icon-folder.gif   Conference Reports for NATAP  
  10th Conference on Retroviruses and Opportunistic Infections
Boston, Mass, Feb 10-14, 2003
Back grey_arrow_rt.gif
Entry Inhibitors, The How and Why of New Agents at Retrovirus: an Update
Reported by Christopher D. Pilcher, M.D. UNC at Chapel Hill Center for AIDS Research
The current range of approved HIV medications includes drugs that inhibit various steps in the virus' life cycle. Most classes of these drugs act inside HIV infected cells: many agents interfere in one way or another with reverse transcriptase (RT)-a viral enzyme which the virus uses to replicate itself within the cell. ("Nucleoside or nucleotide" and "non-nucleoside" RT inhibitors--NRTI and NNRTIs--are chemically very distinct and can work together with powerful synergy against this one target.) Other existing drugs interfere with the activity of HIV protease-which the virus uses to package itself for export out of an infected cell, to begin new rounds of destruction. The third major drug target that has been successfully exploited is HIV gp41-a molecule on the virus' surface that has to change its shape in a specific way in order for the virus to "fuse" with the cell it is attacking and empty its payload inside. By directly binding to gp41, the injectable peptides enfuvirtide (Fuzeon, formerly T-20) and its cousin T-1249 are able to powerfully block the fusion process. Very potent, bioavailable NRTIs, NNRTIs, PIs, and FIs (Fusion Inhibitors) have all been developed.
Some of these agents are more toxic than others, but all have some toxicities which in many cases may accumulate over time.
Moreover, despite the potency of all of these agents, an overwhelming body of evidence proves that even in patients treated with very powerful combinations and maintaining "undetectable" viral loads on therapy the virus is able to continue to replicate and evolve (albeit at a drastically slowed pace). The consequences are
  • evolution of drug resistance in some patients , and
  • persistence of HIV in the body in all. (see reports on reservoirs elsewhere in NATAP website Retrovirus Conference Reports)
In answer to the above challenges, then, the goals of drug development efforts must be to
  • improve the acceptability and tolerability of current drugs
  • find new agents that might have improved activity throughout the body, and
  • find agents that will be active against viruses that have already evolved
resistance to some or all existing drugs---an increasingly common scenario for patients living with HIV and their physicians.
In the quest for agents that might have improved activity compared with older drugs, investigators have wondered if equally potent drugs targeting the parts of the virus life cycle that occur outside the cell might be better than drugs targeting events that take place inside infected cells. Why? Anti-HIV drugs that must be inside the cell to be active can be efficiently neutralized by some cells, using primitive, innate self-defense mechanisms such as "efflux pumps" which sense toxins and eject them outside of the cell. Many experts believe that this kind of "cellular resistance" could be an important reason for the viral persistence and evolution in patients on seemingly potent combination therapies. "Extracellular ART" would completely circumvent this problem. More immediately, any drug working on a separate part of the virus' life cycle would have little potential for "cross-resistance" with NRTI, NNRTI or PIs.
In a summary plenary lecture at this year's Retrovirus Conference (session 23), Eric Hunter from Alabama reviewed the progress of entry inhibitors and prospects for further development of extracellular ART. This lecture emphasized the multiple steps involved in the process of HIV attachment and entry into cells, and explained how each of these steps could be targeted. The first step is HIV's gp120 molecule binding to a CD4 receptor and to a necessary "coreceptor" molecule on the cell surface (CXCR4 and CCR5 are the most common) that must be present on the cell surface for proper attachment. Drugs inhibiting this binding are called attachment inhibitors.
Like a lock needing two separate keys, once both sites on gp120 are bound by CD4 and one of the coreceptors, gp120 flips back out of the way, exposing the virus' previously hidden harpoon molecule, gp41. gp41 is then free to pierce the cell and reel it in close enough to allow virus-cell fusion. Drugs targeting these last steps are called fusion inhibitors (FIs).
The first direct inhibitor of gp41 mediated fusion, enfuvirtide (T20, Fuzeon), has proven active enough to receive FDA approval for use in heavily antiretroviral experienced patients, and the drug (administered only by twice-daily subcutaneous injections) has already made its way into limited expanded access.
At this year's Conference, Roche (Abstract 568) presented the most comprehensive summary to date of the pooled safety and efficacy data from its two pivotal phase III studies, "Toro 1" and "Toro 2" (in which more than 600 highly pre-treated patients received optimized therapy with or without enfuvirtide). As previously reported, the drug was effective at reducing HIV RNA levels by 10-100 fold in the enfuvirtide-treated group in both trials. However, up to 98% of patients had injection site reactions to enfuvirtide in these studies though only about 20% reported "moderate" or worse severity of discomfort associated with local skin swellings. 1-3% patients at every timepoint needed pain medication to manage these problems. Another concern related to frequent injections is the possibility that bacteria could be introduced into the blood. There was a significant increase in the rate of bacterial pneumonia associated with getting enfuvirtide in Toro 1 and 2 and a non-signficant trend to more bacterial infections overall. Also theoretically concerning is the risk for allergic reactions to an injected protein: two cases of proven allergy to enfuvirtide (and two more of suspected allergy) were documented in these trials, though many patients (10% overall) developed elevated eosinophil counts that can suggest allergy without additional symptoms.
Given the very high estimated cost of the new drug (up to $20,000/yr; see related NATAP report) and the common occurrence of these irritating and cosmetically bothersome (if not medically serious) injection site reactions, it is likely that the wide use of enfuvirtide will continue to be limited to very highly motivated patients with few other treatment options. For those for whom enfuvirtide might be very important, however, the information presented on tolerability in the Toro trials should be reassuring in that side effects related to injection were rarely serious, and only led to treatment discontinuation in 3% of treated patients.
Second-generation fusion inhibitors are in the pipeline. T1249 is another injectable Roche peptide extremely similar in design to enfuvirtide but binding to gp41 just downstream of enfuvirtide's gp41 binding site and has recently been tested in phase I/II trials for activity against enfuvirtide-resistant viruses. In a study reported at the Conference on Tuesday (abstract 14Ib), T1249 was given to 23 patients with both genotypic and phenotypic resistance to T20 for 10 days. Reductions in viral load by day 11 were >10-fold for more than half (63%) patients; there was one possible allergic reaction. The activity of T1249 against enfuvirtide-resistant virus may make the use of these fusion inhibitors in sequence possible, for patients failing enfuvirtide. There are no orally available fusion inhibitors in current development, so where we are is where we are likely to be for some time...
Many of the agents previously proposed for use as entry inhibitors have also only been injectable, some requiring hours-long infusions. Progenics' PRO542, a novel tetravalent CD4-Ig fusion peptide and TNX -355-01 (abstract 13) are the latest candidates for this type of injectable therapy targeting areas on gp120 to prevent attachment. Perhaps more exciting, the drug pipeline is now also rapidly filling up with "small molecule" compounds that bind to and inhibit the coreceptors, CXCR4 and CCR5. "Small molecules" can generally be absorbed much better from the GI tract, making oral therapy possible. One such CCR5 inhibitor from Pfizer (UK-427,857; abstracts 12, 546a) is already in phase I dose-ranging trials in humans without any untoward side effects to date.
Two additional CCR5 inhibitors in pre-clinical development were presented at the Conference, and so-far look good for use as pills in people: one from Takeda (TAK 220; abstracts 11, 562) one from Ono (AK602; abstract 10). AnorMed, whose AMD3100 is no longer in development, showed preclinical results for a promising candidate CXCR4 inhibitor AMD070.
A lot of observers are paying very close attention to developments in the attachment inhibitor arena, in part because they have the theoretical potential to result in serious toxicities when used in infected patients. CXCR4 and CCR5 are ubiquitous (everywhere) on cells from many different tissues and have myriad functions including normal growth and development processes. The effects of interfering with the binding sites for these molecules in normal tissues are not known. Perhaps as an example of the hazards of CCR5 inhibition in vivo, the first CCR5 inhibitor to make it into humans ("Schering-C"), was pulled due to its propensity for causing serious cardiac conduction abnormalities in early clinical trials. Another wild card for the use of coreceptor antagonists is the specter of the "coreceptor switch". Viruses can use either CXCR4 or CCR5 more efficiently. It has been commonly noted that many more so-called "X4" viruses are present in patients with late-stage AIDS, while people with very early infection nearly always have dominant "R5"virus populations. The possibility that inhibiting virus attachment to CCR5 would somehow cause a switch to CXCR4-usage-and more importantly, that this switch might hasten patients on the road to AIDS-are a significant worry. To date, most early clinical trials of CCR5 inhibitors have not shown R5ŗX4 switching to take place (at least over a very short time).
Whitcomb and colleagues (poster abstract #557) presented a survey of coreceptor usage among patients enrolled in the Toro trials evaluating the effect of enfuvirtide. The specific technique they used to look at coreceptors (from ViroLogic) had the advantage of detecting mixtures of X4 and R5 viruses, rather than qualifying viruses as "all X4" or "all R5". This study demonstrated clearly that many (24%) of all patients in the Toro trials (a very advanced, heavily pretreated population) had a mixture of R5 and X4 viruses, while 62% were predominantly R5 and a tiny minority were apparently exclusively X4. The sensitivity for detection of X4 variants is not that exquisite using the ViroLogic technique, so the proportion of patients that really have mixtures of virus types was probably higher. In fact, all of the patients having any X4s would likely be classified as "X4" or "SI" using traditional approaches (based on sequencing or MT-2 assay). In other words, these findings suggest that about half of patients that would be candidates for therapy with new coreceptor inhibitors would have large proportions of circulating virus with preexisting resistance from the get-go (no real "switch" required!). While hard core basic scientists have known for years that virus populations are not all-or none, it is cautionary for the rest of us not to pin extravagant optimism on untried therapies. As we have to relearn at every one of these conferences, the drug that makes it all the way through the drug development pipeline is very, very rare. In the case of coreceptor antagonists, it remains maddeningly unclear what potential the entire drug class may have for either efficacy or safety in infected patients.
A number of new classes of antiretroviral agents were discussed at the meeting. In the new agents sessions (oral abstracts 6-14Ib), several novel PI and NNRTI compounds were noted to be into phase I trials, and a novel class of integrase inhibitor compounds called pyranopyrimidines (PDPs) were noted to be potent in the test tube. Most interesting, Mario Stevenson and John Rossi co-chaired a symposium (abstracts 49-52) examining the various aspects of "RNA interference", or RNAi. RNAi is a recently described self-regulatory process that cells use to shut down production of given proteins when they detect excess accumulation of short, double stranded RNA coding for those proteins inside the cell. HIV researchers hit on the idea of trying to introduce short interfering RNA bits into cells to stop HIV production, and as summarized in this symposium, it may work. Dr. Judy Lieberman and others showed surprising evidence that exposure of HIV infected cells to interfering bits of RNA in the test tube could shut down HIV production by these cells. In other words, the RNAs worked even if they were not produced by the cells themselves. In an animal model using a different disease (autoimmune hepatitis), Dr. Lieberman went on to show that interfering RNAs could actually be taken up in large amounts into cells when injected into animals. Together, these preliminary findings may open up an entirely new approach to HIV therapeutics: one can imagine infusing patients with interfering RNAs that would completely shut down HIV replication. The potential for this approach is totally unknown…we will await the updates from these and other labs exploring RNAi with great anticipation over the coming months.
With each new avenue explored, we come a little closer to a long term goal of having tolerable, potent agents capable of limiting virus replication in all patients.