High Adherence May Become Less Essential the Longer a Patient Maintains Viral Suppression on HAART, Findings Suggest

I'm David Bangsberg, at Massachusetts General Hospital and Harvard Medical School. This study is based on a group of people who are homeless or marginally housed in San Francisco, who are on antiretroviral therapy.¹ We're measuring their adherence with random home pill counts, which we have found to be a very objective and reliable measure of adherence -- much better than self-report.

We follow these people over time and we ask the question: Among those patients who are virologically suppressed, what is the risk of virologic failure -- rebound -- as a function of the duration of prior suppression?

David Bangsberg, M.D., M.P.H.

The basic question is: Do you need the same level of adherence to keep your virus suppressed 12 months into therapy as you do the first few months of therapy? We hypothesized that the level of adherence required to sustain viral suppression would decline with time, possibly because the reservoir of latently infected cells declines with suppression.

Not that I know.

We used a statistical approach, followed marginal structural models, to look at each patient every month, look at their level of adherence, and look at the risk of virologic rebound. Then we stratified that as a function of how long someone had been suppressed. You can see that we have a population which is largely people of color, on diverse antiretroviral therapy. There's a high prevalence of injection drug use, crack use and alcohol use. So this is a population at risk for incomplete adherence.

As we looked at them, we broke adherence down into four different levels, and found that, at least for levels of adherence above 50 percent, the risk of virologic failure at any given level of adherence declines with time. Here are people who are taking 50 to 74 percent of their medications. This graph [on our poster] looks at the probability of virologic failure as a function of the number of months of prior viral suppression. Early on, you have about a 50 percent chance of virologic suppression with this level of adherence, and this declines to quite low as you maintain 12 months of viral suppression.

The interpretation is: The goal remains the same. The goal is to achieve as perfect adherence as possible. And the better the adherence, the better the chance of durable viral suppression.

What's the "magic" percentage?

There is no magic number, but with currently available therapies most patients can achieve reliable viral suppression at more modest levels of adherence -- 70 to 80 percent adherence. I want to just be clear that the goal is perfect adherence, but with more potent therapies, the window of adherence that can lead to viral suppression has opened up a bit. What this data suggests is that this window opens up a little bit further. It changes as someone gets deeper into, or has more extended duration of, full viral suppression.

Was this known before?

No.

Is this percentage difference because of the strength of the newer medications?

I think the more potent regimens have opened up this window of adherence, such that you can achieve virologic suppression within a window [of adherence] that's wider than 95 to 100 percent. How does the risk of virologic rebound change over time on a particular regimen? We're controlling for the type of regimen.

What does this mean for short-term interruptions of therapy?

We think that interruptions of treatment appear to be bad, in that you have low-level virologic replication, immunologic stimulation; and I think interruptions should be minimized.

Interruptions early into therapy may have more damage and lead to a greater risk of virologic suppression than interruptions later into therapy. But still, the more interruptions, the greater the risk of virologic rebound.

Are you going to continue following this population?

We are still following this population, and we'll continue to follow this population at least for another year, depending upon our funding.

Were you surprised by the results?

I think that we hypothesized that this would be the case, based on what we know about HIV in latently infected memory cells, and that that population [o cells] declines over many months to years on treatment. Given that declining population, we hypothesized that you might have more of an adherence cushion late into viral suppression than early into viral suppression.

Thank you very much.

Reference

1. Rosenblum M, Deeks S, Van Der Laan M, Bangsberg D. The risk of virologic failure decreases with duration of continuous viral suppression for adherence levels >50%. In: Program and abstracts of the 16th Conference on Retroviruses and Opportunistic Infections; February 8-11, 2009; Montréal, Canada. Abstract 583.