Abstract
IntroductionExtracorporeal membrane oxygenation (ECMO) is an established advanced life support system which allows for prolonged temporary cardiopulmonary support in patients with life-threatening respiratory and/or cardiac failure refractory to maximal medical management. Fungal infections, aspergillus and candidemia in patients on ECMO have been associated with increased mortality. ECMO may alter the pharmacokinetics (PK) of drugs in several ways which include sequestration and increased apparent volume of distribution, leading to sub-optimal drug concentration and thus treatment failure. Antifungal drugs have physicochemical characteristics associated with a higher likelihood of sequestration onto ECMO circuitry potentially leading to a sub-therapeutic drug concentration.
Purpose
The overarching goal of this thesis is to understand whether ECMO support in critically unwell patients impacts antifungal drug concentrations via altering the drugs PK, and, whether predictive modelling utilising population PK and physiologically based PK methodologies can be developed to accurately predict drug concentrations and likelihood of achieving target exposure.
Methods
The research comprised two major components: an ex vivo study and a prospective, observational, non-interventional PK study in adult critically ill patients on ECMO receiving an antifungal drug.
A) The ex vivo experiments employed a blood primed ECMO circuit to investigate sequestration of the commonly used antifungal drugs, voriconazole, posaconazole, and caspofungin.
B) The Clinical PK study: patients were sampled over a single dosing period on
the second day of ECMO treatment, or of an antifungal course where antifungal was commenced whilst the patient was on ECMO. A population pharmacokinetic (PopPK) approach was used to describe the pharmacokinetics of caspofungin in critically ill adult patients on ECMO. Additionally, a physiologically based pharmacokinetic (PBPK) model was developed to predict the PK of caspofungin in adult critically ill patients on ECMO support.
Results/Findings
The ex vivo experiments demonstrated variable degrees of loss of the three antifungal drugs from the blood primed ECMO circuits, with posaconazole exhibiting the greatest loss correlating with both high lipophilicity and high protein binding.
The Clinical PK study: PK parameters of voriconazole in an adult critically ill patient supported on ECMO were measured, the original objective to perform traditional PK modelling was not possible. However, together with other published case reports and studies in adult critically ill patients with or without ECMO, including that from a collaboration in a multi-centre study demonstrate that voriconazole dosing in these patients remains challenging.
The caspofungin PK in 27 adult critically ill patients on ECMO was best described by a 2-compartment PK model with a fixed ECMO and between occasion variability on both apparent Volume of distribution (Vd) and Clearance (CL) as a covariate. Higher than recommended caspofungin loading doses were required to achieve target exposure for C. albicans and C. Parapsillosis but not for C. glabrata. The adult ECMO PBPK model developed for caspofungin successfully translated the ex vivo ECMO findings into a predictive model for drug exposure. Incorporating, a pronounced distribution phase with clearance by OATP1B1-transporter mediated hepatic uptake and biliary elimination.
Limitations
Firstly, a lack of control group of adult critically ill patients not on ECMO, and secondly, inflammatory covariate data was not collected, impacting the assessment of inflammation's influence on antifungal PK in sepsis and septic shock.
Conclusions
The data presented in this thesis confirm that ECMO and critical illness have effects on anti-fungal drug PK and highlights the importance of considering the infecting pathogen when choosing the dose as one dose recommendation is not for all. Antifungal dosing strategies derived from critically ill adult patients not on ECMO is acceptable but should be optimised using TDM. Further studies are required to develop robust models to identify all covariates contributing to their variability. While challenges persist in antifungal dosing for ECMO patients, the developed PBPK model for caspofungin serves as a proof-of-concept, showcasing the value of this methodology in understanding the underlying physiological changes driving drug PK. This innovative approach could be extrapolated to model other antifungal or anti- infective drugs within this patient cohort.
| Date of Award | 20 Sept 2023 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Jeremy Mills (Supervisor), Paul Rutter (Supervisor) & Anna Reed (Supervisor) |