Slow decline in CD4 cell count in a cohort of HIV-infected adults living in Lusaka, Zambia

Abstract

Our understanding of the HIV pandemic and its consequences for human populations is still limited by our lack of insight into the natural history of HIV infection in resource-poor countries. The report by Urassa et al. [1] has contributed to emerging evidence that HIV progression rates in Africa are not so different to those that prevail in Europe and north America [2]. Their data will also help practitioners in Africa to make evidence-based decisions about CD4 cell monitoring in their patients. We have now analysed temporal changes in the CD4 cell counts drawn from a cohort in Zambia. This cohort is smaller but population based, not occupational, and therefore does not show a healthy worker effect. Although we found a similar rate of decline in the CD4 cell count, there is an important qualification to be made about both these datasets. This population-based cohort of unselected adults in Misisi, a shanty compound in Lusaka, has been under follow-up since 1999, and CD4 cell counts have been followed in participants since the initial survey [3]. No antiretroviral drugs were used by any of the participants over the period of the study. Approval was given by the research ethics committees of the University of Zambia and the London School of Hygiene and Tropical Medicine. The initial cohort of 261 adults included 65 HIV-seropositive participants, of whom 12 died, and we were able to obtain repeated measurements (in 1999, 2000, 2001 and 2003) in 24. Among the survivors, the mean age of men (n = 7) was 35 years (SD 7.0) and of women (n = 17) it was 28.6 years (SD 5.7), and the median initial CD4 cell count was 389 cells/μl (interquartile range 255–537). The median initial CD4 cell count was 122 cells/μl among the 12 who died. In the survivors, the mean decline over 4 years was 29 cells/μl per year (Fig. 1). The mean percentage decline from baseline was 30%. In six of those who died and who had had at least two measurements separated by at least one year, the mean decline was 15 cells/μl per year.Fig. 1: Changes in CD4 cell count over a 4-year interval from 1999 to 2002 in 24 survivors with HIV infection. No observations were made in 2002 and where data are missing lines connect from earlier to later observations using a straight line. Although many lines (connecting serial observations from the same participant) overlap, the slow downward trend can readily be seen.Our data indicate that the decline in CD4 cell counts over the period from 1999 to 2003 in adults not treated with highly active antiretroviral therapy was slow, and our estimate of the rate of loss of CD4 cells is in very close agreement with the estimate of 21.5 cells/μl per year from Tanzania [1]. These data support the idea that HIV progression to AIDS and death is slower than at first thought, even in very under-resourced populations living in crowded conditions. The data are unlikely to be skewed by depressed initial CD4 cell counts as a result of intercurrent infection because close analysis of the data depicted in Fig. 1 reveals only three individuals whose CD4 cell counts rose from an initially depressed level. None of these had any evidence of infection around the date of first sample collection in 1999. However, the slow rate of progression observed in both of these cohort studies may not fully explain the observed age distribution of HIV infection seen in sub-Saharan Africa. At this rate of decline it would take over 19 years for newly infected adults in similar populations to progress from a CD4 cell count of 774 cells/μl (the median CD4 cell count in our HIV-seronegative participants [3]) to a count of 200 cells/μl (the level at which AIDS is defined). It is well established that infection occurs in young women at an earlier age than in young men [4], and AIDS-related deaths start to occur in women in their third decade of life [5–7]. It could not have taken 19 years for these young women to progress from infection to AIDS if their infections were acquired sexually after puberty, so we must consider alternative explanations. First, that infections leading to death in the third decade were acquired in childhood. Second, that progression is not constant over time (for example there might be a faster fall in the CD4 cell count immediately after seroconversion), but in the Tanzanian study the rate of decline was similar in seroprevalent and seroincident cases so this is unlikely. Third, that progression in adolescents and young adults could be faster than in older adults. Our study only included four adults under 25 years of age, and in the Tanzanian cohort the adults studied were also mostly over 25 years of age. In most African populations, the majority of the population is under 30 years of age. The rate of decline in teenagers and young adults now needs to be determined specifically in future studies, as this could imply that they need more frequent CD4 cell count measurements or the earlier initiation of highly active antiretroviral therapy to reduce the risk of progression to AIDS.