A session on tuberculosis and HIV at the South African AIDS Conference in Durban last week, sponsored by CAPRISA - the Centre for the AIDS Programme of Research in South Africa - was largely given over to discussing a pilot project at the Prince Cyril Zulu Communicable Diseases Clinic (formerly, the Durban Chest Clinic), which could form the basis for a large scale programme to integrate ARV treatment with TB care.
The pilot study for the planned START trial, as presented by Dr Chris Jack, enrolled 20 patients diagnosed with TB and HIV, 15 of them female, 5 male, to look at the feasibility of combining a once-daily triple combination of ARVs with a fixed-dose four-drug TB treatment regimen. Two patients discontinued after a few weeks and the others continued. 84% achieved undetectable HIV viral load by 24 weeks with two cases of virological failure - showing resistance mutations to all thre drugs used. 17/20 achieved TB cure, which was substantially better than expected, based on past clinic records for similar populations. There were no significant drug toxicities (though patients with a history of pancreatitis or liver disease were excluded from the trial).
The ARV regimen consists of ddI (didanosine), 3TC (lamivudine) and efavirenz, provided by the Indian generic company Cipla in a single blister pack as ‘Odivir’. The TB treatment, for the initial stage of treatment, was with 'Rifafour', containing rifampicin, isoniazid, pyrazinamide and ethambutol. Some patients also received co-trimoxazole prophylaxis, which is in fact recommended for all co-infected people and may be especially important when the CD4 count is below 200 (in this study, 6 of the patients were in that category). In addition, all women who were at risk of pregnancy were required to be on effective contraception, in view of the risks of potential harm to a baby from efavirenz.
During the week, all of the drugs are given at the clinic as directly observed treatment, first thing in the morning. This timing is not ideal, as psychological effects of the efavirenz are liable to affect daily living more than if the drug were taken last thing at night, but is the only practical way to do it. At the weekend, patients take two doses of the ARVs home and assume responsibility for themselves.
The planned clinical trial, the START study, will randomise some patients to take ARV treatment starting as soon as they begin their treatment for TB and compare them to others who will commence ARVs only after finishing 6 months of treatment for TB. The reasons for preferring a later start for ARVs include overlapping toxicities of the drugs used and evidence that rifampicin can reduce the level of efavirenz in the body. Reasons for preferring an earlier start include the high mortality seen among people with HIV during treatment for TB and the possibility that starting ARVs as directly observed therapy may improve the chances of long term success.
A presentation by two medical students highlighted the impact of the pilot treatment study on patients and staff at the clinic, making an important point about infrastructure for ARV delivery in South Africa and other countries. The majority of patients treated for TB in KwaZulu Natal are HIV positive. At present, if they survive to the end of six months treatment for TB, many die soon after, of other infectious diseases. This has demoralised both patients and staff, and led to poor completion rates for treatment. On the other hand, even the limited prospect of expanding access to ARVs represented by the planned trial has given new hope and a stronger incentive to persevere with TB treatment. A high proportion of patients at the clinic express willingness to take ARVs and the availability of ARVs definitely affects their readiness to be tested for HIV.
Reviewing drug interactions, Dr Andy Gray observed that the clinical significance of the reduced blood levels of efavirenz seen when it is co-administered with rifampicin was sufficiently unclear, that clinical trials might be justified. Some clinicians were increasing the daily dose from 600mg to 800mg, but this had been linked to reports of more frequent unwanted effects. It might also be necessary to consider varying the dose of efavirenz according to body weight, since patients under 50kg achieved twice the blood levels of those over 50kg when given the same dose.
Dr Quarraisha Abdool Karim, who introduced the session, observed that treatment for people with HIV who are coinfected with TB would be an essential strategy for targeting ARV treatment to those South Africans who could benefit most from it, and where the benefit to public health would be greatest.
The South African government has rightly stressed the need for committed and adequately trained health care staff to deliver ARV treatment. However, the idea that the service can be put in place before drugs are supplied is questionable. To secure the training and commitment, and for treatment advocacy to be effective, there’s going to be no substitute for real concrete guarantees that the drugs will be provided.
Abdool Karim Q. Integrating TB and AIDS care - a feasible option for resource constrained settings? CAPRISA symposium, South African AIDS Conference, Durban, 2003.
Gray A. Concomitant TB and HIV treatment - more questions than answers. CAPRISA symposium, South African AIDS Conference, Durban, 2003.
Jack C. Starting antiretroviral treatment in adult patients co-infected with tuberculosis in resource constrained settings - the START Pilot Study. CAPRISA symposium, South African AIDS Conference, Durban, 2003.
Matha B. Impact of HIV/AIDS on health seeking behaviours among patients attending a primary care clinic in KwaZulu-Natal CAPRISA symposium, South African AIDS Conference, Durban, 2003.