After the initial enthusiasm about the possibility of eradicating HIV in 1996-97, realism has set in. We are anxiously trying to design strategies to enhance the efficacy of the regimens we are now using, with a view to maintaining the health of our patients for extended periods of time, perhaps indefinitely. In this light, there were 13 presentations that looked at different ways in which this could be accomplished.

In its ongoing studies of patients with maximal suppression of plasma viral load, Markowitz et al [1] showed that ultimate clearance of HIV from cellular reservoirs may proceed with a t1/2 as short as 6 months, although as few as two increases in plasma viral load above 50 copies/mL would increase this figure to near infinity. This observation was supported by the work of the Swiss Cohort study [2] which clearly showed that cellular RNA and DNA (ongoing, low-level viral replication viral replication) was still present in patients with primary HIV infection and a plasma viral load below 3 copies/mL. The speed with which viral load is reduced may be important in addressing these issues. Indeed, in a small group of patients receiving indinavir-based therapies, little or no evolution was seen in the env gene in those whose plasma viral load had decreased most quickly [3]. However, if any level of virologic rebound occurred or if the initial plasma viral load decay was too "slow" (a term that has not yet been precisely defined), evolution in viral sequences took place due to ongoing replication events. This has been confirmed in pediatric patients [4] and in adults receiving a broader range of therapies [5]. It thus becomes clear that the level of antiviral efficacy that must be achieved to eradicate HIV may be well beyond the reach of most of our patients, but may not be theoretically impossible to achieve. It may be that, from a clinical perspective, longer term virologic suppression may, in itself, decrease the probability of virologic rebound over time [6]. Whether this relates to an artificial selection of patients who are more adherent to therapy, experience less toxicity, achieve the highest drug levels, have maintained susceptibility to the drugs in their regimens, or simply patients who have some other unknown biological advantage has not been determined. However, findings such as these give us hope regarding the potential long-term efficacy of the regimens we are currently prescribing.

It should be emphasized that, to date, there have been no reports of successful eradication of HIV from a patient. However, an intriguing case report [7] is tantalizing in this regard. Through a series of interesting circumstances, an individual was informed, two days after the fact, that she had received a transfusion that was infected with HIV. This had been taken from an individual who had recently been infected with HIV, but had not yet developed antibodies to the virus. The recipient was immediately placed on therapy with two nucleoside analogues and a protease inhibitor, and continued for 9 months. The initial sample from the recipient may have contained 3 HIV copies/mL plasma, but all subsequent samples were negative for HIV RNA, DNA and antibodies. Although this likely represents a case of highly effective post-exposure prophylaxis, it also adds to our body of knowledge about our ability to control HIV infection.

It is quite clear that novel approaches must be designed. It is gratifying to see that the combination of zidovudine, lamivudine and efavirenz has maintained its efficacy when studied in a larger numbers of drug-naïve patients. Over 90% of 144 patients receiving this combination had plasma viral loads below 50 copies/mL after 60 weeks [8]. This certainly contributes to increasing the number of therapeutic alternatives available to our patients. But what we really need to make a significant advance in this field is a novel approach, perhaps based on the enhancement of the efficacy the immune response to HIV. In an intriguing study, 8 patients were placed on triple combination therapy for 12 months [9]. With maximal virologic suppression having been achieved for at least 8 months, therapy was stopped for 4 weeks, given for another 6 months, then stopped for 4 weeks once again. The half-life of the viral load rebound increased from 2.2 to 3.2 days after the second rebound, an observation that was taken as a measure of the strength of the anti-HIV immune response having been enhanced by the temporary exposure of the immune system to HIV during the interruptions in therapy. More conventional approaches have been taken to immune system activation with IL-2 [10] and OKT3 [8]. In a group of 8 patients receiving HAART, the addition of IL-2 every month for 9 cycles led to an additional increase of over 250 CD4 cells/mL. Unfortunately, the expected decrease in cell-associated HIV DNA was not observed, and these data do not provide support to the hypothesis that IL-2 could be used to purge the cell-associated viral reservoir more rapidly. Larger studies are underway to address this issue. The combined use of IL-2 and OKT3 as immune stimulants has led to a burst in viral replication in the 3 patients in whom it has been attempted, and this approach must certainly be reconsidered before it is attempted again in clinical trials.

Finally, one of the most exciting presentations was the first on the ability of RRE decoys and antisense RNA to inhibit HIV replication in vitro [11]. In this model, challenging MT-2 cell lines expressing the RRE decoy or the M184I/V antisense with HIV-1 resulted in the inhibition of replication of wild type and lamivudine-resistant viruses up to an MOI of 0.1. These interesting preliminary in vitro results must be followed up, in an effort to help us develop truly useful approaches to gene therapy.