Important note: Information in this article was accurate in September 2002. The state of the art may have changed since the publication date.
The most exciting and eagerly awaited data presented in Barcelona pertained to the treatment of naïve patients. Patients starting therapy for the first time can choose among a number of simple, well tolerated, and highly potent combinations, and those who adhere to therapy shouldn’t have to worry about issues such as drug resistance and salvage therapy for a long time. Nevertheless, treatment failure and drug resistance remain a big concern, especially for highly experienced patients with diminishing options for therapy. A number of interesting presentations on drug resistance and approaches to the treatment of experienced patients were presented in Barcelona.
• Expert prediction of phenotype based on genotype reports
Zolopa from Stanford University presented results of the GUESS study, in which a panel of 12 international resistance experts were asked to predict phenotypic susceptibility based on genotype report [Abstract ThOrB1385]. The study involved 50 randomly selected samples with simultane-ously collected genotype and phenotype analyses. Experts were asked to predict phenotypic susceptibility to all currently available antiretroviral agents using the following fold-change strata: <2.5, 2.5 to 3.9, 4 to 6.9, 7 to 9.9, 10 to 19.9, and >20. In addition, they rated the expected activity of each drug on a 6-point scale.
Susceptibility predictions varied consider-ably, both by class and by drug. For example, accuracy of predictions about NRTI susceptibility ranged from 25% for abacavir to 74% for 3TC. In general, susceptibility was underestimated for all NRTIs other than 3TC. Among the PIs, prediction accuracy ranged from 26% for nelfinavir to 30% for lopinavir/ritonavir. As one would expect, predictions were more accurate for NNRTIs, with about 2/3 of the predictions being accurate for the three drugs in this class. Predictions of drug activity followed similar patterns, although there was stronger agreement among the experts about drug activity than might be expected given the variation in phenotypic predictions.
• Prospective study of resistance testing
The CERT study is yet another prospective study evaluating the utility of resistance testing in patients failing therapy [Wegner SA, et al. Abstract ThOrB1389]. This study compared genotype testing, pheno-type testing, and standard of care (or what used to be standard of care before the widespread use of these tests). Overall, there were no differences in virologic response among the three groups. However pheno-type testing was associated with improved virologic outcome among those who were more heavily treatment experienced, those with prior NNRTI experience, and those with advanced disease. This is in contrast to some of the earlier prospective trials, in which resistance testing appeared to have the greatest utility in patients failing earlier regimens, presumably because they were more likely to have good options for fully suppressive therapy.
• Nucleoside Analog Resistance
Because thymidine analogs (AZT or d4T) have historically been a component of almost all antiretroviral regimens, we are used to thinking of the accumulation of thymidine analog mutations (TAMs) as the sole pathway to NRTI resistance. However, the availability of other NRTIs and concerns about thymidine analog toxicity have sparked interest in thymidine analog-sparing regimens, such as combinations of 3TC or emtricitabine (FTC) plus either tenofovir DF, abacavir, or ddI. As use of such combinations grows, we will undoubtedly begin to see the emergence of different NRTI resistance patterns. Of greatest concern is the K65R mutation, which confers resistance to abacavir, tenofovir, and ddI. K65R is infrequently seen, but data presented in Barcelona and at the preceding Resistance Workshop in Seville suggest that it will emerge more frequently in patients taking non-thymidine analog-containing regimens [Winston A, et al. Abstract TuPeB4600; Alt-Khaled M, et al. XI International Drug Resistance Workshop, Seville, 2002]. The inclusion of a thymidine analog in a regimen containing abacavir, tenofovir, or ddI appears to drive resistance toward the TAM pathway, while preventing the development of K65R. The unanswered question is whether there is an advantage to one pathway over another. Data from the dual-nucleoside era suggested that TAMs might emerge more slowly than K65R, allowing for modification of therapy before significant cross-resistance had developed. However, it is unclear whether these findings can be extrapolated to more suppressive regimens. Resistance data from studies like Gilead 903, where a thymidine analog-containing regimen is being compared with a tenofovir-containing regimen, will be helpful in answering this question.
The M184V mutation confers high-level resistance to 3TC but is known to have a variable effect on the other NRTIs. It increases susceptibility to AZT, d4T, and tenofovir, while decreasing susceptibility to abacavir, ddI, and ddC. The clinical relevance of that loss of susceptibility is thought to be minimal for abacavir, and has been questioned for ddI. Gazzard [Abstract MoPeB3144] and Eron [Abstract TuPeB4486] presented reassuring data suggesting that the clinical impact of prior 3TC experience and/or M184V on ddI susceptibility is minimal.
• Further defining PI resistance
The next protease inhibitor to be approved will undoubtedly be atazanavir (ATV), which has the advantages of being administered once daily and of not raising lipid levels. Resistance patterns are beginning to be worked out for this drug. Colonno presented data on ATV resistance at the Resistance Workshop in Seville, demonstrating the emergence of a novel I50L mutation in 8 of 9 previously naïve patients failing ATV [Antiviral Therapy 2002;7:S4]. Of those, 5 also were found to have the A71V mutation. Interestingly, virus with the I50L mutation remained susceptible or even hypersusceptible to all other protease inhibitors, including amprenavir, for which another mutation at codon 50 (I50V as opposed to I50L) results in a marked decrease in susceptibility. I50L was also found to decrease replication capacity, though this was partially restored by A71V. When ATV is combined with saquinavir, however, 84V is selected rather than 50L.
The strategy of PI sequencing, once widely accepted, has taken yet another hit. It now appears that the I50V mutation, once thought to confer resistance only to amprenavir, is not so benign as it was once thought to be. Specifically, it results in a significant decrease in susceptibility to lopinavir (LPV). Other mutations associated with a decrease in LPV susceptibility are being identified, allowing genotype algorithms to be refined so that they are more predictive of phenotypic susceptibility to LPV [Parkin NT, et al. Abstract TuPeB4613].
• NNRTI Hypersusceptibility
It has been established that NRTI-resistant virus demonstrates phenotypic hyper-susceptibility (HS) to NNRTIs; however, the clinical relevance of this finding has been controversial. Haubrich presented data on 177 patients enrolled in the California Collaborative Treatment Group (CCTG) studies and assessed the association between phenotypic susceptibility to NNRTIs and response to an NNRTI-containing regimen [Abstract ThOrB1388]. Patients were NNRTI-naïve and failing a stable regimen that included NRTIs. NNRTI-HS was defined as a fold-change of <0.4, and was seen in 24% of patients for efavirenz and 20% for nevirapine. It was more common in patients who had at least 3 NRTI mutations. Haubrich found that the mean decrease in viral load 6 months after starting the NNRTI-containing regimen was 1.2 log10 c/mL for those with HS compared to 0.8 log10 c/mL for those without HS (p=0.016), and the difference persisted through 12 months of therapy (p=0.023). CD4 response was also better in patients with HS. NNRTI-HS remained a significant predictor of virologic response in multivariate analyses after accounting for baseline viral load and the number of drugs used in the regimen to which the patient’s virus was susceptible.
There is now little doubt that NNRTI-HS is more than just a number on a phenotype report. However, the clinical implications of these findings are still unclear. While some have argued that NNRTIs should be deferred in order to capitalize on the HS advantage, there are two problems with this approach. First, HS requires the presence of significant NRTI resistance, resistance that we should be trying to prevent by intervening early during failure of an initial regimen. Second, data presented in Barcelona from trials of EFV-based regimens in previously naïve patients [ACTG 384, Gilead 903, CLASS; see Sterling, When to Start ART] demonstrate that such regimens are simple, well tolerated, and highly effective for initial therapy, perhaps the gold standard against which other regimens will be judged . If the “first shot” is still the “best shot,” then it would probably be unwise to ignore the data on initial therapy in order to save these drugs for inclusion in subsequent regimens on the hope that HS will make them more potent than they already are.
Treatment of Experienced Patients
• Tenofovir in experienced patients
Although the big news on tenofovir came from studies in naïve patients (Gilead 903) [see Sterling, When to Start ART], follow-up data on the use of this drug in experienced patients were presented by Pozniak [Abstract WeOrB1266]. Gilead study 907 was a multicenter trial that enrolled 550 treatment-experienced patients who were randomized to continue current therapy or to add tenofovir DF (TDF) to their existing regimen. As noted in previous reports, addition of TDF was associated with a 0.6 log10 c/mL reduction in viral load, both initially and after 24 weeks, when those randomized to continue current therapy were allowed to add TDF as well. In both the 907 and the 902 studies, the best responses to TDF were seen in those without thymidine analog mutations (TAMs), especially if the M184V mutation was present [Margot NA, et al. Abstract ThOrB1390]. Those with multiple TAMs that included the M41L and/or L210W mutation had a poor response to TDF. Because L210W is rarely seen with K70R or T215F but almost always found with T215Y and M41L, it is virtually diagnostic for the unfavorable resistance pattern associated with decreased TDF susceptibility.
This study was important in helping to define the resistance pattern of TDF as well as clinical cut-offs for phenotypic suscepti-bility. In addition, it supported the use of TDF in experienced patients and the strategy of intensification of failing regimens under certain circumstances. However, the success of TDF as a component of initial therapy reported in the Gilead 903 study [Staszewski S, et al. Abstract LBOr17], may cause clinicians to use this drug earlier, rather than waiting to use it for salvage or intensification, when its potency is likely to be lower and less predictable.
• IDV/r vs SQV/r
The MaxCmin1 trial compared two boosted PI regimens, IDV/RTV (800/100 mg bid) and SQV/RTV (1000/100 mg bid) in naïve and experienced patients [Cahn P, et al. Abstract WeOrB1265]. At 48 weeks, virologic suppression was similar in the two arms by on treatment analysis and intent-to-treat (ITT), switch included analysis. However, by ITT switch=failure analysis, 68% of patients on SQV/r had viral loads <400 c/mL vs 53% on IDV/r. Treatment-related grade 3/4 adverse events were also significantly more common among IDV recipients, and lipid profiles were better in the SQV/r arm.
• T-20
The Barcelona conference was a debut of sorts for the investigational fusion inhibitor, T-20, as results of two nearly identical multicenter trials, TORO-1 and TORO-2, were presented in the late breaker session [Henry K, et al. Abstract LBOr19b; Clotet B, et al. Abstract LBOr19a]. The generic name for T-20 is “enfuvirtide,” and the trade name will be “Fuzeon.” (Under the circumstances, I think I’ll just call it “T-20.”) In the TORO trials, patients who had failed all three classes of antiretroviral agents were randomized in a 2:1 fashion to take optimized background therapy (OBT) plus T-20 (90 mg sq bid) or OBT alone. There were approximately 500 patients in each study. At 24 weeks, patients taking T-20 had significantly better virologic responses to therapy (-1.7 and -1.43 log10 c/mL in TORO-1 and TORO-2, respectively, vs -0.76 and -0.65 log10 c/mL, p<0.0001). Viral load suppression to <400 c/mL was observed in 37% and 28% of T-20 recipients vs 16% and 14% of those taking only OBT (p<0.0001). These results clearly indicate that T-20 provides potent virologic activity despite multi-class resistance. However, a new drug used in salvage therapy is typically only as good as the regimen with which it is combined, and T-20 will probably be no exception. T-20 resistance, mediated by mutations in gp41, has been described in patients who fail therapy with this drug. Patients who wait to use T-20 until after they have become resistant to all available antiretroviral agents are unlikely to have a durable response.
• Switch Studies
Martinez presented data from the NEFA trial, in which patients with virologic suppression on PI-containing regimens were randomized to switch from PIs to nevirapine, efavirenz, or abacavir (hence the acronym) [Abstract WeOrB1262]. Overall, there were no significant differences among the three arms with respect to the proportion of patients who maintained viral loads <200 c/mL at 12 months by ITT analysis. However, an as-treated analysis revealed that patients switching to abacavir were less likely to maintain virologic suppression, due mostly to the use of pre-HAART nucleoside analog therapy with presumed NRTI resistance. Patients switching to abacavir had the largest reduction in cholesterol levels, though most patients had normal levels prior to switch. There were no significant differences in fat accumulation or lipoatrophy.
This and other switch studies support the safety of switching from PI-based regimens to NNRTI-based or triple NRTI regimens, but consistently demonstrate the rather unsurprising finding that patients with prior NRTI resistance don’t do well on triple NRTI regimens.
In other switch studies, a switch from nelfinavir to atazanavir was associated with a significant decline in total cholesterol, LDL cholesterol, and triglycerides, as well as an increase in HDL cholesterol [Murphy R, et al. Abstract 9013]. Moyle presented data supporting previous studies [see HHR 2002;14(2):4] that suggested reversibility of lipoatrophy with NRTI switches [Abstract ThPeB7322]. In his study, patients switching from d4T to abacavir experienced a 36% increase in arm fat and a 17% increase in leg fat by DEXA. No increases in subcutaneous fat were observed in patients switching from PIs or EFV to ABC, or even in patients switching from d4T and a PI or EFV to abacavir and AZT.
In the enthusiasm that followed the introduction of protease inhibitors and that ushered in the HAART era, treatment guidelines were aggressive, recommending antiretroviral therapy for patients with CD4 counts <500 cells/mm3 or modest elevations in viral load. In 2001 these guidelines were modified, reflecting a growing reluctance on the part of clinicians and patients to commit patients to life-long therapy prematurely with the accompanying risks of side effects, long-term toxicity, and resistance. As a result of these changes, many patients are now on therapy who were started during the “hit early, hit hard era,” but who would not have been treated based on guidelines in use today. In Barcelona, we heard reports on two studies addressing the issue of treatment inter-ruption in such patients.
Krolewiecki from Buenos Aires presented data on patients treated with stable HAART for at least 6 months whose baseline CD4 counts were >350 cells/mm3 and whose maximum viral load was <60,000 c/mL [Abstract ThOrB1440]. Patients were randomized to continue or discontinue therapy. Patients who discontinued therapy appeared to do well. None developed HIV-related symptoms or AIDS-defining conditions, and the mean decrease in CD4 count was only 14 cells/mm3. Viral load after rebound was within 1 log10 c/mL of the pre-treatment setpoint.
This author presented data from an observational cohort of patients who discontinued therapy with the intention of restarting based on laboratory or clinical parameters [Gallant JE, et al. Abstract ThOrB1439]. At the time of the most recent analysis, 67% the 101 patients in the cohort remained off therapy, after a mean inter-ruption of 74 weeks. Mean CD4 count (most recent off therapy) was 508 cells/mm3. The remaining third resumed therapy after a mean interruption of 34 weeks. The best predictor of time off therapy was pre-treatment CD4 cell count. Using a baseline CD4 count of >500 cells/mm3 as a referent, those with a CD4 count below 200 cells/mm3 were 7 times more likely to resume therapy (p=0.001), and those with a baseline count of 200-350 cells/mm3 were 4 times more likely to resume therapy (p=0.015). There was no significant difference in time off therapy between those with baseline CD4 counts of 350-500 cells/mm3 and those with baseline counts over 500 cells/mm3. Patients who met current DHHS guidelines criteria at the time of treatment initiation were 2.9 times more likely to resume therapy than those who did not (p=0.004), a difference affected more strongly by the CD4 component of the guidelines than the viral load component.
Both studies provide support for an approach that many find intuitive: Patients who didn’t need therapy when they started probably still don’t need it now. There are also longer-term implications for strategies such as pulse therapy, where treatment is interrupted for prolonged periods of time, with the goal of therapy being to maintain the CD4 count above a pre-determined threshold. Clinical trials are in progress to evaluate this approach, though results may not be available for several years.
Another rationale for interruption of therapy is to allow re-emergence of wild-type virus and thereby improve response to salvage therapy. Katlama and colleagues presented data on their GigHAART trial (ANRS 097), a small open-label trial in which patients with advanced disease (CD4 <200 cells/mm3) who were failing therapy (viral load >50 c/mL) were randomized to switch immediately to a multi-drug “mega-HAART” regimen or to defer salvage therapy until after an 8-week treatment interruption [Abstract 263]. Patients who stopped therapy first were twice as likely to have a 1 log10 c/mL drop in viral load or to achieve viral load reduction to <400 c/mL at 24 weeks. This is still a controversial strategy, however, in part because results have been inconsistent, in part because of the potential danger of treatment interruption in patients with advanced disease, and in part because of the lack of durability data [Clotet B, et al. Abstract 264]. Archived resistant mutants are expected to reappear with reintroduction of selective pressure exerted by salvage therapy. Larger randomized controlled clinical trials are in progress to evaluate this approach.
The idea that a “structured treatment interruption” (STI) strategy might help to immunize infected patients against their own virus seems to have died due to lack of supportive data. However, this strategy may still be viable in patients treated during primary infection. Walker presented data on a select group of patients who were diagnosed and treated before Western blot seroconversion [Abstract 259]. Eight of 14 such patients maintained viral loads below 5000 c/mL after at least one STI. Controlled trials are being planned to assess the effectiveness of STI after early antiretroviral therapy.
Still another approach to intermittent therapy involves cycling therapy in order to decrease the cumulative time on treatment, and presumably toxicity as well. In a debate session on treatment interruption strategies, Dybul discussed his experience with the “structured intermittent therapy” (SIT) approach, in which patients with maximal virologic suppression are treated with a “7-days on/7-days off” schedule [Session 261]. Drug regimens have included d4T/3TC/IDV/r and ddI/3TC/EFV. To date there has been no evidence of virologic rebound or drug resistance, though some “blips” have been observed during the off-treatment periods in the IDV/r-containing regimen. It should be noted that when the SIT strategy was studied using longer cycles (2 months on, 1 month off), resistance emerged to 3TC and NNRTIs in patients taking regimens containing those agents [Perrin, et al. Antiviral Therapy 2002;7:S59]. While no recommendations have been made with respect to discontinuation of NNRTI-containing regimens, clinicians should be cautious given the long half-lives and low genetic barrier to resistance with this class of drugs. Temporary substitution of the NNRTI with another agent (e.g. PI or TDF) prior to interruption may be advisable, though this approach has not been studied.
Patients lose treatment options each time they fail an antiretroviral regimen, eventually running out of regimens that can be expected to provide durable virologic suppression. The best way to deal with treatment failure is to prevent it, by carefully preparing and educating patients prior to initiation of therapy, by using potent regimens that patients can take, and by reinforcing adherence at every visit. When failure does occur, it is now clear that allowing continued failure without intervention is a recipe for viral evolution and worsening drug resistance. Intensifi-cation is an option for patients experiencing low-level virologic rebound despite good adherence, as was indicated in the Gilead 907 study. For patients with higher viral loads, the standard of care is to perform resistance testing before selecting the next regimen.
There are a number of new drugs in the pipeline that may be effective in patients who have developed resistance to currently available agents. Some of these drugs have novel mechanisms of action, including T-20 and other agents that inhibit viral entry. Others inhibit reverse transcriptase or protease but have unique resistance profiles, such as tipranavir or some of the second generation NNRTIs. Initial therapy for untreated patients continues to improve, and our hope is that we’ll see begin to see fewer failures and less drug resistance as a result. However, we must continue to search for more effective options and strategies for those experienced patients who do fail.
20020901
JH20020902
©1997,1998,1999,2000,2001, 2002. The Johns Hopkins University AIDS Service, Division of Infectious Diseases. Permission to use and reproduce portions of this newsletter is hereby granted provided that author and publication are fully credited and both copyright and permission notice appear with reprinted material. Inquiries may be directed to Sharon McAvinue, Managing Editor. Website: Johns Hopkins AIDS Service.
AEGiS is made possible through unrestricted grants from Boehringer Ingelheim, iMetrikus, Inc., the National Library of Medicine, and donations from users like you. Always watch for outdated information. This article first appeared in 2002. This material is designed to support, not replace, the relationship that exists between you and your doctor.
AEGiS presents published material, reprinted with permission and neither endorses nor opposes any material. All information contained on this website, including information relating to health conditions, products, and treatments, is for informational purposes only. It is often presented in summary or aggregate form. It is not meant to be a substitute for the advice provided by your own physician or other medical professionals. Always discuss treatment options with a doctor who specializes in treating HIV.
Copyright ©1980, 2002. AEGiS. All materials appearing on AEGiS are protected by copyright as a collective work or compilation under U.S. copyright and other laws and are the property of AEGiS, or the party credited as the provider of the content.
