Population pharmacokinetic analysis of interaction between dolutegravir and rifampin

Proposal
9134

Title of Proposed Research

Lead Researcher

Gary Maartens

Affiliation

Department of Medicine University of Cape Town Health Sciences Faculty

Funding Source

Potential Conflicts of Interest

Data Sharing Agreement Date

06 December 2019

Lay Summary

Tuberculosis is the commonest opportunistic infection associated with HIV in resource-limited settings killing over 25% of HIV-TB co-infected persons1. Rifampicin is the mainstay of anti-tuberculosis treatment and forms the backbone of first-line TB treatment options across the world. Because of the overlapping nature of both epidemics especially in Sub-Saharan Africa, HIV persons co-infected with TB will find themselves on dolutegravir containing regimens and on rifampicin based regimes.Dolutegravir belongs to a class of antiretroviral medicines known as integrase inhibitors. It is broken down in the liver mainly by the enzyme UGT1A1 by the CYP3A4 enzyme to a lesser extent. Rifampicin on the other hand greatly increases the activity of both the UGT1A1 and CYP3A4 enzymes. This means that if both drugs are given together, dolutegravir is broken down much faster in the body since the enzymes responsible for this have an increased activity because of the co-administered rifampicin. Secondly, dolutegravir can be actively pumped out of our body cells by efflux drug transporters such as P-glycoprotein and Breast Cancer Resistance Protein whose activity is also increased by rifampicin.Therefore, co-administration of rifampicin based regimes together with dolutegravir containing regimens for HIV may lead to insufficient concentrations of dolutegravir and pause a risk of treatment failure or development of drug resistance insufficient suppression of the HIV virus The current recommendation for HIV TB co-infected persons on a rifampicin based regimen is 50 mg of dolutegravir twice-daily as opposed to the 50mg once-daily dosing for persons who are not co-infected with TB. However, twice daily dosing presents challenges of compliance and availability of the single pill as opposed to a co-formulated, fixed-dose combination pill especially in resource-limited settings which additionally suffer from very high patient-health care provider ratios. Therefore, understanding how rifampicin affects the metabolism of Dolutegravir assessing the use of other dolutegravir dosing regimens such as 100 mg once daily is necessary especially in Sub-Saharan Africa where dolutegravir is only starting to be used.The objectives of the research would be to• To develop a population pharmacokinetic model for dolutegravir describing the relationship between dose and the resulting drug concentration exposure using clinical trial data.• To develop a model to characterize rifampicin-induced metabolism of dolutegravir • To use clinical trial simulation in combination with specific assumptions and model uncertainties for the generation of the potential range of responses to other dosing regimens of dolutegravir in co-administration with rifampicinReferences 1. Tshikuka Mulumba JG, Atua Matindii B, Kilauzi AL, Mengema B, Mafuta J, Eloko Eya Matangelo G, et al. Severity of Outcomes Associated to Types of HIV Coinfection with TB and Malaria in a Setting Where the Three Pandemics Overlap. J Community Health. 2012 Dec 1;37(6):1234-8.

Study Data Provided

[{ "PostingID": 19940, "Title": "VIIV-ING117175", "Description": "ING117175: a Phase IIIb, randomized, open-label study of the safety and efficacy of dolutegravir or efavirenz each administered with two NRTIs in HIV-1-infected antiretroviral therapy-naïve adults starting treatment for rifampicin-sensitive tuberculosis" },{ "PostingID": 19941, "Title": "VIIV-ING113099", "Description": "Phase 1, open label, two arm, fixed sequence study to evaluate the effect of rifampin and rifabutin on GSK1349572 pharmacokinetics in healthy male and female volunteers" }]

Statistical Analysis Plan

Pharmacokinetic AnalysisPharmacokinetic parameters including but not limited to apparent Clearance, apparent volume of distribution will be determined using a compartmental modelling analysis with the computer program NONMEM 7.4 developed by Beal and Sheiner.Pharmacokinetic analysis will be carried out using actual sampling times as availed in the dataset from the CSDR website Population Pharmacokinetic model building: Modelling will account for fixed effects as well as unexplainable between-and within-subject effects (random effects), A stepwise procedure will be used to find the model that best fits the data. Diagnostic graphical analysis will be performed using R (version 3.6.1). Different compartmental models including a one and two-compartment structural model will be fitted to the data on both linear scale and log domain  The structural model may further be optimized with the incorporation of an oral bioavailability fraction- F, an absorption lag time or transit compartments Inter-individual variability for each model element will be modelled by assuming an exponential error function which maintains positive pharmacokinetic parameters. Inter-individual variability will be assumed to be log-normally distributed with mean zero and variance omega.  The residual error will be modelled by an additive and proportional error structure.Handling concentrations below the limit of quantification. Concentration values that are below the level of quantification (BLQ) will be set to 1/2 of the lower limit of quantification (LLOQ). If there is a series of BLQ datapoints, the last one before the Cmax and the first one after the Cmax will be kept.Assessment of model adequacy. Evaluation of model adequacy and the assumptions used in building the model will be assessed at different stages of model development using the criteria below and model diagnostics• Successful minimization of the model runs• Likelihood ratio test• Inspection of graphical and numerical diagnostics including  Assessment of shrinkage of ETA(η) and EPSILON(ε) Observations vs. population predicted value (PRED)in both linear and log scales with a line of identity and a regression line  Observations vs. individual PRED in both linear and log scales with a line of identity and a regression line  Weighted residuals (WRES) or Conditional WRES vs. Population predicted value Absolute individual WRES (|IWRES|) vs. individual Predictions  conditional WRES vs. time or time after dose  individual predictioons  visual predictive checks A non-parametric bootstrap method or sampling importance resampling (SIR) method will be used to evaluate the degree of bias in the model parameters and generate confidence intervals.Covariate model. Prior to testing of covariates in the model, preliminary plots of the covariates against the parameters will be analysed in R software for possible relationships between the parameters and covariates. Potential covariates to be tested in the model include gender, rifampicin co-administration, age, body weight and ethnicity. The choice for potential covariates is based on biological plausibility, graphical exploration for trends using plots of parameters versus covariates and previously reported covariates in literature.

Publication Citation

A.N. Kawuma, R.E. Wasmann, K.E. Dooley, M. Boffito, G. Maartens, P. Denti Population Pharmacokinetic Model and Alternative Dosing Regimens for Dolutegravir Coadministered with Rifampicin Antimicrobial Agents and Chemotherapy (2022) vol 66
https://doi.org/10.1128/aac.00215-22