May 1, 1998
The changing goals of antiretroviral therapy were highlighted in a state of the art presentation by Dr. Julio Montaner of the University of British Columbia. He reminded the audience that less than two years ago the expert opinion on the goal of antiretroviral treatment was to reduce the viral load to less than 5000 copies/ml and that combination treatment with two drugs was advocated as likely to be able to achieve that goal. Now, it is recognized that the goal of therapy is undetectability in plasma viral load assays that is almost always only achievable in patients who receive more potent combinations, usually involving three drugs. Dr. Montaner did present an analysis of data from Vancouver indicating that a small percentage of patients (10-20%) could achieve an undetectable viral load when a dual nucleoside regimen (e.g. ZDV and 3TC) was used -- however there were no baseline factors that predicted who would be most likely to benefit from such an approach and consequently it could not be recommended.
Dr. Montaner went on to review sobering data from several studies that taken together indicate that the goal of therapy will need to be reducing viral load not just below 500 copies -- the limit of detection on the standard assay used commonly -- but to below the limit of detection on the newer ultrasensitive assay. Indeed sustained long-term suppression of viral replication was seen only in patients whose viral load went below 20 copies/ml. Furthermore, as he pointed out, only 50% of patients treated with antiretroviral combinations reach this target -- no matter what the starting regimen is. Indeed, Dr. Montaner's contention was that results appeared to be similar whether on started with two nucleosides plus a protease inhibitor (such as indinavir or nelfinavir) or two nucleosides plus a non-nucleoside reverse transcriptase inhibitor such as nevirapine. Although there have been no head to head comparisons performed to date, judging across all available studies this appears to be a reasonable conclusion.
As several speakers identified adherence to medications is a major issue, and Dr. Montaner looking to the future speculated that for many patients directly observed therapy would be the key to control of HIV -- in a manner similar to the way this form of drug delivery is the key to tuberculosis treatment. The ultimate need is for drugs that can be given once daily and together and although there are some drugs that can be given once a daily -- the nucleoside analogs, ddI and 3TC, the nucleotide adefovir, and the NNRTIs, nevirapine and efavirenz -- there is as yet no protease inhibitor that can clearly be given once daily. The cost of such an approach would also have to be dealt with since it appears that therapy may be lifelong (as we will see later).
An eloquent discussion of issues related to resistance and testing for resistance was given by Dan Kuritzkes of the University of Colorado. Resistance is almost inevitable if any viral replication occurs in the presence of drug therapy because of the high rates of mutation in the virus during the normal course of replication. As an audience response question indicated, there is a great demand among clinicians to have reliable resistance testing that they can use easily to interpret response to treatment in patients. However there are pitfalls and problems in the two types of tests currently available to clinicians. Phenotypic tests (that determine the behavior of the virus in a laboratory setting when grown in the presence of drug) do indicate drug susceptibility and may be used to test combinations of drugs. However, they are time consuming, expensive, specific only to the strain being tested (and thus may miss resistant virus that is not the dominant strain) and work poorly for some drugs (e.g. ddI and D4T). Genotypic tests (that examine the sequence of the virus or enzyme of interest) are also expensive, can't detect drug-drug interactions and are particularly susceptible to contamination in the laboratory. Furthermore genotypic tests do not predict phenotype and are essentially measures of the average susceptibility of the viral population.
There is very little evidence that resistance testing is clinically useful. Indeed, Dr. Kuritzkes summarized the situation by stating that although resistance testing is very good at explaining why a particular drug or drugs failed, it is not very good at predicting if a regimen will succeed. In general, if a drug is resistant in one of these tests, it is unlikely to be successful; however, the absence of resistance does not predict success -- not only because of the inherent unreliability of the tests but also because mutations in the virus are not the only reasons why patients fail therapy. The bottom line from this presentation was that there was very little clinical utility to resistance testing at present.
All the speakers in the session mentioned adherence to medication as a key issue and for one, Dr. Alexandra Levine of the University of Southern California, it was the subject of her presentation! Although of critical importance in the care of HIV patients, adherence to medications is not, as Dr. Levine pointed out, a new issue. Indeed there is ample information in the medical literature that can be applied to HIV and AIDS from other situations. The first thing to realize is that measuring adherence is hard. Patients generally over report their pill taking and if patients are bad judges, doctors are worse. Indeed, in any chronic illness studies, adherence to medications can vary from 20-80% among patients.
Certain themes are common. Patients' attitudes are key. If a patient understands why a medication is being given, believes that rationale, and derives some positive benefit (e.g. relief of some symptoms) then they are more likely to be adherent. Thus the interaction with the physician, trust in that physician and education given by the physician and other health care workers are key to achieving adherence. Although it is commonly held that persons from ethnic minorities and or drug users are less likely to be adherent, when this has been examined critically it has not been found to be consistently true. However other issues such as psychiatric illness or homelessness have been shown to predict lack of adherence and may need to be dealt with first, before complex medical regimens are introduced. Minimizing complexity would also help, but often is unavoidable. Ultimately therefore, achieving adherence (and therefore a successful treatment that avoids resistance) requires time and effort from the physician to understand the patient as an individual and provide the help and support that person needs.
In the final presentation of the morning, Robert Siliciano of the Johns Hopkins University presented an eloquent review of the current understanding of viral turnover and latency, which unfortunately leaves one with the pessimistic notion that eradication of this virus is unlikely to be successful at least with current therapy. There appear to be several important sites of viral replication. The most common, and the main source for productive viral replication in patients not on treatment is activated CD4+ T cells. Other sites are follicular dendritic cells and macrophages. Virus in all of these compartments is susceptible to potent antiretroviral therapy and replication usually halts rapidly -- within a day in the lymphocytes and two weeks in the other compartments. However, it now appears that true latency (i.e. a situation where viral DNA is integrated into the host cell and remains there without active replication) occurs in a cell population of resting memory T cells. This cell reservoir is small. However, under laboratory conditions, selection of this population from patients' blood can be performed and replication-competent virus can be cultured. In order words, even in patients who have been treated for over two years with highly active antiretroviral therapy, and who have no detectable virus by conventional techniques, virus can be cultured from their resting memory T cells in the laboratory.
This population is small but remarkably stable -- these latently infected cells can be detected very early on after infection even in patients who are treated aggressively during the seroconversion. Their numbers do not decrease with treatment. However, it is also a truly latent population without much replication in vivo. Virus recovered from treated patients remains susceptible to antiretrovirals suggesting that resistance is not being induced with therapy. Thus although these cells may represent the Achilles heel of eradication (since they are designed to live a long time to provide immunological memory), they may not have major in vivo significance for the majority of patients who are being treated. Nevertheless, strategies to eradicate this population are under consideration.