free mass, plasma urea concentration, age, and liver transplantation were significantly related to CrCL, whereas weight and sex were linked to CGR. The modelbased CrCL estimates were superior to standard approaches to estimate CrCL (or glomerular filtration rate) including Cockcroft-Gault, mCrCL, four-variable modification of diet in renal disease (MDRD), six-variable MDRD, and chronic kidney disease epidemiology collaboration as a covariate to describe cefepime and meropenem PKs in terms of objective function value. In conclusion, a dynamic model of creatinine kinetics provides the means to estimate actual CrCL despite dynamic changes in kidney function, and it can easily be incorporated into population PK evaluations.

A population pharmacokinetic model for creatinine with and without ingestion of a cooked meat meal

“…The typical value for creatinine generation rate in healthy volunteers was 67.8 mg/h, which is in agreement with the value of 65.8 mg/h (male, 31 years, 73 kg) calculated based on the equation: baseline CGR, mg/h = (27 -0.173 × age in years) × weight in kg / 24 [10], but higher than the value of 42.8 mg/h (1183 mg/day) and 43.8 mg/h in patients reported by Ullah et al . and Daugirdas et al ., respectively [5,11]. The values for estimated individual creatinine renal clearance were similar to those that have been reported [6].…”

“…A compartmental model based on serum and urine creatinine concentration data from critically ill patients has been previously introduced by Ullah et al . to estimate creatinine clearance [5]. Covariates, including fat-free mass, plasma urea concentration, age, liver transplantation, sex, and body weight were assessed and incorporated into the published model, contributing to more accurate estimates compared to standard approaches [5].…”

“…to estimate creatinine clearance [5]. Covariates, including fat-free mass, plasma urea concentration, age, liver transplantation, sex, and body weight were assessed and incorporated into the published model, contributing to more accurate estimates compared to standard approaches [5]. However, the model assumed a volume of distribution (V d ) of creatinine of 60% of total body weight.…”

“…However, the model assumed a volume of distribution (V d ) of creatinine of 60% of total body weight. This rough approximation ignores variability, which may lead to biased estimates of other PK parameters [5]. Therefore, there is a need to enrich the estimates of creatinine PK parameters with experimental data, in the case of V d including the exogenous administration of creatinine.…”

Model‐Informed Drug Development: Connecting the Dots With a Totality of Evidence Mindset to Advance Therapeutics

“…While discussed in the context of applying PBPK for informing rational dosing of therapeutics in renal impairment, this framework for high‐fidelity models should be broadly applicable for population pharmacology models used for precision dosing—a rapidly advancing area of precision medicine that was the topic of the January 2021 issue of CPT 4 . In complex populations like critically ill patients with sepsis and/or acute respiratory distress, physiological characteristics relevant for drug disposition (e.g., renal function) can dynamically change over time, making precision dosing a moving target, as illustrated by Ullah et al 5 . The authors developed a model‐based approach to quantify creatinine clearance as a time‐varying covariate through population modeling of longitudinal serum and urinary creatinine concentration data.…”