Joel E. Gallant, M.D., M.P.H.
The Hopkins HIV Report - March 1999
As always, the most exciting presentations in Chicago involved the treatment of naïve patients. Nevertheless, a number of important studies addressed the therapy of treatment-experienced individuals. The expanding list of antiretroviral agents and the growing availability of resistance assays have made treatment of experienced patients both more complex and more effective. Some of the studies presented in Chicago may have a significant impact on the way we manage this challenging group of patients.
Treatment of Nucleoside-Experienced Patients
More on nucleoside sequencing: The controversy surrounding sequencing of AZT and d4T continues. Data were presented by Back [Phiboonbanakit, et al., Abstract 487] suggesting that, in contrast to previously presented data, there is no difference in the phosphorylation of d4T between AZT-naïve and AZT-experienced patients. These conflicting in vitro studies now join the several equally conflicting clinical studies, leaving us with a poor understanding of whether there is an appropriate sequence with which to use these two important thymidine analogs. Phosphorylation is not the only proposed mechanism for the possible decrease in d4T efficacy following AZT use, however. A minority of patients taking AZT will develop one of several multi-drug resistant genotypes, and there is evidence that some AZT-resistance mutations carry increased viral fitness [HHR, vol. 10, no. 6; Nov. 1998; p.1]. However, it is not yet known how these findings should be translated into clinical practice.
Nelfinavir (NFV), efavirenz (EFV), or both for NRTI-experienced patients: ACTG 364 was a "rollover" study for patients who were heavily NRTI experienced as a result of prior participation in ACTG 175 and its first rollover trials, ACTG 302 and 303 [Albrecht et al., Abstract 489]. Participants who had viral loads >500 c/ml at entry and were naïve to non-nucleoside reverse transcriptase inhibitors (NNRTIs) and protease inhibitors (PIs) were randomized to receive NFV, EFV, or NFV/EFV plus open-label nucleoside reverse transcriptase inhibitors (NRTIs). At baseline, mean CD4 count was 390 cells/mm3, and median HIV RNA was 7778 c/ml. By weeks 40-48, 74% of those taking NFV/EFV had HIV RNA <500 c/ml compared with 60% of those taking EFV and 35% of those taking NFV (p=0.001, p=0.004 for NFV vs. EFV, p=0.001 for NFV vs. EFV/NFV). In light of the extensive NRTI experience in this group, the superiority of the NFV/EFV is not surprising. However, the explanation for the marked superiority of EFV over NFV is less apparent. Further analysis of the data is necessary to determine whether there was a difference between groups with respect to the number of new NRTIs taken by the participants.
Abacavir in NRTI-experienced patients: ACTG 368 was another rollover trial involving treatment of NRTI-experience patients [Squires, et al., Abstract LB-15]. Participants had received either AZT/3TC or d4T/3TC as part of the control group for ACTG 320, and were randomized to receive indinavir (IDV)/EFV, with either abacavir (ABC) or placebo. In addition, they were randomized to receive IDV at a dose of 1200 mg q12h or 800 mg q8h. The study confirmed the findings of a previous Merck-sponsored study demonstrating the superiority of q8h dosing over q12h dosing of indinavir. In addition, 78% of abacavir recipients vs. 73% of placebo recipients had viral loads below 500 c/ml at week 16, a non-significant difference. This study, along with ACTG 372B (discussed below), emphasizes that abacavir did not improve outcomes in these patients with extensive NRTI experience.
Treatment of Protease Inhibitor-Experienced Patients
Ritonavir/saquinavir salvage therapy after nelfinavir failure: Tebas and Herny reported 60-week follow-up data from their trial of ritonavir (RTV)/saquinavir (SQV) plus two NRTIs in patients who had failed a NFV-containing regimen [Abstract 392]. A durable antiviral response has been demonstrated, with 14 (58%) of 24 patients maintaining HIV RNA <500 c/ml at 60 weeks. As noted in other trials, the viral load at the time of switch was a strong predictor of long-term virologic response. Ten of 12 patients with HIV RNA <30,000 c/ml maintained undetectable viral loads for 48 weeks compared to only 4 of 12 with HIV RNA >30,000 c/ml at switch.
The role of abacavir and nelfinavir after indinavir failure: Scott Hammer presented data from ACTG 372B, a rollover trial from ACTG 320 in which patients with heavy NRTI experience who developed viral rebound on an IDV-containing regimen were randomized to receive efavirenz and adefovir (ADV) with abacavir (ABC) vs. 2 NRTIs and nelfinavir (NFV) vs. placebo [Abstract 490]. In this study, the use of NFV was associated with less viral rebound (57% vs. 79% for NFV placebo, p=0.0016). However, there was no benefit associated with the use of ABC (rebound in 67% vs. 70% for 2 NRTIs, p=0.81). As with ACTG 368, discussed above, these findings again emphasize the lack of utility of abacavir in patients with extensive NRTI exposure. The benefit associated with nelfinavir is somewhat surprising, given that nelfinavir is not generally considered to be a useful "salvage" agent after failure of other PIs. However, as discussed below, it is possible that many of these patients, especially those with relatively low viral loads at the time of switch, had no protease inhibitor resistance at all. This would also help to explain another important finding from this study. Only 43% of those who switched therapy with a viral load less than 15,000 c/ml experienced viral rebound, compared to 81% of those who switched above 15,000 c/ml (p<0.001).
"Intensification" is a widely used term, but there is little agreement about its definition. For some it means the addition of one or more drugs to a new regimen which is successfully suppressing viral replication but which has not brought the viral load below detectable limits. For others it means the addition of one or more drugs at the time of first viral rebound. A third definition, by far the most dubious, is the addition of one or more drugs to any regimen that is no longer working. Arguments in favor of the first strategy include the fact that the use of more drugs is associated with a more rapid decline in viral load, which is in turn associated with a more durable antiviral response. Arguments in favor of the second strategy include the data discussed below suggesting that early rebound may occur despite continued susceptibility to many of the drugs in the regimen. There is little support for the third strategy, however, which is merely the same old sin, the addition of a single drug to a failing regimen, with a new name.
This latter approach may be safe in selected circumstances, however. The addition of hydroxyurea, for example, may sometimes reduce viral load without increasing the risk of drug resistance, since it is not an antiretroviral agent. Another strategy presented in Chicago involved the addition of abacavir to a failing regimen either of AZT/3TC [Rozenbaum, et al., Abstract 377] or of any standard antiretroviral regimen [Vernazza, et al., Abstract 378]. In both trials, the addition of abacavir was associated with significant and sustained reductions in viral load, independent of genotypic 3TC resistance. Such an approach may be relatively safe in some circumstances, since continued use of AZT/3TC or other NRTIs may ultimately preclude the use of abacavir anyway. However, this is probably not the optimal way to use this potent NRTI.
Use of Resistance Testing to Guide Therapy
CPCRA 046 was a randomized study involving patients with evidence of viral load rebound (3-fold increase in viral load) after at least 16 weeks on a 3-drug HAART regimen [Baxter, et al., Abstract LB8]. A total of 153 patients were randomized to one of two groups: in group I (n=78) the patient's clinician received a genotype report together with treatment recommendations from an expert; in group II (n=75) the clinician chose the next regimen based on the patient's treatment history. Clinicians of group I patients were allowed to ignore the expert report if they chose. At the time of switch, genotype reports demonstrated at least one major RT mutation and at least one major protease mutation in 75% of patients; 20% had at least one major RT mutation without protease mutations; and 5% had no major mutations. Although viral load declines were transient, they were more pronounced in group I (-1.17 logs) compared to group II (-0.62 logs, p=0.0001). Antiviral response was associated with the number of active drugs (drugs to which there was no prior resistance) used in the salvage regimen. Those treated with 0-1 active drugs achieved only a 0.1 log drop in viral load, compared to a 1.25 log drop in those treated with at least four active drugs. Less than half of those in group II received at least three active drugs, compared with 86% in group I. Interestingly, viral load declines and differences between groups were significantly higher at participating sites that tended to "adhere" to expert recommendations. Among sites with >80% adherence, viral load declines in group I were - 1.47 logs (vs. group II, -0.47 logs p=.0005) compared with sites with <60% adherence (group I -0.98 logs vs. group II, -0.74 logs).
Because of its design, the conclusions that can be drawn from CPCRA 046 are somewhat limited. It is not possible to sort out the relative contributions of the genotype analysis and the experts' recommendations. It is possible that the results would have been similar had the recommendations been based entirely on an expert review of the treatment history. Nevertheless, this study provides support for the use of resistance assays in designing salvage regimens and emphasizes the importance of the addition of multiple active drugs and of care by experts for HIV-infected individuals.
Resistance in Failing Regimens
In our issue devoted to coverage of last year's ICAAC [HHR, vol. 10, no. 6; Nov. 1998; p.1], I discussed the surprising new finding that many patients failing HAART therapy may have relatively little resistance to the failing drug regimen. A number of studies were presented in Chicago that support these findings. The most important data came from ACTG 343 [Havlir, et al., Abstract LB12]. This was one of three trials testing the induction-maintenance strategy, each of which demonstrated that "de-intensification" of a HAART regimen leads to virologic failure. In ACTG 343, patients who achieved viral suppression on a regimen of AZT/3TC/IDV were randomized to continue triple-drug therapy or to switch to either AZT/3TC or IDV monotherapy. As has been reported previously, 23% of those taking IDV or AZT/3TC developed viral rebound compared to only 4% of those taking AZT/3TC/IDV. Resistance testing was performed on nine isolates from patients failing IDV monotherapy and 17 isolates from patients failing triple therapy. None of the patients with viral rebound on IDV had resistance to any protease inhibitor. Duration of failure ranged from 1 to 3 months, and although viral loads tended to be low, they ranged from 1,000 to over 100,000 c/ml. Of the 17 patients failing triple-therapy, none had phenotypic resistance to protease inhibitors, and only one had a significant protease mutation by genotype analysis (M46L). However, 14 (82%) had the M184V mutation and phenotypic 3TC resistance. These findings cannot be explained by cessation of indinavir therapy, as most patients had detectable IDV levels.
In a presentation by Holder from Merck Research Labs [Abstract 492], IDV resistance was present in only 21% of patients failing AZT/3TC/IDV in Merck 054 and in 22% of those failing IDV/EFV in Merck 067. 3TC resistance and EFV resistance were found in 74% and 92% of patients, respectively. Similarly, in the Trilege trial, another induction-maintenance trial, no AZT or IDV resistance mutations were observed in 58 patients failing maintenance therapy with AZT/3TC/IDV, AZT/3TC, or AZT/IDV [Descamps, et al., Abstract 493]. The M184V mutation associated with 3TC resistance was seen in most patients, including those on AZT/3TC who were not failing therapy.
A number of hypotheses have been proposed to explain these surprising findings, none of which is entirely satisfactory. Nevertheless, these findings suggest that the practice of changing all the drugs in a failing regimen may not be necessary and that resistance testing could be used during early failure to identify which drugs should be changed and which should be continued. This would be welcome news, given the paucity of antiretroviral agents that are unique with respect to resistance. However, we should interpret these findings with caution, since currently available resistance assays are known to be insensitive to resistant mutants present in small quantities. It should be noted that while these findings are now consistent across three studies, we still don't know whether or not the strategy of selectively altering a failing regimen based on resistance data will be effective.
There's Failure and then There's Failure
Complete and prolonged suppression of HIV replication remains the primary goal of antiretroviral therapy. It is clear that this goal is readily achievable in adherent patients for whom appropriate therapy is chosen. With careful selection of treatment regimens, this goal may also be achievable using second and even third regimens. However, patients who have failed multiple antiretroviral combinations will eventually run out of options, and the goal of undetectable virus will become unrealistic. Fortunately, there is growing evidence for a "disconnect" between virologic failure and immunologic or clinical failure. Deeks presented results from an observational study from San Francisco General Hospital demonstrating that although virologic failure blunts further increases in CD4 count, the CD4 count remains elevated long after the onset of viral rebound [Abstract 494]. The mechanism for this "disconnect" is not entirely clear, though it may have to do with the decreased viral fitness associated with some resistance mutations.
Coming away from Chicago, a number of themes stand out with respect to treatment-experienced patients:
Clearly, the best news at this important conference came from trials in naïve patients. Ultimately, the best way to deal with the problem of drug resistance will be to prevent it from happening in the first place. We can do that by providing highly effective, well-tolerated, and convenient drug regimens to patients only after they have been counseled about the critical importance of strict adherence. In addition, we must continue to approach antiretroviral therapy strategically, always aiming for success but at the same time planning for the possibility of failure.
990301
JH990302
©1999. 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.
The original of this article can be found at http://www.hopkins-aids.edu/publications/report/mar99_2.html
This information is designed to support, not replace, the relationship that exists between you and your doctor.