Research question:

To compare treatment outcomes and early relapse rates of the standard 6-month regimen with two alternative treatment regimens for the treatment of tuberculosis, using an open-label 3-arm clinical trial.

If shown to be as effective, the trial regimens have potential to replace currently recommended first-line regimens for patients with TB in resource-poor settings, since they are less onerous to take and supervise.

Trial regimens

 Control regimen: 2 months of daily ethambutol, isoniazid, rifampicin and pyrazinamide followed by 4 months of daily rifampicin and isoniazid (2EHRZ/4HR). This is the standard 6-month first-line regimen in use in Zimbabwe. Trial regimens: 2 months of daily ethambutol, moxifloxacin (400mg od), rifampicin and pyrazinamide, followed by a) 2 months of twice weekly moxifloxacin (400 mg) and rifapentine (900mg) (2EMRZ/2P2M2) orb) 4 months of once weekly moxifloxacin (400mg) and high dose rifapentine (1200mg) (2EMRZ/4P1M1)

Rationale

 Tuberculosis and HIV are two of three major diseases in the developing world; the incidence of new cases of tuberculosis has increased dramatically in recent years, due in large part, to co-infection with HIV. Effective short course regimens of chemotherapy for the treatment of pulmonary tuberculosis have been evaluated in numerous controlled trials worldwide.

These are currently 6-8 months in duration. When adequately administered, they are capable of cure rates (defined here as bacteriological cure at the end of treatment and no early relapse) of 95% or more in patients with drug sensitive organisms. Routine treatment services, however, do not typically achieve such good outcomes. Sub-optimal adherence to treatment is likely to make a major contribution to this difference. Adherence is increased by direct observation of therapy (DOT), but DOT for the full 6 to 8 month adds considerably to the expense and inconvenience incurred by patients and health providers alike and is considered impractical in a many resource-poor countries, including Zimbabwe. If the continuation phase, which starts 2 months into TB treatment, could be administered once or twice weekly instead of daily, this would reduce the number of treatment doses that patients have to take and might also reduce toxicity as well as the cost of treatment. It would facilitate DOT for the entire treatment period, potentially leading to lower failure and relapse rates. If the length of treatment could also be shortened without loss of efficacy, this would also improve adherence to treatment and further reduce costs

The objective is to investigate the potential of two FDA approved drugs, rifapentine (a rifamycin from the same class of compounds as rifampicin) and moxifloxacin (a fluoroquinolone), to allow intermittent doses in the continuation phase and/or treatment shortening while retaining satisfactory treatment outcomes and preventing the emergence of rifamycin mono-resistance. Rifamycin mono-resistance is a particular concern for patients with HIV-related TB, since early relapse with rifampicin mono-resistance was observed among 6 or 30 HIV-positive participants in an American trial of once weekly rifapentine (600 mg) plus isoniazid in the continuation phase. Each relapsing patient had late-stage HIV infection, low CD4 counts and documented co-administration of antifungal azoles. Resistance to rifapentine results in cross-resistance to rifampicin. Ability to prevent the emergence of drug resistance is an essential attribute of TB treatment regimens and so, even at low rate, would preclude further consideration of our trial regimens. In the current trial, our hypothesis is that the combination of a) moxifloxacin plus b) higher dose of rifapentine (to maintain adequate trough levels) will be sufficient to prevent the emergence of rifamycin monoresistance with intermittent dosing. Treatment shortening to 4 months will also be investigated in one trial arm. In mouse models of TB treatment, substitution of moxifloxacin for isoniazid achieved more rapid sterilization than rifampicin, isoniazid, pyrazinamide combinations.

Moxifloxacin has other attributes that may make it particularly well suited as a companion drug to rifapentine, including a similarly long half-life of about 14 hours (the longest of the fluoroquinolones with high activity against M. tuberculosis); greater bactericidal activity against dormant bacilli than isoniazid in vitro; and metabolism by different pathways, making drug-interactions unlikely.

Outcomes

Primary outcome measure

• Combined rate of failure at the end of treatment and relapse by 18 months
• Presence of rifamycin monoresistance in relapse cultures of HIV infected patients
• Occurrence of serious adverse events at any time during chemotherapy

Secondary outcome measure

• Sputum conversion results at two months after the initiation of chemotherapy
• Rate of completion of chemotherapy according to the protocol
• Number of observed doses of chemotherapy ingested
• Any adverse events

Methods

Study design and rationale

An open-label 3-arm clinical trial powered to demonstrate non-inferiority of the two trial arms compared with the standard 6-month TB regimen. Non-inferiority has been chosen because the aim of the trial is not to improve on efficacy, since cure rates are already very good with the standard regimen, but rather to investigate whether a more convenient regimen can still achieve good outcomes. Non-inferiority requires a larger sample size than equivalence testing. The trial is a multicentre study involving newly diagnosed smear-positive TB patients in 4 Southern African countries (Zimbabwe, South Africa, Zambia, and Mozambique).

Two study sites (Beatrice Road Infectious Disease Hospital [BRIDH] in Harare, and Marondera Provincial Hospital) will contribute 270 and 140 smear-positive participants, respectively, from Zimbabwe, to the total study size of 1,250 regional participants.

Inclusion and exclusion criteria

Patients will be eligible for inclusion if they have newly diagnosed pulmonary TB, as evidenced by 2 positive sputum smears, and each of: - i) no previous treatment for TB, ii) aged 18 years and above; iii) a firm home address within the catchment area of BRIDH or Marondera with no intention of moving during the trial period, iv) able and willing to give informed consent to participate in the trial and to give a sample of blood for HIV testing. Exclusion criteria include: - i) co-morbidity (except HIV) that may prove fatal during the study period, or prejudice the response to, or assessment of, treatment, or lead to uncooperative behaviour; ii) TB meningitis; iii) pregnancy or breast-feeding; iv) contraindications to any of the study medications; iv) Hb < 7g/L; or AST or ALT > 5 x upper limit; or creatinine clearance < 30mls/min; or weight < 35kg; v) history of seizures; vi) requires antiretroviral treatment at diagnosis (including CD4 count < 200/mm3); vi) initial resistance to isoniazid, rifampicin, or moxifloxacin

Recruitment

Patients will be recruited from the TB clinics at BRIDH and Marondera using a two-step informed consent procedure, whereby patients will first give written informed consent to eligibility screening (blood count, creatinine; liver function tests, Liver function tests; HIV test, CD4 count if HIV-positive; chest radiography; urine dipstick for glucose; pregnancy test for women of child-bearing age. Eligible patients will be invited to enter the trial after a second written informed consent.

This will introduce a 24hr delay in starting TB treatment, which will be explained to the patient and is unlikely to have significant clinical or public health consequences, given that most patients will have been infectious for several weeks to months before staring their therapy. Patients subsequently found to have initial drug resistance will be withdrawn before the continuation phase, and treated with an appropriate regimen.

Treatment and follow-up

All drugs in the trial regimens will be supplied as loose (not fixed combination) tablets, including drugs in the standard TB regimen. Pyridoxine (10-50 mg daily) will be supplied to all patients in the control arm. Throughout the period of treatment, every dose will be given under direct supervision by a treatment supervisor or domicilary treatment monitor. Patients will be seen in the study clinics at least once weekly during treatment and followed-up for 12 to 14 months thereafter. Sputum for microscopy and culture will be taken at the time of enrolment, monthly from months 2 to 12 inclusive, and at months 15 and 18 months (12 to 14 months after treatment completion, according to regimen). Positive cultures occurring in the month before TB treatment is due to stop, or at any time thereafter, will be investigated as indicating possible treatment failure / relapse with 2 further sputum specimens for culture drug susceptibility testing. Treatment will be modified / restarted if failure / relapse is confirmed by two positive cultures. Any trial arm accruing 2 or more patients with rifamycin-monoresistance will have enrolment discontinued pending further investigations. HIV-positive participants will be started on cotrimoxazole and referred for ART treatment according to international recommendations and local policy.

Patients requiring ART while still on TB treatment will be started on efavirenz-based regimens, with supply of efavirenz in Zimbabwe guaranteed by MSF-Spain (Harare) and MSF-Holland (Marondera). Providing appropriate HIV care to trial participants will be the responsibility of Harare City Health and Marondera Hospital, and not the trial. There are ART clinics at both BRIDH and Marondera Hospital

Specimen export

For the purposes of i) quality control and ii) timely investigation, a 10% sample of sputum specimens collected for culture, a 10% sample of all positive cultures, and a duplicate of all specimen collected to investigate possible treatment failure or relapse will be sent to London by courier for parallel culture and drug sensitivity testing, according to the trial bacteriology protocol. Permission to export sputum specimens is requested for this purpose. Due to the different technologies in place, the London laboratory will be able to give culture and drug sensitivity tests within 7 to 10 days of receiving specimens, whereas the Harare laboratory will take 3 to 8 weeks for the primary culture alone and a further 3 weeks for drug sensitivity testing. Retreatment of patients with failure / relapse will be the responsibility of the trial, and will use predefined regimens chosen according to their drug sensitivity results, so that rapid provision of sensitivity results will directly benefit patients. For purposes of better understanding the pharmacokinetics of the two trial regimens, a sub-sample of 70 patients in each rifapentine/moxifloxacin arm at BRIDH will be asked for separate consent for 3 blood specimens taken 2 ± 0.5, 5 ± 0.5, and either 24 ± 3, or 48 ± 3 hours after dose ingestion during the continuation phase of their treatment (sparse sampling). We request permission from MRCZ and will make separate application to the Research Council of Zimbabwe to export 3mls of frozen plasma (courier on dry ice) for analysis by the Pharmacology Department, University of Cape Town, South Africa.

 Two South African sites (Cape Town and Johannesburg) will also contribute, making a total of 200 participants in each trial arm, and a further 30 participants from Cape Town will contribute specimens for intensive PK sampling (10 samples over 24 hours). Rifapentine and moxifloxacin levels will be analyzed using liquid chromatography-tandem mass spectrometry and structural PK models built for each of the drugs (rifapentine at doses of 1200 mg/week and 900 mg twice weekly, and moxifloxacin at 400 mg/week) to describe the concentrations of each drug over time, in the study population as a whole, including any differences conferred by covariates such as age, sex and weight. There are scientific benefits from having all specimens analysed by a single unit (less measurement variability), and it also a theoretical future benefit for Zimbabwe in having Zimbabwean participants included in this analysis. Patients with failure/relapse will be compared to those with a satisfactory outcome for any differences in standard PK parameters (Cmax, half-life and AUC).

Risks / benefits

The main anticipated risks relate to the study drugs. Moxifloxacin has been used extensively throughout the world as an antibiotic, and is also used for prolonged periods as a second-line TB drug. It has an excellent safety profile. Rare serious adverse effects include neurological toxicity (seizures, psychiatric disturbance) and tendonitis that can progress to rupture. Rifapentine has been widely used in the United States since approval in 1998 for treatment of TB and latent TB infection.

The manufacturer’s recommended dose is 600mg once or twice weekly, but there is increasing experience with higher doses (900mg and 1200mg), which appear to be as well tolerated as the 600mg dose. There is growing recognition that higher doses are preferable, because rifamycin-monoresistance can emerge at the 600mg dose. Rifapentine is closely related, and has a similar adverse-event profile, to rifampicin (the TB drug it will replace in this trial) and is no more toxic. Rifapentine does not cause “flu-like” syndromes due to hypersensitivity on intermittent dosing (unlike rifampicin). Both rifampicin and rifapentine can cause hepatitis, and both cause frequent drug-interactions through induction of the cytochrome p450 system. Experience with rifapentine in pregnancy is limited, and is not recommended. In this trial, pre-enrolment pregnancy testing will exclude pregnant women from entering the trial and women of child-bearing age will be given advice on contraception (rifamycins can lead to failure of oral contraceptive pills). A further risk is relapse with rifamycin monoresistant organisms. The risk is considered small in this trial (because of the use of 900mg /1200mg doses of rifapentine). There will be active surveillance for this event, and suitable alternative retreatment regimens that have shown good outcome in HIV-positive patients will be provided by the trial. All adverse events will be reported, according to GCP guidelines.

Serious adverse events will be reported to the Trial Manager at MRC Clinical Trials Unit (UK) and MRCZ within 24 hours. An independent data safety and monitoring committee (IDMC) will meet every 4 to 6 months during the trial to consider safety data, and to monitor the conduct of the trial with respect to ethical aspects. The IDMC will report to the Trial Steering Committee. Benefits will include the indirect benefits of being a trial participant (TB treatment under trial conditions shows better outcomes than treatment under routine conditions) and, for HIV-infected patients, closer liaison between TB and HIV services than may occur in routine settings. A major benefit to HIV-positive clients taking the trial arms will be that isonizid, which commonly causes or exacerbates peripheral neuropathy especially when taken with D4T, is not part of trial regimens.

Costs and compensation

Patients will receive a meal (2 boiled eggs and bread) before taking rifapentine, and will be compensated for their transport and time (approx 3 to 5 USD equivalent per visit). Insurance against medical costs incurred by trial participants from negligent and non-negligent harm has been taken out with Royal and Sun Alliance (GBP 5,000,000 total liability for all 1,250 trial participants).

Confidentiality assurances

Data will be stored under an allocated study identity number, plus the initials and data of birth of the patient. Files will be kept in locked cupboards, and electronic data will be held in secure databases. HIV testing will be offered with counselling and receipt of results to all patients, according to the national policy of opt-out HIV testing for all TB patients. In the event of patients consenting to provide blood for HIV-testing without wanting to know their results, blood will be run under the identity number only, with results stored separately from other trial data to ensure no breach of confidentiality.

 Conflict of interest

None

Collaborative agreements

Approval was obtained from the ethics committees of St Georges Hospital, London, London School of Hygiene and Tropical Medicine, Biomedical research and Training Institute, Medical Research Council Of Zimbabwe and the Medicines Control Aurthority Of Zimbabwe. Permission to export samples was obtained from Research Council of Zimbabwe. Biomedical Research And Training Institute, Institutional Review Board also approved the study.

Intended use of results

If successful the results of this trial would be of global relevance: shortening TB therapy and once-weekly treatment regimens have been prioritised as goals likely to lead to better global TB control. Implementing a change in the international recommendations for first line TB therapy in resource-limited settings would require consensus, negotiated price reduction, and pre-approval of manufacturers. Mechanisms exist within UN Agencies for each of these steps if this trial is successful. The Chief Investigator and the British MRC have both played an important role in developing the 6-month TB regimes currently in use in Zimbabwe, and were responsible for “Trial A”, reported in 2004 that has resulted in a move away from the 8 month regimen previously in use in Zimbabwe and other African countries (because of high relapse rates). As such the investigators of this trial are unusually well placed to influence future policy and practice.

Funding and trial governance

The study is funded by the European and Developing Countries Clinical Trials Partnership (EDCTP), with the Chief Investigator being Dr Amina Jindani of St Georges University of London; Department of Cellular and Molecular Medicine.

The Medical Research Council (UK) Clinical Trials Unit will supervise data collection and compliance with Good Clinical Practice at each trial site. The local principal investigators are Professor IT Gangaidzo ,Dr Stanley Mungofa (co-PIs for BRIDH site), and Dr Simukai Zizhou (Marondera Hospital)