Tan Zhang (1), Cornelis Smit (2), Catherine M Sherwin (3), Elke H.J. Krekels* (1), Catherijne A.J. Knibbe* (1,4)
(1) Systems Pharmacology and Pharmacy, LACDR, Leiden University, the Netherlands; (2) Department of Clinical Pharmacy, Antonius Hospital Sneek, the Netherlands; (3) Department of Pediatrics, Wright State University Boonshoft School of Medicine/Dayton Children's Hospital, Dayton, USA; (4) Department of Clinical Pharmacy, St. Antonius Hospital, Nieuwegein, the Netherlands. * both authors contributed equally
Objectives:
Contradictory pharmacokinetic (PK) results have been observed between obese adults and obese adolescents for hepatically cleared drugs, with propofol clearance (CL) reported lower in obese adolescents [1], midazolam CL higher[2], and no difference in CL of metformin between obese adolescents and obese adults [3]. For renally cleared drugs such as vancomycin, no in-depth studies on the PK in obese adolescents and obese adults have been performed. Therefore, our study investigates the PK of vancomycin in adolescents and adults with overweight and obesity.
Methods:
Data including 364 concentrations from 125 overweight and obese adolescents (age 10–18 years, total bodyweight [TBW] 28.3–188.0 kg, BMI ≥ 85th percentile according to the CDC growth charts [4]) from a previous publication [5], were combined with TDM data including 393 concentrations from 81 overweight and obese adults (age 29–88 years, TBW 66.7–142.7 kg, BMI 25–41 kg/m2). Selected subjects had normal renal function, and received one or more doses of vancomycin.
A non-linear mixed effect modeling approach was performed to analyze the data using NONMEM 7.4. Covariates that were considered in the analysis were TBW, lean bodyweight, adjusted bodyweight, age, length, sex, creatinine concentration, and renal function estimates calculated by the equations of bedside Schwartz and chronic kidney disease epidemiology. To distinguish between the weight differences related to length and weight differences resulting from obesity, the standard weight for length (which for adolescents corresponds to the weight of the 50th percentile of BMI from the growth chart [4] given the individual’s length, sex, and age [WTfor length, sex, and age] and for adults corresponds to the weight related to the individual length for a BMI of 22 kg/m2 [WTfor length]) and excess weight (WTexcess, TBW minus standard weight for length) were considered as well. For the covariate model selection, the drop in objective function value (OFV) of more than 10.83 (p<0.001), goodness-of-fit (GOF) diagnostics, ETA and CWRES plots split for adults and adolescents, and shark plots were used. The final model was validated by the bootstrap and normalized prediction distribution error analysis (n=1000).
Results:
A two-compartment model with a proportional error model with log-normally distributed inter-individual variability on CL and peripheral volume of distribution (Vd) best described the data. In the base model, visual inspection pointed at an influence of weight-related covariates in adolescents and age as a covariate in adults. In the covariate analysis, WTfor length, sex, and age, WTexcess, and TBW were significant covariates for adolescents’ CL (ΔOFV = -66.634, -28.034, and -62.632, respectively). Because of better GOF diagnostics and ETA plots, implementation of WTfor length, sex, and age on CL using a linear function was preferred. Sex was found to influence CL in adolescents, with CL in boys being higher than that in girls, i.e. 4.83 vs. 3.99 L/h for individuals with a median WTfor length, sex, and age of 47 kg, respectively. Renal function estimates had significant influences on CL in adolescents (p < 0.001) but were not implemented, as only three adolescents contributed to yielding statistical significance. For adolescents, no trends for TBW, WTexcess, age, and length remained. In overweight and obese adults, a negative linear relation between CL and age was found. Besides, CL increased with WTfor length using a power function. The central Vd increased linearly with WTfor length, sex, and age in adolescents, while no covariate effect on Vd was found in adults.
In the final model, CL increased with WTfor length (sex, and age) in adolescents and adults albeit with a different function. In adolescents, boys had higher CL than girls, with adolescents tending to have higher CL than adults. GOF diagnostics, ETA plots, bootstrap, and NPDE showed this model was robust with good predictive performance.
Conclusions:
Our results showed that vancomycin CL in overweight and obese adolescents is generally higher than in overweight and obese adults of the same standard weight for length and with boys having larger CL than girls. WT for length (sex, and age), sex of adolescents, and age of adults are all influential factors on vancomycin CL, and therefore CL in overweight and obese adults is not predictive for CL in overweight and obese adolescents and cannot be directly scaled to overweight and obese adolescents.
References:
[1] Diepstraten, J. et al. CPT pharmacometrics Syst. Pharmacol. 2, 1–8 (2013).
[2] van Rongen, A. et al. Clin. Pharmacokinet. 57, 601–611 (2018).
[3] van Rongen, A. et al. Pediatr. Drugs 20, 365–374 (2018).
[4] https://www.cdc.gov/growthcharts/clinical_charts.htm
[5] Smit, C. et al. AAPS J. 23, 53 (2021).
Reference: PAGE 30 (2022) Abstr 10134 [www.page-meeting.org/?abstract=10134]
Poster: Drug/Disease Modelling - Infection