1. Home>
2. Specialties>
3. HIV/AIDS>
4. Feature

Salvage Therapy

The current recommended therapy for HIV infection is two nucleoside analog reverse transcriptase inhibitors (NRTIs) in combination with a protease inhibitor (PI) (or two) or a non-nucleoside reverse transcriptase inhibitor (NNRTI).¹ These recommendations are based on data demonstrating that a majority of patients on clinical trials will achieve HIV suppression with these potent regimens. However, even in the most successful of these clinical trials, at least 10% to 20% of patients will not achieve durable suppression of viral load. Several observational clinical cohort studies from Europe and the U.S. have described higher rates of virologic failure (20% to 60%) after a year of treatment with various first-line regimens. Consistent independent predictors of virologic failure in these studies were: prior antiretroviral treatment, higher baseline viral load, lower baseline CD4-cell count, and the specific treatment regimen selected. The growing number of patients experiencing virologic failure on antiretroviral therapy must be addressed.

Guidelines for Salvage Therapy

An expert panel convened by the U.S. Department of Health and Human Services developed guidelines for "Possible Regimens for Patients Who Have Failed Antiretroviral Therapy: A Work In Progress." For a patient who has experienced virologic failure on a first-line regimen of two NRTIs and a PI, the guidelines recommend changing to two "new" NRTIs as well as one of the following: (1) another PI, (2) a combination of other PIs, or (3) an NNRTI and another PI (Table 1). A footnote stipulates that "These alternative regimens have not been proven to be clinically effective and were arrived at through discussion ... of theoretically possible alternative treatments and the elimination of those alternatives with evidence of being ineffective. Clinical trials in this area are urgently needed."

A recent descriptive report gives an indication of the clinical application of these salvage therapy guidelines. Deeks and colleagues² performed a retrospective review of medical records of HIV-positive patients at San Francisco General Hospital. They identified 99 patients who experienced virological failure (HIV RNA >500 copies/ml after 16 weeks of therapy) with their initial PI-based regimen using indinavir (N=73), ritonavir (N=23), or nelfinavir (N=3). All patients switched to a second regimen containing a PI that they had not previously taken; 27 patients began a regimen that also contained an NNRTI. After 24 weeks of therapy on the second-line regimen, only 22 patients (22%) achieved HIV RNA <500 copies/ml. Independent predictors of failure included baseline viral load greater than 10,000 copies/ml at the time of switch and failure to initiate a regimen containing an NNRTI. This low virologic response rate clearly indicates that better approaches to salvage therapy are necessary.

Saquinavir to Indinavir

One of the first studies to prospectively evaluate a specific salvage therapy regimen was ACTG 333.³ In this study, patients who had taken at least 48 weeks of saquinavir hard gelatin capsule (hgc) and had not changed their antiretroviral regimens for two months prior to entry were randomized to (1) continue saquinavir hgc, (2) change to saquinavir soft gelatin capsule (sgc), or (3) change to indinavir. At baseline, patients had taken a median of 112 weeks of saquinavir therapy, had a median viral load of 20,900 copies/ml and a CD4 cell count of 222/mm³. The results of an interim analysis showed that at week eight of therapy, the saquinavir hgc group had no change in viral load, the saquinavir sgc group had a minimal change in viral load of - 0.2 log₁₀ copies/ml, and the indinavir group had a blunted virologic response, with a reduction of - 0.6 log₁₀ copies/ml from baseline levels. This study was the first to suggest that the length of prior protease inhibitor therapy may affect subsequent treatment response.

PI to Nelfinavir

With a growing population of patients failing on the first available PIs -- saquinavir hgc, ritonavir, and indinavir -- researchers investigated nelfinavir as a second-line PI. Data from observational cohort studies presented at the 12th World AIDS Congress in Geneva and the 38th ICAAC conference suggested that only 30% to 36% of PI-experienced patients had virologic suppression on regimens containing nelfinavir.

A prospective study of the use of nelfinavir in indinavir-experienced subjects was undertaken in ACTG 372b.⁴ In this study, 94 patients who took AZT or d4T in combination with 3TC and indinavir on ACTG 320 and had viral loads of at least 500 copies/ml were randomized to one of four treatment arms containing efavirenz + adefovir + either: (1) abacavir + nelfinavir; (2) abacavir; (3) 1 or 2 NRTIs + nelfinavir; or (4) 1 or 2 NRTIs. Overall, 29 of 84 subjects (35%) had viral loads <500 copies/ml by week 16. In a factorial analysis, there was a significant difference in virologic suppression between pooled nelfinavir groups (45%) versus groups who did not receive nelfinavir (24%), but no difference between the use of abacavir or other NRTIs. This study showed the benefit of using a PI as part of a second-line regimen.

PI to Dual PIs

In the ABT-462 study, a regimen of concurrent saquinavir and ritonavir resulted in durable virologic suppression in antiretroviral-naive patients. Soon after the report of these results, researchers explored the use of dual PIs as salvage therapy. Three early descriptive reports noted a 22% to 46% rate of virologic suppression in small numbers of PI-experienced patients who received saquinavir and ritonavir. Tebas and colleagues⁵ studied 26 patients with documented virologic failure on a regimen containing nelfinavir who changed to d4T + 3TC + saquinavir + ritonavir. A total of 17 patients (65%) had viral loads <500 copies/ml at week 24, and 14 (54%) continued with virologic suppression over a median of 61 weeks of follow-up.

Similarly, the demonstrated activity of the combination of saquinavir and nelfinavir in antiretroviral-naive patients in the SPICE study subsequently led to its testing as part of a salvage regimen in PI-experienced patients. In abstracts presented at the 12th World AIDS Congress and the 37th ICAAC conference, virologic suppression rates from 8% to 45% were reported on saquinavir/nelfinavir-based regimens from observational cohort studies.

The first prospective, randomized study to compare both of these dual-PI approaches was ACTG 359.⁶ In this study, 277 indinavir-experienced, NNRTI-naive subjects were randomized to receive saquinavir sgc/ritonavir or saquinavir sgc/nelfinavir, together with delavirdine, adefovir dipivoxil, or both. Overall, 77 of 254 (30%) subjects had viral loads 500 copies/ml at week 16. In a factorial analysis, there was no difference between using saquinavir/ritonavir or saquinavir/nelfinavir. However, there was a significant difference favoring delavirdine groups (40% with viral loads 500 copies/ml) over adefovir dipivoxil groups (18% with viral loads 500 copies/ml). Interestingly, there was no difference in virologic response between delavirdine and delavirdine/adefovir dipivoxil combination groups, possibly due to a pharmacokinetic interaction between the drugs. The length of prior indinavir therapy was a significant predictor of the virologic response at week 16.

Newer Drugs

The CNAA2007 study⁷ tested an open-label single-arm study of abacavir, efavirenz, and amprenavir as salvage therapy in 90 patients who had had extensive prior antiretroviral therapy. Of these patients, 72% had taken four to five NRTIs, 44% had taken an NNRTI, and 60% had taken three to four PIs. Overall, 26% of patients had viral loads <400 copies/ml at week 16 of the study. In a subgroup analysis, NNRTI-naive patients with baseline viral loads <40,000 copies/ml had the highest rate of virologic suppression on this study. A significant pharmacokinetic interaction -- coadministering efavirenz and amprenavir lowers levels of amprenavir 39% -- may explain the virologic responses in part.

In ACTG 373,⁸ 54 patients with amprenavir experience were changed to a regimen containing d4T, 3TC, nevirapine, and indinavir. Thirty patients entered the study with viral loads <500 copies/ml, most of whom were taking amprenavir monotherapy. On study, over 80% of subjects had viral loads suppressed below 500 copies/ml by week 12, continuing through week 60 in an on-treatment analysis. Complete follow-up data is forthcoming.

ACTG 398 is an ongoing study of 481 patients with at least four months of prior antiretroviral therapy, allowing up to three PIs, and prior NNRTI-therapy with viral loads 1,000 copies/ml. Patients were randomized to receive open-label abacavir, adefovir dipivoxil, efavirenz, and amprenavir together with either (1) saquinavir sgc, (2) indinavir, (3) nelfinavir, or (4) matching placebo. The study is fully enrolled and results are anticipated.

Investigational Drugs

ABT-378 (lopinavir), an investigational drug, is a PI designed with antiretroviral resistance in mind. Using computer modeling, the compound was shown to have little interaction with the protease enzyme amino acid residue at codon 82, a common site of substitution in patients who have failed ritonavir or indinavir therapy. In vitro, ABT-378 demonstrated potent antiviral activity against some mutant strains of HIV. In addition, when combined with low doses of ritonavir, the pharmacokinetic characteristics of ABT-378 permit twice-daily dosing and achieve substantial increases in drug concentrations to more than 30-fold above the 50% effective concentration for wild type virus. A study in antiretroviral-naive patients demonstrated significant, sustained antiretroviral activity of ABT-378 in a combination regimen.

Benson and colleagues⁹ reported the results from the ABT-765 study, the first study to explore the antiretroviral activity of ABT-378 in PI-experienced subjects. Eligible patients had taken one PI and two NRTIs for at least three months, had never taken more than one PI or any NNRTI, and had viral loads of 1,000 to 100,000 copies/ml on current therapy. At entry, 31 (44%) patients were taking indinavir, 25 (36%) nelfinavir, 9 (13%) saquinavir, 4 (6%) ritonavir, and 1 (1%) amprenavir. Researchers enrolled 70 patients and substituted ABT-378/ritonavir for their prior PI at the beginning of the study, continuing for two weeks. At day 15, patients also substituted two NRTIs (one of which was new for the patient) and added the NNRTI, nevirapine. Over the first two weeks on ABT-378/ritonavir, median viral loads decreased by more than 1 log₁₀ from baseline levels. In an intent-to-treat analysis, 54 of 70 (77%) had viral loads <400 copies/ml at 24 weeks on the new study regimen. Despite the selected patient population, the results appear encouraging. Pentafuside (formerly known as T-20) is an investigational peptide analog of the gp41 protein of HIV. It prevents binding of HIV to the CD4 cell, and is thus the first of a new class of agents: fusion inhibitors. In a study in antiretroviral-naive patients, pentafuside administered parenterally demonstrated antiretroviral activity over 14 days, greatest at the highest dose tested. As an antiretroviral agent with a novel mechanism of action, it was anticipated that this drug would have activity in antiretroviral-experienced patients. The TRI-003 study¹⁰ was a phase II, dose-ranging trial of pentafuside in a patient population with significant antiretroviral experience. At baseline, of the 78 study patients, 99% had taken antiretrovirals, 98% had taken PIs, 41% had taken all four approved PIs, and 46% had taken all three approved classes of HIV drugs. Enrolled patients added pentafuside to their current regimen (if any) at different doses, given either by continuous subcutaneous infusion pump or twice-daily subcutaneous injections for 28 days. The greatest virologic suppression occurred in the pentafuside 100-mg subcutaneous twice-daily group, with a mean 1.2 log₁₀ decrease from baseline viral load at day seven, followed by a rebound back toward baseline levels. The investigators noted the significant, though transient, antiretroviral activity of pentafuside in this highly treatment-experienced patient group.

Additional studies of investigational agents are planned or in progress. Other investigational PIs entering clinical development are tipranavir, L-756,423, BMS 232632, AG 1776, and DMP 450. Other compounds with novel mechanisms of action in clinical development include: PMPA, a nucleotide reverse transcriptase inhibitor (NtRTI), and T-1249, a second-generation fusion inhibitor. Other potential approaches, such as HIV integrase inhibitors, are being explored.

Multidrug Regimens

The use of multiple drugs in patients experiencing virologic failure taking antiretroviral regimens has been termed "multidrug rescue" or "mega-HAART" therapy. Investigators from Germany, Canada, and the United Kingdom have presented their preliminary results using this approach. In these observational studies, regimens with up to nine individual antiretroviral agents, including NRTIs, NNRTIs, PIs, and hydroxyurea, were given with variable rates of virologic suppression. Significant numbers of patients required changes in therapy or discontinued therapy because of adverse events. Miller and colleagues¹¹ tested the multidrug approach and noted that 19 of 24 patients who underwent a drug holiday prior to starting treatment and had a documented shift in the viral population to wild type HIV RNA suppressed to <500 copies/ml. Of concern were increases in viral load and decreases in CD4-cell counts during the drug holiday. Further research will document the risks and benefits of the multidrug approach with or without drug holidays.

Resistance Testing

The use of resistance testing in the selection of antiretroviral regimens, both initial and salvage, remains controversial. Two recent studies have begun to address this issue in salvage therapy: the VIRADAPT study¹² was a prospective, randomized study of 108 patients who had experienced virologic failure on a regimen containing an NRTI and a PI. Patients were randomly assigned to either standard care (N=43) or treatment selected according to genotype (N=65, Visible Genetics assay). At three months of follow-up, the mean change in viral load was -0.46 log₁₀ in the control group and -1.04 log₁₀ in the genotype group (P=0.01); at six months changes were -0.67 log₁₀ and -1.15 log₁₀, respectively (P=0.05). The percentage of patients with viral loads <200 copies/ml was 14% (control group) and 32% (genotype group) at six months (P=0.048). The authors concluded that genotypic testing added significant benefit in choosing a therapeutic alternative, resulting in an improved virological response. A recent update of this study suggested a persistent benefit at 12 months of follow-up in the genotype group.

The GART study (CPCRA study 046)¹³ was a randomized pilot study in patients experiencing virologic rebound after taking at least 16 weeks of PI-containing therapy. A total of 153 patients were randomized to receive either standard care or genotypic testing with expert interpretation based on clinical history. At weeks four and eight, change in viral load was: -0.61 log₁₀ (control group) and -1.19 log₁₀ (genotype group) (P<0.001). At week 12, the difference between treatment groups remained 0.44 log₁₀. The authors concluded that genotypic testing with expert interpretation was a superior strategy in patients experiencing virologic failure on three-drug antiretroviral therapy. Further studies will assess the long-term benefits of both genotypic and phenotypic testing in the selection of salvage-therapy regimens.

Conclusions

Despite the benefits of current antiretroviral therapy, virologic failure is common. Predictors of virologic failure include prior antiretroviral experience, baseline viral load, baseline CD4-cell count, and the specific regimen used. Current guidelines for salvage therapy are based on expert opinion. Retrospective and prospective studies of current salvage therapy approaches have documented short-term virologic response rates of, at best, 20% to 54%. Newer agents may hold more promise, including drugs in existing classes designed with resistance in mind and drugs with new mechanisms of action. Other strategic approaches, including the use of multidrug regimens with or without drug holidays and the use of genotypic and phenotypic resistance testing in guiding the selection of therapy, are under evaluation. Progress in salvage therapy will stem from further basic and clinical research.

— Roy M. Gulick, MD, MPH

Dr. Gulick is Assistant Professor of Medicine and Director of the Cornell Clinical Trials Unit at Weill Medical College of Cornell University.

Published in AIDS Clinical Care September 1, 1999

Citation(s):

1. Panel on Clinical Practices for Treatment of HIV Infection. Guidelines for the use of antiretroviral agents in HIV-infected adults and adolescents. Ann Intern Med 1998 128 1079-1100.

• Original article (Subscription may be required)
• Medline abstract (Free)

2. Deeks SG et al. HIV RNA and CD4 cell count response to protease inhibitor therapy in an urban AIDS clinic: response to both initial and salvage therapy. AIDS 1999 13 35-43.

3. Para MF et al. abstr. 21. Infectious Disease Society of America 35th Annual Meeting, San Francisco, CA, September 13-16, 1997 .

4. Hammer S et al. abstr. 490. 6th Conference on Retroviruses and Opportunistic Infections, Chicago, IL, January 31-February 4, 1999 .

5. Tebas P et al. Virologic responses to a ritonavir-saquinavir-containing regimen in patients who had previously failed nelfinavir. AIDS 1999 13 23-28.

• Medline abstract (Free)

6. Gulick RM et al. abstr. 020. 2nd International Workshop on Salvage Therapy for HIV Infection, Toronto, Canada, May 19-21, 1999 .

7. Eron J et al. AIDS 1998 12 abstr. OP5.2. S9-.

8. Murphy RL et al. AIDS 1998 12 abstr. OP2.4. S9-.

9. Benson C et al. abstr. 7. 3rd International Workshop on HIV Drug Resistance & Treatment Strategies, San Diego, CA, June 23-26, 1999 .

10. Lalezari J et al. abstr. LB13. 6th Conference on Retroviruses and Opportunistic Infections, Chicago, IL, January 31-February 4, 1999 .

11. Miller V et al. abstr. 030. 2nd International Workshop on Salvage Therapy for HIV Infection, Toronto, Canada, May 19-21, 1999 .

12. Durant J et al. Drug-resistance genotyping in HIV-1 therapy: the VIRADAPT randomised controlled trial. Lancet 1999 353 2195-2199.

• Medline abstract (Free)

13. Baxter JD et al. abstr. 61. 3rd International Workshop on HIV Drug Resistance & Treatment Strategies, San Diego, CA, June 23-26, 1999 .