We now know that HIV replication is an extremely dynamic process, with as many as 1010 new viral particles being produced every day. We also know that the replication of HIV reverse transcriptase is a markedly error-prone process. As a result, this high level of viral turnover results in between 104 and 105 mutations at each site in the HIV genome each day(1). It is therefore not surprising that many patients eventually develop resistance to whatever antiretroviral regimen they have been assigned.
Current guidelines for the clinical management of patients with HIV disease have a common goal: to prevent -- or, failing that, to delay -- the development of drug resistance. This goal has not changed in the decade since zidovudine was approved, but the clinical management of patients with HIV disease has changed dramatically. This rapid and radical evolution has consistently outpaced the publication of new clinical guidelines -- which means that practitioners have had to find other ways to keep abreast of changes in this dynamic and challenging field.
The purpose of this article is to help clinicians better understand the particular nature of protease resistance, its clinical implications, and the importance of sequencing these drugs to extract the maximum benefit from the class as a whole, rather from a single agent in that class.
In this context it is important for the practitioner to remember that drug failure does not necessarily equal drug resistance. In vivo, drug failure is suggested by rising plasma HIV RNA levels. In these situations a number of host factors other than resistance -- among them poor compliance, inadequate drug absorption, and advanced immunodeficiency itself -- can contribute to the failure of a particular drug regimen.
Viral factors, including the presence of the syncytia-inducing phenotype of HIV, may also cause drug failure, even in the absence of resistance. However, the emergence of resistance to a specific antiretroviral agent is most likely to be the cause of a regimen's failure.
In the laboratory, viral resistance can be quantified using phenotypic and genotypic assays. Phenotypic resistance refers to the increased ability of HIV to replicate in the presence of a given drug. This type of resistance is usually reported in terms of the inhibitory concentration of the drug necessary to reduce viral replication by 50% or 90% (the IC50 or the IC90). At present these assays are time-consuming and expensive, costing roughly $600-$900 for each agent tested. They can only be done in specialized laboratories, and they take up to several days to complete. For all these reasons phenotypic assays are unlikely to become widely used clinical tools any time in the near future.
Genotypic resistance denotes the specific genomic mutations that are associated with phenotypic resistance. For example, a mutation at position 82 of the protease gene is associated with phenotypic resistance to indinavir -- but not with resistance to either saqunavir or nelfinavir. Several genotypic assays are currently in development. Although these assays are readily reproducible and rapidly obtained -- usually in a single day and at a cost of $400-$600 for the entire battery of tests -- they are limited in their usefulness by the current lack of consistent data correlating mutation patterns with clinical outcome. Until a consensus is reached, these assays should be used only as research tools.
Phenotypic and genotypic resistance has been described for each of the four protease inhibitors that have received F.D.A. approval(2). The patterns for mutations for each of these protease inhibitors are complex: there are more than 20 sites within the HIV-1 protease gene that have been associated with resistance to these agents (Table 1). Although each drug in this class selects for a specific, predictable pattern of genotypic changes, these patterns do overlap -- and where they do, cross-resistance may occur.
|DRUG||CRITICAL MUTATIONS*||SECONDARY MUTATIONS**|
|Saquinavir||48, 90||10, 36, 63, 71|
|Ritonavir||82, 84||20, 36, 46, 54, 63, 71, 90|
|Indinavir||46, 82||10, 20, 24, 32, 54, 63, 71, 84, 90|
|Nelfinavir||30||46, 63, 71, 88, 90|
*Numbers refer to the codon position on the HIV protease gene that has been shown to be critical for the development of phenotypic resistance to each of the protease inhibitors.
**The role of these secondary mutations in the development of protease inhibitor resistance is unclear. Many of these changes may be compensatory and may enhance the virus's ability to replicate once the critical mutations occur. Furthermore, many of these observed changes exist as naturally occurring polymorphisms in untreated patients.
SAQUINAVIR: The first of the protease inhibitors to gain F.D.A. approval selects for several mutations, particularly at codons 48 and 90. A mutation at either codon leads to a 3-to-10-fold decrease in the in vitro susceptibility of HIV to saquinavir. If both mutations occur, susceptibility decreases by a factor of 100(3). Several large-scale trials have demonstrated that saquinavir therapy selects primarily for the mutation at codon 90; the mutation at codon 48 is relatively rare. Saquinavir can also select for changes at sites 10, 36, 63, and 71(4, 5).
INDINAVIR: Our clinical experience with indinavir resistance is extensive. In the early clinical trials of indinavir, the use of low doses of the drug (less than 2.4 grams/day) resulted in dramatic reductions in plasma HIV RNA levels. Because those doses were too low to prevent the development of resistance, reductions in viral burden were transient -- and switching indinavir-resistant patients to higher doses had no effect on HIV RNA levels.
These findings led researchers to conclude that treating HIV-infected patients with suboptimal concentrations of indinavir only served to exert selective pressure on the virus -- which had the paradoxical effect of encouraging the rapid emergence of drug resistance. Data from clinical trials involving other protease inhibitors suggest that this may well be a class effect, and clinicians are therefore cautioned to administer these powerful drugs only at the optimal therapeutic doses.
During the early trials of indinavir, loss of antiretroviral activity (as measured by rising plasma HIV RNA levels) correlated with the emergence of specific mutation patterns within the protease gene(6). Although no single mutation was present in all resistant variants, the presence of substitutions at positions 46 and/or 82 predicted resistance in 29 of 29 isolates. The number of substitutions correlated with the degree of resistance.
The concurrent use of zidovudine appears to reduce the incidence at which genotypic resistance to indinavir develops(7). In one well-designed clinical trial, Merck Study 019, antiretroviral-naïve patients were randomized to receive either ZDV alone, indinavir alone, or the two drugs in combination. After 24 weeks of treatment, patients in the two-drug arm had a relatively low incidence of indinavir resistance (4 of 22 subjects, or 18%). By contrast, patients on indinavir monotherapy had a significantly higher incidence of resistance (9 of 21 subjects, or 43%).
Clinical data on the subjects in ACTG 320 who received the triple-drug combination of indinavir, ZDV, and 3TC suggests that the use of these three agents may further retard the emergence of indinavir resistance (see "The Next Generation of Antiretroviral Agents")(8).
RITONAVIR: Early dose-ranging studies of ritonavir monotherapy found that the use of suboptimal doses of this drug produced the same phenomenon seen with low-dose indinavir therapy: rapid emergence of drug-resistant viral strains(9). Although the literature now describes nine specific amino acid substitutions that confer resistance to ritonavir, an initial mutation at codon 82 appears to be necessary for the subsequent development of other mutations(10). Resistance patterns for this protease inhibitor seem to be similar to the patterns described for indinavir -- which strongly suggests that the sequential use of these agents is unlikely to result in additional therapeutic benefit.
NELFINAVIR: Although our clinical experience with the newest of the protease inhibitors is limited, we do know that viral isolates from 9 of 17 patients who participated in one of the early dose-ranging studies of nelfinavir monotherapy showed evidence of phenotypic resistance to this drug. Genotypic analysis of the protease gene from these isolates revealed that codon 30 was mutant in all of the phenotypically-resistant variants(11). This characteristic mutation for nelfinavir resistance is sometimes associated with changes at sites 46, 63, 71, 88, and 90. In larger, Phase II trials of nelfinavir, all of the study subjects who experienced viral rebound had the codon 30 mutation. In contrast, all patients with persistent viral suppression lacked this mutation.
Significantly, in 55 patients who developed resistance to nelfinavir, the mutations known to confer resistance to other protease inhibitors were seen only rarely or were not seen at all. Furthermore, most of the viral isolates from patients who developed nelfinavir resistance remained susceptible to indinavir or ritonavir(12). The clinical relevance of these observations is not yet clear, but the implication is that nelfinavir may represent an alternative choice for initial protease inhibitor therapy, since it may offer the advantage of reduced cross-resistance to other agents in its class.
When resistance develops to a particular protease inhibitor, that drug loses its potency as an antiretroviral agent. Moreover, the loss of potency appears to be permanent: pharmacokinetic studies show that resistance to a given drug remains in effect up to a year after that drug is removed from a patient's regimen. Thus, the emergence of drug-resistant viral strains reduces the clinician's therapeutic options -- and the patient's chances of achieving clinical stability.
As the data on protease resistance presented here make abundantly clear, the development of resistance to one of the protease inhibitors can confer cross-resistance to one or more other drugs in this class. In theory, at least, the wrong choice of first-line agent can lead to across-the-board protease resistance, whereas the right choice can leave the clinician and the patient with something close to a full range of options, should resistance develop to the protease inhibitor chosen for initial therapy.
Based on the evidence at hand, the clinician should anticipate that resistance will develop, over time, to any of the protease inhibitors. Specific mutations do occur during the course of chronic therapy, even at optimal therapeutic doses, even in fully compliant patients. Therefore, it is critical for clinicians to think in terms of successive regimens that combine protease inhibitors, nucleoside analogs, and non-nucleoside RT inhibitors, rather than a single, once-and-for-all antiretroviral combination.
Clinicians should plan, from the outset, that patients will require an initial drug regimen, a back-up regimen that can be substituted when resistance develops to the first-line therapy, and quite possibly some form of salvage therapy. These serial regimens should be chosen with the utmost care, to reduce as much as possible the likelihood that resistance and cross-resistance will develop (Table 2).
Until prospective studies have determined the proper sequencing of protease inhibitor therapy, clinicians will have to rely on their best assessment of the available resistance data -- and on such factors as drug potency, tolerability, and cost -- when deciding which protease inhibitor to select for initial or subsequent treatment.
1. Coffin JM. HIV population dynamics in vivo: Implications for genetic variation, pathogenesis, and therapy. Science 1995; 267: 483-9.
2. Condra JH, Schlief WA, Blahy OM, et al. In vivo emergence of HIV-1 variants resistant to multiple protease inhibitors. Nature 1995; 374: 569-71.
3. Jacobsen H, Hanggi M, Ott M, Duncan IB, Owen S, et al. In vivo resistance to a human immunodeficiency virus type 1 proteinase inhibitor: Mutations, kinetics, and frequencies. J Infect Dis 1996; 173: 1379-87. [Page 13]
4. Race E, Sheldon JG, Kaye S, Gilbert SM, Moffatt AR, Duncan IB. Mutations associated with reduced sensitivity to saquinavir occur in a minority of patients treated in combination with ddC: Results from a Phase III clinical trial (NV14256). 4th Conference on Retroviruses and Opportunistic Infections, Washington, D.C., January 22-26, 1997. Abstract 600.
5. Sheldon J, Craig C, Race E, Rose J, et al. Reduced sensitivity to saquinavir (SAQ) occurs infrequently, associates only with 48V or 90M and is modest in degree. 4th Conference on Retroviruses and Opportunistic Infections, Washington, D.C., January 22-26, 1997. Abstract 599.
6. Condra JH, Holder DJ, Schlief WA, Blahy OM, Danovich RM, et al. Genetic correlated of in vivo viral resistance to indinavir, a human immunodeficiency virus type 1 protease inhibitor. J Virology 1996; 70: 8270-6.
7. Condra JH, Holder DJ, Schlief WA, et al. Bi-directional inhibition of HIV-1 drug resistance selection by combination therapy with indinavir and reverse transcriptase inhibitors. XI International Conference on AIDS, Vancouver, B.C., Canada, July 7-12, 1996. Abstract Th.B.932.
8. Gulick R, Mellors J, Havlier D, Eron J, Gonzalez C, et al. Potent and sustained antiretroviral activity of indinavir (IND) in combination with zidovudine (ZDV) and lamivudine (3TC). 3rd Conference on Retroviruses and Opportunistic Infections, Washington, D.C., January 28-February 1, 1996. Abstract LB7.
9. Danner SA, Carr A, Leonard JM, Lehman LM, Gudiol F, Gonzales J, Raventos A, Rubio R, et al. A short-term study of the safety, pharmacokinetics and efficacy of ritonavir, an inhibitor of HIV-l protease. N Engl J Med 1995; 333: 1528-33.
10. Molla A, Korneyeva M, Gao Q, Vasavanonda S, Schipper PJ, Mo HM, Markowitz M, et al. Ordered accumulation of mutations to HIV-1 protease confers resistance to ritonavir. Nature Med 1996; 2: 760-6.
11. Patick A, Duran M, Cao Y, Ho T, Pei Z, Keller M, Peterkin J, Chapman S, Anderson B, Ho D, Markowitz M. Genotypic and phenotypic characterization of HIV-1 variants isolated from in vitro selection studies and from patients treated with the protease inhibitor nelfinavir mesylate. 5th International Workshop on HIV Drug Resistance, Whistler, B.C., Canada, July 4, 1996. Abstract 35.
12. Patick AK, Duran M, Cao Y, Ho T, Zhou P, Keller MR, et al. Genotypic analysis of HIV-1 variants isolated from patients treated with the protease inhibitor nelfinavir, alone or in combination with d4T or AZT and 3TC. 4th Conference on Retroviruses and Opportunistic Infections, Washington, D.C., January 22-26, 1997. Abstract 10.
Steven G. Deeks, M.D., and James O. Kahn, M.D., are with the UCSF AIDS Program, San Francisco General Hospital, San Francisco, CA.