icon-    folder.gif   Conference Reports for NATAP  
  16th CROI
Conference on Retroviruses and Opportunistic Infections Montreal, Canada
February 8-11, 2009
Back grey_arrow_rt.gif
World Health Organization CD4 Criteria Fall
Far Short in Predicting Antiretroviral Failure

  16th Conference on Retroviruses and Opportunistic Infections,
February 8-11, 2009, Montreal
Mark Mascolini
Two studies headlined in the opening press briefing of this meeting underlined the often mortally inadequate HIV care still offered in sub-Saharan Africa and the usually insoluble choice policy makers face in speeding antiretroviral access to the millions who need treatment [1,2].
On the one hand, a four-country study involving more than 13,000 people demonstrated the substantial risk of delaying antiretroviral therapy until the CD4 count falls below 200. On the other hand, a 1133-person study in rural Uganda showed that relying on World Health Organization (WHO) CD4-count criteria to signal treatment failure rarely identifies people with actual virologic failure.
The four-country study by investigators from the International Epidemiological Database to Evaluate AIDS (IeDEA) compared age-, gender-, and country-specific non-HIV death rates with mortality in five antiretroviral programs--two in South Africa and one each in Cote d'Ivoire, Malawi, and Zimbabwe [1]. Of the 13,249 people in these group, two thirds were women and the median age was 34. During 14,695 person-years of follow-up, 1177 people (8%) died.
For people who began antiretroviral therapy with fewer than 25 CD4 cells and WHO stage 3 or 4 disease, the death rate exceeded the non-HIV death rate 47.1 times (95% confidence interval [CI] 39.1 to 56.6). For people who started antiretrovirals with at least 200 CD4s and WHO stage 1 or 2 disease, mortality was 3.44 times higher than non-HIV mortality (95% CI 1.91 to 6.17). For people in the latter group who survived the first year of antiretroviral therapy, the death rate was only 1.72 times higher than the non-HIV death rate (95% CI 0.44 to 1.17).
Half of the entire antiretroviral-treated study group had a death rate more than 5 deaths per 100 person-years higher than the non-HIV group.
The IeDEA investigators concluded that "much of the excess [mortality] might be prevented by more timely initiation of antiretroviral therapy."
But starting antiretrovirals in programs that cannot afford viral load monitoring poses grave risks of its own, according to results of a collaborative study by Johns Hopkins University and Makerere University in Kampala, Uganda [2]. Because this HIV program in Rakai, Uganda measures both CD4 counts and viral loads regularly, the investigators could determine how many people who met WHO CD4 failure criteria suffered virologic failure. At the same time, they could calculate how many people with virologic failure met WHO CD4 failure criteria.
WHO CD4 failure criteria are (1) persistent CD4 count below 100, or (2) more than a 50% drop from peak CD4 count, or (3) a CD4 count lower than the first measured CD4 count without a coinfection that might explain the decline. The investigators defined virologic failure with three cutoffs: above 10,000, above 5000, and above 400.
The study involved 1133 people who began their first antiretrovirals between June 2004 and September 2007. Median follow-up measured 20.2 months (interquartile range 12.4 to 29.5 months). In that time, 125 people (11%) met WHO criteria for CD4 failure. Viral load monitoring showed, however, that only 18 of those 125 (14%) had a virologic failure at the 10,000-copy cutoff. Eighty people did endure virologic failure with a viral load above 10,000. But WHO CD4 criteria identified only 18 of them (23%) as failures. Those numbers meant WHO CD4 standards had a sensitivity of 23% and a positive predictive value of only 14% in identifying virologic failure at the 10,000-copy mark. WHO CD4 sensitivity and positive predictive value were also low with a 5000-copy failure cutoff (28% and 8%) and a 400-copy failure cutoff (26% and 21%).
If Rakai clinicians based treatment decisions only on WHO CD4 criteria, they might have needlessly started a second-line regimen in 107 of 125 people. In the 80 people who did suffer virologic failure, clinicians might have delayed a treatment switch too long in all but 18 of them. Delayed switching can open the door to clinical progression as well as to pile-ups of resistance mutations. Other recent work confirms that WHO CD4 yardsticks fall short in predicting virologic failure [3,4]. Abundant research is under way to identify and validate simple and inexpensive viral load assays [5]. This study suggests that research deserves high priority.
1. Brinkhof M, Boulle A, Weigel R, et al. Mortality of HIV-infected patients starting antiretroviral therapy: comparisons with the general population in southern Africa. 16th Conference on Retroviruses and Opportunistic Infections, February 8-11, 2009, Montreal. Abstract 141.
2. Reynolds S, Nakigozi G, Newell K, et al. Evaluation of the WHO immunologic criteria for antiretroviral treatment failure among adults in Rakai, Uganda. 16th Conference on Retroviruses and Opportunistic Infections, February 8-11, 2009, Montreal. Abstract 144.
3. Badri M, Lawn SD, Wood R. Utility of CD4 cell counts for early prediction of virological failure during antiretroviral therapy in a resource-limited setting. BMC Infect Dis. 2008;8:89.
4. Mee P, Fielding KL, Charalambous S, Churchyard GJ, Grant AD. Evaluation of the WHO criteria for antiretroviral treatment failure among adults in South Africa. AIDS. 2008;22:1971-1977.
5. Fiscus SA, Cheng B, Crowe SM, et al. HIV-1 viral load assays for resource-limited settings. PLoS Med. 2006;3(10):e417.