Evaluation of methodologies for estimation of change in systemic drug exposure in renally impaired patients: Elucidation of possible causes to discrepancies in results based on phase I and III data
Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesis
Introduction: Regulatory authorities require certain subpopulations, such as patients with renal impairment (RI) to be studied specifically. This may be done in phase I analyzed with Non-Compartmental Analysis (NCA), and/or as part of phase III utilizing population pharmacokinetic (PopPK) methods. However, it has been suggested that phase I data analyzed with NCA may overestimate the effect of RI, as compared with PopPK analysis.
Aim: This project aimed to investigate causes for the discrepancy previously observed when calculating the exposure increase over different RI groups based on phase I and III data, and to examine the effect of erroneous assumptions made during PopPK model development, which can be of potential benefit in drug development.
Materials and Methods: Phase I and III data were simulated based on PopPK models. Potential causes, related to the methods used, to the over-prediction by NCA were investigated. For phase III data the influence of model misspecification on the estimation of exposure increase in RI was explored.
Results: The observed over-predictions by NCA were suggested to be due mainly to sub-optimal NCA and bias calculations, the latter with respect to creatinine clearance (CrCL) reference value. In PopPK analysis of phase III data, using erroneous structural and/or covariate model may result in severe bias in the estimation of the effect of RI, while disregarding the effect of inter-occasion variability led to low bias.
Conclusions: The previously observed over-prediction by the NCA method appears to mainly be an artefact due to inappropriate methodology. When investigating exposure increase in RI patients using PopPK for phase III data, careful consideration regarding assumptions should be made, especially with lower fraction excreted, as results suggest large bias when an erroneous PopPK model is applied.
Place, publisher, year, edition, pages
2015. , 43 p.
Pharmacometrics, Renal Impairment, NONMEM, R, NCA, Population Pharmacometrics, PopPK
Pharmaceutical Sciences Bioinformatics (Computational Biology)
IdentifiersURN: urn:nbn:se:uu:diva-259519OAI: oai:DiVA.org:uu-259519DiVA: diva2:844567
Subject / course
Master of Science Programme in Pharmacy
Jönsson, Siv, Researcher
Hammarlund-Udenaes, Margareta, Professor