Pharmacokinetics of Atazanavir in Special Populations

Overview

The lack of data relating to the DDI between ATV and RIF is a major limitation to the use of ATV in patients who require treatment for TB. The VirTUAL Workpackage 2 will explore the necessary dose escalation required to overcome this interaction in non-pregnant HIV-infected adults who are virologically suppressed on bPI-based ART, and who are administered RIF as a study drug, not as part of a full TB treatment regimen. As the specific objective of WP2 is to define the dose of ATV, participants taking an alternative bPI will be transitioned to ATV for the duration of that study. However, to extrapolate the results of this study to special populations such as pregnant and postpartum women, children and adolescents and those with other 'special' characteristics such as obesity (BMI \>30 Kg/m2) or malnutrition (BMI \<18.5 Kg/m2) we propose to undertake sparse sampling for pharmacokinetic analysis from individuals who require ATV-based ART for their clinical care. Sparse PK data will be obtained opportunistically from participants in the 'special populations' defined above who are receiving ATV as part of their routine clinical care. Subjects will be identified from clinics including the Joint Clinical Research Center (JCRC) and Infectious Diseases Institute (IDI), Kampala, and from sites including Groote Schuur Hospital and Gugulethu Community Health Centre, Cape Town. The ATV/r data from "special populations" will enable validation and refinement of both the PBPK model (WP1) and the pop-PK models (WP4) of the VirTUAL consortium.

Overview of VirTUAL Consortium Through the VirTUAL Consortium, the investigators aim to define the optimal use of second-line ART regimens in vulnerable populations with TB co-infection. The primary objective is to 'to determine the optimal dose of boosted atazanavir (ATV/r) when used in combination with RIF-based TB treatment in children, adolescents and pregnant or breastfeeding women.' The results from this protocol (VirTUAL WP5) will be considered in the context of the full research programme. This can be summarised as follows: Physiologically-based pharmacokinetic (PBPK) modelling will be developed to understand bPI and RIF Drug-Drug Interactions (DDIs), identifying potential dosing strategies to overcome these in adults and special populations (WP1). This data will inform clinical pharmacokinetic studies exploring the necessary dose escalation of ATV/r, including in the context of high-dose RIF, performed in Kampala (WP2). Intracellular pharmacokinetics will further characterise the DDI (WP3). PBPK and population pharmacokinetics (pop-PK) modelling will be integrated enabling extrapolation to special populations (WP4), and sparse data collection from such populations receiving different combinations of second-line ART and/or TB treatment in Kampala and Cape Town will validate and refine these models (WP5). Capacity building focussing on equipping African scientists with the tools to efficiently define drug dosing in complex populations (WP6), communication and stakeholder engagement (WP7) will increase the application of this methodology to other priority research into pharmacokinetics in special populations. This protocol describes the sparse pharmacokinetic sampling component which forms Workpackage 5 (WP5) of this research programme. The dose-escalation study (WP2) is an interventional trial which will be conducted in 28 healthy, virologically suppressed volunteers who are on ATV-based second-line ART, are aged over 18, have a normal BMI and are not pregnant or breastfeeding, and which will take place at JCRC, Kampala, Uganda. This study is anticipated to commence in mid-2019, and will explore in detail the changes in pharmacokinetics of ATV/r in both plasma and within cells which take place when rifampicin is co-administered, and will evaluate the necessary dose adjustment which is required to concurrently administer ATV/r with rifampicin-based TB treatment. WP2 will generate intensive data on these 28 well characterized individuals who do not have 'special' characteristics. However, there is a paucity of data on ATV/r disposition in patients who are typically excluded from clinical trials, and therefore the observational data from WP5 will be collected from individuals who fall into the listed categories of special populations and who are being treated with ATV/r for their own health. Data from either WP2 or WP5 alone will be amenable to pharmacometric analysis and bring value, but the combined modelling approach using both sets of data in a combined model will allow the most comprehensive evaluation of ATV/r disposition in the wider population, and will enable projections of dosing recommendations for the special populations who require concurrent treatment for TB whilst receiving ATV/r-based ART. Individuals will be identified at routine clinic appointments. They (or their guardians) will be asked to keep a detailed record of dosing times for three to five days prior to their study appointment at which sampling will be performed. If they take their medication in the morning, they will be asked to bring the medication to clinic for observation of dosing. Repeat sampling at subsequent appointments will enable assessment of both between and within-individual variability, for example during periods of rapid childhood growth or during pregnancy and transition back to non-pregnant physiological state. It is aimed to obtain 20 sparse PK profiles in each of the following groups (pregnant, child \<5, child 6-11, adolescent 12-18, obese (BMI \>30 Kg/m2), malnourished (BMI \<18.5 Kg/m2)), with each individual contributing 4 samples per PK visit, and possibly followed up longitudinally during the study (each individual may attend for a maximum of three study visits). Therefore there will be between 7 (6 participants being sampled on 3 occasions and a final participant sampled on 2 occasions) and 20 (if each participant was only sampled on a single occasion) individual participants per group to generate the 20 sampling 'occasions', with a total number of participants across the six groups included in the study of between 42 and 120. This will provide an essential clinical dataset to inform the pop-PK modelling approach and validate the PBPK simulations, describing exposure and pharmacokinetic variability in the populations of interest. The data from this work package will be pooled with the data from the study in volunteers and jointly analysed using PK modelling. Differences in the PK parameters for each of the groups will be investigated. There exist no peer-reviewed published data regarding the transfer of ATV/r into breast milk. Therefore, among women who are enrolled during pregnancy and followed into the postpartum phase, paired breast milk samples will be obtained at the same time as plasma samples.