|Home||Search Abstracts||Browse Sessions||Program Committee||E-mail Abstract Author||View Session|
Wednesday, 1:30 - 3:30 pm
Background: According to the setpoint theory, HIV-1 RNA level remains at stable levels from shortly after infection until a few years before the development of AIDS.† Over the last years, this theory has been critcised. We reconsidered this theory by modelling the joint development of CD4 and HIV-1 RNA level from HIV infection until AIDS diagnosis. We used follow-up data from two cohort studies among homosexual men (N=400), having more than ten years of follow-up, until highly active anti-retroviral therapy became available.
Methods: The joint marker development was fitted via a random effects model. A calendar time effect on the level of CD4 count and HIV-1 RNA level and on the accuracy of laboratory method was included. The effect of use of anti-retroviral therapy (ART) at the individual level was incorporated by allowing the level of both markers to change at the start of monotherapy and bitherapy. Persistent as well as temporary effects of ART were investigated. We also used a random effects model for the marker development in the last four years before AIDS diagnosis. The fitted trajectories were compared with the curves fitted through the scatterplot data.
Results: HIV-1 RNA increased by about 0.1 log per year. No significant change in slope was found in the last four years before AIDS diagnosis. Hence no setpoint in HIV-1 RNA level was found. The scatterplot fit, however, falsely suggested HIV-1 RNA to remain stable more than six months after seroconversion. This difference is explained by the selection mechanisms present in the scatterplot fit. Both monotherapy and bitherapy lowered average HIV-1 RNA levels.† Only bitherapy had a significant effect on CD4 count, but this effect did not persist for more than six months. The level of CD4 and HIV-1 RNA changed over calendar time and CD4 measurements became more accurate over calendar time. The joint model allows reconstruction of missing information on a marker trajectory, using information from the other marker.
Conclusions: The scatterplot data do not provide information on natural history of markers. A random effects model should be used instead. This model showed that HIV RNA level steadily increases more than six months after seroconversion, both at the population level and at the individual level. The setpoint theory does not hold.
Keywords: HIV RNA setpoint; natural history; CD4