Protease Inhibitor Resistance and Salvage Therapy
What to Do When a Patient's Initial Antiretroviral Regimen Begins to Fail
When used as part of a three-drug regimen, protease inhibitor therapy typically results in durable and potent suppression of viral replication. This has led to widespread optimism that HIV infection may eventually become a chronic, manageable disease. However, recent data indicate that a significant proportion of patients do not achieve durable viral suppression with these therapies. Prior therapy, advanced immunodeficiency, and non-adherence are probably the most common determinants of virologic failure, which in some retrospective studies has affected as many as half of all patients.
If viral replication occurs in the presence of a protease inhibitor, drug-specific resistance emerges. Since the protease inhibitors are structurally similar, viral mutations that confer resistance to one agent are likely to confer some degree of cross-resistance to other agents in this class (Table 1). For example, resistance to indinavir invariably results in high-level cross-resistance to ritonavir.
Prospective clinical trials addressing sequential, or salvage, protease inhibitor therapy have not yet been reported. Clinicians must therefore rely on their own experience and on the limited data that have come from the handful of small, often retrospective, studies that have been conducted to date. Several such studies were recently presented in abstract form at three international meetings: the 6th International Workshop on HIV Drug Resistance, Treatment Strategies and Eradication; the 35th Annual Meeting of the Infectious Disease Society of America (IDSA); and the 37th Annual Meeting of the Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC). The results of these studies are summarized below. Readers who want a more detailed analysis of the data will find it at the "HIV InSite" web page (see box).
Salvage therapy after indinavir failure
Several abstracts presented at this year's ICAAC meeting evaluated the activity of ritonavir/saquinavir after the failure of initial indinavir therapy. A group from Spain, led by Dr. Lidia Ruiz, performed a prospective study of ritonavir/saquinavir-containing regimens in 12 patients whose viral load rebounded on indinavir. A median 1 log decline in viral load was seen through week 12 of the study. The durability of this response is unclear.(1)
Similar results were seen in a retrospective analysis of 19 patients who failed indinavir therapy at San Francisco General Hospital. All of these patients had clear evidence of viral rebound while on indinavir-containing regimens, and 11 of the 19 patients modified their nucleoside analogs at the time of the switch to ritonavir/saquinavir. After four weeks of follow-up, a median 1.6-log decline in viral load was observed. Unfortunately, this effect was transient: at week 24, viral load had returned to within 0.5 log of baseline in the majority of patients.(2)
Comment: The strategy of switching to a ritonavir/saquinavir regimen after indinavir failure appears to result in a potent but transient effect on viral load. Considering the cross-resistance among these agents, the lack of a durable response is not surprising. More aggressive regimens, containing experimental therapies, will likely be necessary in order to achieve complete viral suppression in patients no longer responding to indinavir.
Ritonavir/saquinavir after nelfinavir failure
In clinical trials, nelfinavir selects for the D30N mutation in the protease gene. In vitro, virions containing the D30N mutation retain full sensitivity to other protease inhibitors. To some investigators, this observation has suggested that nelfinavir should be a first-line protease inhibitor, with other protease inhibitors reserved for salvage therapy. This hypothesis has not been tested in clinical trials.
At this year's ICAAC meeting Dr. Keith Henry and his colleagues presented their experience with 19 University of Minnesota patients who failed nelfinavir during controlled clinical studies. In a prospective manner, 19 patients with moderately advanced disease and clear virologic evidence of nelfinavir failure were switched to a regimen containing ritonavir, saquinavir, d4T, and 3TC. The majority of these patients had the D30N mutation prior to the switch; the L90M mutation, commonly seen with saquinavir therapy, was found in a few patients. After 16 weeks of follow-up, 9 of 14 patients had an undetectable viral load.(3)
Ritonavir/saquinavir or indinavir after nelfinavir failure
At an ICAAC late-breaker session, retrospective data were presented on 12 patients in the New York area who failed nelfinavir therapy. Patients were switched by their primary care providers to regimens containing either ritonavir/saquinavir (n=6) or indinavir (n=6). Nucleoside analogs were modified whenever possible. After a limited follow-up (median: 8 weeks), a 0.46-log decline in viral load was observed (0.55-log decline in patients on ritonavir/saquinavir; 0.21-log decline in patients on indinavir). Only 3 of the 12 patients achieved undetectable levels of viral load at week 12.(4)
Comment: These two data sets reach inconsistent conclusions. Although the data from the University of Minnesota suggest that salvage therapy after nelfinavir failure may be possible in some patients, the New York experience indicates that failing nelfinavir therapy confers significant cross-resistance to other protease inhibitors. Large prospective studies are needed in order to definitively address the open question of whether protease inhibitors can be used sequentially.
Salvage therapy after saquinavir failure
Saquinavir (Invirase®) has limited activity against HIV, largely due to the drug's very low oral bioavailability. To address this limitation, a soft-gel formulation (Fortovase®) has recently been developed (see "The Next Generation of Antiretroviral Agents -- An Update," in this issue). In clinical trials, both formulations select for the G48V and L90M mutation in the protease gene. In vitro data have suggested that these mutations may not confer cross-resistance to other protease inhibitors, including indinavir.
At this year's IDSA meeting, Dr. Michael Para presented data on behalf of the ACTG 333 team. This multicenter study was designed to investigate the activity of saquinavir soft-gel capsules (saquinavir-SGC) in patients currently receiving saquinavir hard-gel capsules (saquinavir-HGC), and to determine if patients failing saquinavir therapy can successfully switch to indinavir.
In an open-label fashion, 72 subjects on a stable regimen containing saquinavir-HGC were randomized to receive either continued saquinavir-HGC, saquinavir-SGC, or indinavir. Background nucleoside analog therapy was not modified at the time of the switch. After eight weeks of follow-up, subjects switching from the hard-gel to soft-gel formulation had a limited, 0.23-log decline in viral load. In subjects switching to indinavir, a modest 0.58-log decline in viral load was seen. These effects appeared to persist over 24 weeks of treatment.(5)
The in vitro evidence notwithstanding, data presented earlier this year at the International Workshop on HIV Resistance, Treatment Strategies, and Eradication suggest that saquinavir resistance may indeed confer cross-resistance to indinavir. In a study conducted in France, 22 patients were treated with saquinavir and then switched to indinavir. Ten of the 22 patients had a durable virologic response to indinavir, while 12 patients eventually had a rising viral load on the drug. Detailed genotypic and phenotypic studies were performed on this latter group of patients.
Prior to the switch to indinavir, 5 of the 12 patients had a mutation at position L90M (common with saquinavir therapy). All five patients maintained the L90M mutation in the presence of indinavir; none developed the classic indinavir-associated mutation at codon V82A. Six patients had no evidence of genotypic resistance to saquinavir at the time of the switch to indinavir. Surprisingly, after switching to indinavir, typical saquinavir-related mutations (L90M or G48V) emerged, whereas typical indinavir-related mutations, such as V82A, rarely developed.(6) This observation suggests that L90M or G48V were present at the time of the switch, but at levels too low to be detected by the genotypic assay, and that these mutations were rapidly selected for by indinavir.
Comment: In ACTG 333, the lack of a potent response to indinavir after prolonged therapy with saquinavir may be due to cross-resistance or to a durable response to saquinavir at the time of the switch (thus making it theoretically difficult to detect additional declines in viral load). The genotypic resistance data presented by Dulioust and colleagues suggest that viral resistance to saquinavir confers cross-resistance to indinavir in vivo, despite the lack of significant cross-resistance in vitro. Therefore, therapy with two nucleoside analogs and indinavir may not be optimal in patients who have failed saquinavir-containing regimens. Other regimens -- including indinavir plus a non-nucleoside reverse transcriptase inhibitor -- should be considered.
Although prospective, randomized, controlled clinical trials have not yet been reported, failure of an initial protease inhibitor-containing therapy appears to compromise future therapeutic options. After resistance to indinavir or ritonavir emerges, salvage therapy with protease inhibitor-containfing regimens is difficult. Preliminary data indicate that salvage therapy after saquinavir or nelfinavir failure may be possible in some patients; however, this matter requires further investigation.
Possible therapeutic strategies for patients failing an initial regimen are presented in Table 2. Similar strategies were recently proposed by the U.S. Department of Health and Human Services.(7) Based on the preliminary clinical data presented here, clinicians should assume that patients have only one good chance at achieving durable viral suppression with protease inhibitor therapy. The most effective protease inhibitor-containing regimen should therefore be used first.
For Further Information
1. Puig T, Bonjoch A, Ruiz L, Arno A, Sirera G, Romeu J, Clotet B. Usefulness of ritonavir and saquinavir combination therapy for HIV-advanced patients failing on indinavir. 37th Interscience Conference on Antimicrobial Agents and Chemotherapy, Toronto, Canada, 1997. Abstract I-201.
2. Deeks S, Grant RM, Horton C, Simmonds N, Follansbee S, Eastman S. Virologic effect on ritonavir (RTV) plus saquinavir (SQV) in subjects who have failed indinavir (IDV). 37th Interscience Conference on Antimicrobial Agents and Chemotherapy, Toronto, Canada, 1997. Abstract I-205.
3. Henry K, Kane E, Melrome H, Simpson J, Patick A, Winslow D. Experience with a ritonavir/saquinavir based regimen for the treatment of HIV-infection in subjects developing increased viral loads while receiving nelfinavir. 37th Interscience Conference on Antimicrobial Agents and Chemotherapy, Toronto, Canada, 1997. Abstract I-204.
4. Sampson MS, Barr MR, Torres RA, Hall G. Viral load changes in nelfinavir treated patients switched to a second protease inhibitor after loss of viral suppression. 37th Interscience Conference on Antimicrobial Agents and Chemotherapy, Toronto, Canada, 1997. Abstract LB-5.
5. Para MF, Collier A, Coombs R, et al. for the ACTG 333 Study Team. ACTG 333: Antiviral effects of switching from saquinavir hard capsule (SQVhc) to saquinavir soft gelatin capsule (SQVsgc) vs. switching to indinavir (IDV) after prior saquinavir. 35th Annual Meeting of the Infectious Diseases Society of America, San Francisco, September 13-16, 1997. Abstract 21.
6. Dulioust A, Paulous S, Guillemot L, Boue F, Galanaud P, Clavel F. Selection of saquinavir-resistant mutants following a switch from saquinavir. International Workshop on HIV Resistance, Treatment Strategies, and Eradication, St. Petersburg, FL, June 25-28, 1997. Abstract 16.
7. U.S. Department of Health and Human Services, Panel on Clinical Practices for Treatment of HIV Infection. Guidelines for the use of antiretroviral agents in HIV-1 infected adults and adolescents, November 5, 1997.
Steven G. Deeks, M.D., is Assistant Clinical Professor of Medicine, UCSF AIDS Program, San Francisco General Hospital.