EARNEST Rifabutin Pharmacokinetics (PK) Substudy

• Summary Background and aims

The standard anti-tuberculosis drug, rifampicin, has major drug-drug interactions with the cornerstone of second-line therapy in resource-limited settings, lopinavir/ritonavir (Aluvia tablets). Concomitant administration of rifampicin and lopinavir/ritonavir reduces lopinavir/ritonavir levels to such a large degree that HIV control is compromised, and resistance may develop. Therefore international and national guidelines recommend that rifampicin should not be used with lopinavir/ritonavir, and rather recommend that rifabutin be used instead if at all possible.

The current recommendation of using a reduced dose of 150 mg rifabutin thrice weekly when administered with boosted protease inhibitors (bPI) is based on pharmacokinetic studies in healthy volunteers that demonstrate substantial drug-drug interactions which lead to increased rifabutin levels at standard doses. However, there is also concern that with the reduced dose the resulting levels of rifabutin might lead to failure of anti-tuberculosis therapy or TB relapse. Further, there is general concern that the entire class of rifamycins have been sub-optimally dosed throughout the history of anti-tuberculosis treatment. One barrier to using a higher dose of rifabutin with boosted protease inhibitors is the potential for substantial increases in toxicity, in particular the risk of uveitis which is a well-recognised side effect of high dose rifabutin.

This pilot randomised open-label pharmacokinetic substudy will therefore compare 150 mg rifabutin daily versus thrice weekly, both in combination with lopinavir/ritonavir taken as part of second-line ART in the EARNEST trial of second-line antiretroviral therapy, in terms of (i)toxicity (ii) pharmacokinetics of rifabutin, lopinavir/ritonavir, raltegravir participants enrolled from arm B of main EARNEST trial only) and (iii) PK/PD (pharmacokinetic/pharmacodynamic) relationships.

• Trial design

A parallel two group, open-label, multi-centre, randomised controlled pilot trial, embedded within a larger randomised controlled trial (EARNEST).

• Detailed Background

Tuberculosis is one of the most common opportunistic infections in HIV-infected people taking antiretroviral therapy (ART). Concomitant use of rifampicin and many antiretroviral drugs is complicated by drug-drug interactions caused by the potent induction of genes involved in drug metabolism and transport by rifampicin, which can result in subtherapeutic antiretroviral drug concentrations. Drug-drug interactions between rifampicin and ritonavir-boosted protease inhibitors (bPI), the cornerstone of second-line therapy following the WHO public health approach to ART, are more marked than with the non-nucleoside reverse transcriptase inhibitors (NNRTIs) commonly used in first-line therapy, and therapeutic lopinavir concentrations have only been achieved with substantially increased doses of lopinavir/ritonavir or ritonavir ("super-boosting"). However, this has lead to high rates of hepatotoxicity in healthy volunteers and HIV-infected adults. The importance of the lopinavir/ritonavir "backbone" is such that risks of loss of HIV control and development of resistance with concomitant administration of rifampicin are considered too large, and thus international and national guidelines recommend that rifampicin should not be used with lopinavir/ritonavir, and rather recommend that rifabutin be used instead if at all possible. This alternative strategy of replacing rifampicin with rifabutin is followed in resource-rich settings. Pharmacokinetic studies in healthy volunteers demonstrated a substantial increase in rifabutin levels when given in the standard dose of 300 mg daily together with bPI, and so the current dosing recommendations are 150 mg rifabutin thrice weekly when administered with bPI. However, there is also concern that with the reduced dose, the resulting levels of rifabutin might be inadequate (for example, 9/10 patients had low maximum concentration values for rifabutin below the TB minimum inhibitory concentration in a study of Boulanger et al), and that this, together with intermittent dosing of the companion antituberculosis drugs, could lead to failure of anti-tuberculosis therapy or TB relapse. A recent PK study in 16 African patients has confirmed that rifabutin levels are low at the standard of care dose when coadministered with boosted lopinavir. Further, there is general concern that the entire class of rifamycins have been sub-optimally dosed throughout the history of antituberculosis treatment. One barrier to using a higher dose of rifabutin with boosted protease inhibitors is the potential for substantial increases in toxicity, in particular the risk of uveitis (inflammation of the middle layer of the eye) which is a well-recognised side effect of high dose rifabutin. A second potential concern is that a higher dose of rifabutin may also increase lopinavir levels, albeit to a lesser extent than the interactions in the opposite direction. Studies on rifabutin-PI drug-drug interactions in healthy volunteers have been hampered by the occurrence of adverse events which are unacceptable in a healthy population, whereas in clinical management of patients with TB the need to optimise therapeutic efficacy may lead to a higher threshold for discontinuing therapy or reducing dose.

EARNEST is an open, parallel group, randomised controlled trial in 1277 HIV-infected adults and adolescents switching from first- to second-line antiretroviral therapy (ART). The trial is being conducted in fourteen centres and five countries (Malawi, Uganda, Zimbabwe, Kenya and Zambia). Enrolment started in April 2010, and finished by April 2011. All EARNEST participants have a second-line regimen based on a boosted protease inhibitor (bPI), Aluvia, following WHO guidelines. The randomisation is to whether this bPI is supported by the standard of care 2NRTIs (Arm A), a drug from the new integrase inhibitor class, raltegravir (Arm B), or raltegravir for an induction period of 12 weeks only, then moving to bPI maintenance monotherapy (Arm C). Each participant will be followed for 144 weeks, and the total duration of the trial will be 4 years (trial closure end 2013).

This substudy will enrol only patients who have initiated second-line bPI-containing ART within EARNEST.

• Those patients who develop TB during EARNEST follow-up. All patients in EARNEST are receiving lopinavir/ritonavir-based regimens and will therefore initiate rifabutin-containing TB treatment as per guidelines (rifampicin is contra-indicated). These patients will be enrolled into the sub-study as soon as possible after starting TB treatment.
• Those patients who are already established on rifabutin-based TB treatment at the time of sub-study screening and who have at least 10 weeks remaining before completing their course of TB treatment.

It is anticipated that a substantial minority of EARNEST participants will develop TB during trial follow-up, or will already be on maintenance anti-TB therapy at substudy entry. The fact that patients will be followed long term as part of EARNEST provides a unique opportunity to collect both short and longer-term data on incidence of adverse events, as well as pharmacokinetic data, within the context of the larger EARNEST trial.

• Substudy objectives

Given the concerns about potential under-exposure to rifabutin with the currently recommended dose of 150 mg thrice weekly with bPI, this pilot randomised open-label pharmacokinetic substudy will therefore compare 150 mg rifabutin daily versus thrice weekly in combination with lopinavir/ritonavir taken as part of second-line ART in the EARNEST trial, in terms of:

(i) toxicity (ii) pharmacokinetics of rifabutin, lopinavir/ritonavir, raltegravir (participants enrolled from arm B of main EARNEST trial only) (iii) PK/PD (pharmacokinetic/pharmacodynamic) relationships.

• Substudy hypothesis

The dose of rifabutin currently recommended for concomitant administration with lopinavir/ritonavir (150 mg 3 x week) can be increased to 150 mg daily without significantly increasing toxicity and whilst providing improved rifabutin exposure.