Louis Sandra (1), Anne Smits (2), Karel Allegaert (2), Thomas Bouillon (1)
(1) Drug Delivery and Disposition, Department of Pharmaceutical and Pharmacological Sciences, KU Leuven, Leuven, Belgium (2) Woman and Child, Department of Development and Regeneration, KU Leuven, Leuven, Belgium.
Objectives: Propofol is frequently used for induction and maintenance of anesthesia and procedural sedation in (pre)term neonates and infants. Being a UGT1A9 and CYP2B6 substrate, maturational on top of scale effects have to be accounted for to describe its pharmacokinetics in this population. The current population PK models fall short in terms of suitability for extrapolation, plausibility and/or predictive capacity in the target population [1-3]. We would like to demonstrate that independently accounting for GA and PNA instead of aggregation of these metrics into postmenstrual age (PMA) improves the description of the pharmacokinetics of propofol in (pre)term neonates and infants.
Methods: An analysis dataset was compiled from 3 previously published studies (Allegaert 2007 [1], Smits 2016 [4] (PK unpublished), Sepulveda 2011 [5]). The dataset contains 837 arterial samples from 107 individual PK profiles (preterm: 53 (49.5%), neonate: 66 (61.7%)). Patients were aged from 0-2.0 years (median 0.022 years, GA 24.6-40.1 weeks, PNA 0-104 weeks) and weighed between 0.58-11.44 kg (median = 3.050 kg). 2 to 9 arterial blood samples per PK profile were available. A three-compartment model with bolus/infusion input and linear elimination, allometrically scaled parameters (fixed coefficients [7]) and a maturation term on elimination clearance was used to describe the concentration-time course of propofol. Investigated maturation terms, were: i) sigmoid Emax model based on PMA [3,6], ii) Richards’s model [7] based on PMA, iii) i) + accelerated maturation after birth [6] and iv) a power model based on GA + asymptotic exponential model based on PNA. The analysis was performed using MONOLIX 2018R2 [8]. Objective function value (OFV) expressed as -2 log likelihood (-2LL), the Akaike information criterion (AIC), prediction corrected visual predictive checks (pcVPC) and standard errors of the parameters were used for judging goodness of fit/selection of models and covariate inclusion/deletion.
Results: The relevant models are displayed in Table 1. Separation of intrauterine and postnatal maturation yielded the lowest OFV. Since V1 negatively correlated with age, an effect of PNA on V1 was added to this model, which further improved the fit.
Table 1: Population parameters and OFVs
|
Model |
TV (RSE%, IIV) |
|
|
|
|
|
Maturation parameters |
OFV |
|
|
|
All: fixed allometric coefficients |
CL [L min-1 70kg-1] |
V1 [L 70kg-1] |
Q2 [L min-1 70kg-1] |
V2 [L 70kg-1] |
Q3 [L min-1 70kg-1] |
V3 [L 70kg-1] |
|
-2LL |
AIC |
|
|
(1) Emax mat. based on PMA |
1.63 (5.5, 0.462) |
12.3 (10.3, 0.491) |
4.21 (6.92, 0.352) |
42.4 (3.52, 0.182) |
0.502 (6.14, 0.421) |
228 (7.94, 0.514) |
PMA50 = 40.9 wks |
93.52 |
123.52 |
|
|
(2) Emax mat. based on PMA + “acceleration term” |
1.78 (5.21, 0.379) |
12 (10.2, 0.464) |
4.21 (6.49, 0.362) |
42.1 (3.45, 0.18) |
0.491 (6.31, 0.433) |
217 (8.99, 0.58) |
PMA50= 39.4 wks |
66.41 |
100.41 |
|
|
(3) intrauterine (GA) and postnatal (PNA) maturation |
1.58 (4.96, 0.379) |
12.2 (10.6, 0.502) |
4.11 (6.08, 0.367) |
42.2 (3.58, 0.214) |
0.496 (6.42, 0.435) |
233 (7.99, 0.49) |
GMAX= 0.314 |
62.54 |
94.54 |
|
|
(4) (3) + effect of PNA on V1 |
1.56 (5.97, 0.38) |
19.8 (8.18, 0.292) |
4.16 (6.19, 0.367) |
41.9 (3.94, 0.22) |
0.511 (6.26, 0.436) |
246 (8.14, 0.513) |
GMAX= 0.334 |
43.92 |
77.92 |
|
All parameters of all models were scaled allometrically: (WT/70)^0.75 for clearances, (WT/70) for volumes. (1) Emax maturation model on clearance: MAT = PMA^HILL/(PMA50^HILL + PMA^HILL). (2) As (1) multiplied with BIRTH = (1+FBmax*(1-exp(-PNA*log(2)/T1/2a)))/(1+FBmax). (3) MAT = GMAX*(GA/38)^b + (1-GMAX)*(1-exp(-PNA*log(2)/T1/2b)). (4) Additional covariate effect of PNA on V1: exp(-PNA/52*betaV1).
Conclusions: Independently accounting for intrauterine (driven by GA) and postnatal (driven by PNA) maturation improves the description of propofol pharmacokinetics in a dataset with a relevant proportion of premature neonates. Accounting for the observed relationship of weight normalized V1 and PNA further improves the fit. However, it is more than likely that this “age” effect is caused by a change of body composition in this rapidly growing population. We would embrace efforts to extend the work of Al-Sallami et al [9] formalizing the estimation of fat free mass in children “down” into the neonatal population. In our opinion, the development of predictive equations accounting for developmental changes in postnatal body composition is crucial to improve the predictive properties of pharmacokinetic models in this variable and vulnerable population.
References:
[1] Allegaert K, Peeters MY, Verbesselt R, Tibboel D, Naulaers G, Hoon JN De. Inter-individual variability in propofol pharmacokinetics in preterm and term neonates. Br J Anaesth 2007; 99:864–70.
[2] Wang C, Peeters MY, Allegaert K, Blussé van Oud-Alblas HJ, Krekels EH, Tibboel D, Danhof M, Knibbe CA. A bodyweight-dependent allometric exponent for scaling clearance across the human life-span. Pharm Res. 2012;29:1570-81.
[3] Eleveld DJ, Colin P, Absalom AR, Struys MMRF. Pharmacokinetic e pharmacodynamic model for propofol for broad application in anaesthesia and sedation. Br J Anaesth 2018; 120: 942–59.
[4] Smits A, Thewissen L, Caicedo A, Naulaers G, Allegaert K. Propofol Dose-Finding to Reach Optimal Effect for (Semi-)Elective Intubation in Neonates. J Pediatr. 2016; 179: 54-60.
[5] P. Sepulveda, L. I. Cortınez, C. Saez, A. Penna, S. Solari IG and ARA. Performance evaluation of paediatric propofol pharmacokinetic models in healthy young children. Br J Anaesth 2011;107:593–600.
[6] Anderson BJ, Holford NHG, Anderson BJ. Tips and traps analyzing pediatric PK data. Pediatr Anesth 2011;21:222–37.
[7] Richards AFJ. A Flexible Growth Function for Empirical Use. J Exp Bot 2016;10:290–300.
[8] Monolix version 2018R2. Antony, France: Lixoft SAS, 2018. http://lixoft.com/products/monolix/
[9] Hesham Saleh Al-Sallami, Ailsa Goulding, Andrea Grant, Rachael Taylor, Nicholas Holford SBD. Prediction of Fat-Free Mass in Children. Clin Pharmacokinet 2015;54:1169–78.
Reference: PAGE 28 (2019) Abstr 8892 [www.page-meeting.org/?abstract=8892]
Poster: Drug/Disease Modelling - Paediatrics