In the absence of a cure for HIV/AIDS, it is quite possible that advances in our therapeutic approach to this infection will come from a better understanding of the immunologic response of the host, and from the development of novel strategies to enhance it. As such, it is quite heartening to see an entire section of a conference on HIV drug resistance and treatment strategies devoted to the subject of immune reconstitution.

In a study of 18 individuals who received NRTI therapy, it appeared that the development of drug resistance mutations in the reverse transcriptase (RT) gene (at codons 41, 74, 184 and 215) led to increased cytotoxic T-lymphocyte (CTL) recognition of these altered virions (Samri et al, Abstract 188) [1]. The clinical significance of this potentially adaptive response -- where the host "picks up" when the drugs fail -- is unknown.

This approach to recruiting specific anti-HIV immunity has been studied more formally in the context of structured treatment interruptions (Garcia et al, Abstract 189) [2]. In a group of 4 patients with maximal virologic suppression who underwent three 8-month treatment interruptions over 96 weeks, the viral set point showed a definite downward trend (0.6-1.5 log₁₀ copies/mL lower than pre-therapy levels) with each treatment interruption. This was associated with a strong and broad HIV-1-specific CTL response and a strong CD4 lymphocyte proliferative response to HIV-1 antigens.

It is now quite clear that the types of responses we can expect in patients exhibiting a maximal virologic response to HAART are quite heterogeneous. In a small study of 10 individuals, the number of cells leaving the thymus (recent thymic emigrants, or RTEs) and naive T-cells seemed more related to age and potential size of the remaining RTE cell pool than to any other characteristic of the treatment or baseline CD4 cell counts (Franco et al, Abstract 191) [4]. It has been argued that the number of RTEs may relate more to increased naive T-cell division (presumably due to HIV-induced immune activation; Hazenberg et al, Abstract 193 [5]). However, in some studies, there does seem to be an association between an early increase in thymic volume (at weeks 12 and 24 after the initiation of HAART) and subsequent T-cell repopulation (Rubio et al, Abstract 196) [6]. Additional information has been generated in 19 patients experiencing virologic breakthrough on therapy (Lecossier et al, Abstract 192) [7]. In these individuals, the maintenance of a higher CD4 cell count in the presence of ongoing viral replication was associated with thymic output (as measured by RTEs), which was itself related to younger age (a known correlate of higher residual thymic function). Further, an insightful analysis was performed in 45 patients who exhibited a maximal virologic response to protease inhibitor (PI)-based HAART therapy, including AZT/3TC with either IDV or SQV-SGC, (the CHEESE study, Back et al, Abstract 195) [8]. The observed exponential decrease of the regeneration rates as a function of age (5.0% and 3.6% per year for naive CD4 and CD8 cells, respectively) is in close agreement with the reported exponential decay of the volume of the thymic epithelial space (4.1%/year in adults above 20 years of age). Although these data are quite relevant to our understanding of the pathogenesis of HIV infection, their clinical significance depends on our ability to design effective treatments based on this information.

To evaluate the potential benefit of a non-specific immune-based intervention, a group of 11 patients were evaluated after stopping successful HAART therapy that had been administered with (n=6) or without (n=5) concomitant IL-2 (van Lunzen et al, Abstract 194) [9]. All patients experienced a rebound in plasma viral load off therapy, with no evident virologic or immunologic benefit of the IL-2 (as measured by CD4 cell counts, and T-cell responses to mitogens and recall antigens). This is somewhat disappointing. An Australian study of 20 patients at the earliest stage of HIV infection showed a more vigorous recovery of naive CD4 T lymphocytes in response to HAART therapy, as compared with 21 patients having more established infection, even though both groups had become infected in the previous 6 months (Smith et al, Abstract 197) [10]. It may thus be more efficient to evaluate immune-based interventions in a group of patients who are most likely to exhibit a measurable response to them, such as those with acute/early infection.

In conclusion, we are becoming aware of the relevance of the host's immune response in the control of HIV infection in vivo, and we have developed reproducible ways of measuring this phenomenon. It is hoped that the insightful of specific and non-specific therapies in carefully selected hosts will allow us to make progress in this important field.