Fall of HAART
Considered to be the pre-eminent infectious disease meeting in the world, the 39th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC) was held in San Francisco this fall, from September 26 to 29, 1999. In the absence of a World AIDS Conference this year, this was the first major venue for HIV researchers and pharmaceutical companies to present their data since the Retrovirus Meeting last January. Just one month later, the Seventh European Conference on Clinical Aspects and Treatment of HIV-Infection took place in Lisbon, Portugal. Given the short time between the meetings, the HIV-related presentations were virtually identical -- only the accents were different. Because so much information was presented, we have decided to combine our coverage of the two conferences in this and the next issue. [This is part of the reason for the delay in our production timetable. The November issue should follow in a couple of weeks. We hope to be back on schedule early next millennium.]
Trouble Draining the Reservoir
There were no major surprises at either conference. Eradication of the infection, the Holy Grail of AIDS research, remains beyond HAART's grasp. As has been recounted in the pages of Treatment Issues numerous times before (e.g., see July/August 1999, pages 14-15; July/August 1998, pages 2-3), HIV can hide, latent, in resting memory T-lymphocytes. The virus harbored in these cells remains capable of reigniting widespread infection as soon as antiretroviral therapy is removed. And indeed, several groups at both conferences detailed how the virus re-emerges quickly in the majority of patients who opt to discontinue therapy, even when they have had years of almost complete viral suppression (although there were some notable exceptions -- more on this in our next issue).
Hopes were that this reservoir of latently infected cells would decrease over time, but at each successive conference we hear that the number of quiescent infected cells remains fairly constant. Partly, the size of the pool is unchanged because the memory cells are long-lived. Now data from both the labs of David Ho, MD, at the Aaron Diamond Research Center and Anthony Fauci, MD, at the NIH show that the pool is also reseeding itself.(1) In other words, viral replication continues albeit at an extremely low rate. On one hand, this could be seen as good news, because if we could keep the pool from reseeding itself, it would eventually become depleted. But could treating "harder" stop this virtually undetectable replication, or are we at the limit of what antiviral drug therapy can achieve?
There was news of incremental advances on a number of other fronts. New data were presented on the comparative efficacy of protease inhibitor and protease inhibitor-sparing regimens (more in our next issue). In addition, there have been some minor advances in the clinical management of antiretroviral toxicity. Finally, there are new theories concerning the potential causes of HIV-treatment related fat redistribution -- although the etiology of the condition is beginning to look positively Byzantine. There were also presentations on a growing number of interventions aimed at treating or preventing the condition.
Barbarians at the Golden Gate
Despite these advances, there are increasing concerns that many patients in North America and Western Europe may have already reaped the maximum benefits of HAART. While it is true that even those who experience virologic breakthroughs on HAART often sustain stable elevations in CD4 cell count for months and sometimes years, there are subtle indications that this immunologic cushion against disease progression is beginning to wear thin. One report from San Francisco General Hospital suggested that the incidence of one old foe, CMV retinitis, might be on the increase.(2) The number of cases at the site has risen from a low of four in 1997 (compared to 48 just two years earlier, but before the widespread availability of HAART) to ten in 1998 and the trend has worsened this year. Many of these patients had CMV that became quiescent after initiating HAART. However, upon experiencing virologic failure and drops in CD4 count to below 50 cells, CMV began to re-emerge. The paper's author Mark Jacobson, MD, recommends restarting CMV prophylaxis in such patients.
How common an occurrence is HAART failure? According to unpublished estimates from HIV Cost and Services Utilization Study (HCSUS), cited in one presentation by Julio S. G. Montaner, MD, as of last fall over one-third of all HIV-positive adult patients on HAART in the US have a viral load of 20,000 copies/ml or higher. The tally could only be worse one year later. This figure would not be so frightening if there were an infinite number of HAART regimens. Yet, despite the approval of 14 drugs to treat HIV, cross-resistance limits one's options.
After one antiretroviral combination has failed, how does one choose what to use next? Clearly this depends on a number of factors, including treatment history, the reasons for failure, how soon after virologic failure therapy was switched and, of course, the patient. Over the next couple of months, we hope to explore the role for drug resistance testing (geno- and phenotyping) in making treatment decisions.
Resistance testing is especially important for choosing one's next regimen because most salvage therapy studies conducted to date provide little practical guidance on sequencing strategies. For example, we still have few clear answers about when a second protease inhibitor-based regimen can be used successfully after the virus has become resistant to the previously used protease inhibitor. Yes, a number of studies have addressed the subject, but their results are generally unenlightening because of poor trial design or because our understanding of antiretroviral therapy has evolved since the study's inception. For example, now we know that the development of resistance to just one agent in a regimen can be enough to allow the virus to escape antiviral suppression, but most of the early protease inhibitor sequencing studies did not take this into account. As a result, they reported widely varying rates of response depending upon how soon after virologic failure therapy was changed and how high the patient's viral load was upon switching regimens.
One can argue that a delayed switch and a high viral load were probably surrogates for protease inhibitor resistance. In such cases, only a very low percentage achieved viral suppression to undetectable levels with their second HAART regimen. When the virus has developed resistance to a protease inhibitor, response rates on a second protease inhibitor-based regimen ranged from 30% to 40% in a number of studies reported at the recent conferences. For example, in a small study of sixteen salvage therapy patients presented in Lisbon, only 44% had more than a 1.0 log drop in viral load (a much lower threshold for success than "going undetectable") on a regimen of amprenavir/abacavir/ddI/hydroxyurea.(3) Phenotypic testing suggested that seven were susceptible to amprenavir, but only four of those responded. Whether newer agents such as ABT-378/r can do any better awaits clinical demonstration, although there have been studies that show responses can be enhanced by the addition of a non-nucleoside reverse transcriptase inhibitor (NNRTI) -- as long as the patient is NNRTI-naive (see "A Cornucopia of New Drugs" in this issue).
It is even more unlikely that you get a second chance at using an NNRTI. Since the K103N mutation that causes cross-resistance to all the currently marketed NNRTIs is so ubiquitous, the outlook has always appeared bleak for sequencing these drugs, notwithstanding the hopes of some pharmaceutical companies. For example, an analysis of the efavirenz expanded access program reported disappointing responses for patients switched to efavirenz after failing other NNRTI-based regimens.(4) Three-month data were available for 28 patients who entered the program with prior NNRTI experience. After starting their new regimen, twelve had viral loads below 400 copies/ml, but not all had evidence of NNRTI-resistance. Those who did faired poorly. For the most common mutations, only one of eight patients with both K103N and Y181C, none of three with Y181C and three of six with K103N went undetectable. Similarly, a study in Lisbon found only that only 11% of nevirapine failures had viral loads sustained below 400 copies/ml six months after switching to an efavirenz based regimen.(5)
So thousands of patients have now failed multiple HAART regimens, or rather HAART has failed them. Many are now infected with viral strains that have developed multiple mutations that confer resistance to all of the currently marketed antiretrovirals. To construct new HAART regimens capable of producing sustainable reductions in viral load, these patients need new compounds unique enough to treat their drug resistant virus. There are a host of new drugs in development, and we devote the bulk of this issue to them (see "A Cornucopia of New Drugs" in this issue).
Unfortunately, there simply are not that many novel agents far enough along in clinical development to be available for people in need of salvage therapy soon. For these individuals, clinicians and researchers are devising new approaches to therapy. One is simply more drugs. By the time patients are on their third regimen, it is not uncommon for them to be taking four antiretrovirals, but now they may be taking five or six (sometimes called mega-HAART). In particularly dire situations, some are using up to eight drugs, most of which they have used before, in an approach called "giga-HAART."
Others are experimenting with strategies such as structured treatment interruptions (STIs), formerly referred to as drug holidays. STI, which was the big story at both conferences, has become a catch phrase for any planned treatment interruption, but the point to these interruptions depends upon the clinical or research context. For instance, STIs can be used to check whether a patient's immune system can control whatever low level of virus remains after years of undetectable viral loads.
On the other hand, in salvage therapy patients, it is postulated that this technique might "undo" some drug resistance. The theory is based upon basic principles of evolution. In order to evade the effects of therapy, the virus must make certain compromises that slow down its ability to reproduce. When an antiretroviral regimen's selective pressure is removed, wild type virus (the strains most prevalent before therapy is introduced) should become predominant again because it grows faster than the drug-resistant strains. In the most optimistic scenario, wild type virus would out-compete drug resistant strains to the point where they would be starved out; but this seems unlikely given the diversity of the immunologic ecosystem. Numerous mutant strains are likely to coexist because they are each effective at exploiting their respective niches.
Nevertheless, there is a possibility that something close to a complete shift to wild type virus did occur in some patients (but not all) in one German study on the use of STI before mega-HAART.(6) The report created a stir, however, because the stakes of treatment interruption were so high and the results somewhat mixed. For now, taking very advanced patients off a failing regimen before beginning a new one will remain highly controversial until more evidence becomes available.
This article was provided by Gay Men's Health Crisis. It is a part of the publication GMHC Treatment Issues. Visit GMHC's website to find out more about their activities, publications and services.