Something to Measure
Propelled by Results From Multiple Studies, Viral Load Monitoring Lunges to the Forefront of HIV Medicine
Optimization — and Access — Remain
Recently, several very powerful and sensitive tests for measuring the presence of HIV RNA in the blood have become available to clinicians and patients (although primarily to those with insurance to cover their exorbitant price). The FDA has approved one viral load test, the Roche PCR (“Amplicor”) test, for diagnostic and prognostic use. Two other tests, Chiron’s “Quantiplex” branched-chain DNA (bDNA) test and Organon Teknika’s nucleic acid sequence-based amplification assay (NASBA) are also available, though not yet FDA approved. The Roche test can detect down to about 200 RNA copies per cubic milliliter (mm3) of plasma, and the Chiron and Organon Teknika tests detect down to about 400-500 copies. Newer, second and third generation assays can detect virus lower still: down to around 25 copies/ml.
Viral Load for Assessing Prognosis
In Vancouver, multiple lines of evidence converged to demonstrate that viral load was clearly of great prognostic value in assessing a given HIV-infected individual’s risk of progression and death over a two-, five- and ten-year period. The most compelling evidence was presented by John Mellors in a retrospective survey of 1,601 men from the Multicenter AIDS Cohort Study (MACS) using viral RNA levels in 1985 to predict clinical outcome in 1995. For 855 patients who died of AIDS the average baseline RNA count was 24,200; for 993 who developed AIDS it was 19,145; for 749 who remain alive it was 4,426 and for those who did not develop AIDS it was 3,636. Those entering with viral load over 30,000/ml had a 13-fold increased relative risk of AIDS and an 18-fold increased risk of death. Risk of progression and death grew steadily with increasing viral load. CD4 remained a useful predictor, especially for patients with lower CD4 cells. Mellors stated that viral load and CD4 should be used together in assessing prognosis.
Seven other papers confirmed the association between higher viral load and more rapid progression in both adults and children. Of note, many children were found to have higher viral loads than adults, which may explain why some children progress so rapidly. Several studies tried to correlate maternal RNA levels with the risk of transmitting HIV to offspring. While higher plasma RNA did indeed correlate with higher rates of transmission, there was no critical cut-off below which the risk of transmission was greatly reduced. Therefore, mothers with high and low viral load should be offered antiretroviral therapy to interrupt vertical transmission. While higher plasma RNA levels correlated with higher levels of HIV in semen and vaginal fluids, lower levels did not guarantee that the genital fluids were non-infectious. Patients with low plasma RNA levels are clearly still infectious.
Viral Load for Assessing Treatment Response
Virological analysis from several studies demonstrated that patients experiencing a reduction in viral load during the initial weeks of treatment obtain later clinical benefits from this reduction (VA 298, ACTG 116B/117, ACTG 175, CPCRA 007, Delta, Abbott late-stage ritonavir study, Roche saquinavir/ddC study). Thus, the ability of viral load changes to predict clinical benefit from a regimen appears well-founded. However, no study has yet attempted to maintain viral load below a certain critical “threshold,” or to use viral load levels to trigger a change in therapy. Such viral load-based strategy trials will be an important research priority in coming years.
Viral Load for Medical Management
Although viral load has emerged as a critical tool in the diagnosis, prognosis and treatment of HIV disease, many patients and providers lack access to these tests or knowledge about how to use them. The province of British Columbia adopted a viral load-based treatment strategy in May 1996, and has secured a price for the Roche Amplicor test which is less than half that in the USA ($63 in Canada vs. $150 in the USA). British Columbia recommends treatment for those with CD4 below 500/mm3 or those with viral load above 10,000/ml, with the goal of reducing viral load by at least 50%. The International AIDS Society released recommendations to initiate therapy when viral load exceeds 5,000-10,000 copies and CD4 counts are falling, or when viral load is over 30,000 regardless of CD4 levels. They suggest that treatment should reduce HIV levels by at least 0.5 log, to below 5,000 copies, or to undetectable levels (three very different goals), and recommend use of viral load tests twice at baseline, within 3-4 weeks of changing therapy, and every 3-4 months or in conjunction with CD4 cell tests. These guidelines are based more on guesswork than on research, and were immediately outdated by the end of the Vancouver conference, where five studies showed that viral levels could be lowered to undetectable levels for up to 48 weeks using five different regimens in five different populations, ranging from AZT+3TC+ritonavir in acutely-infected individuals (Markowitz), AZT+3TC+nelfinavir in AZT-naive individuals (Ho), AZT+3TC+indinavir in AZT-experienced subjects (Gulick), ritonavir+saquinavir in AZT-experienced subjects (Cameron) and, perhaps most surprisingly, AZT+ddI+nevirapine in AZT-naive subjects. David Ho demonstrated that, at least in one of these studies, viral load was below 25 copies/ml, and possibly at zero, 24 weeks after beginning triple-drug therapy.
Unanswered Questions about Viral Load
While viral load is an exciting and powerful tool, the jury is still out on how best to use it, how often to measure it, and what sort of changes in viral load levels should be attempted in using antiretroviral therapy. Only 2% of the body’s HIV resides in the blood; the rest dwells in the body’s lymphoid tissues and in the brain. Most available drugs do not penetrate the blood-brain barrier, so there is a danger that the virus, cleared from the blood and even possibly the lymph nodes, will hide out in the brain until therapy is removed, and then re-emerge to reinfect the immune system. Moreover, CD4+ cell increases which accompany viral load reduction do not fill all the holes in the immune repertoire which are caused by HIV-induced immunosuppression. In Vancouver there were many anecdotes of people who stopped prophylaxis or maintenance when their CD4+ cells rose ostensibly out of a danger zone; some of these people later developed opportunistic infections. Immune recovery from HIV-induced damage appears to be incomplete, at least in the period measured, in spite of impressive viral load reductions.