FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2802196725> ?p ?o ?g. }

• W2802196725 endingPage "1630" @default.
• W2802196725 startingPage "1628" @default.
• W2802196725 abstract "Paracetamol (acetaminophen) is commonly used to manage mild and moderate pain in neonates 1. There has been extensive work exploring the pharmacokinetics of paracetamol in neonates, but few studies have also evaluated the metabolites in conjunction with the parent drug. A recently published population pharmacokinetic model successfully characterized the pharmacokinetics of paracetamol and its metabolites in the plasma and urine of preterm and term neonates and indicated that both weight and postnatal age played an important role in describing the interindividual variability 2. The FDA suggests that an external validation is the most ‘stringent’ form of validation for a pharmacokinetic model 3. Therefore, this analysis sought to validate the previously published model with an external cohort of preterm and term neonates to verify the extrapolation of this model to broader clinical settings. Patients for the validation dataset were recruited from the Gasthuisberg Neonatal Intensive Care Unit, following approval of the study by the local ethics board of the University Hospital, Leuven, Belgium. Informed consent was obtained from the parents. The validation dataset included a total of 793 measured urine and plasma concentrations (353 plasma paracetamol, 132 urinary paracetamol-glucuronide, 154 urinary paracetamol-sulfate and 154 urinary paracetamol concentrations) from 54 neonates with a median postconceptual age (sum of gestational age and postnatal age) of 36 weeks (range: 27–60) and a median weight of 2.5 kg (range: 0.5–6.3) 4. Thirty of the neonates only had plasma paracetamol samples available. None of the patients had plasma metabolite concentrations measured. The published neonatal paracetamol metabolite model was implemented using non-linear mixed effects modelling software (NONMEM v 7.3; ICON Development Solutions, Elliot City, MD, USA) as previously described 2. Briefly, the model had a single compartment for plasma paracetamol, 3-plasma compartments for the metabolites (paracetamol-glucuronide, paracetamol-sulfate and the combined oxidative pathway metabolites paracetamol-cysteine and paracetamol-N-acetylcysteine) and 4-urine compartments (one for paracetamol and each of the metabolites). Paracetamol and metabolite concentrations were simulated for each individual neonate in the external dataset using first order conditional estimation with interaction while fixing the parameters to those previously determined (NONMEM MAXEVAL = 0 POSTHOC). Concentrations were only simulated for times at which there were measured concentrations. Normalized prediction distribution errors (NPDE) were calculated by simulating 1000 datasets and comparing the predicted concentrations to observed concentrations using NPDE in NONMEM 5, 6. If the external validation is successful, the data from the NPDE should have a theoretical mean of 0 and be normally distributed with a variance of 1. Statistical analysis in R (version 3.3.3, R Foundation for Statistical Computing, Vienna, Austria) was used to evaluate if the NPDE had a normal distribution and a theoretical mean of 0 with a variance of 1 6. The results of the analysis found that the mean NPDE was −0.28, which was statistically different than 0 as assessed by the Wilcoxon signed rank test (P < 1e-28). Additionally, the variance of the NPDE was 0.7, which was statistically lower than 1 as assessed by the Fisher test for variance (P < 2e-11). Finally, the Shapiro–Wilk normality test indicated that there was not an equal distribution around 0 (P < 2e-16), which can easily be visualized by the NPDE histogram in Figure 1A. A clear bias was found over concentration range (Figure 1B,C), with a marginal bias over time (Figure 1D). Following this finding, the data were further subdivided into paracetamol and the metabolites of each plasma or urine compartment to evaluate if the majority of the bias arose from a single compartment (data not shown). Recalculation of statistics showed there was not a single analyte or compartment that could account for the evident bias, suggesting it was shared among all analytes in the model. There are a few notable differences between the original model cohort and the external validation cohort. One of the limitations of the available external dataset is that none of the plasma metabolite concentrations were available. It is possible that the model heavily relies on the concentrations in the plasma compartments, resulting in the bias seen in the external validation. Though the original model contains over 50 parameters, we do not believe the model was over-parameterized based on the low shrinkage of the original analysis. Secondly, urine samples for the external dataset were collected by catheterization or by plastic collector. Conversely, the original urine collection was done by collecting urine from the diaper, which has more variability in its measurement. It is possible that the differences in the collection of urine samples is contributing to the bias. However, as previously noted, dividing up the data into plasma and urine samples did not result in a successful external validation, indicating it is not only the urine data that led to the bias. Finally, there may be differences in the methods used to quantify the concentrations between the two datasets that are contributing to the bias. Both datasets used HPLC though the original dataset used MS/MS for detection and the external dataset used UV detection 4, 7. Additionally, the original dataset had a lower limit of quantification (LLOQ) of 0.05 μg ml−1 while the data from the external dataset had a LLOQ of 2 μg ml−1, which is to be expected with the two types of detection used. Both methods were successfully validated with very similar intraday and interday coefficients of variation, suggesting these two analytical methods were very similar in their variation. Furthermore, if the majority of the bias came from concentrations that were very low, it would support the notion that the differences between the assays contributed to the bias. However, the bias over the concentration range was not larger at the smaller concentrations (Figure 1C), suggesting the differences in bioanalytical method did not contribute significantly. This analysis sought to externally validate a recently published paracetamol and metabolite population pharmacokinetic model and highlights the importance of understanding the limitations of external data when attempting to validate complex models. While the current analysis did not validate the model further investigation into the source of the noted bias is warranted. C.M.T.S. is a Senior Editor of BJCP and J.N.v.d.A. an Executive Editor of BJCP. The other authors have no competing interests to declare. The authors would like to thank Dr. Karel Allegaert for providing the data used to test the external validity of the model. The initial clinical trial data used for the external validation was supported by National Institutes of Health grants from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (R01HD060543, to J.N.v.d.A.) and the National Center for Advancing Translational Sciences (UL1TR000075, to the Children's National Health System), and by a contract for analytical laboratory services from McNeil Consumer Healthcare (Division of McNEIL-PPC, Inc., Fort Washington, PA, USA). Collection of the data to populate the external dataset was supported by the Dutch Top Institute Pharma project number D2-104, and the Fund for Scientific Research, Flanders (1700314 N Fundamental Clinical Investigatorship)." @default.
• W2802196725 created "2018-05-17" @default.
• W2802196725 creator A5020515788 @default.
• W2802196725 creator A5052819935 @default.
• W2802196725 creator A5057776173 @default.
• W2802196725 creator A5075917872 @default.
• W2802196725 creator A5083033053 @default.
• W2802196725 date "2018-04-17" @default.
• W2802196725 modified "2023-10-16" @default.
• W2802196725 title "Evident bias in a paracetamol metabolite population pharmacokinetic model applied to an external dataset" @default.
• W2802196725 cites W1565106207 @default.
• W2802196725 cites W1957718823 @default.
• W2802196725 cites W2063049671 @default.
• W2802196725 cites W2120725688 @default.
• W2802196725 cites W2402000540 @default.
• W2802196725 cites W2735224076 @default.
• W2802196725 doi "https://doi.org/10.1111/bcp.13588" @default.
• W2802196725 hasPubMedCentralId "https://www.ncbi.nlm.nih.gov/pmc/articles/6005626" @default.
• W2802196725 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/29667221" @default.
• W2802196725 hasPublicationYear "2018" @default.
• W2802196725 type Work @default.
• W2802196725 sameAs 2802196725 @default.
• W2802196725 citedByCount "1" @default.
• W2802196725 countsByYear W28021967252019 @default.
• W2802196725 crossrefType "journal-article" @default.
• W2802196725 hasAuthorship W2802196725A5020515788 @default.
• W2802196725 hasAuthorship W2802196725A5052819935 @default.
• W2802196725 hasAuthorship W2802196725A5057776173 @default.
• W2802196725 hasAuthorship W2802196725A5075917872 @default.
• W2802196725 hasAuthorship W2802196725A5083033053 @default.
• W2802196725 hasBestOaLocation W28021967251 @default.
• W2802196725 hasConcept C105795698 @default.
• W2802196725 hasConcept C112705442 @default.
• W2802196725 hasConcept C126322002 @default.
• W2802196725 hasConcept C149782125 @default.
• W2802196725 hasConcept C2777477808 @default.
• W2802196725 hasConcept C2778722691 @default.
• W2802196725 hasConcept C2908647359 @default.
• W2802196725 hasConcept C33923547 @default.
• W2802196725 hasConcept C71924100 @default.
• W2802196725 hasConcept C98274493 @default.
• W2802196725 hasConcept C99454951 @default.
• W2802196725 hasConceptScore W2802196725C105795698 @default.
• W2802196725 hasConceptScore W2802196725C112705442 @default.
• W2802196725 hasConceptScore W2802196725C126322002 @default.
• W2802196725 hasConceptScore W2802196725C149782125 @default.
• W2802196725 hasConceptScore W2802196725C2777477808 @default.
• W2802196725 hasConceptScore W2802196725C2778722691 @default.
• W2802196725 hasConceptScore W2802196725C2908647359 @default.
• W2802196725 hasConceptScore W2802196725C33923547 @default.
• W2802196725 hasConceptScore W2802196725C71924100 @default.
• W2802196725 hasConceptScore W2802196725C98274493 @default.
• W2802196725 hasConceptScore W2802196725C99454951 @default.
• W2802196725 hasFunder F4320337472 @default.
• W2802196725 hasFunder F4320337611 @default.
• W2802196725 hasIssue "7" @default.
• W2802196725 hasLocation W28021967251 @default.
• W2802196725 hasLocation W28021967252 @default.
• W2802196725 hasLocation W28021967253 @default.
• W2802196725 hasOpenAccess W2802196725 @default.
• W2802196725 hasPrimaryLocation W28021967251 @default.
• W2802196725 hasRelatedWork W1492229869 @default.
• W2802196725 hasRelatedWork W1964759623 @default.
• W2802196725 hasRelatedWork W1998356222 @default.
• W2802196725 hasRelatedWork W2027216937 @default.
• W2802196725 hasRelatedWork W2030336508 @default.
• W2802196725 hasRelatedWork W2056938528 @default.
• W2802196725 hasRelatedWork W2065722033 @default.
• W2802196725 hasRelatedWork W2081460129 @default.
• W2802196725 hasRelatedWork W2132691174 @default.
• W2802196725 hasRelatedWork W2957242515 @default.
• W2802196725 hasVolume "84" @default.
• W2802196725 isParatext "false" @default.
• W2802196725 isRetracted "false" @default.
• W2802196725 magId "2802196725" @default.
• W2802196725 workType "article" @default.