Longitudinal Analysis of Quantitative Virologic Measures in Human Immunodeficiency Virus-Infected Subjects with ⩾400 CD4 Lymphocytes: Implications for Applying Measurements to Individual Patients

The natural variability of quantitative virologicmeasures among human immunodeficiency virus (HIV) type 1-infected persons was prospectively studied in 29 untreated persons with >600 CD4 cells/ILL and in 15 persons receiving zidovudine monotherapy who had 400-550 CD4 cells/ILL at study entry. Cell- and plasma-associated infectious HIV-1, provirus, and virion RNA were deter­ mined monthly as were numbers of CD4 and CD8 cells. HIV-1 replication varied widely among subjects with similar CD4 cell counts. The within-individual variability was significantly less than the variability between subjects for all virologicmeasures. Plasma virion HIV-1 RNA levels had the least variability. A mathematical model was devised to assess whether a potential therapeutic intervention significantlyalters peripheral HIV-1 load. The model indicated that three measurements of plasma RNA would be outside the 95th percentile for the expected change in an individual due to natural variability. This approach can be used to accurately assess a therapeutic intervention among persons with low plasma HIV-1 titers. A number of studies have found that human immunodefi­ ciency virus (HIV) type 1 persistently replicates during the course of infection [1-5]. Both acute and chronic infections are associated with detectable levels of virus in the peripheral circulation [6-8]. Although lymphoid tissue is the primary tissue site of HIV -1 replication, cross-sectional studies have shown an inverse correlation between a high titer of HIV-1 in peripheral blood mononuclear cells (PBMC) and plasma and low CD4 cell counts [9-13]. While some studies have shown a correlation between increases in virus load and the progres­ sion of infection [14-18], few have quantitated HIV-l load among untreated clinically asymptomatic individuals over time, and little is known about the natural variation in these virus load measurements. We performed detailed analyses of the proviral, viral RNA, and infectious load in the peripheral blood of a group of HIV-l-infected subjects with >400 CD4 cells/ f.LL in order to develop a model by which one could measure in a single person the degree of change indicative of virologic progression or significant inhibition by a therapeutic inter­ vention.