Treatment Strategies:

Salvage Therapy and PreventingFailure

The first part of this report discusses treatment options if you have failed protease inhibitor therapy. This discussion includes new protease inhibitors in development and studies about to begin which offer new regimens. But, it may be in your interest to delay making a treatment change until research can identify which regimens will be successful in suppressing protease resistant virus. It is also important to consider that additional options will become available in the relatively near future. However, if you want to make a treatment change now, some of the best options are reviewed below. The second part of the article discusses treatment strategies for limiting or preventing resistance, treatment failure, and cross-resistance.

The AIDS Clinical Trials Group (ACTG) is the branch of the National Institutes of Health that conducts AIDS studies. In July, the ACTG will begin two, new 48-week studies for individuals who have failed protease inhibitors.These two studies will explore new experimental treatments that will hopefully suppress protease resistant virus. The studies are new in the sense that we don’t know how effective they will be in suppressing resistant virus.

ACTG 398 is for individuals who have failed single or multiple protease inhibitors. The study offers individuals a 5-drug regimen consisting of 1592U89+ efavirenz+adefovir (PMEA) in addition to protease inhibitor therapy. The protease part of the regimen will consist of either 141W94 (amprenavir) alone or 141W94 in a double protease combination with nelfinavir, indinavir, or Fortovase (SGC saquinavir). Investigators plan to enroll 460 study participants.

ACTG 400 is for individuals who have failed only nelfinavir. Participants will be offered a 4 or 5 drug regimen consisting of efavirenz+2 NRTIs, plus protease inhibitor therapy. The protease therapy will consist of 1 or 2 drugs, either ritonavir+saquinavir, indinavir, 141W94, or indinavir+141W94. This study plans to enroll 300 participants.

Both of these studies are expected to begin in July. There are 32 ACTG sites throughout the US. Any of these sites can offer this study if the site wants to. A list of sites with contact information is on page 10. If you are interested, it may be important to contact the site quickly for information.

Adefovir, efavirenz, 1592U89, and 141W94 are new drugs that have not yet been approved by the FDA and therefore are not commercially available. Except for 141W94, they are available through an expanded access program sponsored by the pharmaceutical company developing the drug. The contact information for these expanded access programs is available on the front page of this newsletter. An expanded access program for 141W94 is targeting July for startup. Under consideration is a unique plan to offer various regimens incorporating 141W94.

The ICC (Inter Company Collaboration) is a coalition of the major HIV associated pharmaceutical companies. They are planning to begin a study in May or June, of novel treatment regimens for individuals who have failed protease inhibitor therapy. Originally, the study was supposed to begin in February, but it has been delayed. The study will be open label and will be available at 7 sites listed on page 10.

Nelfinavir Failure. If nelfinavir is the only protease inhibitor you’ve failed, a study conducted by Keith Henry and reported at the 5th Conference on Retroviruses and Opportunistic Infections suggests that a ritonavir+saquinavir regimen may work. In the study, 13/19 individuals who failed nelfinavir were undetectable (<500 copies/ml) after 6 months. Further details about the study are on page 20. ACTG 400, described above, is a study for individuals who’ve failed nelfinavir.

Using 6 or More Drugs. In NATAP Reports’ January ‘98 issue, we reported the findings of Dr. Cassy Workman from Australia. She explored the concept of recycling previously used NRTIs for 12 individuals. She used a 6-drug regimen for individuals who failed protease inhibitors. The 12 people had previously used all available nucleosides and three protease inhibitors including indinavir, ritonavir, and saquinavir. However, they had never used nevirapine or nelfinavir. The 6-drug regimen Workman used was d4T, 3TC, ddI, nevirapine, nelfinavir, and saquinavir. Three individuals were intolerant to components of the regimen that they had previously been intolerant to. However, 9/9 individuals remained on therapy and had undetectable viral load, <400 copies/ml, at week 12. Twelve weeks is too soon to draw conclusions, but follow-up information will be reported.

At the First International Salvage Therapy Workshop in New Orleans in April ‘98, Dr. Schlomo Staszewski of Goethe University in Frankfurt, Germany, discussed some of his experience in clinical practice using "Mega HAART" regimens for individuals who’ve failed therapy, including protease inhibitors. He used six, seven, and eight drug regimens that included one, two, or three protease inhibitors and recycled NRTIs. He showed data on several individuals who responded with "dramatic" and "durable" six month viral load reductions to undetectable. However, he said the side effects of using so many drugs could make the regimens difficult to tolerate. He suggested using a maintenance regimen consisting of fewer drugs after viral load was reduced to undetectable by the "Mega HAART" regimen. However, one of the individuals rebounded to detectable after switching to a maintenance regimen. He switched the person back to the Mega HAART regimen and they went back down to undetectable. Maintenance regimens are still being studied.

It has been suggested that phenotypic with or without genotypic resistance testing might be useful in sorting out which drugs a person might be resistant to. The drugs that resistance developed to could be eliminated from a Mega HAART or salvage regimen. But, some doctors and researchers are not convinced the results from resistance testing are reliable in all instances.

New Protease Inhibitors. In addition to 141W94, two other new protease inhibitors are in development. All three hold a measure of promise that they may be active against protease resistant virus because their resistance profile is or may be different. At this point it is premature to know if they will be effective, but it is likely they will have to be used as part of a potent multi-drug combination.

141W94 is furthest along in development of the three new PIs, but it is uncertain if the drug will be effective against protease resistant virus. So far, studies conducted indicate that if protease resistant virus will be sensitive to 141, the drug wil have to be used as part of a potent multi-drug combination. By the time the FDA reviews 141 for accelerated approval, which is expected in the latter part of ‘98, additional resistance and cross-resistance data will be available. At the 5th Conference on Retroviruses and Opportunistic Infections, preliminary 16-week data was reported on double protease combinations containing 141. ACTG studies 398 and 400 will be the first to explore using these combinations in individuals who’ve failed protease inhibitors. The preliminary pharmacokinetic data for the combination of 141+indinavir shows that indinavir increases 141’s drug concentration in the blood, thereby possibly making it a potent therapy. This combination is used in both ACTG 398 and 400. Results from pharmacokinetics studies for the combinations of 141 with saquinavir, ritonavir or nelfinavir are not yet available. It is expected that Glaxo Wellcome will offer an expanded access program for 141 during July ‘98.

Pharmacia’s and Upjohn PNU-140690 and Abbott’s ABT-378 are both protease inhibitors about to enter pilot studies to see if they can suppress protease resistant virus. Upjohn has reportedly just started recruiting for a small study evaluating the effectiveness of PNU-140690 in individuals who’ve failed protease therapy. In one arm, only indinavir or ritonavir failures will receive PNU-140690. A second arm is for individuals who have experience with multiple protease inhibitors. Participants in the second study will receive PNU-140690 if they’ve failed their current protease therapy and had previous protease therapy. The objective is to make a preliminary observation to see if individuals who’ve failed protease inhibitors will be sensitive to PNU-140690. If all goes well, the hope is that by this summer, a larger study will begin exploring PNU-140690’s effectiveness for individuals who have failed a PI.

Abbott’s second generation PI, ABT-378, is currently in phase II studies in HIV infected individuals. ABT-378’s resistance profile has been reported to lack the 82 mutation. This mutation is important for ritonavir and indinavir resistance. In vitro, ABT-378 was significantly more potent than ritonavir. Based on its different resistance profile and its potential added potency, it is hoped that this drug will suppress protease resistant virus. It is expected that ABT-378 may be able to suppress virus up to 20 fold resistant to ritonavir. This may have implications for suppressing virus resistant to other protease inhibitors. ABT-378 will be combined with a small amount of ritonavir (50 or 100 mg) because ritonavir greatly enhances the blood levels of ABT-378. This enhancement allows ABT-378 to be dosed twice daily. Once a day dosing is being considered by Abbott, but appears unlikely. Studies so far have shown that ABT-378 will likely be able to be taken with or without food. Various dosing regimens are now being explored and a pilot study is about to begin exploring its use in individuals who’ve failed protease therapy.

PD178390. PD178390 is a protease inhibitor which is still in an early stage of development as it enters advanced stage preclinical studies, but so far it has some appealing characteristics. The developer, Parke Davis, reported at the Retrovirus Conference that the drug has displayed an encouraging resistance profile in lab studies. It was active in lab experiments (in vitro) against single and double mutants known to contribute to ritonavir, saquinavir and indinavir resistance. And it had little P450 3A4 inhibition. The drug also had 7 hours above the EC95 in dogs. Although encouraging, high expectations should be tempered because obstacles that prevent development do often emerge during these early stages.

PMPA. Results from a preliminary 30-day study reported at the Retrovirus Conference showed that an oral version of PMPA dosed once a day reduced viral load 1.22 log in a small group (n=28) of HIV infected individuals. This study used a lower dose than the 15-day study conducted last summer in which PMPA was administered intravenously; the higher dose showed more antiviral potency. Gilead has said that as soon as safety is realized, higher doses will be explored. PMPA is a nucleotide and is not expected to be cross resistant to protease inhibitors, making it a potentially useful drug for individuals who’ve failed protease therapy. Gilead Sciences is planning studies to explore its role in treating protease resistant virus.

In the study presented at Retrovirus, baseline CD4 was 375 and viral load was 31,600 copies per ml. Dr. Stephen Deeks reported that PMPA had potent activity in SIV-infected macaques and has a prolonged intracellular half-life of 15-30 hours. Deeks reported that bioavailibility of PMPA was 41% without food (reduced to 27% with food).

PMPA appears to have a preferable resistance profile. Investigators conducted a resistance mutation analysis at baseline and day 35 for the individuals taking 150 and 300 mg/day. They reported that no patient developed a mutation. Five patients had the 3TC M184V mutation at baseline. The virologic response was similar to those without M184V. It was reported that HIV-1 strains and clinical isolates resistant to AZT, ddI, 3TC and adefovir (PMEA) remained susceptible to PMPA. Similar to adefovir, clinical isolates with the 3TC associated M184V mutation had several fold increased susceptibility to PMPA.

Investigators reported the following grade 3 or 4 adverse events as possibly or probably associated with PMPA, of 28 total study participants: 5 (18%) had increased CK, 3 (11%) had increased AST/ALT (liver function tests), and 1 (4%) had peripheral neuropathy.

T-20: A Fusion Inhibitor.
T-20’s mechanism of action is different than that of other current therapy approaches. It binds to the HIV-1 particle and prevents it from fusing and entering the CD4 cell. Fusion and entry inject the components necessary for viral reproduction into the CD4 cell. If the process can be prevented, the genetic material (RNA) for the virus cannot enter the CD4 cell and the virus cannot reproduce itself. This mechanism, if successful, can prevent new infection. Antiretroviral therapy works differently. After viral entry into a CD4 cell, reverse transcriptase inhibitors and protease inhibitors work to inhibit the reproduction process, but the components for viral reproduction are in the cell.

At the September 1997 IDSA Conference, data from a small preliminary phase 1/2 study on 16 HIV infected individuals receiving various doses of T-20 were presented. This was the first data reported on T-20 in HIV infected individuals. There were only four persons in each dose group. Participants were both treatment naive and experienced (9/16 were treatment experienced). Their baseline mean CD4 count was 319 and baseline mean viral load was 5.10 log (about 125,000 copies/ml). Participants were required to stop taking all antiretroviral therapy 15 days prior to starting T-20. Participants received a single dose of T-20 by bolus intravenous infusion, followed by a 2-day washout period and then 14 consecutive days of T-20 twice daily, every 12 hrs. No drug-related adverse events and no dose limiting toxicities were reported during the treatment period. As you can see from the data below, there was a dose dependent decrease in viral load and a dose dependent increase in CD4.

At the highest does tested in this study, 100 mg every 12 hours, viral load was reduced 1.50 log at day 14. The CD4 increase was 52. All 4 participants receiving the 100 mg dose were <500 copies/ml (undetectable) using bDNA, but 2/4 were undetectable using the Roche Amplicor test.

Trimeris is the company developing T-20. Currently, it is unlikely they can develop an oral administration for T-20 due its size and molecular composition. Alternatively, they want to develop use of a subcutaneous infusion pump delivery system. They are working with a company called MiniMed who has marketed a similar system used by diabetics for delivering insulin. The company believes a continuous infusion would permit a steady flow of T-20 without peaks (Cmax) or troughs (Cmin) in blood levels.

The company is about to start the first study to evaluate this pump system. The study would be an attempt at salvage therapy for 40 individuals who have failed protease inhibitor therapy. Initially, participants would receive T-20 for 10 days to evaluate if they were achieving the same response as seen in the diabetic study. After the 10-day period, a multi-drug antiretroviral regimen including efavirenz+nelfinavir+saquinavir would be added.

If the subcutaneous infusion system is effective, additional studies are planned. Company officials believe resistance would be less of a concern than with current antiretroviral therapies because the site where T-20 binds is well conserved and continuous blood levels of drug distributed by the pump system would eliminate drug level troughs, thereby inhibiting the development of mutations and consequent resistance. Initially, their goal is to test T-20 as a salvage therapy. They believe the pump system would alleviate some of the compliance issues associated with taking current treatment regimens. The company is concerned that individuals may be reluctant to use such a delivery system, so they will be surveying a cross-section of HIV infected individuals in major cities to evaluate people’s feeling about this. NATAP will report information as it develops. The first study is expected to start during June ‘98 at three sites.

Hydroxyurea (HU). HU is a drug used for various cancers and has received much attention lately. To have antiviral activity against HIV, HU must be combined with a specific type of NRTI. A number of studies have shown that HU can be an effective treatment when combined with ddI, ddI+d4T, and ddI+a protease inhibitor. Previous studies showed it was not effective when combined with AZT. At the Retrovirus Conference, Dr. Franco Lori reported potentially significant results from his study of HU+ddI+a protease inhibitor. The results are reported inside this issue. Future studies exploring HU in combination with other NRTIs including d4T alone, 1592U89, FddA, and adefovir are expected.

HU has some unique properties including its ability to increase the effectiveness of ddI. Data suggests that combining HU with ddI may overcome ddI resistance. Bristol Myers has stated that resistance to HU has not been observed in its 35 years of use. There are some potential adverse effects of HU. In a study conducted by R. Rossero and R. Pollard at University of Texas-Galveston, neutropenia (lowering of WBC count) to below 700/uL developed in patients who had pretreatment absolute neutrophil counts below 1700/uL, but it was reversible upon withdrawal of HU. Of 31 NRTI experienced individuals receiving HU+ddI+d4T with a CD4 count about 230, four individuals developed progressive peripheral neuropathy or elevation in pancreatic enzymes.

HU blocks cell activation. As a result, HU may be effective in treatment during early disease because it may limit virus targets and viral production. Due to HU’s mechanism of action, CD4 increases may be slight or non- existent when used with only ddI or ddI+d4T. However, when HU was combined with ddI and a protease inhibitor in the Franco Lori study, CD4’s increased appreciably.

After failing protease inhibitor(s) you may want to consider using HU in a salvage therapy. HU in a potent multi-drug combination that could include a double protease combination and 1592U89, NNRTIs, and/or adefovir is one option. However, using HU in this way is experimental. We do not yet have data from studies using HU with 141W94, ritonavir, nelfinavir, adefovir, 1592U89 or with double protease combinations. Another option for salvage therapy is to stop the current failing regimen and use a regimen of HU+ddI or HU+ddI+d4T. Hopefully this regimen would hold viral load stable until enough new drugs are available to put together an effective multi-drug regimen. Since resistance to HU doesn’t seem to develop, and since resistance to ddI may not be a concern, it is possible that a reduction in viral load using HU+ddI+d4T could remain stable for a prolonged period, even if viral load is detectable.

Treatment Strategies to Prevent Failure

T-Cells (CD4s) Can Remain Elevated After Viral Load Rebound on Protease Therapy. When should a person switch their regimen? There is not an easy answer. A few leading researchers think it may be acceptable to maintain viral load at a low but detectable level of, for example, 5,000 copies/ml. However, viral load can increase rather than stay stable. Furthermore, most experts agree that the goal of therapy should be lowering viral load to undetectable. At the January ‘97 4th Retrovirus Conference, Dr. Dale Kempf, of Abbott Labs, reported findings from ritonavir studies suggesting that the lower the nadir (the lowest level reached) of viral load, the greater the durability of viral load suppression. Since then, data from several studies have suggested that lowering viral load to <50 copies/ml may be crucial to durable suppression of viral load. Adequate and durable suppression of viral load in other virus compartments besides the blood (CSF, brain, lymph tissue, genital secretions, etc) may be or is likely associated with sustaining viral suppression in the blood to <50 copies/ml. See Reducing Viral load to <50 copies/ml on page 5.

Recently, there has been a report that despite viral load rebound on protease therapy, CD4s can remain elevated. At the 5th Retrovirus Conference, Dr. Stephen Deeks reported data from a retrospective study of 79 patients in his clinic. Deeks reported that despite a median duration of nine months since developing evidence of virologic failure, the median CD4 count remained 103 cells above the level prior to starting protease therapy. Virologic failure is defined as the two most recent viral load assays >500 copies/ml using bDNA, both measured after 20 weeks of therapy. However over time CD4 decline should be expected if viral load remains elevated. An elevated viral load implies ongoing viral replication, which means added resistance and mutations.

At the First International Salvage Therapy Workshop in New Orleans April 16-18, Dr. Robert Schooley offered an explanation for the elevated CD4s despite virologic failure. Viral replication may be low enough in these individuals not to lose CD4s. It is suggested that a new equilibrium or set point has been established between the virus and the immune system. It’s also been suggested that the fitness, or ability, of the virus to replicate has been reduced by the presence of certain protease inhibitor mutations.

There probably is benefit to maintaining a CD4 increase to a certain level despite viral load rebound. Does this mean that there is a benefit to remaining on a therapy after viral load rebound? Read on.

Monthly Monitoring of Viral Load. Many leading researchers think it is potentially more beneficial to detect viral load rebound immediately and switch therapy immediately thereafter. Only two studies have started to address this approach, but additional data should be collected over time because a number of ACTG studies will be incorporating it into their study designs. ACTG 347 switched 141W94 monotherapy failures quickly to a 4-drug regimen including indinavir+nevirapine. The preliminary data is encouraging; about 85% are undetectable at about 4 months. Inside this issue is a report on the study. Dr. Joel Gallant reported data from a small study at Retrovirus showing that individuals responded better to a ritonavir+saquinavir regimen following failure of a protease inhibitor when participants had a lower viral load at the time of switching. These results suggest that detecting viral load rebound immediately and switching right away breeds better responses to therapy. Again, this study is reviewed inside this issue in greater detail.

Monitoring your viral load closely by taking a viral load test every month may be the best way to detect a rebound in viral load. If you detect a rebound you probably want to confirm it with a second test. If you can detect the rebound quickly you can then switch your protease therapy immediately. It is possible that such an action might limit or prevent protease inhibitor cross-resistance.

The rationale for this approach is based on several accepted ideas. A rebound in viral load means virus replication is increasing. Mutations to the drug you’re taking and consequent resistance to that drug results from virus replication while taking a drug. If you discontinue your protease inhibitor as soon as viral load rebound occurs, you may limit the opportunity for the virus to develop mutations. Initial mutations might develop, but secondary mutations may not. Theoretically, the potential for cross-resistance may be limited or prevented. Although there is only a small amount of data supporting this approach, acceptance is growing among researchers that this approach may offer the best opportunity to respond to a subsequent protease therapy after failing the current one.

Reducing Viral Load to <50 copies. As stated above, there is increasing evidence that reducing viral load to <50 copies/ml will yield greater durability of viral load suppression. In efavirenz studies (see report inside), participants with undetectable viral load by a sensitive measure are more likely to remain suppressed during the course of the studies, extending out to as much as about 12 months. Also, participants in nevirapine study 1046 did not remain under 200 copies/ml unless they had <20 copies/ml. These data are from NNRTI 3-drug combination studies. Except for the French study reported below, similar data have not yet been reported from protease studies. But, at the International AIDS Conference in July ‘98, it is expected that Agouron will report data from their 511 study of nelfinavir+AZT+3TC on the durability of antiviral effect when a person goes below 50 copies/ml as opposed to not going below 50 copies/ml.

At the Retrovirus Conference, a French research group reported their study findings, concluding those individuals with <20 copies/ml had greater durability of viral load suppression than those with between 20 and 200 copies/ml. The French research group studied 349 individuals who started a three drug protease inhibitor containing regimen between April ‘96 and October ‘96. They were followed for a median of 15 months. Of the group, 53% were on an indinavir regimen and 47% were on a different protease regimen. The median baseline CD4 and viral load were 97 cells and 100,000 copies/ml. Also, 92% were treatment experienced while 8% were naive. At the third month, 37% (n=129) had <200 copies/ml (Roche Amplicor assay), and 10.3% (n=36) had <20 copies/ml using the ultrasensitive RT-PCR test. In this study, viral escape among virological responders was defined as having a HIV RNA <200 copies/ml, followed by a viral load that becomes >1000 copies/ml on two consecutive tests, despite continued therapy. After a median follow-up period of 15 months, viral escape occurred after three months in 7.1% (2/28) of those who had a viral load of <20 copies/ml and in 34.1% (27/79) of those who had viral load >20 copies/ml.

The researchers found baseline HIV RNA was associated with HIV RNA reaching below 20 copies/ml after three months (median 15 months follow-up). Of study participants with baseline HIV RNA <100,000 copies/ml, 14.3% (25/175) had viral load <20 copies/ml after 3 months and 6.4% (11/174) of individuals having >100,000 copies/ml had viral load <20 copies/ml after 3 months (p<0.05). When they analyzed baseline CD4 stratified to <97 cells or >97 cells, they did not find any correlation with durability of viral load response.

Is Rapidity of Viral Load Decline Associated with Durability?

A research group from Amsterdam reported preliminary data from a study they initiated. They feel that lowering viral load to as low as possible as quickly as possible might be associated with better antiviral activity. They compared 3, 4, and 5-drug protease containing regimens in a partially retrospective analysis and found that the 5-drug regimen caused a more rapid decline in viral load. It remains uncertain if more rapid decline in viral load is associated with better clinical effect in the long term, so the investigators will continue to follow the patients for durability of viral load response and long term clinical effect. It is suggested that the more rapid viral load declines or the steeper the slope of decline the more quickly replication may be stopped or inhibited. Possibly, the more quickly viral load declines the more likely a person may be to reach <50 copies/ml.

Rapid decline of viral load can be influenced by previous antiretroviral drug experience, the potency of a regimen, and baseline viral load. The lower a person’s baseline viral load when they begin a new regimen, the more likely they will reach <200 or <50 copies/ml.

Participants’ median baseline CD4 cell counts were 190, 370 and 360 cells in the 3, 4, and 5 drug regimens, respectively. Median baseline viral loads were about 100,000 for all three groups. Patients were selected from three drug trials:

• Three drug regimen (NUCB2019): 15 treatment naive participants who started with AZT/3TC/ritonavir. The data from these patients was drawn retrospectively because NUCB2019 was conducted prior to this clearance study

• Four drug regimen (ADAM study): 35 treatment naive individuals who started with d4T/3TC/nelfinavir/ saquinavir. This study is being conducted by the Netherlands group

• Five drug regimen (ERA study): 8 treatment naive and 1 experienced individuals who started with AZT+3TC+ 1592U89+ indinavir+nevirapine

Results. Table 1 contains the plasma/serum HIV RNA values at the sampling time points for all three groups. The values in the table represent the amount of virus found. For example, in the 5-drug group at baseline, the median HIV RNA was 4.9 log. After 8 weeks of therapy the median HIV RNA was 1.7 log, meaning viral load was reduced by 3.2 log. But, the viral load decline during the first week of therapy for all three groups was greatest (-1.8 log) in the 5 drug group. See Table 1

Safety. Only 2/10 patients receiving the 5 drug regimen had no medication-related subjective or objective complaints. In 3 patients, severe allergic reactions necessitated discontinuation of 1592 nevirapine; one of these patients left the study. Nausea was a frequent complaint and has led to some further adaptations of study medications. The authors concluded that individuals taking the 5 drug regimen experienced more side effects than those taking the triple regimen.

Based on the HIV RNA levels at day 0, 7, and 14 the mean slopes of virus decay were greater for the 5-drug regimen than for both the three and four drug regimens. There was no difference between the three and four drug regimens. The virus half-life was shorter for the five drug regimen than the four or three drug regimen. There was no difference between the virus half-life in the three or four drug regimen.






Table 1. Median log HIV RNA