Drug Resistance and Virulence
January 31, 2003

Clinical trials are a highly structured treatment setting that may not be readily translated into clinical practice. For example, subjects in clinical trials are often drug naive, have low CD4+ cell counts, do not have diarrhea, and are expected to adhere well to medical recommendations. Patients in clinical practice often have none of these characteristics. To determine the effectiveness of potent protease inhibitor therapy in clinical practice, Dr. Steven Deeks (UCSF) in collaboration with Dr. Grant retrospectively reviewed all patients treated with potent protease inhibitor therapy between March 1996 and October 1997 at San Francisco General Hospital’s AIDS clinic. Virologic failure, defined as two consecutive viral loads >500 HIV-1 RNA copies/ml of plasma, occurred in 53% of subjects. Risk factors for virologic failure included high pretreatment viral load, low baseline CD4 count, and failure to alter the antiretroviral regimen when protease inhibitor therapy was added. These data indicate that virologic failure occurs in many patients in clinical practice based on currently available antiretroviral regimens. These data also indicate that treatment with potent protease inhibitors must be combined with other antiretroviral agents to avoid drug resistance.

Whether such virologic failure portends clinical failure has been an active area of research. Jason Barbour, a UCSF bioinformatics graduate student working in the laboratory, analyzed trends in blood CD4 cell counts in persons who had virologically failed therapy. As reported by other groups, CD4 cell counts rose after the initiation of antiviral therapy, possibly representing sparing of lymphocytes or lymphopoiesis during periods of viral suppression. Surprisingly, CD4 cell counts continued to increase among many subjects who had suffered virologic failure. This suggests the hypothesis that HIV-1 replicates less virulently in the presence of antiviral therapy, possibly because of decreased viral replication or a disassociation between viral replication and disease, as we have observed in the natural hosts of SIV (see below). More detailed analysis indicated that subjects with detectable viremia still had levels of viremia that were much lower than pretreatment baselines, supporting the hypothesis that immune preservation after virologic failure is due to decreased viral replication rather than to evolution toward less virulent replication. This work also has had direct clinical implications by altering the way that clinicians interpret viral load measurements during therapy. While the absolute level of viral load was widely used to guide therapy in the past, the analysis of our clinic data indicated that the change in viral load relative to pretreatment baseline was more predictive of CD4 trends than the absolute level of viral load. Comparison of viral loads during treatment with pre-treatment levels has now become the standard of medical care.
To test whether continued drug selection was required to maintain partial virological and CD4 responses after long-term virological failure, a randomized clinical trial was performed by Dr. Deeks in collaboration with Dr. Grant. Careful monitoring of viral load after interruption of antiretroviral treatment among viremic persons indicated that viral load increased 0.4 log10 immediately after antiretroviral interruption but before any change in drug resistance. This increase in viremia likely reflects continued drug activity against partially resistant viral populations. After 2–16 weeks, a second rise in plasma viral load occurred concomitantly with an overgrowth by drug-sensitive viral populations. These drug-sensitive viral populations had higher replication capacity than baseline drug-resistant populations in a single-round replication assay developed by ViroLogic, a biotechnology company based in South San Francisco. The rise in drug-sensitive viremia was associated with a dramatic decrease in CD4 counts, a clinically important measure of immune status. This study showed that partial virological responses after long-term virological failure of therapy were dependent on continued drug selection pressures that had residual activity against the resistant viral population and maintained selection for viral populations with diminished capacity for replication.

Diminished replication capacity of drug resistant HIV-1 had been suggested in laboratory assays involving indicator cells and constructed viruses, but direct evidence of markedly decreased fitness in humans had not been confirmed. The study of subjects stopping antiretroviral therapy provided an opportunity to observe naturally occurring HIV-1 populations competing in the clinically relevant ecosystem, namely human beings. For this research, the ratio of drug resistant and drug sensitive HIV-1 in the populations was assessed at frequently using a differential probe hybridization assay developed at Bayer Diagnostics, a biotechnology company in Emeryville, California. This detailed viral genetic analysis revealed rapid overgrowth of resistant virus in most subjects after therapy was stopped, yielding estimates of the relative fitness of drug resistant HIV-1 in vivo between 50 and 90 percent. The extent of viral impairment measured in humans correlated with changes in viral load, demonstrating clinical relevance. Further, viral fitness correlated with differences in replication capacity measured using the rapid assay, suggesting novel uses for viral fitness assays in the medical management of AIDS.

 Tell a friend
 Printer friendly version