The development and recent licensure by the FDA of new molecular techniques designed to detect circulating virion-associated HIV RNA in plasma has enabled investigators to study the viral dynamics and pathogenesis of HIV and has allowed clinicians to carefully monitor the therapeutic efficacy of new antiretroviral drugs. Recent publications and data from the 11th International Conference on AIDS have shown that as many as 10 billion new HIV virions are produced per day with a half life in plasma of 6 hours. The life cycle of the virus from infection of one cell to the production of new progeny which infect the next cell is approximately 2.6 days. This extraordinarily high level of viral replication, viral mutation rate, cell destruction, and cell replacement has led to a dramatic shift in the clinical management of HIV-infected patients and, in particular, the use of antiretroviral therapy.

In a recent report in Nature: Medicine [1996;2:625-629], Saag and colleagues provide a review of the use of viral load tests in the management of HIV-infected individuals (see page 13, Snapshots of the Literature). These are listed as interim recommendations since few data are available from clinical studies to indicate the exact viral load parameters for instituting or changing antiretroviral therapy. It is generally agreed, however, that once therapy has been started, the minimal decrease in HIV RNA should be greater than a 0.5 log decrease (a threefold linear decrease) from baseline. Most effective therapies appear to achieve at least a one log decrease, and triple therapy has been shown to achieve 2-3 log decreases. Because of the development of resistance, these dramatic decreases in viral load are often lost over time, but the goal is to still keep viral load lower than 10,000 copies per ml. How low is adequate is not known, but therapeutic advances discussed at the meeting are aiming for levels that are undetectable (< 500-1000 copies/ml).

In Scott Hammer's plenary presentation, he noted that baseline HIV-1 RNA levels do predict progression to a very high degree. This statement was supported by a number of scientific presentations including John Mellors' MACS data and Neil Graham's data on viral load in the Baltimore MACS and ALIVE cohorts described below. Changes in RNA levels in response to antiretroviral therapy, however, do not fully explain the treatment benefits, but still provide a useful tool for evaluating initial response to treatment in order to determine both when a patient should be placed on therapy and when an antiretroviral regimen has lost its benefit.

John Mellors of the University of Pittsburgh updated his recent findings (Science 1996) on the prognostic value of viral load measurements. These data were derived from blood samples taken from patients enrolled in the MACS. Mellors et al tested 1,604 blood samples taken at enrollment or at seroconversion from members of the MACS cohort, then followed the natural history of the disease among those patients. HIV RNA was measured by the second generation bDNA assay. Patients were divided into quartiles based on level of HIV RNA measured at baseline and these were correlated with disease progression. The lowest quartile was separated into those patients with viral loads less than the limit of detection and those who had detectable virus. Results showed a striking separation between all five groups and the risk of disease progression. Patients with the highest viral load at baseline have the highest risk of disease progression compared to those in all other quartiles, with the risk following a stepwise pattern from highest to lowest amount of RNA measured. Over 90% of patients with undetectable virus survived more than 10 years, or the entire length of the study, while those with the highest viral load on entry survived approximately four years. CD4 cell counts were also independent predictors of risk, although they were not as predictive as the level of viral RNA. The best predictive value was obtained when viral load was combined with CD4 cell count.

These data were complemented by data presented by Neil Graham of The Johns Hopkins University, who reported on viral cultures taken from patients enrolled in the Baltimore MACS cohort and the ALIVE study of HIV-infected IV drug users in Baltimore. In summary, those patients with the highest viral load as measured by culture methods also had the highest risk of disease progression and death, decreasing in a progressive fashion to the lowest risk among individuals with negative cultures. Those with the highest viral load had almost four times the risk of progression to AIDS after two years than those with the lowest load. Data from these two particular studies demonstrate that viral load measurements are applicable to gay men, IV drug users, and women, the latter two of which were part of the ALIVE cohort. It was further suggested that the RNA detection methods would probably be more sensitive than the culture method.

Additional studies examining the utility of HIV viral load included a number of studies which demonstrated that HIV RNA decreased more dramatically in combination therapy than in monotherapy in both the Delta studies and CPCRA 007. In both studies, a one log decrease in HIV RNA at six months resulted in a significantly lower risk of disease progression or death. In several other studies, it was evident that pregnant women with markedly elevated RNA levels were more efficient transmitters of HIV to their offspring. However, none of the studies was able to identify a threshold which was predictive of transmission. For example, the likelihood of transmission increased 18-fold for every 10-fold increase in viral load. However, some women with low viral loads were still found to transmit HIV, suggesting that other factors besides viral load may be important in predicting transmission. In one of the late-breaking sessions, data from the women-infants' transmission study (WITS) demonstrated that elevated RNA copy number at birth was predictive of in utero infection compared to intrapartum infection and that 90% of infants infected intrapartum have undetectable RNA copy number or, on average, less than 500 copies/ml. By one month of age, RNA rises to a median of 315,000 copies/ml. Infants with RNA levels >315,000 copies/ml had a 5-fold greater risk of progression to AIDS or death compared to infants below that level. None of the infants with levels below 100,000 copies/ml progressed to AIDS or death. Without therapy, RNA viral levels remained relatively high in contrast to adults throughout most of their lives. These were the first data presented on a large cohort of infants, and they clearly suggest that aggressive antiretroviral therapy in infants may be beneficial in decreasing disease progression.