Much of the media attention surrounding the Conference on Retroviruses and Opportunistic Infections in Atlanta this week surrounded a single case report, of a toddler that was functionally cured of HIV through rapid and aggressive treatment starting within days after birth.

From a public health perspective, however, the more remarkable news may be the attempts to find standard ways of curing HIV in anyone who is infected, including eradicating the virus in patients with longstanding infections.

At the opening plenary session, Robert Siliciano, of the Johns Hopkins University School of Medicine, reminded his audience "how remarkable it is that the field has advanced to a point where eradication is being discussed" for a disease that was once an all-but-certain death sentence. "Nevertheless," he added, "the practical challenges remain daunting."

When highly active antiretroviral therapy, or HAART, was first introduced, it was thought that it might be curative. But that hope was not to be fulfilled. Instead, Mario Stevenson, of the University of Miami Miller School of Medicine, told the audience at a talk on "Research on Viral Reservoirs, Persistence, and Cure," in every single individual that stops therapy, HIV returns, no matter how long it has been suppressed. "HIV-1 has figured out a way to hang in there by its fingernails," Stevenson said.

The reason is that while HAART prevents the infection of new T cells, it does not affect T cells that are already infected. Most of those cells die off on their own soon enough, with their life spans measured in days.

But some are longer-lived. And every now and then, HIV will manage to infect a cell just as it is going from an active to a memory state, resulting in a virus that is integrated but not transcribed – a situation Siliciano called "a perfect recipe for persistence."

The toddler whose cure with drugs was reported at CROI this week was likely cured because the virus did not have time to establish such a reservoir. Other patients have been cured by bone marrow transplants from donors whose T cells are naturally resistant to HIV infection.

But in its attempts to cure patients using only drugs, Stevenson said, "the world is wrestling with what would appear to be a simple task: flushing out the virus from its hiding places, wherever those may be."

Stevenson said among the challenges is to figure out what the virus is actually doing during so-called latency. There are two competing ideas, though Stevenson noted that they are not necessarily mutually exclusive.

One is that the virus is hiding in resting memory T cells. But "the less popular view is that there may be some degree of residual replication," in which new viral particles are produced, but at a rate that is too low to make any sort of clinical difference.

The two theories are different in one important way – namely, what they predict will happen if antiretroviral therapy is intensified. If there is no viral replication going on in the first place, adding an extra drug to prevent replication will have no effect on the reservoir. But if there is HIV replication going on that is not detected by current assays, then intensifying therapy – in effect, working harder to suppress replication – could be a useful strategy.

At the conference, a clinical trial testing the addition of integrase inhibitor Isentress (raltegravir, Merck & Co Inc.) showed changes that suggested the intensification was having an effect on HIV replication, even though such an effect could not be detected by the currently used methods.

The fact that it took very specific measurements to see an effect of the trial illustrates a point that Stevenson made in his talk, where he expressed skepticism about whether cure trials as a whole are actually ready to measure progress toward a cure.

"It's a little bit of the cart before the horse," Stevenson said. "We're using investigational agents to try and eliminate these reservoirs, and I'm not sure that the actual assays . . . that we're using to gauge the impact of these agents are really up to the task."

Clinical trials trying to get at the reservoir would do well to make sure that they are actually using assays that will tell them what they need to know. "Otherwise," Stevenson said, "we may miss short-term differences in reservoir size after an intervention" that provide subtle but important evidence that a treatment is actually having an effect.