CASCADE Study Shows Antiretroviral Therapy Is Beneficial above CD4 Cell Count of 350, Less Clear above 500

		SUMMARY: Participants in the large CASCADE cohort who started antiretroviral therapy (ART) with a CD4 cell count in the 350-500 cells/mm3 range -- in accordance with current U.S. treatment guidelines -- had a reduced likelihood of death than those who deferred until later, investigators reported at the XVIII International AIDS Conference last week in Vienna. Starting above 500 cells/mm3, however, did not reduce the risk of progression to AIDS or death in this analysis.

By Liz Highleyman

A growing body of evidence suggests that early antiretroviral therapy, before the immune system sustains significant damage, leads to better outcomes, including protection against opportunistic illness, non-AIDS conditions such as cardiovascular disease and cognitive impairment, and overall mortality. On the other hand, however, starting early could cause problems related to long-term drug-related toxicities and resistance.

A large randomized study known as START is just getting underway, and a recent smaller analysis from Haiti offered some of the first data from a randomized trial indicating that earlier ART is beneficial. Most evidence on this question, however, has come from observational cohort studies.

As described by Joseph Eron and fellow investigators, the CASCADE Collaboration is an ongoing natural history study of HIV seroconverters from more than 2 dozen clinical cohorts in Europe, Australia, and Canada followed since the start of the combination ART era in the late 1990s.

The CASCADE "When to Start" analysis looked at outcomes according to whether participants started or deferred treatment. The study included 9455 patients enrolled between January 1996 and May 2009. At the time of enrollment they were at least 6 months past seroconversion, did not have an AIDS diagnosis, and had not yet started ART.

Most participants (78%) were men (56% men who have sex with men). The average age at the time of seroconversion was 30 years and the estimated duration of infection was 1.3 years. People who started treatment had better prognosis in some respects (including less likely to have a history of injection drug use or to be coinfected with hepatitis B or C), but poorer prognosis in others (including higher viral load, shorter duration of infection, and slightly lower average CD4 cell count).

Participants were analyzed in 161 sequential nested cohorts of people enrolled during a single month. The researchers compared those who started ART (defined as any three-drug antiretroviral regimen) versus those who deferred for that month, looking at how many developed an AIDS-defining illness or died from any cause. Participants who remained alive could join the next monthly cohort, so some were members of multiple cohorts. They were followed for about 5 years on average for a total 52,268 person-years of data.

Researchers described the study as "intent-to-treat in spirit." Outcomes among people who started therapy and those who deferred during a given month (regardless of what they did later) could not be directly compared since participants were not randomly assigned to one strategy or the other.

A limitation of the analysis is that it did not include serious non-AIDS-defining conditions -- such as cardiovascular, kidney, and liver disease -- that occur more often among HIV positive compared with HIV negative people in the ART era and contribute to shorter life expectancy. It also did not account for treatment interruptions after starting ART or therapy that did not produce full viral suppression.

Participants were divided into 5 categories according to CD4 cell count: 0-49 cells/mm3 (n = 183), 50-199 cells/mm3 (n = 1521), 200-349 cells/mm3 (n = 4459), 350-499 cells/mm3 (n = 5527), and 500-799 cells/mm3 (n = 5162). People with more than 800 cells/mm3 were not included in the analysis because they contributed too few events for a meaningful statistical analysis.

Results

	Overall, 812 people (8.6%) progressed to AIDS during the study period.
	A total of 544 participants (5.8%) died during the study.
	Benefits of starting ART were clearly evident among people with lower CD4 counts, but less so at higher levels:
	0-49 cells/mm3: 55 events per 1000 person-years (PY) among those who started ART vs 193 per 1000 PY among those who deferred; adjusted hazard ratio (HR) 0.32, or 68% reduction in risk; 50-199 cells/mm3: 22 vs 57 events per 1000 PY; adjusted HR 0.48, or 51% risk reduction; 200-349 cells/mm3: 19 vs 29 events per 1000 PY; adjusted HR 0.59, or 41% risk reduction; 350-499 cells/mm3: 17 vs 21 events per 1000 PY; adjusted HR 0.75, or 25% risk reduction; 500-799 cells/mm3: 15 vs 19 events per 1000 PY; adjusted HR 1.10, or essentially equal risk.
	Patterns were similar when looking at deaths alone.
	Analysis of 3-year cumulative event rates emphasized the differential benefit according to CD4 count, including calculation of the number of patients who would need to start treatment to prevent 1 case of AIDS or death:
	0-49 cells/mm3: 17% cumulative risk for those who started ART vs 47% for those who deferred; number needed to treat = 3; 50-199 cells/mm3: 6% vs 21% cumulative risk; number needed to treat = 7; 200-349 cells/mm3: 6% vs 10% cumulative risk; number needed to treat = 21; 350-499 cells/mm3: 3% vs 6% cumulative risk; number needed to treat = 34; 500-799 cells/mm3: 5% cumulative risk for both groups; number needed to treat = infinite.
	For death alone, the respective numbers need to treat to prevent 1 death were 6, 14, 74, 71, and 239 patients, respectively.

The researchers concluded that starting antiretroviral therapy with a CD4 count < 500 cells/mm3 "appears to reduce risk compared to deferring at baseline."

At CD4 counts of 350-499 cells/mm3 and above, however, "absolute risk [was] very low in [the] short term (1 year)," and at CD4 counts of 500-700 cells/mm3, there was "no apparent benefit to [ART] initiation for the larger population of patients with CD4s in this range."

These findings support current U.S. treatment guidelines recommending ART initiation when CD4 count falls below 500 cells/mm3. Given that the analysis left out non-AIDS conditions, however, the issue of whether treatment is beneficial above this level remains unresolved.

In the words of session moderator Sharon Walmsley from the University of Toronto, "We know to start [ART] early, but how early we still don't know."

Investigator affiliations: University of North Carolina at Chapel Hill, Dept of Epidemiology, Chapel Hill, NC; EpiQuest Sciences, Inc, Libertyville, NC; Medical Research Council Clinical Trials Unit, London, UK; McGill University, Department of Epidemiology, Biostatistics, and Occupational Health, Montreal, Canada; North Carolina State University, Department of Statistics, Raleigh, NC; GlaxoSmithKline, Worldwide Epidemiology, Research Triangle Park, NC; Vanderbilt University Medical Center, Nashville, TN; University of North Carolina at Chapel Hill, School of Medicine, Chapel Hill, NC.

7/27/10

Reference
M Jonsson, JS Fusco, SR Cole, and others. HAART initiation and clinical outcomes: insights from the CASCADE cohort of HIV-1 seroconverters on "When to Start." XVIII International AIDS Conference (AIDS 2010). Vienna, July 18-23, 2010. Abstract THLBB201.