When to Start HIV Therapy
Doctors give their opinion
on the continuing debate
by Enid Vázquez
When is the best time for people with HIV to start therapy?
No one knows for sure, but the estimate has gone up and down over the years. Now at CROI, and previously at ICAAC (see the January/February Positively Aware), a large study continues to show that you have a greater chance of survival if you start HIV therapy earlier, when you have more than 500 T-cells.
According to a revision of U.S. HIV treatment guidelines in December 2007, the suggested time for starting therapy is when T-cells drop to 350. Previously, the guidelines suggested that therapy begin when T-cells drop to 200. There are circumstances when it’s recommended that a person without symptoms begin therapy before reaching 350, for example, during pregnancy or in the presence of kidney disease or hepatitis. (Visit www.aidsinfo.nih.gov.)
The NA-ACCORD study (North American AIDS Cohort Collaboration on Research and Design) found a 60% greater survival rate for people who started anti-HIV treatment above 500 T-cells, compared to those who waited. This survival rate was determined after the study adjusted for potential confounding factors, such as viral load or hepatitis C infection. An observational study like NA-ACCORD is prone to confounding—real life issues that can make its statistics invalid. The study excluded, however, injection drug users. The observational period was 10 years. The study was published in the April 1 issue of The New England Journal of Medicine (NEJM). According to an accompanying editorial, “The strengths of the study notwithstanding, the results of the NA-ACCORD study cannot be considered definitive evidence that everyone with HIV should start receiving antiretroviral therapy.”
Another observational study presented at this year’s CROI, the ART Cohort Collaboration (“ART” stands for antiretroviral therapy), also found a larger survival benefit to starting therapy earlier. The benefit, however, only went up to 400 T-cells. Beyond that, there was no difference in survival. (The Collaboration looked at increases of 50 T-cells at a time, while NA-ACCORD only compared individuals in the 350 to 500 T-cell range, and then those people above 500.)
There’s a lot to consider, however. Mari Kitahata, M.D., of NA-ACCORD, noted that, as research continues to show the negative effect of HIV on the body, such as inflammation and permanent damage to the immune system, the hope that earlier therapy can increase survival continues to grow.
At the same time, newer HIV drug combinations are much more tolerable, making them less of a problem for starting therapy in the first place.
With all the complications of this scenario, Positively Aware polled HIV specialists for their take on data pointing towards earlier treatment. See also News Briefs on page 16 for a new study, START, designed to help answer this question, and the Fall 2008 issue of RITA! at www.centerforaids.org.
Ross Slotten, M.D., Klein & Slotten Medical Associates, Chicago
I believe in earlier treatment. HIV is one of the few treatable diseases where we choose to wait to treat rather than treat as soon as someone is diagnosed with the infection. For example, if someone is found to have syphilis, we don’t advise the person to return in six months to see what happened without treatment. Similarly, we should not delay treating HIV infection, at least in theory. The development of lipodystrophy has been the main obstacle to early treatment. However, newer combinations, especially those without AZT and Zerit, do not appear to be associated with this terrible syndrome. What we don’t yet know is what other long-term side effects might arise from antiretroviral therapy. The fact that abacavir may be associated with an increased risk of coronary artery disease is a real concern. Nevertheless, it is unlikely that any side effect will ever surpass the deadly consequences of untreated HIV infection. Moreover, treating HIV also reduces transmission of the virus to uninfected people, which is another potent argument for initiating therapy early in the course of the disease.
Frank M. Graziano, M.D., Ph.D., University of Wisconsin Hospital and Clinics, Madison
I have always believed the following:
- If there is someone who comes to me and has found out he/she has been infected within the last 2-3 months, I will always start them on therapy (I don’t find very many of these).
- If someone comes to me and tells me they want therapy no matter what their CD4 count or viral load is (high or low) or no matter what I say, I will start them as long as we do enough education to ensure they understand what they are getting into. I figure they know something I don’t know (not many ask for this).
- If someone comes in with a rapidly falling CD4 count (even 900 going down to 500) over a relatively short period of time, I will start them.
- Same with viral load. I feel uncomfortable with someone having a viral load greater than 100,000 for a long time.
- Basically, I make a decision based upon each individual who is sitting in front of me (not just guidelines).
All of this does not answer the question. Based on the recent data from the new pathogenesis of HIV and the role of inflammation in this process, I believe that we should be starting patients (with some of the newer HIV drugs) within the first few times we see them—whether high or low CD4 count. I wish I could tell you that I do this. The thing that stops me is the uncertainty of what will happen with long-term use of the drugs. I believe that there are side effects with every drug we use (it is just the severity of the problem seen). We need more clinical studies now to see how to approach this problem. My honest feeling is that I will soon start having patients on HIV medications at much higher CD4 counts than what we have now.
I am uncertain I answered the question scientifically. It comes more from 28 years of experience with treatment.
Dr. Graziano provided the doctor’s perspective in this year’s Annual Positively Aware HIV Drug Guide.
Joel Gallant, M.D., M.P.H., Johns Hopkins University School of Medicine, Baltimore
We know from both observational and clinical trial data that people should be treated if they have AIDS, severe HIV-related symptoms, or CD4 counts below 200 cells/mm3. More recently, observational studies led guidelines committees to recommend therapy at CD4 counts of 350 cells/mm3. Now, we have data from the NA-ACCORD cohorts showing a 70% lower all-cause mortality in people who started treatment at CD4 counts between 350 and 500 cells/mm3 compared to those who waited until their CD4 counts fell below 350 cells/mm3, and a later analysis from the same study showing a 60% lower mortality in those who started with CD4 counts above 500 cells/mm3 vs. those who waited until their CD4 count had fallen below 500 cells/mm3. NA-ACCORD may be the most important observational study to help answer the question of when to start, but as with any observational study, it has potential weaknesses: most importantly, that people who start therapy early may live longer for reasons that have nothing to do with when they started. They may be the type of people who are less likely to die simply because they keep themselves healthy.
Many HIV experts feel we need a large randomized trial to answer this question. However, such a trial would be expensive and would take years to complete. During the course of the trial, changing practices and guidelines might make it impossible to enroll. In the end, it might help to answer the question of whether the best CD4 threshold is 350 or 500 cells/mm3, but how important is that distinction? For a newly diagnosed 25-year-old, starting therapy at 500 might mean 1–2 years more treatment than starting at 350, but that’s 1-2 years out of 4-5 decades of treatment over a lifetime.
In addition, it’s possible that our focus on CD4 thresholds is misguided. Some of the problems caused by HIV infection are present from the first day of infection, regardless of CD4 count. There is evidence of heightened levels of inflammation, immune activation, and tendency to develop blood clots, and these problems are more closely related to viral load than to CD4 count. These changes could help explain the increased risk of so-called “non-AIDS-related events,” such as cancers, heart attacks, and liver disease, that are now getting so much attention.
as research continues to show the negative effect of HIV on the body, the hope that earlier therapy can increase survival continues to grow.If you ask most HIV experts what they would do if they were infected, they usually say, “I’d start treatment regardless of CD4 count and viral load.” I believe that antiretroviral therapy should be considered in every patient. Some people should wait: mainly those who simply aren’t ready, willing, or motivated to take and be adherent with therapy. Some would argue that long-term non-progressors should wait, although we know that despite their high CD4 counts and low viral loads, non-progressors still have higher levels of inflammation and immune activation than people with undetectable viral loads on ART.
HIV infection is unique among infectious, transmissible diseases in that, despite being treatable but fatal if left untreated, our approach has been not to treat until there is evidence to support treatment. For any other such disease, the “default” would be to treat, and the burden of proof would be on those who wanted to wait. How did we end up with such a backwards approach to therapy? The reasons are historical: we began with no treatment at all; then we had fairly ineffective therapy; then we had effective but difficult and toxic therapy. Finally, in the third decade of the epidemic, we developed effective, well tolerated, and fairly non-toxic treatment. Had we begun with the treatment we have today, we probably wouldn’t be spending so much time arguing about when to start.
Dr. Gallant is the author of 100 Questions & Answers About HIV and AIDS and a frequent contributor to TheBody.com. He is also editor-in-chief of the HIV Guide for Johns Hopkins, which includes an interactive forum. Visit www.hopkins-hivguide.org.
