Although there is no question that the use of IPT could reduce the burden of TB in people with HIV overall, the results from the Botswana IPT trial pose almost as many questions as they answered. So after the findings were presented, researchers, policy makers and patient advocates debated how the results should be interpreted, and expressed somewhat diverging views how they should affect policy.
WHO currently recommends that isoniazid should be provided to people with HIV when active TB is excluded, and IPT is one of the three I’s at the heart of its TB/HIV strategy, along with intensified case finding and infection control.
However the results of the Botswana study are likely to influence WHO recommendations very swiftly.
Policy in the works
“The results of the Botswana IPT trial are very timely because we are currently revising the policy of intensified case finding (ICF) and isoniazid preventive therapy (IPT) to re-conceptualise the implementation of IPT — not as a stand-alone intervention but intimately linked to ICF,” said Dr Haileyesus Getahun of WHO’s STOP TB Department.
WHO is holding a meeting in the last week of January 2009 to reformulate the ICF/IPT policy. As one of the most rigorous trials on IPT to date, the Botswana IPT study will have a significant impact on the new recomendations.
“Clearly the Botswana study, a randomised controlled trial, definitely falls into the strongest area,” said Dr Getahun, describing how the WHO currently weighs evidence in its guideline development process. “Once we’ve graded this evidence, we will consider the strength of the recommendation. But it’s not only the evidence but also the operationality, the implementation and other issues that will be considered.”
Numerous studies have shown that six to nine months of IPT can lower the risk of active TB — but IPT has been a hard sell to national TB programmes, despite the existing WHO recommendation. In fact, even some of the thought leaders in HIV have been questioning IPT.
“I’m not sure IPT is worth the trouble to be quite honest,” said Dr Francois Venter of Johannesburg Hospital at the HIV Implementers’ Meeting in Windhoek Namibia this year. “We really need to critically re-evaluate IPT. It has distracted national governments and TB programmes like nothing I’ve seen before — I believe it is a long time coming that people start questioning whether this is how we should be spending our resources.”
“IPT has not been a very popular intervention,” said Dr Helen Ayles of ZAMSTAR during a session at this year’s Union World Conference on Lung Health.
The short duration of IPT’s benefit: will IPT have to be given continuously?
Previous studies in Africa have reported that IPT’s benefit lasted at least a year and up to two and a half years.1, 2
“Why is the durability so much shorter in the Botswana trial compared to these other two trials?” asked Professor Gavin Churchyard. He pointed out that the two other trials were in the pre-ART era and that after treatment completion, patients were followed only periodically every three to six months (so to some extent the Botswana study may simply have detected cases earlier). But the more frequent clinic visits in the Botswana study might also put people at risk of exposure to TB.
“It makes one wonder whether transmission within the health services is contributing to more rapid loss of durability in the Botswana trial,” Prof Churchyard said, “which underscores the need for having infection control in our health services.”
The smaller study by Martinson et al found no difference between continuous IPT and short course preventive therapy in the intent-to-treat analysis. However, adherence was an issue in that study and the per-protocol analysis suggested that continuous IPT was more effective than any of the other preventive therapy arms.
Another study that could have a bearing on how this finding is interpreted has just been completed by the Tuberculosis Research Centre in India. The study compared a six-month course of isoniazid/ethambutol to continuous IPT — but the results will only be presented at the Conference on Retroviruses and Opportunistic Infections next February.
One downside of continuous IPT could be an ongoing risk of toxicity. While the Botswana IPT study suggested IPT was fairly well tolerated, and that most severe adverse events were seen in the first several months of IPT, Dr Getahun noted that Martinson et al reported a higher rate of severe toxicity in the continuous IPT arm.
Does the study provide reassurance to national programmes that have been reluctant to roll out IPT?
“Five years ago, I got tuberculosis... I had to suffer from TB disease when this might have been prevented,” said Carol Nyirenda, a patient advocate living with HIV and working with the Community Initiative for Tuberculosis, HIV/AIDS and Malaria plus other related diseases (CITAM+) in Zambia. But she said that national programmes are even refusing to provide IPT on an opt-out basis (when people with HIV request it). “I strongly believe this withholding of life-saving drugs is not ethical – we need to move from discussion to implementation urgently. Our lives are at stake.”
“What does it take to actually change practice on the ground?” asked Mark Harrington of the Treatment Action Campaign. “I think that it is obvious [the WHO guidelines] didn’t really have much of an effect on practice, and all the same issues are brought up no matter what the data say.”
But this study may change things, according to Dr Jeremiah Chakaya, formerly the national TB programme manager in Kenya and presently co-chair of the STOP TB Partnership.
“In Kenya, even though a lot of study indicated that IPT was useful, a lot of us clinicians did not believe in it, we thought it was too difficult, we thought it was impossible. But like some of the people in this room here, I’m a very recent convert of IPT,” said Dr Chakaya. “Many of us were concerned that in routine programme settings it may be difficult to exclude active tuberculosis and that in fact a lot of TB could leak through the TB screening procedures and inadvertently lead to isoniazid monotherapy and increased isoniazid resistance. What this study has shown us, is that in fact isoniazidresistance does not seem to increase in this study.”
In addition, he said that the study addressed several other key concerns of programme managers by showing that fairly high adherence was possible, and that IPT appears relatively safe.
But the study wasn’t designed to address operational concerns of managing ‘well’ patients, as noted by Dr Francois Venter at the HIV Implementer’s Conference in Windhoek this year.
“Wellness programmes are so shocking, not just in developing countries but in developed countries. We know that we cannot retain people within the system unless we have them on antiretrovirals — study after study tells us that — and yet we take these people, we diagnose them and we try to insert them into a system where all we have to offer them is isoniazid. And outside of very specialised programmes, every programme I see seems to haemorrhage people out of the system unless you put them on antiretrovirals — and we need to engage with that before we start offering IPT in my opinion.”
Dr Chakaya suggested that programmes might instead have to think about delivering IPT at another level of the health system — moving IPT out to the community and relying much more heavily on “community health care workers, patient support groups and things like that. But we need to remember that the more complex the intervention the less likely that the intervention will be taken to scale.”
The role of tuberculin skin tests
But the Botswana IPT findings showing no significant benefit of IPT in people with negative TSTs could wind up making IPT much more complex.
“I think one of the take-home messages - at least for me is: it would be difficult to justify the continual provision of IPT without a TST or an alternative to identify patients most likely to benefit,” said Dr Themba Moeti (the man who launched the Botswana IPT programme).
“I think it is clear that if TST-negatives do not benefit - and study after study shows that TST-negatives do not benefit from IPT - then TST-negatives should not be given IPT,” said Dr Chakaya. “It doesn’t seem right for us to know what science is telling us and do different things just because they are easier to do.”
Javid Syed of the Treatment Action Group argued that even a small benefit to TST-negatives could justify the routine prescription of isoniazid to people living with HIV, regardless of TST status.
“In fact, 15% of protection was provided to TST-negatives. And for HIV-positives, 15% is not a figure we can ignore,” he said.
However, the study did not demonstrate that degree of benefit with any level of confidence, nor did a Cochrane Group meta-analysis.
“All of us would have wished that the data would have shown incontrovertibly that everyone benefited and we don’t have to bother with TST, but the data are what they are… and to disregard this is scientifically and intellectually somewhat dishonest,” said Dr Ken Castro of the CDC. “And we’ve been reminded not to disregard the 14% who may have benefited from IPT if TST-negative but [this ignores] the higher side effects in that group — so where’s the balance in terms of risk-benefit?”
The bigger problem is that TSTs do appear to be difficult to implement in many resource-limited settings.
“TST is an obstacle to IPT," said Dr Kerrigan McCarthy of Johannesburg’s Reproductive Heath and HIV Research Unit earlier this year at the South African AIDS Conference.3
“Firstly there’s the hassle of the supply chain management of TST - quality tuberculin, syringes, PPD RT23, cold chain management.”
In addition, Dr McCarthy said that Johannesburg clinics reported wide variations in the proportion of TSTs interpreted as positive, indicating lack of consistency in performing and interpreting the test. Health care workers arrived at different conclusions about whether the size of the skin reaction, or induration, indicated a positive result or not.
“Quality control of TST is impossible! If the quality of a test cannot be controlled, that test should not be used.”
Indeed, many programmes have had difficulty keeping tuberculin for the tests in stock, nurses have to be trained on how to correctly interpret the result (especially in people with HIV), and people have to return to the clinic within a few days to have it read —which can be a challenge in settings where the clinic is far away, transport is limited, and people have to work to survive and can’t spend much time going to, and waiting in, a busy clinic. In fact, data suggest at least a quarter never come back for their readings — which means that many people who could benefit never come back for treatment. This is why WHO currently recommends that IPT be given to everyone with HIV if they live in a setting with a very high burden of TB.
“Tuberculin (for the TSTs) needs a cold chain and refrigeration. With HIV care and treatment being decentralised into the most peripheral facilities, how are we going to reconcile that?” said Dr Getahun.
“TST is a real barrier to the implementation of IPT,” agreed Professor Churchyard. “In South Africa, as an example, we had absolutely negligible IPT uptake when we had TST as a criteria for enrolling. A year ago, we took IPT out of the revised guidelines, which have now been rolled out and in the last year the number of people put on IPT has tripled.”
Requiring TSTs will mean more expense, training and supervision — and could simply become the new excuse not to offer IPT.
It’s a conundrum, especially for treatment advocates, who are put into the awkward situation of having to weigh the risk-benefit for the HIV community overall versus the benefit for the individual. Treating everyone would reduce the risk of TB by around 56% (marginally higher than the benefit offered by ART in the study) — but it would also result in treating three quarters of the population unnecessarily, and placing them at risk of toxicity. Perhaps that choice (preferably informed) ought to be left up to each individual.
References
[1] Quigley MA et al. Long-term effect of preventive therapy for tuberculosis in a cohort of HIV-infected Zambian adults. AIDS 15(2): 215-222, 2001.
[2] Johnson JL et al. Duration of efficacy of treatment of latent tuberculosis infection in HIV-infected adults. AIDS 15: 2137-2147, 2001.
[3] McCarthy K et al. INH prophylaxis: Lessons learned during implementation in Inner City Johannesburg: 2008. CREATE Three I’s symposium, 4th South African AIDS Conference, Durban, South Africa, 2009.