AIDS 2006: XVI International AIDS Conference

2006年8月13－18日

加拿大安大略省多伦多

August 13 - 18, 2006, Toronto, Ontario, Canada

Antiretroviral Treatment Strategies for Adults
David Cooper, MD

Strategy trials with immediate clinical implications constituted the overarching theme of antiretroviral therapy trials presented at the International AIDS Conference (IAC) held in Toronto, Ontario, Canada, July 13-18, 2006. Continuing a theme that began at the Conference on Retroviruses and Opportunistic Infections (CROI) earlier this year, structured treatment interruptions were a focus of this conference as well. Other important trials examined the use of a boosted protease inhibitor to start or simplify antiretroviral therapy (ART), and other trials continued to examine the optimum combination of antiretroviral agents (ARVs) to use as initial therapy.

Structured Treatment Interruptions

Structured, or "strategic," treatment interruptions (STIs) offer the potential to spare patients some of the adverse effects, cumulative toxicity, and financial burden of ART. These are critical considerations, given that many patients might use antiretroviral drugs for decades after they start therapy. Therefore, it was reasonable to evaluate whether a treatment interruption strategy was associated with any changes in the rate of disease progression or toxicity. The SMART study was designed to address this question and enrolled over 5000 patients, making it the largest ART clinical trial ever conducted. Investigators randomized patients with CD4+ cell counts > 350/microliter (mcL) to either use continuous ART to maintain the viral load as low as possible (virologic suppression [VS] arm, n = 2752) or to take ART episodically (discontinuation [DC] arm, n = 2720). CD4+ cell counts were used as guides for starting and stopping therapy; ART was discontinued until the CD4+ cell count fell below 250, and then ART was resumed when the CD4+ cell count rose above 350. To the surprise of many, an interim analysis showed a statistically significant increase in clinical endpoints, including deaths, in the treatment interruption arm. The study was terminated prematurely, and the results were presented at CROI.[1] Of particular note, there was an elevated risk for major cardiovascular events, such as myocardial infarction, in the DC group.

Analyses continued to attempt to explain the reasons for the results, and these data were presented at IAC. Wafaa El-Sadr and colleagues[2] presented analyses that compared the hazard ratio in subgroups in the study arms in order to determine whether particular factors would provide insight into the reasons for the adverse outcomes in the DC arm, or whether there were subgroups of patients who benefited from the interruption in ART. Factors that were evaluated included:

· baseline HIV disease stage, including prior AIDS-defining condition, CD4+ cell count, CD4%, viral load, and coinfection;

· history of ART use, including number of ARVs used and duration of ART; and/or

· demographic characteristics such as race/ethnicity, sex, age, and history of intravenous-drug use.

The result of these analyses was that no subgroups were found in which the DC arm was superior to the VS arm. Two examples of the data from these analyses are shown in Figures 1 and 2. The conclusion was that STI, according to the SMART study design, "should not be recommended for any subgroup of patients."

Figure 1. Risk for opportunistic disease or death in subjects in the SMART trial randomized to the DC vs VS study arms analyzed by nadir CD4+ cell count. Adapted from El-Sadr et al[2] with permission.

Figure 2. Risk for opportunistic disease or death in subjects in the SMART trial randomized to the DC vs VS study arms analyzed by baseline CD4+ cell count and CD4%. Adapted from El-Sadr et al[2] with permission.
During the treatment interruption, CD4+ cell counts fell and viral load increased. To what extent do these factors explain the results in the SMART study? According to the analyses presented by Lundgren and coinvestigators[3] for the SMART investigators, the CD4+ cell count (Figure 3) and viral load recorded proximal to the clinical event were associated with a greater risk for opportunistic disease (OD) and death. Together the latest CD4+ cell count and the viral load did explain part, but not all, of the increased risk for clinical events (Figure 3).

Figure 3. Comparison of adjusted and unadjusted hazard ratios (HR) for OD and death, OD, and non-OD death in the DC/VS study arms. Adapted from Lundgren et al[3] with permission.

The greatest increment in the incidence of clinical events was among the people who had the highest CD4+ cell counts (Figure 4), and it is those people who suffered the most damage from the DC strategy. Some researchers have hypothesized that there might be a massive inflammatory response resulting from virologic rebound after sustained virologic suppression, and such an inflammatory response might also explain the increased risk for heart attacks. Another analysis of the SMART data showed that the DC strategy did not improve the quality of life by a variety of measures, including general health perception.[4]

Figure 4. Incidence of opportunistic disease or death by latest CD4+ cell count stratum in the DC vs VS study arms. Adapted from Lundgren et al[3] with permission.

The findings of the SMART study were further supported by the results of the DART study,[5] which used a defined period in the STI arm, with 12 weeks on ART followed by 12 weeks off ART. The study randomized 813 patients to the STI or continuous therapy arm. Although there was no difference in the number of deaths during the median follow-up period of 51 weeks, the hazard ratio for clinical World Health Organization (WHO) stage 4 events was 2.6 for the STI arm, and the study investigators concluded that the STI strategy cannot be recommended. In contrast, the results of the BASTA study, reported by Maggiolo and colleagues,[6] appeared to support the use of STI, at least in selected patients. Although this study was done quite carefully, there were 76 patients in the STI arm and 38 in the continued HAART (highly active antiretroviral therapy) arm. In contrast to the statistical power of the SMART study, with over 5000 patients, this study was very underpowered and the results could happen by chance.

At this time, the overall conclusion is that the strategy of treatment interruptions cannot be recommended. In the SMART study, no subgroups could be found that benefited from such interruptions. Although further thought will be given to the underlying reasons for the results observed to date, neither CD4-guided nor time-defined STI has a place in clinical practice.

Initial Therapy
HAART in the DART Study

In keeping with the fact that this is truly an international conference with the theme "Time to Deliver," Munderi and coworkers[7] presented data on the 2-year survival and causes of death among patients (n = 1819) enrolled in the Entebbe, Uganda, cohort of the Development of ART in Africa (DART) trial. Survival and other clinical parameters of patients in the DART cohort were compared with those of a matched pre-ART historical cohort (n = 658). Patients in the DART cohort were randomized to receive ART consisting of coformulated zidovudine/lamivudine + efavirenz or coformulated zidovudine/lamivudine + tenofovir. The median CD4+ cell count at enrollment was 93 cells/mcL in the DART cohort and 75 cells/mcL in the historical controls. Approximately one third of patients in each cohort had CD4+ cells < 50/mcL. Despite the low CD4+ cell counts, there was a dramatic effect in the DART cohort (Figure 5). Overall,