To Start, or Not to Start?

Leading Advocates and Physicians Offer Their Thoughts on a New Study

July/August 2010

"The drugs we use today are safer, easier, and more convenient than the ones we used then. Their toxicities may be small compared to the long-term risks of HIV infection."

START (Strategic Timing of Anti-Retroviral Therapy) is a large, six-year international study which will enroll 4,000 treatment-naive individuals who have a CD4 count greater than 500. The purpose of this randomized, controlled study (the gold standard) will be to find out if the chance of developing a serious illness or of developing AIDS is less if patients start taking HIV medicines at a time when their CD4 cell count is still fairly high.

PA asked several leading physicians and advocates to give us their thoughts on START, the pros and cons, and what some of the issues are, especially in regards to enrollment of the study here in the U.S. with the recent changes to the Department of Health and Human Services (DHHS) treatment guidelines on recommending when to start HIV therapy.

To learn more about the study, and to find sites in your area that are enrolling, visit, click on "Search for Clinical Trials," and enter "START HIV."

Joel Gallant, M.D., M.P.H.

Joel Gallant, M.D., M.P.H.

Professor of Medicine and Epidemiology at the Johns Hopkins University School of Medicine's Division of Infectious Diseases, and Associate Director of the Johns Hopkins AIDS Service

A clinical trial to determine when to start ART sounds like a no-brainer. Trials give us definitive answers, without the bias and confounding that plague cohort studies. But will START give us the answers we're looking for?

START will take a long time and will be hard to recruit. People are already divided over the issue -- some embracing the "hit early" approach, others wanting to wait. It's a rare combination of clinician and patient willing to put the question up to the toss of a coin, and opinions will solidify further over the next few years. The new guidelines only make it harder to enroll a trial in which participants must be willing to be less aggressive than what is now considered standard of care.

Deciding to start therapy has always been a question of weighing risks and benefits. We know that HIV replication is bad for you, even at high CD4 counts, but starting early didn't make sense 10 to 15 years ago, when treatment had inevitable toxicity. The drugs we use today are safer, easier and more convenient than the ones we used then. Their toxicities may be small compared to the long-term risks of HIV infection.

"No one knows the correct answer and it may be that there is not a simple, single answer at all, but a much more complicated approach."

Speaking of long-term risk, let's say START shows no benefit to early therapy over a five to six year period. Does that prove we should defer ART? Not if you believe that untreated HIV infection has long-term consequences. For decades, we thought of HIV infection as a disease of immunosuppression: your CD4 count fell, and you became vulnerable to opportunistic infections and malignancies. But that wasn't the whole story. Even at high CD4 counts, HIV causes chronic inflammation and immune activation that can increase the risk of diseases we didn't before think of as being HIV-related, such as heart disease and non-AIDS-defining malignancies. HIV replication in the brain can cause cognitive deficits that won't necessarily get better with therapy. Deferring therapy may not cause problems in the short-term, but will it increase the risk of heart disease, lymphoma, or dementia 10-20 years from now? START can't answer that question, so it won't satisfy those of us who believe that HIV infection is always bad for you and should be treated.

Many view the decision to start ART as a decision that requires solid evidence, waiting being the natural default in the absence of data. But is that just an accident of history? In the early '80s we had no treatment. Then we had lousy therapy until the mid-'90s, when we got therapy that was effective but difficult and toxic. It's no wonder we needed clinical trials to prove we should treat people who were doing well. But if we'd had the drugs we have now in 1986, would anyone have insisted on a clinical trial before recommending early treatment? The decision not to treat is still a decision.

Prevention has been a dismal failure: People have abandoned condoms. People who proclaim the virtues of abstinence end up having babies or getting syphilis. Vaccines may be decades away. Circumcision is great if you're a heterosexual African man. Microbicides haven't worked yet, and PrEP isn't ready for prime time. Treating people with HIV may be the best form of prevention we have, and it's the only one that improves the health of the people who use it.

I would encourage those who haven't made up their mind about early ART to enroll in START, and I'll be very interested in seeing the results. In the meantime, I'm taking an "opt-out" approach to antiretroviral therapy; recommending ART unless there's a good reason not to treat.

Dr. Gallant is a member of the DHHS Guidelines Panel. The opinions he expresses in this piece are his own, and do not necessarily reflect those of the other guidelines panel members.

David Wohl, M.D.

David Wohl, M.D.

Associate Professor of Infectious Diseases and Co-Director of the AIDS Clinical Trials Unit at the University of North Carolina (see "Ready, Set ... When?")

I think a randomized trial comparing the strategies of initiation of HIV therapy above and below a CD4 cell count of 500, if feasible, would be of great interest. However, there remains the feasibility issue (regardless of guideline recommendations), the cost of this study, and the time until results will flow from such an effort. In addition, we should understand that those enrolling in such a trial may well be somewhat like those who opted to start HIV medication in the NA-ACCORD cohort, that is, folks diagnosed early in infection, motivated to consider starting therapy, and with perhaps better health-maintaining behaviors. Therefore, a randomized trial does not eliminate all biases and may not include those more likely to meet up with hard adverse endpoints (like death).

The December 2009 DHHS treatment guidelines, in my personal view, make it clear that there is considerable debate in the HIV community regarding starting therapy above a CD4 cell count of 500. The panel, like the rest of the world, was split on the issue. Those patients and clinicians who believe the data support a benefit of HIV treatment at higher counts may be less inclined to participate in such a study, as will those who want to avoid and delay therapy until absolutely necessary. Those who are unsure are perfect for the trial and there are many who remain in this category.

Lastly, we should avoid hyperbole and oversimplifications. The recent statements in support of the START trial that are e-circulating have a partisan tone and have made inaccurate assumptions regarding the change in the treatment guidelines that do not serve them or the trial well. The scientific and advocacy communities need to maintain an open mind regarding the when to start question. No one knows the correct answer and it may be that there is not a simple, single answer at all, but a much more complicated approach that takes into account a number of individual variables. Until more definitive data are available, no one should be faulted for taking stock of the information, albeit incomplete, and making a decision on what is best for them or their patient. That is called best judgment.

Bob Huff

Bob Huff

Treatment Activist

According to research presented at the 2010 CROI, San Francisco has already made significant progress in expanding treatment access, which may have resulted in lowering community viral load rates. It makes sense that they press forward to try and achieve even more success. The potential payoff for the campaign would be halting the spread of new HIV infections in the city and surrounding areas -- an important and amazing outcome if they succeed.

I understand why some are worried that the professional consensus about when to start treating HIV might shift without a basis in high quality evidence from a clinical trial. People are concerned that the opportunity to enroll and conduct such a trial might be lost -- and we will never be sure if the benefits of treating HIV whenever it is diagnosed outweigh the risks. I agree that a large clinical trial (the START study) of this question is essential, but I don't think communities should be restricted from doing what they think is best. And as evidence accumulates that viral suppression restricts transmission, this is a factor that individuals may also consider when deciding to begin treatment.

Randomized, controlled studies are the best way we know to produce evidence to support making critical medical decisions. The evidence about the best time to start treating HIV with antiretroviral drugs is still ambiguous. If immediate treatment prevents cumulative damage from ongoing HIV replication, then the right trial should show that. If the long term use of ARVs is more harmful than the untreated infection, then the trial should shed light on that as well. There are several theories, and many opinions, and some are convinced that early treatment is the way to go. But nothing changes medical consensus as decisively as results from a randomized trial. That's why it is essential that the START trial be given every opportunity to enroll and proceed and why doctors and people with HIV should do what they can to support START. The information produced by START will impact how millions of people with HIV around the world will be treated for decades to come.

The reports presented at CROI on lowering community viral load helped solidify my support for the idea of treating HIV whenever it is diagnosed (and not waiting for a CD4 count trigger). I hope Project Inform also reports on its experiences with communicating what it learns about the range of individual experiences with quality of care, protection of rights, and respect for people with HIV as this program goes forward. If universal testing and treating become the standard of care in the future, other cities and organizations will look to the experiences in San Francisco, Washington D.C., and other pioneer cities to learn what worked and what didn't.

Tim Horn

Tim Horn

President and Editor-in-Chief,

START naysayers argue compellingly that it takes, on average, three years for someone with HIV infection to progress from 500 CD4 cells to 350 CD4 cells in the absence of treatment. Considering that most people living with HIV will spend decades on antiretroviral treatment regardless of when they start, what difference does three years make?

A big difference, actually.

True, three years seems puny when considering the overall length of time a person will ultimately remain on antiretroviral therapy. But to someone newly diagnosed with HIV, they are three very important years. It is an intense period of adjustment for many people living with HIV -- a time to prepare for a long life with HIV and the lifetime of treatment that makes this possible.

"For nearly 15 years, people living with HIV have been pulled in different directions."

More times than not, newly diagnosed people enter care with much more on their plates than an immune-deficiency virus taking up residence in their bodies. Many are still reeling from shock, fear, stigma, and shame. Many have not yet learned to communicate effectively with health care providers. Many still lack a basic understanding of HIV, including its treatment and prevention. Many need to be lined up with vital support programming such as housing, mental health services, and drug addiction counseling. Finally, many need to begin taking appropriate steps to manage risk factors for co-morbidities including diabetes and cancer, along with liver, cardiovascular, kidney, and pulmonary disease.

The onus is on START to confirm, once and for all, whether antiretroviral therapy should be given top billing in the long list of priorities virtually all newly diagnosed individuals face upon entering care. If this randomized, controlled trial confirms profound disease-free survival benefits associated with early treatment -- without increases in the risks of side effects, adherence problems, and drug resistance -- the heightened priority will be justified. If, however, START ultimately concludes that treatment can be safely delayed until CD4s hit 350, clinicians and patients can remain confident in approaching care in a stepwise fashion.

For nearly 15 years, people living with HIV have been pulled in different directions based on (often persuasive) theories regarding the big treatment questions. START raises one of the most important questions of all. We owe it to ourselves, and to those who will inevitably come after us, to answer it.

Cal Cohen, M.D., M.S.

Cal Cohen, M.D., M.S.

Research Director of Harvard Vanguard Medical Associates and Community Research Initiative of New England, and Clinical Instructor at Harvard Medical School

Ever since we have had HIV treatment, there has been discussion about the benefits versus the risks and burdens of these medications, such as side effects or changes to lab tests. However, when HIV antivirals were given to people with significant damage to their immune system -- as measured by a CD4 count below 200 -- the benefits of treatment were dramatic and obvious. There was no doubt that someone on antivirals would clearly have a decrease in their chance of developing AIDS-related illnesses and death.

Two things happened. One, better and safer HIV treatments were developed, and two, there was a multi-year era of studying treatment interruption -- a strategy designed to preserve the benefits of these medications to the immune system, but also giving the body time off of them to minimize their cumulative side effects. That led to the definitive SMART study -- a large study comparing continuous treatment to a strategy of treatment interruptions -- using a CD4 count of 250 as the time to restart treatment. It is widely known that this study was stopped earlier than planned, as treatment interruptions using this strategy led to an increase in the risks of serious medical conditions such as heart disease and malignancies. While these events were rare, they were important enough to lead to the now accepted conclusion that once started, it is best to maintain treatment.

So, in the past year, groups of HIV experts met on behalf of the Department of Health and Human Services (DHHS) and reviewed all of these data plus other important studies and came to two conclusions. First, the committee recommended that anyone with a CD4 count between 350 and 500 start antivirals. Second, that while half of the experts felt that anyone with a CD4 count above 500 should start treatment, the other half said they weren't convinced. This should inspire researchers to do the definitive study comparing the two approaches.

"The short answer is that we all have opinions about what is best to do now, since without definitive data, this is the best we can do."

But if the guidelines panel all agreed about starting between 350 and 500, why doesn't the START trial change the starting criteria to be in that same range? One answer is that the study is consistent with the panel. But more importantly, the data about starting treatment with a CD4 count above 500 has some important flaws and gaps -- and the panelists were split. It is important to note that these same data informed the discussion and decision about what to do between 350 and 500. And just as there are no definitive data with a CD4 count above 500, there are no definitive data between a CD4 count of 350 to 500 either. Thus, the START trial is addressing the data gaps in both ranges and will provide the definitive answer.

Note that for anyone with a CD4 count above 350, the issue in starting HIV treatment is no longer a question about preventing AIDS-related illnesses by starting ART. These are fortunately rare in this higher CD4 range. The question is the issue of serious medical events, highlighted by the SMART study. Will HIV treatment minimize these events in people with these higher counts? This is the focus of the START trial. If most people are likely to be doing well without treatment for a few years, should we still start treatment? And what about the impact of HIV and the timing of treatment decisions on brain function, on bone, on blood vessels, on lungs, and even on transmission of HIV? What about the long-term side effects of even our best treatment now? Is it better to start everyone with HIV on treatment, to provide some benefits for some or maybe all of these outcomes? Or can someone postpone treatment for a few years, and then start with a good CD4 count without having lost any ground by waiting?

The short answer is that we all have opinions about what is best to do now, since without definitive data, this is the best we can do. For some, the answer about starting treatment is that therapy is right for anyone below 500; for others it is anyone at any CD4 count. And either of these may turn out to be the case. If treatment was perfect and all bad outcomes were reduced with treatment, of course we would recommend it to everyone at any CD4 count. But the challenge is to separate what we believe and want to believe from what we will learn from a study, where we can demonstrate the actual benefits and risks rather than just predict what we'd see if we had that study. And that is the role of the START trial. To provide the definitive data so we can move beyond a comparison of opinions and varying certainty, to an era in which we can see what does happen.

Got a comment on this article? Write to us at

This article was provided by Positively Aware. It is a part of the publication Positively Aware. Visit Positively Aware's website to find out more about the publication.

No comments have been made.

Add Your Comment:
(Please note: Your name and comment will be public, and may even show up in
Internet search results. Be careful when providing personal information! Before
adding your comment, please read's Comment Policy.)

Your Name:

Your Location:

(ex: San Francisco, CA)

Your Comment:

Characters remaining:


The content on this page is free of advertiser influence and was produced by our editorial team. See our content and advertising policies.