AEGiS-GMHC: Protease Inhibitors: Overview and Analysis

Important note: Information in this article was accurate in 1994. The state of the art may have changed since the publication date.

Protease Inhibitors: Overview and Analysis

Gay Men's Health Crisis: Treatment Issues, Volume 8 no. 2 - April, 1994
Dave Gilden

This article was compiled by Dave Gilden with the assistance of Ben Cheng, Rick Loftus and others. Much of the information contained below was obtained through private conversations with company sources who did not want to be identified.

Protease inhibitors are the one class of new, promising anti- HIV therapies likely to be available in the near future. It is hoped that by attacking HIV from a different direction, the new compounds will represent a great step forward in treating AIDS.

Dr. Margaret Johnston, one of the chief AIDS researchers at the National Institutes of Health, recently told the Wall Street Journal, "I really have my fingers crossed because after the protease inhibitors, I don't see much in drug development."

At least a dozen companies are now developing protease inhibitors. In preliminary lab experiments, some of their products have exhibited great potency against HIV. Yet Johnston would do well to keep her fingers crossed. Protease inhibitors have been beset by several serious practical problems. Now that larger trials are beginning, Treatment Issues presents a review of this fresh class of drugs and the strategies of the companies that are developing them.

What are Protease Inhibitors?

Protease inhibitors act at a late stage in HIV's replication process, after the copy of HIV's genes within a cell have initiated mass production of new viral components. The proteins within HIV's outer envelope first come out of the cell's protein factory in a long connected string that must be cut up. HIV protease is an enzyme that acts like a chemical scissors. It snips out the parts for the virus core so that they reassemble in the proper configuration. This snipping process occurs as newly formed virus particles bud out of the cell. Protease inhibitors lock onto and sheathe the protease's active site (the blade of the scissors, as it were). In doing so, they disrupt the creation of mature virus particles

HIV protease is quite small and as such makes an ideal target to stop HIV replication. It is a member of the aspartyl- protease enzyme family[1] that also includes the major human enzymes renin (which regulates kidney activity) and pepsin (which digests protein in the stomach). The human enzymes have a somewhat different, asymmetric structure, though, and should not be affected by a molecule aimed at the HIV protease.

Early studies confirmed that the HIV protease enzyme is essential for viral infectivity and replication[2] and that the virus particles produced without protease are faulty and cannot infect other immune system cells.[3,4]

As was hoped, protease inhibitors effectively prevented the production of infectious virus in chronically infected cells.[5] Current anti-HIV drugs such as AZT act against another HIV enzyme, reverse transcriptase, which helps the virus insert its genes into newly infected cells. They do not influence viral production in the chronically infected cells, in which HIV has already taken up.

Difficulties in Developing Protease Inhibitors

As usual there is a great leap between theory and practice. Researchers encountered many early obstacles to developing protease inhibitors.

First, a given compound may have been a potent inhibitor of protease in chemical experiments, but it had poor or no anti- HIV effects in cell cultures.[6] Secondly, poor oral absorption or rapid breaking down of the compound in the liver made it necessary to administer the protease inhibitor intravenously or ingest large amounts of it orally. Thirdly, unexpected toxicities, particularly in the liver where the compounds are metabolized, appeared in animal and human safety studies. Lastly, most of the protease inhibitors are very hard to manufacture, a problem that exacerbates the difficulties in maintaining adequate amounts in the body.

It is still unclear whether there will be any drug interactions between protease inhibitors and other medications and whether this will affect drug absorption and action. As most drugs are metabolized in the liver, such interactions are likely. Further studies are needed to understand how to integrate protease inhibitors into overall therapy for HIV disease.

Viral Resistance

The rapid emergence of viral resistance to AZT and other reverse transcriptase inhibitors (ddI, ddC, d4T), along with their toxicities, limits their long-term usefulness in the treatment of HIV. Protease inhibitors may be less prone to this failure. It is thought that HIV is less able to alter the structure of the protease enzyme than that of reverse transcriptase. Resistance to protease inhibitors may be slower to develop and at a lower level than for reverse transcriptase inhibitors.

But resistance has already been observed for some protease inhibitors.[7] Combining a protease inhibitor with a reverse transcriptase inhibitor may delay development of resistance. In test tube experiments, when a protease inhibitor was combined with AZT or ddC (or alpha interferon) additive or synergistic anti-HIV interactions were seen.[8] It is also hoped that the protease inhibitors will have antiviral activity against HIV strains that are resistant to AZT.

Summary of the Status of Individual Protease Inhibitors

Most major issues surrounding protease inhibitors boil down to one main question: Do economically affordable treatment regimens produce adequate blood levels of the drug? At the point when blood levels of the drug are at their lowest point (this is the trough level), concentrations must be well above those needed to stop nearly all viral replication. At that same time, the level of drug in the blood must be well below toxic concentrations.

(To put this in technical jargon: Trough levels of the drug should be multiples of the IC90 level, and the therapeutic index should be high. IC90 is the drug concentration that shuts down 90 percent of HIV activity. The therapeutic index is the ratio between a toxic dose of a drug and an effective one.)

If effectiveness can be ensured without compromising safety, the chance of drug-resistant HIV strains or cumulative toxicity emerging is diminished.

Drug developers are gradually overcoming some of the obstacles to an effective protease inhibitor. More potent and more orally available substances are slowly moving into clinical trials. In the following sections, we describe the status of the most advanced protease inhibitors and how well their industrial sponsors have coped with the problems inherent to this new therapeutic strategy.

Hoffmann-La Roche

Hoffmann-La Roche's "saquinavir" (formerly Ro 31-8959) was the first anti-HIV protease inhibitor tried in humans. The Phase I safety and pharmacokinetic studies were done in England, France and Italy.

Roche's molecule was one of the original series of compounds that were very potent against HIV but not very bioavailable when taken by mouth. Absorption of saquinavir is only 4 percent. Saquinavir also turned out to be extraordinarily difficult to manufacture. The seventeen steps involved take nearly a year to complete.

Results from the European studies show that this drug is very well tolerated with almost no side effects. At the highest dose given (600mg three times a day), saquinavir showed some antiviral activity (there were increases in CD4 cell counts, a drop in viral load and a decrease in HIV p24 protein).[9] However, the maximum response seemed to occur four to six weeks after starting therapy. In the French study, CD4 cell counts were essentially back to baseline by week sixteen. Better antiviral activity was observed when the Roche drug was combined with AZT. Then, CD4 cell increases were higher and sustained past sixteen weeks, the cutoff point for the trial analysis.[10]

Current Studies:

Another combination trial is being conducted by the NIH's AIDS Clinical Trial Group (ACTG 229). The trial compares saquinavir plus AZT versus saquinavir plus AZT plus ddC versus AZT plus ddC alone. Additionally, a trial of saquinavir versus AZT versus the two combined is in progress at Mt. Zion Hospital in San Francisco. Data from these trials are expected very soon.

Planned Studies:

Two multicenter pivotal phase III trials should begin very shortly:

1. A US 40-center study of 1200 HIV-positive volunteers with CD4 counts between 50 and 300 and prior, unsatisfactory experience with AZT. To be eligible, volunteers also must not have been on ddC, ddI or d4T for more than two weeks. There are four arms to the trial: Saquinavir at 600mg three times a day; ddC monotherapy; and two combination arms (ddC plus either 200 or 600mg saquinavir three times a day). Treatment will continue for 48 weeks. The primary endpoint will be time to first AIDS-defining clinical event or death. Immunologic and virologic lab values will be secondary endpoints. This trial is already underway at a few sites. For more information call 800/526-6367.

2. An international study of 1800 HIV-positive volunteers (600 in the US, 1200 abroad) without prior experience with AZT. Treatment will continue for 18 months. This trial is similar to the US trial in design except that AZT will be used instead of ddC. Roche plans to start the international study in March.

Analysis:

The Roche phase III trials has raised a number of criticisms. The first of these concerns uncertainty about the optimum dose of saquinavir. The impact previously seen on lab values (CD4 count, etc.) is rather minor and temporary. Moreover, a suboptimum dose of the drug may increase the likelihood of resistance developing. Given saquinavir's lack of toxicity so far, dosages could be increased to attain maximum benefit. Why didn't Roche perform a small multiple dose-escalation trial before the initiation of these large multinational trials?

Dr. Thomas Merigan at Stanford University is now embarking on such a dose-escalation study. Doses two to four times the amount Roche is administering in its phase III trials will be examined. Dr. Merigan describes the trial as a quick, four month look at sixteen volunteers (CD4 count range: 200-500) to see how larger amounts of saquinavir affect virus levels and the rise of drug-resistant strains. (Drug resistance is known to have occurred in one Italian patient and may be the reason for the drug's very transient effect in the French single agent study). The results could push Roche to develop a second generation, more advanced protease inhibitor, or at least a version of saquinavir that is more absorbable, a process that will take two to four years to complete.

Roche would be hard-pressed to expand production quickly to supply everyone in a large trial, let alone all the HIV- positive people in the world, with the higher doses Dr. Merigan is trying out. But the main reason for not moving on to larger doses with the present formulation appears to be the drug's prohibitive manufacturing cost. The price of treatment with the high dose regimen would be more than the market could bear, Dr. Merigan contends.

Suppose, though, that Merigan finds that saquinavir produces significant benefit at the higher doses while Roche's large phase III trials detect just a mediocre benefit. Will the FDA approve the drug at the lower (1800mg/day) dose? There is the risk that a two-tiered standard of care will arise, in which only the richest receive the really effective high dose.

Large numbers of people receiving too small a dose is likely to spread resistant HIV strains far and wide, eventually destroying saquinavir's usefulness. An analogous phenomenon is occurring with tuberculosis.

It could be, too, that the large phase III trials will not really be large enough or long enough to document benefit if that benefit turns out to be slight. While Roche and the FDA sort out the issues, there at present are no plans to institute an expanded access program for the drug. The problem once again seems to be the difficult, expensive production process. For now, you can only obtain Roche's protease inhibitor by entering a clinical trial - that is, you risk getting AZT or the still more problematic ddC, either alone or in combination with only a modest amount of saquinavir.

The development of these two trials occurred without a great deal of interaction with community representatives. This could well be due to the history of conflicts between Roche and AIDS activists. It is time to put these issues behind. Roche is simply too big a player in AIDS drug development to be allowed to develop a critical AIDS therapy without continued dialogue with community representatives. Broad ranging discussions are in everyone's interest.

Merck

Dr. Roy Vagelos, Merck's chairman, has targeted the development of a useful anti-HIV agent as both a corporate and personal goal. To date, Merck's programs (vaccines, monoclonal antibodies, and reverse transcriptase inhibitors) have yielded disappointments. Merck researchers discovered the structure of the HIV protease molecule, and protease inhibition is now the cornerstone of their HIV research effort, a program designed to get answers quickly.

The first results on Merck's protease inhibitor (L-735,524) were very encouraging. In fact, they were the best data for any anti-HIV drug to date. Blood levels of people on L- 735,524 showed a vast decrease in HIV activity. The detectable amount of HIV core protein (p24) went down to zero, and the level of HIV genetic material (RNA) in the blood was reduced by 99 percent. Significant CD4 count rises and weight gains also occurred.

Four of the first patients on L-735,524 have continued with the drug, and, unfortunately, the follow-up data are not so good. Blood tests are available for three of the individuals. In all three cases, HIV RNA started to rebound after week twelve and by six months was back at the original, baseline level. CD4 counts and p24 levels remained improved, but the latest measurements show a worsening in those values, as well. (Both of these parameters reflect immune system competence. The level of free p24 depends on the amount of antibody the immune system produces against that protein.) The reason for this relapse is almost certainly due to the development of drug-resistance. L-735,524 is administered every four hours, and the minimum concentration in that four hour period is considered barely sufficient. (It equals the IC 90 level.) Merck has never tried to push their medication to see what the maximum tolerable doses are. The company stopped escalating the dose when the first signs of altered liver function (elevated levels of transaminase and bilirubin) developed.

Analysis:

Merck has responded to this new development (the HIV RNA analysis only became available at the end of January) by boosting the dosage of L-735,524 for all participants in its current 60-person trial. (The dose has been raised from 200 or 400mg six times a day to 600mg six times a day, the highest Merck dares go.) Larger clinical trials have been put off indefinitely while Merck tests L-735,524 in combination with AZT and ddI.

Either of these strategies might put off the time when resistant mutants become predominant in the body. But Merck researchers were hoping for a "home-run" drug that completely stopped HIV. Still, the Merck drug seems to produce a profound antiviral effect - better than either AZT or the Roche protease inhibitor. This drug could provide significant clinical benefits to patients. There is a risk that Merck will too quickly abandon a compound that represents a tangible improvement over past drugs. The company should make sure that this does not happen. Trials of L-735, 524 in targeted populations, such as patients with advanced disease who cannot take AZT, could be initiated without delay. Merck reports that it is actively working on the development of other protease compounds in addition to L-735,524, although it is not believed that any of them have gone into animal tests yet. Despite some disappointments, the company has been able to initiate trials and get answers in a rapid period of time. In addition, it has maintained a continued and open dialogue with community representatives about its entire drug development program. These are no small accomplishments.

Abbott

Abbott is now on its third protease inhibitor. The first, A- 77003, was discontinued last summer. That substance had to be delivered by a central catheter 24 hours a day and caused severe vein irritation without achieving much benefit. In February 1994, Abbott halted development of its oral protease inhibitor, A-80987. Last fall, an eight-week, four- arm efficacy trial of A-80987 found that the high and medium dosages (given in six daily doses) were associated with a reduction in p24 blood levels. However, there was no reduction in viral load and some liver toxicities were reported.

Abbott has concluded that protease inhibitors can work and is continuing to search for a nontoxic version. The company is now preparing for human testing of its third compound, A- 84538. This compound is about as potent in suppressing HIV replication as AZT is in susceptible viral strains. Moreover, it is equally active in strains that are AZT-resistant. But A-84538 can cause liver and eye problems. Extensive ocular exams will be required in the trial.

Analysis:

Abbott seems finally to have arrived at a very bioavailable formulation - 80 percent is absorbed when fed to rats - but A-84538 is so unstable that study participants will not be able to take it home with them. They will have to visit the clinic every day to get their daily dose in a liquid solution.

Abbott may have a while to go before developing a protease inhibitor that has antiviral activity and relatively low toxicity. The company's program is small, although it does include some high quality researchers.

Abbott has been tight-lipped about its plans. In the past, its relations with the AIDS community has been contentious due to the plodding pace of its AIDS drug development programs. The present protease inhibitor program, with all its false and confusing starts, threatens to repeat past troubles.

Searle (Monsanto)

Searle has two lead protease inhibitors. The first, SC-52151, appears to be less attractive than the second generation compound, SC-55389A. SC-52151 has less bioavailability, a shorter half-life in the body and less distribution in tissues due to protein binding. In addition, shifts in anti- viral potency were observed after HIV was exposed to SC-52151 for one year in the lab, implying the development of resistance. No toxic effects have been seen in animals receiving 600mg per kilogram of body weight for four weeks. Lab tests with chronically infected cells found that SC-52151 is about ten times as potent in suppressing HIV as AZT. Searle claims to be advancing as fast as possible with its molecules. Phase II studies of SC-52151 are beginning in Berlin. SC-52151's bioavailability rate when taken orally is 20 percent, much higher than for Roche's protease inhibitor, but the drug has to be administered in a 70 percent alcohol solution. Clinical trials of SC-55389A will begin later in 1994. Oral bioavailability for this substance is 70 percent in dogs, one of the highest rates seen so far.

Analysis:

Levels for SC-52151 fall to the subtherapeutic levels (below the IC90 threshold) between dosages. This is not encouraging. HIV might rebound relatively quickly in this situation. It will be interesting to compare SC-52151 with the human data for the second generation compound, SC-55389A. With its longer half-life in the body and less protein binding, SC- 55389A should reach higher concentrations than its predecessor. Additionally, it can be given in a pill form. If the data confirms its superior profile, SC-55389A will become Searle's lead compound.

Vertex/ Burroughs Wellcome

Vertex is a small biotechnology company based in Cambridge, Massachusetts and founded in 1989 by Joshua Boger, previously the senior researcher at Merck. Vertex's claim to fame is rational drug design. This approach integrates biology, biophysics, and medicinal chemistry to create highly specific small-molecule drugs based on the atomic structure of proteins involved in disease. Its three main areas of drug development are HIV, drug-resistant cancer and hemoglobin disorders (sickle cell anemia and thalassemia).

Burroughs Wellcome Co. has signed a five-year HIV protease inhibitor development deal with Vertex involving as much as $32 million plus a royalty payment based on sales. Wellcome will fund the clinical trials and receive worldwide marketing rights to the compound.

Analysis:

Wellcome's investment is good news - both for Vertex and for people with HIV. For Vertex, the deal essentially removes most of the development risks. If the deal reflects the quality of the preclinical data (so far unreleased), Vertex's molecule, VX478, may have terrific stability and safety in the body. Also, the molecule's structure is simple and easy to manufacture. (It is rumored that when Wellcome management were shown the structure at the end of negotiations, it sold the deal.)

Vertex officials claim their drug has the potential to pack the "big wallop" that would knock out HIV. They say that the minimum level of VX478 between doses is ten to one hundred times more powerful than the Merck drug's minimum. For added protection, they envision starting combination drug trials from the very beginning - no doubt with Burroughs Wellcome's AZT.

Human trials have been temporarily put on hold while Vertex's new partner evaluates trial design. But the compound is now fully funded and in the hands of an experienced development team. Considering Burroughs Wellcome's investment and track record, one can predict that the development program will be large and aggressive.

Agouron

Agouron is a small biotechnology company based in La Jolla, California. Like Vertex, it specializes in designing therapeutic molecules. Agouron has developed three non- peptide molecules that inhibit HIV protease in laboratory tests. Using its structural design approach, the company has altered the molecules in ways that change their potency, bioavailability and solubility in lipids (to improve their ability to cross the blood brain barrier). Final preclinical kinetics and toxicology will be performed during the first half of 1994, and a decision will be made on which compound will go into humans.

The oral bioavailability in animal models ranges between 20 and 50 percent. The compound appears to have a good toxicity profile - toxic levels of each compound are many multiples of the effective (IC90) concentration. The company predicts that twice-daily administration will be sufficient for these drugs. But very little can be said from the preclinical data. Once again, the great hurdle will be whether any of these compounds achieves a minimum concentration in the body high enough to stop HIV replication.

Analysis:

Agouron has approximately $30 million in cash and is burning approximately $15 million annually even before it starts human trials with these compounds. A corporate partner may now be difficult to find because of the number of protease inhibitors entering clinical trials. If Agouron's compounds look good in preclinical testing, it would be a tragic error to slow their development due to a shortage of cash.

Other Companies

Many other institutions have protease inhibitors in development, including Pfizer, Ciba-Geigy (CGP-53437), Hoechst-Bayer (Hoe/Bay-793, which is being dropped but may lead to better compounds), Upjohn (U-75875), Sanofi (SR- 41476), Kyoto Pharmaceutical University in Japan (KNI-272), Nippon Mining, Parke-Davis, and, reportedly, SmithKline Beecham, Syntex, and the National Cancer Institute. Some of these compounds are not yet ready for human trials, and others are being improved.

Glaxo reported last summer that their penicillin-based protease inhibitors had pharmacokinetics problems and had been canceled. A Du Pont Merck protease inhibitor (DM-323) was recently canceled, too, but a second generation molecule is under development.

Conclusion

Protease inhibitors have a definite theoretical appeal as a therapeutic strategy. The first data indicate that they do have some efficacy - at least as good as reverse transcriptase inhibitors like AZT - although there are a number of practical problems.

For many uses, protease inhibitors may represent an improvement on reverse transcriptase inhibitors like AZT. But the ethics of clinical trials is such that they will be tested against the standard of care, which means AZT for some patients and ddI or ddC (and d4T, should it be approved) in others. Protease inhibitors will be unavailable for the foreseeable future to those who wish to avoid the toxicities of AZT and similar medications. To fulfill protease inhibitors' promise any time soon, research on them must be accelerated, with the most promising compounds given priority development.

Getting Quicker Answers About Protease Inhibitors Protease inhibitors are the only new AIDS therapy on the verge of acceptance. Quick public access - or rapid abandonment if protease inhibitor research reaches a dead end - should be a high priority for government and industry. These steps could be taken to facilitate development every step of the way:

1. Establish clear, identifiable standards for FDA approval Evaluation of protease inhibitors begs for standardization. The FDA needs to make clear what it will accept as proof of efficacy. The same old problem of surrogate markers has come up again. As it stands now, we may have to wait for data on clinical markers (worsening of symptoms or death) to demonstrate efficacy. Yet a number of tests that measure viral load now can give a very rapid picture of how HIV is responding to therapy. Researchers, government officials, and community representatives need to reach a consensus about how to evaluate these compounds and which particular measurements of viral burden, tests such as polymerase chain reaction (PCR), "quantitative competitive" PCR and branched chain DNA, are most reliable.

2. Pursue partial cures. Many companies are looking for the "home-run" protease that eliminates HIV, but development of partially effective therapies, the "singles" and "doubles," must not be abandoned. These latter compounds could be more potent and less toxic than the nucleoside analogs and help a great many individuals. When necessary, the National Institutes of Health (NIH) should step in and provide funding for trials of these drugs in special populations. These populations include people with advanced AIDS, where a particular drug may bring temporary but important antiviral effects and clinical benefits.

3. Initiate combination trials. It is doubtful that protease inhibitors will ever be a cure for AIDS all by themselves. They may be used with a wide variety of other antiviral drugs (including reverse transcriptase inhibitors) and immune modulators. Other compounds such as various cytokines and antioxidant nutrients could be part of the package. A legitimate evaluation of protease inhibitors requires that they be tested as real people will use them - not alone or only together with a given company's other products. Once we have a better idea of protease inhibitors' potential, the government could sponsor massive trials that compare people on their usual, multifaceted regimens with people following both their normal regimens and taking one (or several?) protease inhibitors.

4. Test several protease inhibitors together. It would be relatively easy and inexpensive to find out how the different protease inhibitors act together - both in lab cell cultures and soon thereafter, in people. This would provide important information as to whether the use of different protease inhibitors could impact the development of resistance. For instance, researchers might begin testing the Roche compound and the Merck compound together in lab cells to see how they interact. This work could be sponsored by the pharmaceutical industry's Inter-Company Collaboration for AIDS Drug Development or by the NIH.

5. Assist small, cash poor companies. Some companies are pushing ahead just because they have the resources to do so while some underfunded companies are languishing even though they may have a better compound. The present climate for investment in biotechnology is not good; the government has a responsibility to see that financial considerations are not the main arbiters of which protease inhibitors become available to the public. When necessary, public funds could be made available to further research on promising products (with, of course, the public assured of a fair return on its investment).

6. Government should work with industry. NIH officials should meet with all the companies currently developing HIV protease inhibitors to underscore the importance of developing effective anti-HIV therapies that could be available within the next two years. Industry and government can work together to maximize the chances for success with this class of compounds. - D.G./D.G.