Reaction to the trial, published this year in the New England Journal of Medicine, has split the South African HIV/AIDS community with a number of HIV clinicians criticising the study design.

But Professor Salim Abdool Karim, Pro Vice-Chancellor at the University of KwaZulu-Natal, says that it is easy to judge his trial with the benefit of hindsight.

‘€œBut when the trial was designed, there were no clear guidelines about whether TB and HIV could be treated together, mainly because of concerns about how patients would tolerate the drugs, how the drugs would interact and concerns about ‘€˜Immune Reconstitution Syndrome [a resurgence of latent infections],’€ said Abdool Karim.

Between 2005 and 2008, Abdool Karim and colleagues conducted a trial on 642 patients on TB treatment to establish when they should start antiretroviral treatment.

The trial, ‘€œStarting ARV therapy at three points in TB therapy’€ (SAPiT), divided patients into three groups. The first two groups started ARVs while on TB treatment, while the third group only started ARVs once they had completed their six-month TB treatment.

Twenty seven people (12.7 percent) died in the ‘€œsequential arm’€ where ARVs were delayed until after TB treatment was finished. This was double the deaths in the other two groups (5.8 percent) which integrated TB treatment and ARVs.

But initially more deaths took place in the integrated arm of the trial. Most of the deaths in the sequential arm only happened after six to ten months when all patients had already finished their TB treatment. In other words, the benefit of combining TB treatment and ARVs took some months to emerge.

Bio-ethicists Sean Philpott and Udo Schuklenk described the trial as ‘€œdeeply flawed’€ and estimated that there had been around 10 ‘€œpreventable deaths’€ in the trial on the website of the Hastings Centre bioethics forum.

‘€œThe failure of this research ethics committee to recognise the clinical, ethical, and legal deficiencies in this study is shameful and, we hope, will be investigated by the relevant South African authorities,’€ charges Philpott, associate professor of bioethics at Union Graduate College in the US and Schuklenk, Research Chair in Bioethics at Queens University in Canada.

A number of HIV clinicians have also criticised the trial. Their key concern is that the patients in the ‘€œsequential arm’€ had low CD4 counts (average 140) yet their ARV treatment was delayed for six to nine months, despite ARVs being recommended for people with CD4 counts below 200.

Dr Francois Venter, head of the Southern African HIV Clinicians Society, acknowledged that it was an ‘€œawkward situation, as I have great respect for the researchers involved’€.

But Venter said that the study design was ‘€œbad’€ and that the researchers could have got similar results if they had recruited a bigger group of patients with CD4 counts above 200.

‘€œThere are a number of questions that need to be answered to clear the air,’€ said Venter. ‘€œWere there senior HIV clinicians looking after the patients? Why was there such a long delay in patients in the sequential arm getting ARVs? Why did the KwaZulu-Natal ethics board approve the study?’€

This week, Dr Douglas Wilson, head of clinical services at Edendale Hospital, and Dr Graeme Meintjes from the University of Cape Town criticised the trial in a letter to the New England Journal of Medicine.

‘€œThe South African National Department of Health guidelines at the time of the study recommended initiating antiretroviral therapy in patients with CD4 counts of less than 200 cells per cubic millimetre after they have completed two months of tuberculosis treatment,’€ said Wilson and Meintjes.

‘€œThe SAPiT study did not offer the standard of care to patients in the sequential-therapy group, resulting in advanced immuno-suppression that remained untreated for up to nine months and higher mortality.’€

In response, Abdool Karim said: ‘€œThe World Health Organisation guidelines, which were the basis of the 2004 South African guidelines, categorically state that recommendations on the initiation of antiretroviral therapy in tuberculosis are ‘€˜provisional’€™ and ‘€˜pending ongoing studies’€™, since the ‘€˜optimal time to initiate [antiretroviral agents] in patients with [tuberculosis] is not known’€™.’€

Doctors overseeing those on the trial could put them on ARVs at any time if they showed signs that they needed it, and this happened in seven percent of cases.

International TB experts have also defended the trial.

‘€œThe criticisms of SAPiT are unjustified, in my view, and represent a retroactive application of the very knowledge that the study generated,’€ said Dr Richard Chaisson, a world renowned TB expert, Professor of Medicine and International Health Director of the Johns Hopkins University Centre for Tuberculosis Research in Baltimore.

‘€œThe SAPiT trial has provided extraordinarily important data that will ultimately save thousands of lives.   When the study was designed and conducted, there was no reliable information on the timing of antiretroviral therapy in patients with HIV and TB,’€ added Chaisson.

‘€œAt that time, the major concerns of clinicians were immune reconstitution syndrome (IRIS), drug interactions and additive drug toxicities.   As a result of these concerns, almost no patients with HIV/TB were receiving antiretroviral therapy, not only in South Africa, but in the US and Europe.  

‘€œIn our clinics here in Baltimore, almost all clinicians opted to wait until TB therapy was completed before starting ART to avoid complications.   Most clinicians believed that a delay in starting treatment was not detrimental.   In fact, a mantra of the time was ‘€˜ART is never an emergency, but TB therapy is’€™. The SAPiT study clearly showed that earlier initiation of ART was safe and lifesaving, and that IRIS was not the serious problem that most clinicians had anticipated it would be. ‘€œ

Professor Scott Hammer, head of Columbia University’€™s Division of Infectious Diseases, said the trial was ‘€œethically designed and executed and has contributed truly important data to the HIV-tuberculosis co-infection field’€.  

UNAIDS adviser Dr Catherine Hankins said it was ‘€œjustifiable to run the sequential arm in the manner that it was done at that time for several reasons’€.

‘€œFirst, it was common practice at the time in most TB programmes to defer treatment until at least the intensive phase was over, if not to the end of TB treatment,’€ said Hankins.

‘€œSecond, the HIV status of many TB patients was not known because many TB programmes were not even offering HIV testing to their patients, despite recommendations from WHO and UNAIDS in 2004. Third, WHO needed unequivocal evidence to make clear recommendations about starting all people with TB on antiretroviral drugs as soon as possible ‘€“ and the SAPIT study has now provided exactly that. ‘€“ Health-e News Service.