Achieving undetectable viral load is the primary objective of anti-HIV therapy, but for some people it’s an unattainable goal. But now, as PAUL KIDD explains, there’s encouraging news for those who can’t quite get there.
Undetectable. It’s the magic word of HIV treatment, a powerful shibboleth that draws a stark line between those who are responding to treatment, and those who are branded ‘treatment failures’.
For almost a decade now, undetectable viral load has been the ‘gold standard’ of HIV treatment. Study after study has shown that people who manage to lower their viral load to undetectable levels respond better to treatment and do better in the long term.
But not everyone manages to maintain consistently undetectable viral load. Most positive people have occasional unexplained rises in viral load (so-called ‘blips‘) and some people never get to undetectable at all. Faced with detectable viral load, patients and clinicians have to make a judgement about whether that individual’s treatment is failing, whether a change in therapy is needed, or whether to stay with the current regime. And despite all the research, making those judgement calls remains more art than science, often guided by gut feeling, instinct, or naked hope.
But recent research has started to deepen our understanding of viral load measurements, and while the goal of undetectable viral load is unlikely to change any time soon, there’s encouraging news about the significance (or otherwise) of low-but-detectable viral load.
The protease moment
The word ‘undetectable’ first entered our collective vocabulary at the watershed 1996 International AIDS Conference in Vancouver. That was the year that Dr David Ho, of the Aaron Diamond Research Centre in New York, stunned the world with the first HAART success stories. ‘The protease moment,’ as it’s come to be known.
On stage in Vancouver, Ho told the hushed crowds of the stunning success of the first protease inhibitors, and conjured up a trio of new and enticing terms to describe the future of HIV treatment: ‘viral eradication’, Ho said, was possible if you ‘hit hard, hit early’ and maintained ‘undetectable viral load’.
It was mind-boggling stuff, and it earned Ho the accolade of being Time magazine’s ‘Man of the Year’ in ’96. What we didn’t know then was that HIV could hide away in viral reservoirs, which would keep the goal of viral eradication out of reach, and we hadn’t yet seen the long-term side effects of combination therapy, something which has taken the gloss off the ‘hit hard, hit early’ approach to treatment, but undetectable remains very much with us today.
An imperfect tool
The importance of viral load testing to modern HIV medicine cannot be overstated. Together with CD4 counts (which have been used since the early days of the epidemic), viral load tests provide an invaluable tool for measuring an individual’s health, monitoring response to treatment, and predicting the future development of disease.
But viral load is an imperfect tool. It can tell us a lot, but it cannot tell us everything we need to know. And that magic word – undetectable – means different things in different contexts. The first viral load tests could measure down to about 400 or 500 copies per millilitre; these days the ‘ultra sensitive’ test, which has a limit of 50 copies/ml, is routine. There are expensive, specialised tests out there that can measure viral loads as low as two copies, but because of HIV’s ability to evade detection, even with the most sensitive test, ‘undetectable’ does not mean ‘zero’.
Viral load levels can be a predictor of HIV disease progression, especially in people who have never taken HIV treatment. This means that viral load can be useful in deciding when to commence anti-HIV treatment.
The MACS study enrolled 1600 HIV-positive gay men in the early years of the AIDS epidemic, collecting and freezing blood samples at six-monthly intervals. In the mid-1990s, these stored blood samples were analysed for viral load levels, and the results were compared with the medical histories of the participants, many of whom had since died. There was a strong correlation between higher viral load levels (above 55,000 copies/ml) and the risk of developing AIDS-related illness, among people with the same CD4 count.
While this study was an important advance in our understanding of HIV/AIDS, its conclusions need to be read with a note of caution. The results were adjusted for CD4 count, and CD4 count remains the single best predictor of HIV disease progression.
People with higher CD4 counts are very unlikely to develop AIDS-related illnesses, regardless of their viral load. Because of this, most doctors today use CD4 counts, not viral load levels, as their primary guide in deciding whether to recommend starting treatment. But a high viral load can be a warning sign that the CD4 count is on the verge of dropping, so by looking at both the CD4 count and the viral load, doctors and patients can make more informed decisions.
Where viral load tests come into their own is in measuring response to treatment. While the jury is still out on whether people with higher viral loads at the time they start treatment are less likely to respond well (different studies have produced conflicting answers to this question), viral load tests are invaluable in measuring the effectiveness of treatments in controlling the virus.
Most doctors recommend having viral load tests at the time that treatment is started (or changed) and again in about a month. Viral load measurements one month after treatment are strong indicators of the prospects of longer-term success: a large European study published last September1 found that people who failed to get to undetectable within four weeks were much less likely to be undetectable after six months. As explained below, having a low-but-detectable viral load may not be a major cause for concern, but undetectable viral load remains the gold standard and should be pursued, especially if you have other treatment options to explore.
Blips
If your viral load has previously been undetectable, the emergence of a test result in the detectable range can be a major headache. The first thing you and your doctor will want to do is determine whether the rise is a sign of treatment failure or just a ‘blip’.
Viral blips – isolated detectable viral load results in people who have previously been undetectable – are surprisingly common. The number of people in clinical trials who experience one or more viral blips ranges between about 25 and 40 percent. Blips may be a sign of transient rises in viral load, perhaps coinciding with a bout of illness, they may be the result of testing error, or they may be an early warning sign of impeding treatment failure.
Testing errors in viral load tests can and do occur, and may be more common with the more sensitive viral load tests which are now in common use. Doctors at the Chelsea and Westminster Hospital in London reported in 2002 that of 249 blood samples retested after returning apparent blips, 59 percent turned out to be testing errors. Because of this, most doctors recommend repeating the viral load test to confirm the result before taking any action to respond to what may be a testing error.
The significance of viral blips has been the subject of much debate and research. Some studies have concluded that viral blips are normal and no cause for concern, while others have suggested that people who experience repeated blips are more likely to develop sustained viral rebound in the future.
In a study2 presented at the Retrovirus conference in 2001, researchers suggested that people who experience more than one viral blip (defined as a viral load of between 100 and 500 copies/ml) are more likely to go on to develop treatment failure. People who changed treatment after their third blip were much more likely to return to undetectable viral load than those who did not. The risk of treatment failure was higher in people who had bigger blips.
If you experience a viral blip, deciding what to do can be difficult. Do you act early and switch to a new treatment combination, and run the risk of having to deal with side effects, or do you ‘wait and see’ whether your small rise in viral load goes back to undetectable?
A small US study3 provides some support for the ‘wait and see’ approach. This study followed 79 positive people who had viral loads between 50 and 500 copies/ml and who did not change therapy. After three years, 50 of the 79 (63 percent) had either returned to undetectable levels or stayed below 1000 copies, and in most cases had seen a rise in CD4 counts.
While this small, short-term study doesn’t reduce the importance of trying to get to undetectable if you can, it does offer some breathing space for people who have very low viral load and don’t want to switch treatments.
Low, but not undetectable
There is also encouraging news for those who are unable to get to undetectable, but who are able to get their viral load low.
A French study published in 2003 examined the medical records of 3736 HIV-positive people who had been treated with a protease inhibitor based HAART regimen and who had all achieved undetectable viral load4. After a year on treatment, 71 percent of the participants had had consistently undetectable viral load, 10 percent had viral load levels between 500 and 5000 copies, and 19 percent had viral load levels over 5000. The investigators looked at the three groups to determine whether there was any difference in the numbers of people who had either a decrease in CD4 count or the development of an HIV-related illness, and found no significant difference between the undetectable group and the ’500 to 5000’ group when they were followed for over a year.
Another study, conducted by American researchers and reported in 2004, looked at the relationship between viral load and the likelihood of progression to AIDS or death in people with HIV5. As with the French study, the 3010 participants in this study were placed into one of three groups based on their viral load, this time after six months of treatment: undetectable (under 400 copies/ml), low viral load (400 to 20,000) or high viral load (over 20,000). After following the participants for more than four years, the researchers found no significant difference between the undetectable and low viral load groups, but among those with viral load over 20,000, the outlook was clearly worse.
The results of these two studies suggest that for people with detectable, but low-level, viral load, perhaps as high as 20,000 copies/ml, the risk of developing AIDS-related illness is no greater than for people with undetectable viral load, at least in the medium term.
This is obviously encouraging news for those people who are unable to achieve or maintain undetectable viral load levels.
In the longer term, researchers still believe that persistently detectable viral load, even at relatively low levels, could lead to resistance and eventual treatment failure, so most doctors will still want to consider a treatment change if there are options available, in the hope of getting the viral load down to undetectable levels. But for positive people with fewer options or those who prefer not to change their treatment due to side effect concerns, these studies provide some welcome relief.
References
1 Smith CJ et al Use of viral load measured after 4 weeks of highly active antiretroviral therapy to predict virologic outcome at 24 weeks for HIV-1-positive individuals. J Acquir Immune Defic Syndr 37: 1155-1159, 2004.
2 Greub G et al. Low level HIV viral rebound and blips in patients receiving potent antiretroviral therapy. Eighth Conference on Retroviruses and Opportunistic Infections, Chicago, abstract 522, 2001.
3 Natural History of Patients with Low-Level HIV Viremia on Antiretroviral Therapy. Vincent Lo Re; Leanne Gasink; Jay R. Kostman; et. al. AIDS Patient Care and STDs. 18(8):436-442.
4 Abgrall S et al. Clinical and immunological outcome in patients with human immunodeficiency virus infection, according to virologic efficacy in the year after virus undetectability, during antiretroviral therapy. Clinical Infectious Diseases 37 (on-line edition), 2003.
5 J Acquir Immune Defic Syndr. 2004 Sep 1;37(1):1147-1154.
