The Third International Conference on Drug Therapy in HIV Infection -- held in Birmingham, U.K. from November 3 to 7, 1996 -- reflected the ongoing vigorous disputes over how to apply the new anti-HIV combination therapies from a European point of view. That point of view is difficult to characterize because some countries are following a relatively conservative strategy whereas healthcare providers in other countries are much more aggressive in starting treatment with protease inhibitors plus other drugs in highly potent "knockout" cocktails.
The British tended to be the most conservative. Ian Weller, M.D., a noted researcher at the University of London and the chair of the conference's scientific committee, said at the opening press conference that if a new patient came to him with a CD4 count of 190 and was otherwise healthy, the only thing he would prescribe was prophylaxis for pneumocystis carinii pneumonia. In an interview at the end of the conference, Dr. Weller elaborated that he was more concerned about the possible long-term toxicities of the new antiviral agents than about the dangers of irreversible damage to the immune system if HIV was left unchecked -- a danger he dismissed as "speculative." At the press conference, Dr. Weller's view was disputed by Dutch and Swedish colleagues, Dr. Pieter Reiss and Dr. Erik Sandstrm. And Joep Lange, M.D., of the University of Amsterdam went so far as to declare during the conference, "Any significant viral load should lead to therapy."
Paralleling these different national attitudes, Dutch health authorities have agreed to fund protease inhibitors for early HIV infection whereas the British government, whose National Health Service is deeply in the red, has yet to provide any new funds for these drugs. Therapy in the U.K. is mostly limited to people with CD4 counts under 350.
A critical issue in this debate is how durable is the response to anti-HIV therapy. Outside of the initial, or acute, phase of HIV infection, no one expects elimination of the virus, just chronic management of the disease. Precisely when and under what conditions breakthrough infections of therapy- resistant HIV occur remain to be seen. David Cooper, M.D., of Sydney, Australia, warned in an interview at the end of the conference, "We risk repeating the experience of AZT -- in which people's hopes were raised only to be dashed a few years later." In contrast, Douglas Richman, M.D., during one of the conference's final presentations held out the hope that with enough drugs (at least three new ones at a time in a given patient), the drop in HIV replication would be so great that drug resistance could be put off indefinitely in many people.
There were even disagreements about how to prove treatment concepts: at the conference's opening session, the American Richard Schooley, M.D., who chairs the National Institutes of Health's AIDS Clinical Trials Group, gave an impassioned plea for ending the practice of conducting definitive drug trials based on comparisons of physical disease progression. He argued that these huge, long-running comparative trials are a waste of resources, leading to results that are irrelevant because of more recent advances in treatment and the poor correlation between trial results and individual patients' experience. Dr. Schooley said, "There is no need to keep showing the superiority of better viral load reductions." He did advocate a trial that evaluated individualizing treatment regimens to keep patients' HIV levels undetectable and allow them to stay "ahead of HIV replication and resistance." The next day, the British statistician Thomas Peto, Ph.D., countered with the argument, "Viral load is okay for phase II [preliminary efficacy trials] and for checking 'me too' drugs for equivalence, but not for proving out new classes of drugs or establishing enhanced efficacy."
Central to the debate on the relative merits of data on disease progression and viral load is the state of HIV infection in the lymph nodes and other lymphoid tissue. Viral load tests monitor HIV levels in the blood, but the major population of virus and the damage that the virus wreaks is concentrated in the lymphoid tissue. Though not formally reported at the Birmingham conference, a paper then in press by a University of Minnesota team cast a shadow over the conference proceedings. That paper (published a week later in the journal Science: Ashley Haase et al., Science. Nov. 8, 1996; 274(5289):985-989) used tonsil biopsies (much easier to perform than lymph node extractions) to produce graphic photos of the location of HIV virus in lymph tissue. It turns out that there are 1,000 to 10,000 times more HIV in lymph tissue than in blood. Plasma viral load does indeed correlate with the number of HIV-infected cells in the lymph tissue, and these seem responsible for most of the free virus appearing in the blood. But, 90% to 99% of the HIV particles in the lymph have been trapped on the surface of the follicular dendritic cells (FDC) in that tissue. Although coated with antibody, that virus is still infectious, and its longevity is unknown. As long as there is still some HIV sticking to the FDCs, HIV infection apparently can rebound should antiviral treatment be terminated.
The University of Minnesota investigators are now using their technology to analyze tonsil biopsies taken from participants in a Dutch trial of the AZT/3TC/ritonavir combination, and the first results from that analysis were reported at the conference. Participants in the Dutch trial saw their viral loads drop by 99.8% (2.7 logs) at 16 weeks. (There were 33 treatment-naive individuals, with a median CD4 count of about 155 and median plasma viral load of about 210,000. They were assigned to one of two open-label arms, which turned out to have equivalent results: start all three drugs right away or delay the ritonavir for three weeks.) By week 24, no trial participant had detectable HIV in their blood plasma (limit of detection was 238 HIV RNA copies/ml). In the first six patients whose tonsil tissue was analyzed, all had no detectable HIV at week 24.
This report caused a sensation because of the advanced disease status of the individuals involved. Virus levels in the lymph tissue, including the FDCs, clearly have been greatly reduced by six months, but Keith Henry, M.D., of the University of Minnesota group, noted in a telephone interview, "The detection limit for this assay is 30 copies per milligram of tissue. Saying that levels went below detection, doesn't mean that there is not still a fair amount of virus."
Other conference reports also reported lymph tissue clearance in parallel with reductions in plasma viral load: Alain Lafeuillade, M.D., of Toulon, France, used a four-drug combination (AZT/3TC/ddI/saquinavir -- with ketoconazole added to reduce the breakdown of saquinavir in the liver) that reduced lymph node HIV levels by 99% in 15 patients with baseline CD4 counts averaging 264.
A similar parallel decay in blood and lymph node HIV levels was found in 15 participants from an international study of AZT/ddI/nevirapine (participants' baseline CD4 counts averaged 376 -- see the June/July, 1996 Treatment Issues article, Nevirapine Surprise for details of this trial).
In the lymph node substudy, three of the four fully compliant participants on triple therapy had undetectable lymph node viral loads after over a year of therapy, but the limit of detection with bDNA assay used was 10,000 copies of HIV RNA per milligram of lymph tissue, a fairly high amount.
AZT/ddI/nevirapine could leave substantial amounts of replicating and increasingly resistant HIV in the lymph nodes. This may be why HIV levels in the blood seem to be creeping up, judging by the one-year overall trial results. At six months, those on AZT/ddI/nevirapine had undetectable plasma viral loads (limit of detection: 200), compared to 57% after one year.
If AZT/ddI/nevirapine is not potent enough for many people, maybe the ritonavir/saquinavir protease inhibitor combination will do. Additional results, out 12 to 20 weeks, were released in Birmingham for a pilot study of 140 people with CD4 counts of between 100 and 500 and no prior protease inhibitor exposure. Due to ritonavir's inhibition of liver metabolism, blood levels of saquinavir are raised 20-fold by the combination. After 12 weeks of treatment, median viral loads declined by up to 1,000-fold (3 logs) in those receiving any of the four different combination regimens, with no regimen being clearly superior to the others. In the two trial arms with data now available for week 20, viral loads continued to decline a little further, and the median drop now clearly exceeds 1,000-fold. This compares with a mean viral load drop of 25-fold (1.4 log) for AZT/ddI/nevirapine at 28 weeks. For those out to 20 weeks in the ritonavir/saquinavir trial, the median viral load now measures only about 20. Detecting that small amount is quite a trick considering that the limit of detection is 200 for the PCR assay used to evaluate viral load in the trial's participants. The nevirapine triple combination trial used the same assay and counted everyone testing out at below 200 as having a viral load of 200.
It turns out that the ritonavir/saquinavir study included actual viral load values obtained down to 20, even though the test was not considered accurate at these levels. The ritonavir/saquinavir investigators effectively (and tacitly) imposed a lower limit of 20 rather than 200, and by this contrivance, allowed their treatment to achieve an extra ten-fold (one log) apparent drop in viral load. It is therefore impossible to take their results at face value. According to John Sninsky of Roche Molecular Systems, which developed the PCR assay, the test is not appropriate for low viral loads because its variation at these levels has not been established.
Roche has developed an ultra-sensitive test, now available for research purposes, that is accurate down to 20 HIV RNA copies per milliliter of plasma. That was the test recently used to recheck the lowest viral loads in the nevirapine trial, and these data were summarized in Birmingham. With the ultra-sensitive PCR assay, 51% of the trial participants receiving AZT/ddI/nevirapine had undetectable viral loads at one year, just 6% less than with the standard PCR. Using a lower limit of detection would have increased the observed viral load drop for the nevirapine triple combination by a considerable amount -- with Roche's ultrasenstive test, the mean viral load drop was 2.25 logs (178-fold) at six months and 1.8 log (63-fold) at one year. In the ritonavir/saquinavir trial, it will be difficult to evaluate the trend in HIV levels between six and 12 months because of the unknown, but assumed to be large, amount of error in the standard PCR assay at low viral loads.
It remains that ritonavir/saquinavir is one of the most powerful combinations, and many doctors are trying to use it as salvage therapy for people failing other therapies. It is not for everyone, though. The trial data reveal a huge range in responses to the two protease inhibitors, which may be partly due to interpersonal differences in the extent that ritonavir raises saquinavir levels. At a press conference in Birmingham, Richard Haubrich, M.D., of the University of California San Diego described the experiences of seven of his private patients taking ritonavir/saquinavir, two of whom did not improve to any appreciable degree. In the discussion following Dr. Haubrich's account, it was commonly felt that the population in whom ritonavir/saquinavir was less likely to provide benefit includes those with higher viral loads (100,000 or above), concurrent illnesses or prior treatment with ritonavir or saquinavir -- or a lengthy history of HIV treatment in general. (Other, as yet undefined, personal characteristics no doubt also play a role.) For such people, the combination of the two protease inhibitors is probably not sufficient. They need at least a third new drug to which their HIV is not yet resistant.
In regard to new drugs, the news from the Birmingham conference was mixed. There was a detailed description of the mutations that arise after 18 months in trial participants on AZT/3TC in the pivotal American trials for the combination. Earlier reports claimed that resistance to 3TC somehow reduced resistance to AZT, but this analysis found extensive dual resistance to the two drugs, involving both classic AZT mutations and several novel ones. Cross-resistance to ddI or ddC did not appear to be a major issue, though, as has sometimes been feared to arise from 3TC resistance.
In one session, Emilio Emini of Merck (makers of the protease inhibitor Crixivan) pronounced that general cross-resistance exists for nucleoside analogs and it is now appearing for protease inhibitors. Later on, Amy Patick from Agouron Pharmaceuticals, which is developing the new protease inhibitor nelfinavir (Viracept), disagreed. Citing her study of resistance to nelfinavir in 16 people with up to one year on nelfinavir, she claimed that cross-resistance rarely occurs between nelfinavir and other protease inhibitors. This observation needs further confirmation with more patients and longer follow-up. Further, it has not been determined at all whether people starting to fail on other protease inhibitors can switch to nelfinavir without encountering drug resistance on the part of their HIV.
Emilio Emini warned, "If selection for resistant virus occurs, future therapeutic options may be very limited." Multi-drug resistant HIV is an increasing problem as mutations accumulate. These mutations both provide for drug resistance and compensate for any slowing of viral replication due to the resistance mutations. Fear of forcing HIV on such an evolutionary pathway was a decisive reason for the many Birmingham attendees who wanted to put off therapy in "healthy" people, at least until more information is available.
The mortality results of ACTG 193A, an early large anti-HIV combination therapy study in very advanced disease, were finally released at the December meeting of the AIDS Clinical Trial Group. The study compared AZT/ddC, AZT and alternating ddI, AZT/ddI and AZT/ddI/nevirapine in 1,314 subjects with a median CD4 cell count of 20 -- most of whom had prior treatment with nucleoside analogs. The study found that there were significantly fewer deaths on the three drug arm (118) as compared to AZT/ddC (142) or AZT alternating with ddI (148). However, the reduction in deaths was not statistically significant when the triple therapy was compared to AZT/ddI (128 deaths) -- and the AZT/ddI arm, in turn, was not significantly better than the two arms of the study with the poorest performance. There was also a trend toward longer median survival on the triple combination.
If anyone is particularly interested in these outdated results -- since therapy including protease inhibitors has been shown to be dramatically better than nucleoside analog combination therapy in this population -- the data should nevertheless be viewed with skepticism, since 74% of the patients went off study treatment for reasons other than death, drug toxicity or completion of the study. (The median time on treatment was only 45 weeks in the study, which lasted roughly three years.) Ongoing subset analyses and comparisons of the effect on progression may yet yield some interesting minutiae, but this lengthy trial serves to buttress Robert. Schooley's polemic in Birmingham (see main story above) about the futility of clinical endpoint trials in the current environment.
- Theo Smart
Nevirapine generally lowers blood levels of indinavir according to newly released data from the nevirapine/indinavir drug interaction study. The results are not completely unexpected because nevirapine is known to induce liver activity that breaks down drugs such as the protease inhibitors. Though significant, the effect on indinavir blood levels is not dramatic, and varies from person to person -- in fact a couple of trial participants actually saw increases in blood levels of indinavir.
The study was in 24 HIV-positive adults (19 available for analysis) who were treated with standard doses of indinavir for seven days before receiving nevirapine. Nevirapine was administered according to the dose-escalation schema recommended to reduce the incidence of drug-related rash: 200 mg a day for 14 days, followed by 200 mg twice a day. Data to day 36 was available for the drug interaction analysis.
Nevirapine reduced the mean peak levels of indinavir by only 11%, but the mean area-under-the-curve concentrations (a measure of the drug's total presence in the body between doses) and trough levels (minimum level between doses) were significantly reduced -- by 28% and 38% respectively. There was a slight reduction in nevirapine area-under-the-curve concentrations. Both drugs were well tolerated together. Of the five patients who discontinued, three were due to toxicity -- two in response to nevirapine- related rashes, and one participant developed kidney stones on the first day of indinavir treatment.
The degree of interaction was slightly greater than hoped, but virtually identical to the interaction observed between nevirapine and saquinavir. The stakes are higher with indinavir, though, since a substantial number of people who develop resistance to this drug appear to lose susceptibility to most of the other protease inhibitors available or in development. Increasing the indinavir dose to 1,000 mg three times a day may achieve blood levels of indinavir closer to what is achieved by standard dose in the absence of nevirapine. Upping the dose costs more, though, and runs the risk of more side effects given the variation of indinavir blood levels among individual patients.
Of course, the combined antiviral effects of the two drugs may substantially reduce the incidence of resistance. This study is collecting resistance data, but the median starting viral load was very low (4,000 copies per ml) as many of the volunteers were on nucleoside analog background therapy before and during the study. Virologic data out to day 96 is available on sixteen patients (three more have discontinued the study). Fourteen have viral load below 400 copies, and two have rebounding viral load. Clearly, larger trials will be needed to adequately address the effectiveness of this combination.
Data from an open-label study of indinavir/nevirapine/3TC will be presented at the Conference of Human Retroviruses this January.
- Theo Smart
For the first time ever, the Food and Drug Administration (FDA) Antiviral Advisory Committee found itself unable to recommend approval of an anti-HIV drug. Amid complaints that the data from studies of delavirdine (brand name Rescriptor) were contradictory and inconsistent, the committee reached a four-to-four tie vote on Pharmacia and Upjohn's application to market this nonnucleoside reverse transcriptase inhibitor (similar in action to the approved NNRTI nevirapine) for use in combination with other antivirals. Committee chair Fred Valentine, M.D., illustrated his panel's mixed feelings when he declared, "Right now, I don't believe the data supports approval, though I do believe the drug probably deserves it." It remains to be seen what final action FDA officials will take on the company's application.
The case for approval was based primarily upon data from three studies, two large clinical endpoint studies (which measure disease progression in terms of opportunistic infections and death) and one surrogate marker study (which checked only on CD4 count and viral load changes) -- see the box for a detailed summary of the clinical endpoint studies. Although there were modest positive changes in CD4 cell counts and viral load in the pivotal clinical endpoint studies, in the third study, ACTG 261, there was no significant difference in the surrogate marker changes between AZT/ddI/delavirdine and AZT/ddI. (But these are very preliminary results. ACTG 261, which is just winding up, is a one-year, 550-person trial in people with CD4 counts between 100 and 500 and little or no previous anti-HIV treatment.) The worst stumbling block was that in one of the large trials, the transient surrogate marker benefit provided by the addition of delavirdine to ddI monotherapy did not translate into less disease progression.
The Advisory Committee also was concerned that the data on drug interactions with protease inhibitors were derived from studies in HIV- negative people, so there was no clear data on the safety or effect of delavirdine with protease inhibitors in HIV-positive patients over time. (For more specifics on the interaction of delavirdine and protease inhibitors see September 1996 Treatment Issues article, Delavirdine/Protease Inhibitor Interactions. Delavirdine inhibits the breakdown of many drugs in the liver, including the protease inhibitors. This may raise drug levels dangerously in some circumstances.
Even Committee members who favored approving delavirdine were underwhelmed by the trial results. "I am dismayed at the modest data seen in the studies especially when compared to AZT or ddI monotherapy," said Wafaa El-Sadr, M.D., of Harlem Hospital. Nevertheless, she and others felt that there was a role for the drug in combination therapy, particularly in patients who could not tolerate other options.
Diane Murphy, M.D., of the University of Florida commented, "Maybe these were just the wrong drugs to use in combination. There was always a positive trend in CD4 cell counts and viral load - It would be wrong to [block approval] because of the clinical endpoint study," she said.
It is now recognized that few anti-HIV drugs produce more than a temporary effect when used alone or in suboptimal combinations.
Nevirapine, a very similar drug to delavirdine, looked much better when it was combined with two other drugs in previously untreated patients. There are similar studies underway for delavirdine, but no data are available yet.
A final issue that plagued the panel was the absence of any data on use of this drug in the pediatric population. During the public comment session of the hearings, pediatric activists from across the nation complained that pharmaceutical companies are ignoring this population, and that the FDA has been complicit by not requiring companies to investigate drugs in children with HIV. The most moving statements at the hearing came from HIV-infected mothers of such children, many of whom said that their health was improving on the new HIV combination regimens and that they could no longer stand by and watch their children go untreated. The mothers demanded that the FDA not countenance applications for drug approval in adults from companies that were not submitting data in pediatric populations as well.
Trial 017 compared the ddI and delavirdine combination to ddI plus placebo in a group of patients with a mean CD4 count of 135 cells/mm3 (entry CD4 count between 0 and 300) and an average viral load of 5.8 logs (630,000 HIV RNA copies/ml of plasma). All patients had prior AZT experience, and 25% had received ddI in the past. Three hundred ninety volunteers were enrolled, of whom 13% were women, 32% non-white and 11% injecting drug users. The study compared the rates of death, clinical progression defined as time to AIDS or death, and disease progression (defined as time to AIDS, death or 50% drop in CD4 cell count).
The peak viral load reduction (0.8 logs, or 84%) at week four was significantly greater for the delavirdine/ddI arm, but did not persist and was not different at subsequent points in time. After 92 weeks of follow-up, there was no significant difference in clinical efficacy between ddI/delavirdine and ddI alone. The number of deaths were similar in both arms (ddI arm: 61, combination arm: 66). Yet the number of opportunistic infections while on study medication were fewer for the combination arm (n=63) as compared to ddI monotherapy arm (n=90).
Trial participants with entry CD4 counts greater than 50 had a slightly better CD4 and viral load response. This enhanced effect may have been due to higher delavirdine plasma concentrations: Trial participants who achieved nevirapine concentrations of at least 7.5 micrograms/ml had better results, and these people tended to be the ones with an initial CD4 count above 50.
Another trial, 021, which compares AZT plus placebo to AZT plus one of three doses (200 mg, 300 mg, or 400 mg three times per day) of delavirdine, remains active and has not been terminated. This trial enrolled over 700 patients with CD4 counts between 200 and 500 (mean ranging from 325 to 340, depending on the trial arm) and an average viral load of 5.2 logs (160,000), of whom 60% were antiretroviral naive, 14% women, 31% non-white and 10% injecting drug users. Over 75% of patients in each arm had a detectable delavirdine plasma concentration. In terms of CD4 responses, after 62 weeks, the groups with the two higher doses had approximately a 20 point CD4 count difference as compared to the lower dose group or the AZT monotherapy group. In addition, there was a sustained 0.5 log difference in HIV RNA between the 300 mg or 400 mg delavirdine dose groups and the two other groups at 62 weeks of follow-up. In this trial, the rate of treatment emergent medical adverse events was similar in all arms, although again the rate of drug-related rashes was greater in the delavirdine groups (22% to 31% as compared to 16% in the AZT monotherapy group).
Future studies will combine delavirdine with AZT/3TC, as well as protease inhibitors, and will investigate the use of delavirdine in preventing maternal-fetal transmission and needlestick exposures. A five-drug combination trial including hydroxyurea also is being considered.
- Gabriel Torres, M.D.
Attacking new critical points in the HIV lifecycle is vital to finding cheaper, more effective and resistance-defying antiviral therapies. On November 20 to 21, a symposium, "Novel HIV Therapeutic Strategies," covering such innovative approaches was held in McClean, Virginia, sponsored by the Cambridge Healthtech Institute. Attendees reviewed targets for drug therapy revealed by recent progress in the understanding of how HIV infects cells and replicates in those cells, interacting with cellular proteins and the immune system in the process. Many of the compounds mentioned have been covered previously in Treatment Issues, but there were several truly novel drug leads that have received little attention in the past.
The reverse transcriptase enzyme, which converts HIV's RNA gene set into DNA before the virus inserts itself into cells nuclei, often has been compared to a hand, complete with a palm, fingers and thumb. The fingers and thumb grab a hold of the nucleotide building blocks and attach them to the growing strand of HIV DNA, lying in the "palm" of the enzyme. This analogy was first made several years back, when the three-dimensional structure of the reverse transcriptase enzyme was discovered.
Knowing the enzyme's structure could have important implications in the design of new, more potent reverse transcriptase inhibitors. Ten years ago, discovery of the three-dimensional structure of the protease enzyme led to the discovery and development of protease inhibitors.
A group at State University of New York (SUNY), Buffalo has used reverse transcriptase structure to design a chemical, DNP-poly-A, that fills the entire nucleotide binding region of the enzyme, in a fashion similar to how protease inhibitors fill the active site of the protease enzyme. When DNP- poly-A was added to HIV-exposed cells, the cells did not become infected.
DNP-poly-A is a sequence of nucleotides (an oligonucleotide). Usually, if strands of such material are found floating around in the blood or outside of protected areas like the cell nucleus or mitochondria, they are treated by the body as noxious garbage or enemy viral genes and quickly destroyed by special enzymes. Most oligonucleotide drugs similarly are broken down before they can reach their targets. The researchers from SUNY do not believe that this will be a problem with DNP-poly-A, because the drug was stable even after prolonged incubation in cell cultures with a variety of enzymes known to cut up strands of nucleotides.
Next, the researchers administered the drug as an injection to mice with murine leukemia virus, which has a reverse transcriptase enzyme very similar to HIV's. Viral load dropped to undetectable levels in the mouse blood, and there was little toxicity. One problem with the compound, is that even though the team has modified it to increase its absorption from the digestive tract, it still looks as though DNP-poly-A will have to be administered as an injection.
There is no word yet if any pharmaceutical company is interested in developing this compound.
Ribosomal frameshifting is a process unique to viruses. It occurs as the viral genetic material is translated into proteins that eventually will be incorporated into new virions. In HIV, frameshifting is carefully orchestrated to produce an exact ratio of 'gag' gene proteins (the core proteins of the virus) to 'pol' gene proteins (the reverse transcriptase, integrase and protease enzymes). If the virus produces too much gag in relation to pol, or visa versa, new infectious virions cannot be properly assembled.
A start-up biotech concern, RiboGene, Inc., has developed a laboratory test that screens for products that affect frameshifting either by inhibiting it (no pol enzymes) or inducing it (little gag core protein). Another researcher present at the conference, Louis Henderson of the National Cancer Institute, suggested yet another mechanism for interfering with frameshifting to disrupt the delicate balance of viral production. Premature frameshifting will prevent production of the last portion of the gag precursor protein. This would eliminate HIV's "zinc fingers" (see Treatment Issues October, 1995 article, Zinc Fingers - The Next Antiviral Target, pages 7-8), without which the virus cannot package its genetic material into new virions.
RiboGene's test screen has found several distinct classes of chemicals that induce frameshifting, though none have been found that inhibit the process. One of these compounds, RG 501 blocks viral replication in cell cultures at doses that do not kill or slow the growth of these cells. The company is currently pursuing analogs of RG 501 in animal models and is in talks with a number of major pharmaceutical companies interested in buying or co-licensing the compounds.
One of the few compounds reviewed at the meeting that has already made it into human trials is Paracelsian's PN355, now named androvir. The compound is derived from an herb used for colds and the flu in Chinese and Ayurvedic medicine. Androvir inhibits HIV replication in cell cultures and appears to trigger the death of HIV-infected cells.
Androvir's mechanism of action is complex. When a virus or cancer takes over a cell, the cell life cycle is altered in such a way as to encourage the production of viral materials and cancer-promoting proteins. In other words, the virus shifts the cell's gears into drive. The virus recruits certain cellular factors to do this. Androvir blocks the activation of these cellular factors, which slows this viral or cancer-driven cell maturation, thereby stopping viral replication. Since androvir's target is cellular, the virus may not be able to mutate to escape the drug's effect.
The company has now studied the drug at doses up to 60 mg per kg per day in over 30 volunteers with HIV or cancer. One study in HIV-infected patients was recently completed at Bastyr University in Seattle, Washington. Viral load was reduced by 38% (which is not a significant amount given the variability of viral load tests), and CD4 count increases averaged about 30% over baseline. Dr. John Babish, chief science officer and senior vice president of Paracelsian also said that there were several unusual changes observed very early within the study, such as rapid, dose-related improvements in serum cholesterol levels, which returned to baseline when therapy was discontinued. The company plans future studies to find a more effective dose.
Anecdotal observations and laboratory studies indicated previously that human chorionic gonadotropin (HCG) might have an anti-KS and anti- HIV effect. See Treatment Issues July/August 1995 article, Human Chorionic Gonadotropin, page 15, and Pamela Harris, M.D., in AIDS Patient Care and STDs, March, 1996, concerning systemic KS treatment with massive amounts of HCG.
Now, the October 24 New England Journal of Medicine has reported on the first human trial in Kaposi's Sarcoma for this pregnancy-related hormone. Investigators documented that intralesional injection of a formulation of HCG reduces the size and occasionally leads to complete regression of the treated KS lesions.
The new results come from two small studies performed by Parkash Gill, M.D., and colleagues. In the first 24-patient study, three lesions per patient were treated -- one with a placebo, and two with HCG at doses of 250, 500, 1,000 or 2,000 IU (12 lesions at each dose) three times per week for two weeks. A significant dose-effect was seen, with one, five, five and ten of the 12 injected lesions responding at each dose, respectively. The non-responding lesions at the highest were in one patient that stopped treatment after one dose due to pain at the site of injection. Complete tumor regression was observed in five of the lesions treated at the highest dose, two at the 1,000 IU dose, and one each at the two lower doses. However, some of the placebo- treated lesions located near treated lesions had responses as well.
Because of the effect on nearby lesions, the researchers ran a second study with twelve patients who were randomized to have two lesions treated with either 2,000 IU HCG or placebo. None of the placebo patients responded, while ten of the twelve HCG-treated lesions responded. Five out of six lesions were shown to be complete responses by biopsy. In each of the trials, the side effects were mild, and included dizziness, nausea, headache, anxiety and pain at the site of the lesion.
Before conducting the clinical studies, Dr. Gill's team, working in collaboration with Dr. Robert Gallo, screened four commercial HCG preparations to determine which had the greatest anti-KS effect in the test tube. The greatest activity was found in Wyeth-Ayerst's preparation of HCG called A.P.L., but this may not be because it had the highest concentration of HCG.
Dr. Gallo observed, "The active material is not the normal HCG molecule, but something that accompanies it in the crude material, in the pregnant women's urine." (Commercial HCG preparations are derived from the urine of pregnant women.) The active component may be a peptide produced by the degradation of HCG.
"One of the most important concerns about this paper is that people will rush out and start taking HCG. The problem with doing that is that the amount of this crude fraction varies from commercial source to commercial source. Even the most active preparations vary by lot. It probably depends on what stage the urine is collected during pregnancy -- there is probably more of it early in the course of pregnancy,"said Dr. Gallo.
Whatever the active component, it appears to induce apoptosis (a form of cell suicide induced) in the lesion. Dr. Gallo believes that this component probably acts as both a growth and rejection factor early in the growth of the fetus. He also promises that more will be published on the anti-HIV activity of the HCG extract in the near future.
- Theo Smart
In October's Treatment Issues, we looked at the current record of IL-2 research. This immune stimulant may help to ensure stability and even reinforcement of available immunity by increasing the sheer numbers and concentration of existing CD4 cells, but it does little to influence the quality of immune defenses. The response against new opportunistic infections could in this way be strengthened and accelerated, but the response against previously encountered microbes, including HIV itself, likely will show no signs of recovery. HIV presumably has killed off most of the cells that became activated to fight off prior infections, leaving holes in the immune system's repertoire of responses.
But then again, who knows how much IL-2 helps? The data on IL-2 remain extremely sketchy after all these years. Clifford Lane, M.D., at the National Institutes of Health, who has conducted the only long-running IL-2 trial in HIV,1 insists that opportunistic infections only occur at the traditional CD4 counts among people on IL-2 therapy. This statement implies that the OIs are restricted almost entirely to nonresponders since IL-2, when successful, keeps CD4 counts from falling to a perilous level. The basis for Dr. Lane's assertion is unconvincing, though -- only seven actual opportunistic infections among 60 volunteers who started IL-2 at different timepoints in the trial and have been on therapy for varying lengths of time. Meanwhile, there has been little sign of reduction in HIV levels, except perhaps in a few people with early HIV infection -- and even then, only over the very long-term. Since general immune stimulation via IL-2 or other agents has not offered much hope of directly restricting HIV, researchers all along have been considering ways to construct a strong specific anti-HIV immune defense. Therapeutic vaccines are the major strategy that has been proposed to preferentially increase surviving CD4 cells that could orchestrate a new defense against HIV.
After introduction of new foreign proteins (antigen) to the body, naive CD4 cells sensitive to portions of that protein eventually are activated and multiply to construct new immune defenses. The therapeutic vaccine concept is to improve on the immune defense against a particular already-existing infection by inoculating pieces of the infectious agent's protein presented in a way to trigger new naive cell activation. An immune-enhancing adjuvant is frequently used to help this process along. Just this fall, investigators presented evidence that therapeutic vaccines can reduce the extent and duration of genital herpes outbreaks in people with frequent eruptions2 and in infected guinea pigs.3
Most prominent researchers have always been skeptical of the validity of the therapeutic vaccine approach for HIV, though. David Ho, M.D., director of New York's Aaron Diamond AIDS Research Center, echoed a common argument when he quipped, "Vaccines are not promising because the body already sees lots of HIV antigen. Adding a little extra is not likely to make any difference."
There is also a serious objection concerning viral diversity: Even if an induced immune response is effective, won't HIV merely mutate to rearrange the bit of viral protein that triggers the attack?
And indeed the field of therapeutic vaccines for HIV is littered with failures, despite clear demonstrations that the vaccines provoke new immune responses against HIV. The most extensively tested product has been a gp160 (HIV envelope protein) inoculant made by bioengineered insect cells, developed by MicroGeneSys. This gp160 had the amino acid sequence, but not the folded structure, of the envelope found on a particular laboratory strain of HIV (LAV). Periodic injections in people with early HIV infection were found to induce new varieties of anti-gp160 antibodies as well as strong immune cell proliferation when white blood cells taken from inoculated HIV-positive volunteers were exposed in the test tube to more MicroGeneSys gp160.4 Though clearly "immunogenic" the MicroGeneSys vaccine completely flopped in a placebo-controlled trial coordinated by the Walter Reed Army Institute of Research. After following 608 volunteers (with a starting CD4 count of at least 400) for three to five years, researchers concluded that the bimonthly injections of the vaccine were safe, but no difference existed between the placebo and vaccine groups in terms of occurrence of opportunistic infections, drop in CD4 count, rise in plasma HIV levels or other measures of disease progression.5 A similar 278 person, three-year Canadian trial had equally negative results.6
Recent reports at various conferences have also detailed the failures of three other, more naturally structured vaccine products (Immuno AG's gp160 produced by mammalian cell cultures, a hybrid canary pox virus with a gp160 envelope, and British Biotech's virus-like particles consisting of yeast protein particles coated with HIV p24 core protein). A published report on an earlier trial involving Genentech's gp120 (an envelope component) therapeutic vaccine also just appeared.7 The trial, involving 573 volunteers with initial CD4 counts over 600 and followed for 15 months, also could detect no difference in terms of CD4 decline or HIV plasma viral load reduction between the vaccine and placebo groups.
Yet hope springs eternal in those interested in this approach. After the failure of the MicroGeneSys vaccine, Fred Valentine, M.D., of New York University commented, "We still have an enormous amount to learn about what makes a protective immune response. I'm willing to have a more open mind than most folks. If these vaccines don't work, let's try some new ones."
One of the surviving therapeutic vaccines is the so-called Salk immunogen (sponsored by the Immune Response Corporation under the brand name Remune). First proposed ten years ago by Jonas Salk, this vaccine consists of inactivated HIV stripped of its envelope. HIV's internal proteins are more conserved (less variable), making for a vaccine capable of creating immune responses that interact with a wide variety of HIV strains and perhaps less subject to mutational escape by the virus. Surrounding the HIV particles in the vaccine is a special mineral oil adjuvant known as Incomplete Freund's Adjuvant (IFA) that is supposed to increase processing of the vaccine material by the immune system. Past trials have shown at best slight improvements in HIV levels and percent of CD4 cells along with enhanced response to p24 antigen over a one year period8 (in 103 volunteers with starting CD4 counts above 550). There has been no proof of real clinical improvement, and one long-term observational report of 25 of the original volunteers who received Remune (a median of eight inoculations over three years) noted that eight had died. But unlike the sorry state of IL-2 testing (see Treatment Issues November 1996 article, Outrunning HIV to Protect Immune Defenses, Remune's corporate sponsor has invested in a large clinical endpoint trial to confirm the product's value.
This trial, launched after protracted controversy with the FDA over manufacturing and protocol details, will include 3,000 HIV-positive participants with CD4 counts of between 300 and 549. They will be vaccinated every three months for three years with either the Remune vaccine or IFA alone (the control arm). Even if Remune proves to have little benefit, the trial is not very risky for participants in that they may take any antiviral therapy they please, and Remune seems to have little toxicity. Other immune-based therapies, such as IL-2, are not allowed. It should be noted, too, that enrollment in the Remune trial will make individuals ineligible for most other trials. A small preliminary pediatric trial and an adult expanded access program (for those who cannot participate in the clinical endpoint trial) also exist -- for further information on any of these projects call 800/684-8624.
In Los Angeles, the AIDS Research Alliance is in the middle of a 24- person, one-year trial of "HGP-30" in people with HIV and CD4 counts ranging from 50 to 750. HGP-30 is a portion of HIV p17 core protein that is considered particularly immunogenic and little subject to mutational alterations. This 30 amino acid peptide is joined to "keyhole limpet hemocyanin," or KLH, which originated in a type of mollusk and contains many sequences of its own that are highly immunogenic. In a study presented at last summer's International Conference on AIDS, researchers reporting on a preventive HGP-30 vaccine in 11 HIV-negative volunteers found evidence (in cell culture assays) of increased anti-HIV antibody production or T-cell proliferation in six and four of the volunteers, respectively.9 Although KLH's broad immune stimulation may be helpful in a preventive vaccine, that stimulation risks triggering heightened HIV replication when used in a therapeutic vaccine.
Another pilot therapeutic vaccine trial is winding up at Duke University and is sponsored by the National Institute of Allergy and Infectious Diseases' special program for expediting innovative therapies, the Division of AIDS Treatment Research Initiative (DATRI). The vaccine agent consists of a special mixture of peptides representing known immunogenic sequences (epitopes) from two regions of gp120 (C4 and V3) plus the same IFA adjuvant used in Remune. The selection of the peptides is based on a theory that antibody and cytotoxic lymphocyte (CTL) responses to certain HIV epitopes in humans suppress other immune responses that are necessary later on, as HIV alters its envelope protein to escape immune attack. This phenomenon, observed in experiments with mice,10 may be the reason that the immune system eventually loses control of HIV. The C4-V3 peptide mixture is designed to avoid this occurrence while containing epitopes representing the major varieties of HIV found in North America and Europe. An analysis of the trial results should be available this winter, but the investigators are skeptical: "The likelihood of success with therapeutic vaccines is decreasing," said John Bartlett, M.D., of Duke.
Inclusion of the new highly active antiretroviral drug combinations (HAART) could be essential to the success of therapeutic vaccines. As is probably also true with IL-2, HIV replication may have to be maximally suppressed before the cells responsible for new immune responses can replicate without impediment and without stimulating additional HIV replication. Maximal HIV suppression could prevent the appearance of escape mutants no longer susceptible to the therapeutic vaccines. HAART's virtual elimination of HIV might be necessary before the benefits of the new immune responses elicited by the vaccines will have an observable effect -- they might be best at preventing breakthrough infections by drug resistant virus.
There are at present two upcoming trials that plan to combine a therapeutic vaccine with HAART. At New York University, Dr. Valentine has drawn up a protocol with the Remune vaccine plus AZT, 3TC and the protease inhibitor indinavir. He commented, "This trial is directed toward basic immunogenicity questions in people who are quite healthy. We are asking whether you get new and better immune responses with less HIV, the logic being the HIV screws up the immune response." The eight-month study, expected to commence this winter, will enroll 150 participants randomized to receive either Remune alone, IFA plus AZT/3TC/indinavir, or Remune plus the triple combination.
A second exploratory trial has already started at Stanford University and when fully enrolled will include 30 HIV-positive volunteers with CD4 counts of between 200 and 400. After a nine-week introduction to a d4T/ddI/ritonavir combination, the trial participants will be assigned to receive either the vaccine or placebo for a total of seven monthly injections. This trial, ACTG 246, also has a placebo-controlled, 16-person subset for people with CD4 counts over 500, and these volunteers will not be taking anti-HIV drugs. The vaccine administered in this trial, Immuno AG's gp160, is similar to one reported this fall to have failed in a 208-person trial covering a similar population.11 A higher dose and a different version of gp160 (based on a more representative strain of HIV) are being used this time.
In the completed trial there was a tendency for vaccine recipients on anti-HIV drugs to experience deeper reductions in plasma HIV levels than the reductions experienced by those taking anti-HIV drugs alone. The Stanford trial started out as a study of how HIV mutates around immune defenses. HAART was added to the protocol when it became available to see how greatly boosting CD4 count improves immune responsiveness.
Finally, researchers at the Food and Drug Administration recently announced further tests in mice of a new concept for a therapeutic vaccine.12 The HIV-derived immunogen is a segment of HIV gp120 protein (an 18 amino acid peptide from the V3 loop) that contains epitopes capable of exciting both antibody and CTL responses. This amino acid string is coupled to the walls of heat-inactivated Brucella abortus bacteria (a cow pathogen that causes undulant fever in humans). This bacterium is capable of eliciting both antibody-producing B-cells and CD8 cells (including CTLs) that attack virus- infected cells in a manner partially independent of the CD4 T-helper cells. (Some other bacteria can do this too; inactivated B. abortus just happens to be relatively innocuous in humans. This type of immune stimulation is a result of properties of the polysaccharides on bacterial cell walls.)
If it works, the Brucella abortus based vaccine might have special therapeutic properties in people whose immune response to HIV is waning due to the virus-induced destruction of their CD4 T-helper cells. The latest mouse experiments found that vaccination created a pool of anti-HIV CTLs even in mice whose CD4 T-helper cells had been artificially depleted. Previous experiments had also found production of both interferon gamma and certain types of antibodies (immunoglobulin G2a) that are associated with the "TH1" immune profile. This cytotoxic lymphocyte oriented pattern of immune response is thought to be particularly effective against HIV. Many questions obviously remain concerning the safety as well as the formulation of this type of vaccine. It is not clear which HIV-derived peptides coupled with Brucella abortus, if any, would achieve an effective defense against real HIV infection in humans. Here, too, there is no way to rationally design a therapeutic vaccine, since what constitutes a protective immunity against HIV remains unclear.
The immune system suppresses other viral infections, though, and it seems to contribute to the longevity of most HIV-infected long-term nonprogressors. Perhaps further observation of such nonprogressors will indicate how to improve the naturally arising immune defense against HIV to create an immunity analogous with what develops in other diseases. Despite generating new immune responses, therapeutic vaccines' record in controlling HIV has been unsuccessful so far. This failure does not rule out other attempts using either radically altered versions of present vaccine designs or completely different strategies -- if only there was some guide as to how to proceed.
1. Kovacs JA et al. New England Journal of Medicine. Oct. 31, 1996;
2. Straus SE et al. Thirty-sixth Interscience Conference on Antimicrobial Therapy and Chemotherapy. September 15-18, 1996; abstract H083.
3. Harrison CJ et al. Thirty-sixth Interscience Conference on Antimicrobial Therapy and Chemotherapy. September 15-18, 1996; abstract H121.
4. Valentine FT et al. Journal of Infectous Diseases. June, 1996; 173(6); 1336-46.
5. Birx D et al. Eleventh Interantional Conference on AIDS. July 7-12, 1996; (abstract TU.A.275).
6. Tsoukas C et al. Eleventh Interantional Conference on AIDS. July 7-12, 1996; (abstract Tu.A.274).
7. Eron J et al. Lancet. Dec. 6, 1996; 348(9041):1547-1551.
8. Trauger RJ et al. Journal of infectious Diseases. June, 1994; 169(6):1256-64.
9. Kahn J et al. Eleventh International Conference on AIDS. July 7-12, 1996; (abstract Th.B.944).
10. Golding Basil et al. Journal of Virology. June, 1995; 69(6):3299-33017.
11. Goebel F.-D. et al. Third International Conference on Drug Therapy in HIV Infection. Nov. 3-7, 1996; (abstract OP6.4).
12. Lapham C et al. Journal of Virology. May, 1996; 70(5):3084-92.
The 1996 elections will mean big political changes for AIDS in Washington. The balance of power between the parties remains the same, but mass retirements in the Senate and shake-ups at the committee level mean a more difficult climate for AIDS in the new Congress. This session of Congress, the 105th, will no doubt revisit all the unresolved AIDS controversies of the last one.
Debates likely to take place include the structure and final budget authority for AIDS research at the National Institutes of Health, where the preservation of the independent Office of AIDS Research in its current form is in question. FDA reform is certain to be taken up again, too, and especially with FDA commissioner David Kessler resigning, an outcome that favors rapid drug development with little consideration for patient protection is all the more probable. The struggle over providing adequate, ensured funding for the state AIDS Drug Assistance Programs and for Medicaid, which are facing increased demands due to the new HIV drugs, also will continue. (And remember that the new FDA commissioner will have to be approved by the Senate Labor Committee and then the full Senate.) Lastly, AIDS advocates shiver over the prospect of having to fight the Coburn Bill. Introduced into the last session of Congress by Rep. Tom Coburn, Republican of Oklahoma, who is expected to introduce it again in some form, the Coburn Bill contains a slew of measures for mandatory HIV testing and registration of those who test positive.
The Senate has changed the most since the elections. Republicans increased their majority in this chamber by two seats. The split is now 53-47, with Democrats maintaining power to protect a filibuster by blocking a cloture vote, which requires 60 votes. Many moderate Republicans and AIDS allies retired from office, making this session's Senate more ideologically conservative than its predecessor. In total, 14 senators retired in 1996, resulting in the highest number of "open seats" since the 1914 elections.
Two committees with jurisdiction over most AIDS programs, the Appropriations and Labor committees, have new chairmen and several new members. Since the vast majority of legislative action is made at the committee level, these changes will have a huge impact on AIDS issues.
The Labor Committee is the main authorizing committee in the Senate for most AIDS programs, including AIDS research, the Ryan White CARE Act, AIDS prevention, etc. Authorizing committees create federal programs, and determine their structure and purpose. Formerly chaired by moderate Republican Nancy Kassebaum of Kansas, who retired, the committee now will be lead by James Jeffords of Vermont. Sen. Jeffords appointment is perhaps the best news AIDS advocates have had all month. He is a liberal Republican, an AIDS ally, and he often bucks his party's leadership on important issues.
Other developments on the Labor Committee are not so welcome. Three new Republican members on the committee, Mitch McConnell of Kentucky, Mike Enzi of Wyoming, and Tim Hutchinson of Arkansas, are very conservative and not known for their leadership on AIDS. (Neither New York nor California, the two states with the largest AIDS epidemics, are represented on the Labor Committee.) Also, a new Public Health Subcommittee, which may have jurisdiction over some AIDS programs, has been created. It will be lead by Bill Frist, a conservative Republican from Tennessee, who also happens to be a heart surgeon.
The Appropriations Committee is the other key committee for AIDS. This committee controls the federal purse strings, allocating money to federal programs each year. Formerly chaired by moderate Republican Mark Hatfield of Oregon, another retiree, it is now led by Ted Stevens, a conservative Senator from Alaska. AIDS advocates will have their work cut out for them trying to influence Sen. Stevens. Sen. Stevens is joined on the committee by three conservative colleagues: Kay Bailey Hutchinson of Texas, Larry Craig of Idaho, and Lauch Faircloth of North Carolina. Barbara Boxer, the California Democrat and a leader on AIDS issues, will join the committee minority. Arlen Spector, a moderate Pennsylvania Republican, will keep his chairmanship of Appropriation's Labor Subcommittee, which is also good news for AIDS advocates.
House Republicans were chastened in this election, but they still maintain a majority. The final tally in the House is 227 Republicans and 207 Democrats, the slimmest margin in the House in the past 40 years. Several right-wing members, particularly from the Northeast and West Coast were defeated. The best news in the House races was the narrow victory of Democrat Linda Sanchez over Bob Dornan in Orange county, California, ending a long and ugly legislative career. A would-be Presidential candidate and the House's leading homophobe, "B-1 Bob" was known both for his reactionary theatrics and his devotion to expensive military gizmos. His departure will remove a source of continuous tension in the House.
Fewer changes have occurred in the House committees than in the Senate. The two committees crucial for AIDS in the lower chamber, the Commerce and Appropriations Committees, retain most of their old character. The leadership of both remains unchanged.
The Commerce Committee is the authorizing committee in the House for the majority of AIDS bills. Thomas Bliley, a Republican from Richmond, Virginia, will stay in charge of the committee, and most key members won re- election. Henry Waxman, the Los Angeles Democrat, stepped down as the ranking member on the Health And Environment Subcommittee, which oversees most AIDS programs, for the ranking post in the Government Oversight Subcommittee. This is probably the biggest blow for AIDS advocates. His AIDS leadership at the subcommittee level will not easily be replaced. Meanwhile, Eliot Engel, a Bronx Democrat and another AIDS advocate, joined the committee minority, increasing New York City's representation on the committee to three.
The House appropriations committee is also largely unchanged. Bob Livingston, the New Orleans Republican, remains as chairman, although he may face a leadership fight from fellow Republican Joseph McDade of Pennsylvania. Rep. McDade was forced to step aside as chairman in the last Congress because he was under indictment for campaign finance violations. His recent acquittal of these charges means he is eligible again for a leadership post.