What do treatment as prevention study results mean for treatment?

The findings of HPTN 052, the randomised study that recently demonstrated that providing early antiretroviral therapy (ART) to HIV positive individuals reduced transmission of the virus to their partners, confirmed what many researchers had long been saying — that treatment of people living with HIV may be one of the most effective ways to reduce the spread of this virus.

Evidence has been accumulating ever since the advent of combination antiretroviral therapy that people with suppressed or undetectable HIV viral loads, especially those on therapy, are a great deal less likely to transmit HIV than untreated persons. But the evidence provided by HPTN is undeniable — early treatment profoundly reduced transmission by 96% and the study had to be terminated early because it was considered unethical to allow the placebo arm to continue.

aidsmap.com has published extensively on the background on the treatment as prevention, including the rationale, cohort studies, modelling and controversies in a series of linked articles. However, there are a number of issues specific to treatment in sub-Saharan Africa and other low-income settings that need to be more fully addressed.

Although the findings of HPTN 052 were hailed as a ‘game-changer’ for global HIV prevention by UNAIDS, some advocates are uncomfortable with the phrase. Firstly, people working in the field needed little convincing of the effects of treatment as prevention — but advocates are hoping that this study provides overwhelming evidence to convince funders to step up and provide the resources to reach targets for universal access to HIV prevention, testing, care and treatment (in a considered way that also strengthens global health in general).

The evidence may not be persuasive enough for some. Already some critics have pointed out that the findings only apply to people who have been identified as having HIV and put onto ART, and that, at a population level in sub-Saharan Africa, where most remain untested, and people with early infection remain undetected and highly infectious, the effects of treatment may not be as profound.

Furthermore, it is not a once-off intervention like circumcision, and some economists have suggested that providing life-long ART is too expensive an intervention to consider, when there are many other issues competing for funding in an era of budget cutting. As advocates, we need to be aware of these arguments, and realise there is still much for us to do to push our agenda.

But to the extent that these findings revitalise the fight against HIV, they are a potential ‘game-changer’ for treatment programmes. Indeed, without effective treatment programmes, any ‘game-changing’ effects of HIV treatment on prevention will not materialise.

The preliminary results of the study leave a number of unanswered questions regarding the impact of treatment on both new infections and on the health of people who receive earlier treatment. Indeed, it raises some immediate questions on the effect of early treatment on both the health — and the human rights — of people living with HIV.

What does this study tell us about the possible benefits to people with HIV of early treatment?

For treatment as prevention to work, the people taking the drugs need to be convinced that early treatment is in their interest.

Does this study provide convincing evidence that treatment should begin earlier than current guidelines recommend?

It should be emphasised out that early treatment in this study was defined as having between 350 and 550 CD4 cells, a stage at which other cohort studies have shown some health benefits, such as a delay in the progression to AIDS, but have not been able to conclusively demonstrate a survival benefit. As the first randomised study to study antiretroviral treatment in less advanced HIV infection HPTN 052 is likely to add to the evidence of early ART’s benefits.

HPTN 052 showed a very significant reduction in the risk of extrapulmonary tuberculosis among those who received earlier treatment, and also a trend towards less morbidity and mortality in the early treatment arm. Until these results are analysed in more detail we do not know whether the trend in morbidity events alone was significant, or what volume of health service utilisation was avoided by earlier treatment. These data may become available by the time of the International AIDS Society conference in Rome in July, or in a paper now being prepared for journal publication.

Nevertheless, the TB finding alone is likely to strengthen arguments for prioritising early treatment for HIV-positive people in order to prevent new TB cases.The finding that TB was significantly less frequent in people who were treated earlier extends the finding of the CIPRA HT 001 study, which also found a reduced rate of tuberculosis among people who started treatment at a CD4 count between 200 and 350, when compared to treatment deferred until later. This finding adds further weight to the view that earlier antiretroviral therapy is a very powerful means of TB control in settings where the burden of both HIV and TB is high.

It has also been argued that earlier treatment will reduce the risk of developing non-AIDS defining serious events such as cancers, heart disease or liver failure.

A recently published study comparing rates of non-AIDS defining serious events in the United States and Botswana found that the rate of such events was around twice as high in the Botswana cohort, with the difference driven chiefly by cardiovascular events and by non-AIDS-defining malignancies. This finding could be interpreted as illustrating the consequences of late treatment initiation, but further confirmatory studies are needed from other resource-limited settings before we fully understand the burden of non-AIDS defining serious events and the most important correlates.

It is unclear if HPTN 052 will be able to provide detailed information about the impact of earlier treatment on these outcomes in low and middle-income settings. Although the study collected data on these outcomes from 2007, it may not have sufficient statistical power because the study closed much earlier than expected. This may also have limited its power to detect differences in mortality and in morbidity. More information about the impact of earlier treatment on these outcomes will come from the Strategic Timing of Antiretroviral Therapy (START) study, which is comparing treatment initiation at CD4 counts above 500, or at 350.

We also need to know more about long-term toxicity of antiretroviral therapy in people with higher CD4 cell counts. HPTN 052 began recruiting participants in 2005, and initially used AZT/3TC plus nevirapine, efavirenz or atazanavir in the first-line regimen. HPTN 052 will be able to provide more information about drug toxicity in people with higher CD4 counts than most other major clinical trials, and will thus be a valuable source of information about the safety of earlier treatment in non-Caucasian populations.

The questions about the risks and benefit of early treatment also need to be considered in the context of the rights and needs of people with HIV. Although many people with HIV would agree that preventing transmission to sexual partners is an important reason for taking HIV treatment, there is a risk that urging earlier treatment onto people with HIV places the responsibility for transmission squarely on people living with HIV, rather than recognising that their sexual partners have an equal responsibility to protect themselves. This is particularly problematic in countries where there is growing willingness to use the criminal law to prosecute people with HIV for transmitting HIV – or even for failing to disclose their HIV status. How long will it be before legislators – or judges - conclude that failing to take available treatment should be considered as contributory negligence in cases of HIV transmission?

Challenges in translation

Any attempt to expand treatment still faces a number of bottlenecks, many of which are attributable to a lack of trained staff and a lack of health systems information. All of these are likely to be well known to readers of HATIP.

Performance at each of these stages needs to be maximised in order to achieve anything approaching the 96% reduction in transmission rates seen in HPTN 052.

• Improving rates of diagnosis without coercion: Even though awareness of the benefits of antiretroviral therapy is spreading in the worst-affected countries, reluctance to test for HIV is still widespread, and the median CD4 count at the time of diagnosis remains very low. Improving the acceptability of testing for HIV, dispelling the stigma associated with HIV and informing people about the effectiveness of treatment will be critical steps in scaling up treatment so that it has an impact on new infections as well as illness and death. Models of community-wide testing that respect the rights of people not to test, and which ensure confidentiality, will be critical in achieving widespread uptake of testing.
• Linkage to care and retaining people in care prior to ART: referral from testing facilities and retention in care of untreated people continue to pose challenges for many treatment programmes. In some settings barriers such as co-payments for treatment and transport to clinics will pose problems.
• Capacity to initiate patients onto treatment: Even if rates of testing and diagnosis are high, this will have little or no effect in prevention terms unless people who are eligible begin treatment promptly. This is critically dependent on the capacity of the health system to start patients on treatment, and is likely to require task-shifting in order to expand the number of health care workers who can initiate and / or monitor treatment. Efforts to delegate prescribing or monitoring of antiretroviral treatment to nurses and other non-physician health care staff are at variable stages across sub-Saharan Africa, and will need much greater emphasis if health systems are to reach ambitious targets for treatment enrolment.
• Proportion of eligible persons on therapy: a survey of cohorts in 8 PEPFAR focus countries in sub-Saharan Africa showed that between 2005 and 2008 the median CD4 count at treatment initiation was 135 cells/mm³, indicating that any preventive benefit of ART would occur only after a long period of undiagnosed infection had elapsed. The proportion of persons on treatment will obviously depend on the CD4 threshold for starting treatment, and who is eligible to start treatment at what CD4 level.
• Proportion of patients retained in care: Some treatment programmes report high rates of loss to follow-up in patients who have already started treatment. In some cases patients are lost because they started treatment too late to save their lives, but in many cases features of the treatment programme are to blame.
• Proportion failing therapy: Although treatment failure and the risk of drug resistance has been held up as the great flaw in the `treatment as prevention` model, treatment failure in people who remain engaged in care is likely to be less of a problem than loss to follow-up. Nevertheless, a better understanding of which adherence interventions are most effective, and in which settings, will be critical in limiting rates of treatment failure. Earlier treatment carries with it the risk that if people fail treatment, they will require more costly second or third-line regimens. Early treatment without strong adherence support programmes and follow-up mechanisms poses the risk of earlier treatment failure and correspondingly greater costs in the future.

For all these reasons UNAIDS last year launched a new concept for thinking about treatment scale-up and the role of treatment as a prevention tool – Treatment 2.0. This has five components, or `pillars`:

• Creating a better pill and diagnostics: a fixed dose combination that has a low risk of toxicity, low monitoring requirements and a low risk of resistance would aid expansion of treatment, as would diagnostics that could be used anywhere to reduce the burden on health systems.
• Treatment as prevention as part of a combination prevention strategy.
• Stop cost being an obstacle by driving down the cost of treatment and monitoring
• Improve uptake of HIV testing and linkage to care
• Strengthen community mobilisation in order to improve the engagement in care of populations at high risk of HIV infection.

Treatment 2.0 provides an important aspirational goal for advocates, scientists, donors and policymakers, but the more immediate questions of how to address the bottlenecks that keep people off treatment and untested need to be resolved.

In a report to the South African parliament this week, Health Minister Aaron Motsoaledi showed that addressing the bottlenecks is, first of all, a matter of political will.

Although things are far from perfect in South Africa, the country has made dramatic progress in the past year towards scaling up treatment. As a result of a national campaign that has begun to treat the epidemics of HIV and TB as the country’s number one health problem after years of political denial and bureaucratic torpor, South Africa has made dramatic progress in less than a year.

.

SOUTH AFRICA'S PROGRESS TOWARDS UNIVERSAL ACCESS

 For treatment to have any impact as a prevention method, this is the minimum level of political commitment and organisational response that will be necessary to achieve high levels of diagnosis and treatment coverage.

A more comprehensive approach that aims to test and treat an even larger proportion of the population will be tested in a number of field implementation trials now being planned, with the intention of evaluating the impact of early and widespread treatment on new infections, behaviour and HIV-related morbidity and mortality. The studies will also evaluate the cost-effectiveness of offering wider treatment, and the acceptability of offering the HIV test on a universal basis.

These include studies in Uganda and KwaZulu-Natal, together with studies that embed antiretroviral therapy within more comprehensive packages of `combination prevention`. (See list here).

One challenge for future studies will be defining the components of a package of combination prevention. Contrary to the assumptions of many policymakers and donors, we still have remarkably little idea of the effectiveness of non-biomedical prevention measures in varying settings, or of how to tailor packages of combination prevention to local epidemic settings.

To take just one example, what should be the balance between investment in biomedical approaches such as circumcision and structural approaches such as empowering women and girls, and does this vary according to HIV prevalence? Although circumcision may result in a rapid reduction in risk of infection for men because it is a one-off procedure, it will not reduce the risk of infection for women and girls until it has delivered a long-term reduction in HIV prevalence for men.

Conversely, although economic empowerment for women and girls might reduce vulnerability to HIV infection by a variety of mechanisms, this is likely to be a long-run approach that cannot deliver rapid reductions in new infections.

Understanding how to strike this balance is part of the very complex research agenda that we face in learning how to develop combination prevention, and requires attention not just to the local epidemic – what UNAIDS describes as `know your epidemic` - but also attention to the local context, which means everything from understanding what influences people to change their behaviour, to the ways in which gender roles are understood and talked about in a specific society.

Another important research issue is understanding how scaled-up community-based campaigns to offer counselling and testing for HIV can be integrated with activities that improve TB case-finding across the whole community, expand the provision of insecticide-treated bednets and effective treatment for malaria, and improve the diagnosis and treatment of other common infections and diseases. The community mobilisation necessary for treatment as prevention to succeed also offers huge opportunities for other health gains to be achieved. We hope to look in more detail at some examples of these approaches later in the year.

Cost

A shift towards earlier treatment – and more widespread treatment – would obviously have enormous cost implications.

However there are signs that a possible policy shift is not being dismissed on grounds of cost. Despite the pressure on US foreign aid budgets, the US Global Health Initiative is already looking at how it might assess the cost-effectiveness of treatment’s prevention impact.

“We are in the midst of beginning a rigorous internal USG [U.S. government] discussion on just that question,” said US Global AIDS Ambassador Goosby said (see report). “What I’m now committed to doing is shepherding the dialogue within USG quickly so we will know what we are up against. I think this is a study that reaches the level [that] it needs to be fully looked at, needs to challenge the way we are doing business.”

In an interview with the Science Speaks TB/HIV blog Anthony Fauci, Director of the United States National Institute of Allergy and Infectious Diseases at the National Institutes of Health, said:

“If you don’t put in the money now, sooner or later you are going to have to pay at the end of the spectrum. And if you do that, at the end of the day, the amount of money spent is going to be much more. I am totally sensitive to the financial constraints we have, but I believe in the big picture of things, investing more money now is the way to go.”