Optimisation of Neonatal Antimicrobial Therapy Using Pharmacokinetic

199 Pages·2016·5.31 MB·English

Optimisation of Neonatal
Antimicrobial Therapy Using
Pharmacokinetic-
Pharmacodynamic
Modelling
Eva Germovˇsek
A dissertation submitted in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
of
University College London.
Institute of Child Health
University College London
November 2015
2
I, Eva Germovˇsek, confirm that the work presented in this thesis is my own.
Where information has been derived from other sources, I confirm that this has
been indicated in the work.
Abstract
Bacterial infections, namely sepsis and meningitis, are among the major causes
of morbidity and mortality during the neonatal period. In an era of increasing
antimicrobial resistance, when few new types of antibiotics are being developed,
antimicrobial therapy needs to be optimised to ensure that adequate doses are
given. At the same time, since renal function is immature in neonates, the dosing
regime needs to be designed to minimise toxicity.
The studies described here aimed to address the following questions: what
is the appropriate way to scale drug clearance in the paediatric population; how
can treatment be individualised and optimised to help improve the therapeutic
drug monitoring of gentamicin; what meropenem dose should be recommended
for neonates and infants with sepsis or meningitis; and finally how can a mod-
elling approach be used to facilitate the definition of neonatal sepsis. The above
questions were addressed using distinct strategies. An extensive comparison of
published models for scaling clearance was performed. Population pharmacoki-
neticmodelsusingdatafromlargegentamicinandmeropenemstudiesinneonates
were developed, and then either implemented in provisional software, or used to
make dose recommendations, respectively. Also, in a preliminary study, item re-
sponse theory models were applied to pharmacodynamic data from neonates with
sepsis.
The use of allometric weight scaling with a postmenstrual age driven sig-
Abstract 4
moidal maturation function was recommended as a standard approach for scaling
clearance. The population pharmacokinetic model developed using gentamicin
data showed that specifically timed trough levels are not needed for therapeutic
drug monitoring. The results of the meropenem study imply that the current
recommended dosing regimen for neonates is appropriate for susceptible bacteria.
Finally, the proof-of-concept study suggested that metabolic acidosis provided the
most information about the sepsis status of neonates.
Acknowledgements
There are many who I would like to thank for providing their support and for
making my PhD experience a pleasant one.
Firstly, I would like to thank all the participating neonates and infants, as
well as their parents and hospital staff, without whom I would not have been able
to perform any of the analyses presented in this thesis.
I am very grateful to Joseph Standing for his supervision during my PhD, as
well as his patience and guidance, and for introducing me to the academic world.
I would also like to thank Nigel Klein for his support during the last three years.
ToSebastianUeckertforgivingadviceandprovidingassistancewiththeitem
response theory models.
I have received funding from the NeoMero study, part of the European
Union Seventh Framework Programme for research, technological development
and demonstration, from Action Medical Research, and from the UCL IMPACT
PhD studentship.
I would also like to thank: Charlotte Barker, Rollo Hoare, Hannah Jones,
Felicity Fitzgerald, Julia Kenny, Liam Shaw, Deji Majekodunmi, Ronan Doyle,
Sonia Melo Gomes, Vania de Toledo, and other students and staff at the Institute
of Child Health, and members of the London Pharmacometrics Interest Group,
for their kindness and support.
Acknowledgements 6
Also, to Judit Szalay, Andrea Csomor, Esther Odoom-Opoku, Veronika
ˇ
Sibelja, Alenka Gjuran, Petra Kovaˇciˇc, and Nina Intihar for their friendship and
for cheering me up when I needed it.
Finally, to my family for always being there for me.
Contents
1 Introduction 20
1.1 Neonatal infections . . . . . . . . . . . . . . . . . . . . . . . . . . 20
1.2 Pharmacokinetic-pharmacodynamic modelling . . . . . . . . . . . 21
1.3 Approaches for analysing data from multiple subjects . . . . . . . 23
1.4 Statistical modelling . . . . . . . . . . . . . . . . . . . . . . . . . 25
1.5 Evaluation of non-linear mixed-effect models . . . . . . . . . . . . 29
1.6 Developmental pharmacology . . . . . . . . . . . . . . . . . . . . 30
1.6.1 Pharmacokinetic differences . . . . . . . . . . . . . . . . . 30
1.6.2 Pharmacodynamic differences . . . . . . . . . . . . . . . . 32
1.7 Pharmacokinetic-pharmacodynamic relationship of antimicrobial
agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
1.8 Aims and Structure . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2 Comparison of methods for scaling clearance in neonates, infants
and children 36
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.2 Aim . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.3 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.3.1 Search for published models for scaling clearance . . . . . 40
2.3.2 Data collection . . . . . . . . . . . . . . . . . . . . . . . . 41
2.3.3 Comparison of different models for size and maturation . . 44
2.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
2.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Contents 8
2.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3 Pharmacokinetic model for treatment individualisation 55
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
3.1.1 Gentamicin pharmacodynamics . . . . . . . . . . . . . . . 55
3.1.2 Therapeutic drug monitoring of gentamicin . . . . . . . . . 58
3.1.3 Gentamicin pharmacokinetics . . . . . . . . . . . . . . . . 61
3.1.4 Previously published population pharmacokinetic models . 62
3.1.5 Creatinine . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
3.2 Aim . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.3 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.3.1 Study population . . . . . . . . . . . . . . . . . . . . . . . 69
3.3.2 Non-linear mixed-effects model building . . . . . . . . . . 72
3.3.3 Model evaluation . . . . . . . . . . . . . . . . . . . . . . . 75
3.3.4 Comparison with published models . . . . . . . . . . . . . 77
3.3.5 neoGent software . . . . . . . . . . . . . . . . . . . . . . . 78
3.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
3.4.1 Study population . . . . . . . . . . . . . . . . . . . . . . . 78
3.4.2 Non-linear mixed-effects model building . . . . . . . . . . 79
3.4.3 Model evaluation . . . . . . . . . . . . . . . . . . . . . . . 82
3.4.4 Comparison with published models . . . . . . . . . . . . . 86
3.4.5 neoGent software . . . . . . . . . . . . . . . . . . . . . . . 88
3.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
3.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
4 Pharmacokinetic-pharmacodynamic modelling for population-
level dose recommendation 93
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4.1.1 Meropenem pharmacodynamics . . . . . . . . . . . . . . . 94
Contents 9
4.1.2 Meropenem pharmacokinetics . . . . . . . . . . . . . . . . 95
4.1.3 Cerebrospinal fluid . . . . . . . . . . . . . . . . . . . . . . 96
4.1.4 Previously published population pharmacokinetic models . 97
4.2 Aim . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
4.3 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
4.3.1 Study population . . . . . . . . . . . . . . . . . . . . . . . 98
4.3.2 Non-linear mixed-effects model building . . . . . . . . . . 100
4.3.3 Probability of target attainment . . . . . . . . . . . . . . . 103
4.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
4.4.1 Study population . . . . . . . . . . . . . . . . . . . . . . . 104
4.4.2 Non-linear mixed-effects model building . . . . . . . . . . 107
4.4.3 Probability of target attainment . . . . . . . . . . . . . . . 112
4.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
4.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
5 Pharmacodynamics of neonatal sepsis 123
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
5.1.1 Item response theory models . . . . . . . . . . . . . . . . . 123
5.1.2 Efficacy of gentamicin . . . . . . . . . . . . . . . . . . . . 126
5.1.3 Defining neonatal sepsis . . . . . . . . . . . . . . . . . . . 126
5.2 Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
5.3 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
5.3.1 Study population . . . . . . . . . . . . . . . . . . . . . . . 127
5.3.2 Non-linear mixed-effects model building . . . . . . . . . . 128
5.3.3 Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
5.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
5.4.1 Study population . . . . . . . . . . . . . . . . . . . . . . . 132
5.4.2 Non-linear mixed-effects model building . . . . . . . . . . 132
Contents 10
5.4.3 Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
5.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
5.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
6 Conclusions 142
6.1 Further work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
Appendices 147
A Search and screening procedure 147
B NONMEM control file for the final gentamicin model 148
C Individual plots of observed and predicted gentamicin trough
concentration 152
D R code for predicting the time when plasma concentration of
gentamicin goes below 2 mg/L 153
E An example output of the neoGent software 157
F NONMEM control file for the final meropenem model 158
G NONMEM control file for the final item response theory model161
H Colophon 164
References 165