Exploring the Benefits and Limits of Highly Active Antiretroviral Therapy

Highly active antiretroviral therapy (HAART) has thus far failed to fulfill the great hope of eradicating human immunodeficiency virus 1 (HIV-1).^{1 - 2} Not only does a reservoir of long-lived latently infected cells appear to persist for years despite apparent complete viral suppression, but as the study by Dornadula et al³ reported in this issue of THE JOURNAL shows, continuing low-level viral replication is likely to be occurring in most persons with HIV-1 apparently suppressed with HAART. However, a second article in this issue by Whitcup et al⁴ indicates that despite the apparent failure of HAART to completely control viral replication, current regimens lead to meaningful immune reconstitution and have a substantial beneficial effect on an opportunistic pathogen that only recently had been one of the most difficult to manage with chemotherapy directed against it.

Ongoing viral RNA expression and virus evolution in peripheral blood mononuclear cells and lymphoid tissue of HAART-treated patients have been reported by a number of laboratories^{5 - 11} ; thus, the present findings of Dornadula et al are not entirely unexpected. One preliminary report suggests that episodes of low-level transient viremia are commonly detected in patients treated with HAART who are monitored frequently, even when currently available assays are used.¹² Dornadula et al now show that low levels of plasma HIV-1 RNA indicative of residual viral replication could be detected in each of 22 patients whose HIV-1 was believed to be completely suppressed (viral load of less than 50 RNA copies/mL). Using a sensitive polymerase chain reaction assay involving hybridization to a phosphorus 32–labeled probe, they were able to quantify an HIV-1 gag RNA fragment to 5 copies/mL in plasma, revealing a mean plasma RNA level of 17 copies/mL (range, <5-42 copies/mL). Nearly half also had low-level detectable HIV-1 RNA in genital secretions. Similar results have also been reported by Yerly et al using a different sensitive assay for plasma HIV-1 RNA.¹³

The findings of Dornadula et al are striking but should be interpreted cautiously. The sensitive assay developed by Dornadula et al needs further validation, and it will be important to determine if the amount of HIV-1 RNA fragment measured by their assay parallels viral protein expression or levels of infectious virus in plasma.^{14 - 15} Some studies suggest there may be much lower levels of HIV-1 infectivity than RNA in plasma, for example, because some virions may be replication incompetent. It is also not clear what the clinical consequences of this low-level viral replication might be. However, it is worrisome that drug-resistant virus eventually may emerge even with only this low level of replication. This has not yet been observed^{6 ,16 - 17} but further study now seems warranted. Also important is an evaluation of the extent to which incomplete adherence to complicated treatment schedules may have contributed to the persistence of viremia.

The study by Whitcup et al, on the other hand, demonstrates the substantial benefit of HAART: when persons with HAART-induced increases in CD4 cell count to more than 0.15 × 10⁹/L discontinued their maintenance therapy for cytomegalovirus (CMV) retinitis, none of the patients experienced a recurrence of retinitis after a mean of 16 months without anti-CMV therapy. Although the study did not include direct measures of anti-CMV immunity, some of the patients developed an increasingly recognized immune recovery inflammatory disease associated with HAART, in this case, uveitis. In another study, HAART alone has been associated with clearance of CMV viremia without anti-CMV therapy.¹⁸ That CMV-specific immunity can be recovered with HAART alone makes sense since treatment with HAART is known to lead to increases in memory and naive lymphocytes that might be able to respond to ongoing antigen production.¹⁹

An important question is why HAART does not lead to recovery of immune responses to HIV. Numerous studies report that immune responses to HIV-1 decrease with the incomplete suppression of viremia provided by HAART.^{20 - 23} These dichotomous responses are not as paradoxical as they may first appear. It can be hypothesized that the level of HIV-1 antigenemia that occurs when plasma RNA levels are less than 50 copies/mL in plasma is unlikely to be enough to sustain these responses. Indeed, it seems likely that HIV-1 proteins are expressed from only a fraction of the RNA detected in plasma, whereas CMV replication may provide more ongoing antigenic stimulus to drive anti-CMV immune responses. Determining levels of viral protein antigens, as well as RNA, will test this hypothesis.

The finding of residual viral replication in the setting of HAART is not sufficient to change treatment recommendations, particularly when current regimens seem to be successful by measures such as providing persistent augmentation of immunity to CMV and other opportunistic pathogens. The suggestion to intensify treatment regimens will require more data and will require weighing the possible benefits vs toxic effects, tolerability, and interactions of these more complex drug regimens, as well as the ability of patients to adhere to them. It also will be important to determine whether a clinical benefit can be derived from more frequent or sensitive viral load monitoring than is presently standard, although the data from Dornadula et al suggest that clinical trial data should be analyzed using more sensitive viral load assays.

At this point, HIV eradication is not feasible with chemotherapy alone. However, there are many examples of immune control of persistent viral infections, most notably the herpesviruses.²⁴ These viruses are not eradicated but are held in check by effective immune responses. Emerging data in HIV infection indicate that some persons achieve immune control of the virus without the need for antiretroviral therapy through persistent HIV-1–specific immune responses.²⁵ In addition, early treatment of acute HIV-1 infection can lead to marked augmentation of such HIV-specific immune responses.^{26 - 27} The practical implication of the studies by Dornadula et al and Whitcup et al is that although HAART alone may not be enough to stop HIV replication, it does offer hope for boosting immunity to HIV-1, as already occurs for CMV, thereby providing increased containment of HIV replication. Future efforts need to be directed at developing combination strategies involving both antiretroviral therapy and immunotherapeutic interventions. The next major advances are likely to come from a combination of these 2 approaches.