The Body Covers: The 39th Annual Meeting of the Interscience Conference on Antimicrobial Agents and Chemotherapy
Managing HIV Disease for Life --
New Issues Challenges, and Opportunities
September 26, 1999
In this talk, Dr. D'Aquila summarized the salient aspects relevant in understanding the issues involved in resistance of HIV to antiviral treatment. The topics addressed include updates on the transmission of resistant HIV, mutations which are selected during virologic escape from triple drug ART, considerations of cross resistance, and evaluations of resistance testing in managing patients.Abstract: The Science of Resistance Comes of Age
Primary resistance mutations are generally those that occur first on a nonsuppressive regimen, and result in a decrease in the sensitivity of HIV to the drug. They may reduce replication capacity in the absence of the drug, although some mutations do not result in this loss of "fitness." Secondary mutations are those that usually happen if HIV is allowed to continue replicating in the presence of the drug. These mutations may further improve on fitness as well as increase the IC50 required for reestablishing suppression with the same agent. Further, these secondary mutations increase the likelihood of within-class cross-resistance.
Within the nucleoside inhibitors, multiple mutation patterns have been observed. While there is some overlap among the agents, there are some distinct patterns as well. Recent controversies have been over the occurrence of mutations initially described with AZT and how often they manifest when using stavudine instead. While these mutations have been reported in those on d4T, the impact of these mutations to the phenotypic susceptibility to stavudine remains minimal, and thus the relevance of these mutations remains uncertain. Within the nonnucleoside RT inhibitors, there appear to be two patterns, one which includes the K103 mutation resulting in cross class resistance, and one which does not. However, the ability of efavirenz to inhibit HIV which has mutations other than K103N remains unclear, and the ability to "sequence" nonnucleosides remains a point of continued controversy.
The prevalence of multidrug resistant HIV has been document. The impact of transmission of these mutations has been reported, including a decrease in the response rate to commonly selected regimens. A slower than expected fall on an initial regimen should prompt the clinician to consider underlying resistance as a plausible mechanism if such a response is observed in the case of a patient who is judged to be fully adherent with the regimen. Resistance testing might help to define the degree of preexisting resistance, as well as help to define which regimens might still be maximally active. As an example, Dr. D'Aquila described the results in one patient who was treated with a regimen of AZT, 3TC, and indinavir. In the second month, the viral load had only gone from 6 logs to over 4 logs RNA. The virus in this patient was sequenced, and found to have some degree of resistance to this regimen. The regimen was replaced the one containing stavudine, didanosine, abacavir, efavirenz, and hydroxyurea, and there was a prompt reduction in RNA to below 50 copies in the next 100 days on therapy.
Differences among the PI's were reviewed, in terms of contrasts in cross resistance. Data from study CCTG 575 were reviewed. (Haubrich, ICAAC 99) In this study, phenotypic sensitivity and cross resistance was tested in patients with viral escape on a nelfinavir containing regimen, versus those who had been on indinavir. The results showed that nelfinavir resistant HIV had less cross resistance seen to other PI's as compared to indinavir resistant HIV. The former evidenced phenotypic resistance to primarily just nelfinavir, with about 10% or less cross resistance seen to the other agents. However, indinavir resistant isolates also demonstrated reduced sensitivity to most of the other agents, retaining sensitivity to primarily to only saquinavir (and amprenavir to a lesser extent). Further, data suggest that the D30N mutation results in decreased replicative fitness in the absence of the drug, as compared to the effect of an L90M mutation which has less of an impact on fitness.
Thus, there are several settings where resistance testing can play a useful clinical role. As described above, there is the scenario of identifying resistance in an untreated individual. In addition is the situation of identifying which agents might have significant resistance in an individual who has viral rebound, and the added ability to identify compounds to which HIV might still be sensitive. This latter situation has several studies which address the role of testing in this circumstance. Two retrospective studies have demonstrated the predictive potential for resistance testing. Zolopa studied the role of resistance testing in addition to clinical predictors in determining the success of the dual PI combination of ritonavir/saquinavir in patients who had virologic rebound after nelfinavir. These data showed that most of the success of this regimen was predicted by the mutations observed in the protease gene, and there was little added prediction accuracy when including clinical factors (Zolopa, Antiviral Therapy 1999, Suppl 1, 37.) Specifically, those patients who had either no PI mutation, or just a mutation at position 30 were more likely to have viral suppression when compared with patients who had other mutations, including changes at position 90. Similarly, data about the efficacy of abacavir in nucleoside experienced patients shows a clear correlation with the degree of resistance to both AZT and 3TC.
Data from two prospective studies confirmed the potential for these tests to assist in improving the success after virologic escape on a regimen. One study, known as Viradapt, was a randomized study of patients with viral rebound on a PI based regimen, in which clinicians were either given or not given the results of a genotype when deciding on the next regimen. The results showed that there was a significant increase in thelikelihood of viral suppression to below 200 copies in those who had genotype data available, with about 30% success on this arm as compared to 14% success in the control arm. In a follow up, the cohort who did not get a genotype at baseline were given a genotype after month 6 on their new regimen. The data after this crossover showed that 30% of this cohort were now able to achieve virologic suppression. Further data about the value of this test came from an analysis of costs of care during this study. While the genotype group had the additional costs of this test, there was a tradeoff in reducing the costs of antiretroviral therapy in this group, and thus the overall costs at the end of twelve months were similar in the two groups. Data from the CPCRA GART study were also reviewed. This study also showed the improved in viral suppression in the group where genotype data were available, and an analysis demonstrated that the benefit of the genotype was to guide the practitioner in the ability to use more active agents in treating patients. An analysis of the specific PI mutations impacting success showed that the L90M and T69D mutations were correlated with less suppression, where as D30N was again associated with an increased likelihood of success in the next regimen.
There are several remaining challenges in the use of these tests. One is a better understanding of whether mulitple resistance mutations are linked in the same isolate, or whether mutations are present in a population of virus. This analysis would be helpful in that multidrug salvage is based on the concept that while there are multiple mutations present in the new quasispecies that emerged from drug pressure, there may still be value to using agents to which resistance is measured. This would be the case if some of the isolates were resistant to each of the medications, but that no single isolate was resistant to all of the agents in use. If this could be measured, it could be that some medications would still have some activity in controlling part of the viral population, and a multidrug approach would be designed so that each of the mutated viruses was targeted by at least two new agents selected for use.
In summary, resistance testing has helped in our understanding of how regimens fail. Studies have also shown that such testing can play a role in increasing the likelihood of success when making a switch after viral rebound. Further, data continue to show that there are differences among the agents in terms of the mutations they select for, and the impact of these mutations on the success of subsequent options.
Authored by: Richard D'Aquila, M.D., Associate Professor of Medicine, Harvard Medical School
This article was provided by The Body PRO. Copyright © Body Health Resources Corporation. All rights reserved.