Slow Progress On HIV Vaccines
by Luis Santiago
This winter's Conference on Advances in AIDS Vaccine Development in Bethesda, Maryland was the eighth such meeting sponsored by the National Institute of Allergy and Infectious Diseases (NIAID). The yearly event brings together investigators from NIAID's HIV vaccine testing centers, these centers' community advisory boards, independent researchers and pharmaceutical companies.
Important new developments in basic science were announced this year, in areas like the mucosal immunity and vaccine delivery mechanisms. Researchers were excited about a new product, the canary pox-gp120 prime-boost vaccine, which seems to trigger an impressive cellular immune response not seen with previous candidates. But the extremely short list of vaccine candidates under development and the paltry attendance by pharmaceutical companies signaled a lack of interest and investment in the field of vaccines.
One manifestation of this lack of input is that some of the most central issues remain unresolved. Researchers have yet to agree on the so-called "correlates of protection," what sort of measurable immune response -- blood-borne versus mucosal, antibody versus cellular -- a vaccine should trigger to confer protective immunity to HIV. Vaccines against other diseases have been developed without settling the "correlates" question, but human testing of HIV vaccines remains bogged down by this controversy. Meanwhile, scientists have not created a generally accepted animal model for more direct testing of vaccine-generated protection.
Neutralizing Antibodies in Chimps and Humans
In June, 1994, following a recommendation of a special advisory panel, NIAID decided not to proceed with large-scale human testing of two vaccine products, both of which were genetically engineered versions of the HIV envelope protein gp120. Among the panel's most important arguments was the finding that the antibodies generated by these products were unable to neutralize "primary isolates" of HIV, that is, virus obtained from infected people (not laboratory strains).
At the time, questions were raised about the validity of the test used to measure viral neutralization. Notably, this assay could not detect neutralizing antibodies in the blood of several vaccinated chimpanzees that were shown to be protected from HIV infection when injected with the virus.
At the vaccine conference, Susan Zolla-Pazner, Ph.D., from New York University Medical Center, reported the results of two blinded studies involving a new, more sensitive test. Both the new and the conventional neutralization assay work by mixing the virus with antibodies and putting them in cell cultures to determine if, and to what degree, the presence of antibodies prevents the infection of cells. The conventional assay adds PHA to the cultures, which powerfully stimulates cell proliferation and makes the cells much more prone to infection with HIV. The new assay, known as the "resting cell assay," uses no such stimulant.
With the new test, Dr. Zolla-Pazner's group showed that antibodies present in the blood of three vaccinated, protected chimpanzees did indeed neutralize HIV and that these antibodies persisted for more than a year. Furthermore, her data demonstrated that there is an "absolute" correlation in the chimpanzees studied between protection from HIV and the presence of such neutralizing antibodies.
In the second study, blood from sixteen human HIV-negative recipients of several gp120-based vaccines was tested using the resting cell assay to detect neutralization against an HIV-1 subtype B primary isolate. A vaccine recipient showing greater than 50 percent neutralization (at a specific dilution level) was defined as "reactive." Although studies with the conventional assay have never been able to detect neutralizing antibodies for primary isolates, ten out of sixteen or 62 percent of the vaccine recipients were reactive based on the resting cell assay results.
It is now clear that the gp120 vaccines can induce neutralizing antibodies. Whether the neutralizing antibodies protect against infection may depend on whether CD4
cells are in an activated or resting state when exposed to HIV. However, the conventional assay may not accurately reflect the immune system even when activated, so it may be too high a standard for any vaccine.
The possibility of reconsidering the negative decision made two years ago remains slim, mainly because the gp120 products alone do not elicit significant cellular responses, as some of the new candidates do. Still, Genevax, a Genentech spin-off, and Chiron/Biocine, which own the two major gp120 products, continue to collaborate with Thai and U.S. authorities on large trials in Thailand.
Infections despite Vaccination
The gp120 vaccines are clearly not completely protective. There have been widespread reports of documented new HIV infection among participants in the existing gp120 vaccine studies. A preliminary analysis of these cases of acute HIV-1 infection in individuals participating in Phase I and II trials of the gp120 vaccines was presented in January at the Third Conference on Retroviruses.
Biochemist John P. Moore from the Aaron Diamond AIDS Research Center in New York described nineteen trial participants who had become infected out of 596 total enrollees. In fourteen of these cases, individuals had received the full course of immunization with four receiving the Genentech MN strain gp120 vaccine, four the Biocine SF-2 strain gp120 vaccine and the remaining two getting placebo.
Researchers concluded that the vaccine-induced antibodies were transient, type-restricted and of lower magnitude than those induced by actual HIV infection. In one case presented as an example, the individual was infected right after a booster shot, which is usually when the level of antibodies is highest. All vaccine breakthroughs were typical of community isolates, genetically different than MN and SF-2. No effect of pre-vaccination on viral load has been detected so far, and there were no special qualities that distinguished these cases from non-vaccinated controls.
The ALVAC Vaccine
Many researchers believe that a successful HIV vaccine must be able to stimulate at least two different kinds of immune responses: a humoral response, which is essentially the production of antibodies, and a cellular response, mainly the production of cytotoxic lymphocytes (CTLs). Antibody production helps the body rid itself of free virus while CTLs destroy infected cells.
The vaccine products furthest ahead in clinical studies (Genentech's and Chiron/Biocine's versions of gp120), stimulate strong antibody responses, but they do not elicit the production of CTLs. A new vaccine product, recently renamed the ALVAC vaccine, seems to be able to elicit both a cellular and a humoral response. This vaccine consists of Pasteur-Merieux-Connaught's live canary pox virus (harmless to humans) carrying portions of several HIV genes (for eliciting the cellular response) plus a "booster" of Chiron/Biocine's gp120 (to trigger the humoral response).
ALVAC is still in early safety and immunogenicity studies, but the data have created a lot of enthusiasm among the vaccine research community. By the time of the vaccine conference, close to 50 percent of the trial participants in one of the studies had new cellular responses to HIV. If this pattern is sustained, it is very likely that a larger Phase II study will be initiated in the U.S. early next year, with even larger Phase III studies in 1998.
HIV usually enters the body through the mucosal surfaces lining the reproductive and gastrointestinal systems. It is likely that CTLs in the bloodstream are not enough to achieve full protection from sexual exposure to HIV. The activation of specific mucosal responses will be required, too.
A study reported by Max Essex, director of Harvard University's AIDS Institute, underscored the importance of ensuring mucosal protection, whether through vaccines or topical anti-HIV microbicides. Researchers experimented with isolates of HIV-1 type B recovered from gay Americans and type E recovered from heterosexually infected Thais. The scientists introduced these two isolates into two different immune cell cultures: peripheral blood mononuclear cells (PBMCs, mostly lymphocytes and macrophages) and Langerhans' cells (an antigen presenting cell existing on the skin and oral and genital mucosa, but absent from the rectal mucosa).
Both HIV subtypes replicated similarly in PBMCs, but type E alone replicated at particularly high rates in the Langerhans' cells. Since Langerhans' cells are located on mucosal surfaces, they can be infected there directly. PBMCs are subject to viral attack only if HIV reaches lower tissue layers, most commonly by crossing mucosal membranes through abrasions or lesions.
Scientists have identified a specific type of antibody, known as IgA, as the main actor in defending mucosal surfaces. When an infectious agent reaches certain areas in the mucosal membranes called Peyer's patches, cells known as the M cells initiate a complex process resulting in the massive production and secretion of IgA. These antibodies migrate to other mucosal tissues where they block mucosal infection and eliminate the invading pathogen before it can enter the bloodstream.
Studies in monkeys show that systemic HIV immunity (usually accomplished by injecting a vaccine into the blood) does not automatically grant protection from mucosal infection. In human studies, systemic immunization has failed to produce HIV-specific IgA in external secretions, particularly saliva. In order to develop vaccines that can effectively trigger this line of defense, a vaccine must be able to reach the Peyer's patches and stimulate the M cells to generate IgA against HIV. Researchers are studying different vaccination routes (oral, nasal, vaginal, rectal), as well as novel delivery mechanisms (microspheres, bacteria/viral vectors, liposome encapsulation) that can effectively do this.
Oral and Lymph Node Vaccines
A study presented by Marian R. Neutra, Ph.D., from Harvard Medical School and Children's Hospital, looked into the potential benefits of rectal or vaginal vaccination, since it is thought that oral immunization may not give optimal rectal and vaginal mucosal protection. In mice, rectal vaccination increased the induction of secretory IgA in both the rectal and the vaginal/cervical regions. But only vaginal, not rectal, immunization raised IgA in women's reproductive tract.
AVEG (AIDS Vaccine Evaluation Group) Study 023 is a human trial focusing on mucosal immunity. In this safety/immunogenicity study, participants are "primed" with an intramuscular injection of United Biomedical's peptide immunogen followed by oral boosting with the same peptide combination encased in time-release "microparticles." The trial's main objective is to compare this regimen to other oral/intramuscular and oral/oral regimens previously studied.
All 35 participants in AVEG 023 had mucosal reactions, and some anti-viral responses were found in pre-seminal fluid and feces. Two other AVEG mucosal immunization protocols (027 and 028) are planned to evaluate different vectors and routes of administration and their effect on mucosal immune responses.
There also has been positive experience injecting vaccines directly into lymph nodes to stimulate mucosal immunity. This technique is known as "targeted lymph node immunization," or TLN. A British study described at the conference administered a vaccine against SIV to thirteen macaque monkeys. Seven animals were inoculated in the iliac (groin) lymph nodes, three intradermally, and three nasally/rectally. Eight monkeys functioned as controls.
All animals were rectally exposed to SIV. Of the seven animals inoculated directly into the lymph nodes (TLN vaccination), four resisted infection and the remaining three had low viral loads. None of the animals vaccinated via other routes, and just one of the untreated animals were able to resist infection. The TLN-vaccinated animals had higher antibody levels in rectal washings and urine and far higher T-cell proliferation than those vaccinated by other means.
New Goal: Prevent "Disease," Not Infection
The traditional view of an HIV/AIDS vaccine required that it provided "sterilizing immunity," which means that it would prevent an individual from becoming infected with the virus. Vaccines for other viral diseases only prevent the development of acute illness, not infection itself, but this approach has been perceived as extremely risky in the case of HIV.
Data from studies of HIV plasma viral load (level of virus in the blood) and its relationship to disease progression, have led some to openly challenge this fear. At the vaccine conference, John Mellors, M.D., of the University of Pittsburgh, presented an analysis from the Multicenter AIDS Cohort Study (MACS) relevant to this debate. The MACS analysis looked at the HIV "setpoint," which is the HIV level attained after primary infection and reflects the balance reached between virus activity and the immune response.
The MACS investigators observed that a high HIV setpoint is associated with rapid disease progression, while a low setpoint is associated with slow progression. If a drug or vaccine is developed that can push down this initial steady state, disease progression might be slowed to a very low rate, even though an individual would remain chronically infected. Since viral load also is associated with infectiousness, a vaccine that merely limited the HIV setpoint would have substantial epidemiological impact by reducing the rate of HIV transmission.
The most immediate implication for HIV vaccine research is the possibility of using plasma viral load as an endpoint measurement for vaccine evaluation studies in primates. A candidate product that does not prevent infection but keeps viral load at a minimum would qualify for further testing in humans.
From "Vaccine Trials" to "Preventive Trials"
In another retreat from the absolutism of sterilizing immunity, NIAID is now talking of "preventive trials" as opposed to "vaccine trials." Purportedly, this new emphasis is a recognition that behavioral modification and physical and chemical barriers are also keys to preventing HIV infection.
Peter Piot, from the United Nations AIDS Program, reported at the conference that the acceptance of female condoms has been higher than expected and called for a stronger push for their distribution. U.S. AIDS Policy Coordinator Patsy Flemming emphasized the need to strengthen research on vaginal or rectal microbicides against HIV.
Dr. Sharon Hiller from the University of Pittsburgh reported on the vaginal ecosystem and its impact on HIV infectivity. She identified several species of lactobacilli, a type of bacteria present in healthy vaginal environments, whose presence is directly correlated to a decreased incidence of gonorrhea or bacterial vaginosis. Moreover, these lactobacilli (L. crispatus and L. jensenii) produce natural microbicides that have been shown to kill HIV.
Studies of commercial sex workers in Thailand have found an association between the lack of these lactobacilli and increased prevalence of HIV. In the U.S., it has been found that African-American and Latino women are more likely to lack these lactobacilli, and Dr. Hiller is developing a program that offers them along with physical and chemical barriers, such as condoms and nonoxynol-9 (N-9). Dr. Hiller stressed the need to insure that microbicides do not adversely affect the vaginal ecosystem. High doses of N-9 have been found to cause genital ulcerations which increase the risk of HIV infection.
Dextran sulfate, a sulfated polysaccharide, is another candidate microbicide. It is currently in safety studies, both alone and in combination with nonoxynol-9. A second candidate mentioned at the vaccine conference was the buffer gel known as Reprotect, which lowers pH levels in the vaginal walls (lower pH, or greater acidity, is thought to kill HIV). A large number of potential microbicides including other sulfated polysaccharides and extracts from pomegranates, are in preclinical testing.
Partnership with Industry
When the decision was made in 1994 not to go ahead with large-scale trials of gp120, some of the companies involved complained of being "cheated" by the government. They had invested money under the assumption that their vaccines would be "appropriately" tested if they showed any potential. These complaints have little merit since the government's refusal to proceed had strong scientific rationale. It is a fact, though, that many companies are wary of entering the AIDS vaccine field without some safeguards for their investments. A number of companies have dropped out of the field altogether.
In his Special Plenary lecture at the vaccine conference, NIAID chief Anthony Fauci, M.D., called for "nontraditional partnerships" between government and industry to give new impetus to HIV/AIDS vaccine research. NIAID's strategy mainly consists of establishing concept-specific milestones or targets through negotiation between government and industry. If the targets are achieved, the product moves to the next phase.
The new ALVAC vaccine has become a test case for this strategy. NIAID, Pasteur-Merieux and Chiron/Biocine have an agreement to move ALVAC into larger Phase II trials if at least 90 percent of the participants in the initial Phase I studies develop humoral responses (antibodies), and at least 30 percent develop cellular responses (CTLs). At least one study must show too that the ALVAC regimen protects nonhuman primates (gp120 by itself already has protected chimpanzees from getting infected with HIV).
In a follow-up meeting at the While House, Vice President Al Gore held a discussion with NIAID officials and representatives from eleven pharmaceutical and biotech companies on ways to speed up the development of vaccines, microbicides and other anti-HIV products. Whether these negotiations lead to a real turnaround remains to be seen. The stakes are enormous for the millions of people around the globe at risk of contracting the disease, most of whom will never be able to afford the extremely expensive antiviral treatments now entering the AIDS marketplace.
The (Viral) Burden of Motherhood
by Dave Gilden
One of the lasting controversies emanating from the Third Conference on Retroviruses and Opportunistic Infections held this winter (see Treatment Issues, February, 1996) was the relationship between a pregnant woman's plasma HIV levels (viral load) and transmission of the virus to her baby. With the Roche PCR and Chiron bDNA assays for measuring viral load under review by the FDA (see Treatment Issues article Pediatric Treatment Update; March 1996), viral load assessment may soon play a major role in patient management. The disagreements that unexpectedly arose over this reputed association have important implications for how well these tests can be relied upon for predicting the future course of disease and the influence various treatments will have on that course.
Four studies on mother-to-child (vertical) HIV transmission were presented at the Conference, with three so newly completed that they appeared among the event's "late breakers." These studies looked at whether higher HIV levels in blood plasma correlated with greater mother-to-child transmission. They sought to ascertain whether a threshold viral load existed below which HIV transmission did not occur. A related question was also under examination: whether an AZT-induced reduction in a woman's viral load leads to a reduced risk of her baby contracting HIV before and during birth.
The first report (abstract S25) was presented by Yvonne Bryson, M.D., of the University of California Los Angeles, and has recently been published.
1The UCLA investigators found that women who transmitted HIV to their babies prior to or during birth tended to have higher viral loads at delivery than those that did not (a median of 94,000 copies of HIV RNA per milliliter of plasma for the transmitters versus 4,600 copies/ml for the nontransmitters -- a highly significant difference). No one with a viral load below 25,000 copies/ml transmitted HIV, and a viral load at delivery of greater than 50,000 was the single best predictor of transmission among the many variables analyzed. AZT-treated nontransmitting mothers generally experienced reduced viral loads during pregnancy (an eight-fold, or 0.9 log, median decrease), and their viral loads at delivery had a distribution similar to that of untreated nontransmitters.
The three late breaker presentations were not so sanguine about viral load's predictive power. Richard A. Koup, M.D., reporting for the Pediatric AIDS Foundation's Ariel Project, described the findings that viral load was a poor predictor of HIV vertical transmission in a group of 151 pregnant women (LB2). The transmission rate in this cohort was thirteen percent. (Eighty-three percent of study participants were taking AZT.) There was a trend toward higher viral load in the HIV-transmitting mothers -- transmitting mothers' viral loads averaged 5,200 copies/ml, compared to 2,600 for the nontransmitters -- but there was a large overlap between the two groups and the difference was not quite statistically significant. Vertical transmission occurred even in women with HIV levels so low that their virus was unculturable.
The Woman Infant Transmission Study Group (WITS) also found that there was only a weak association between viral load and vertical transmission (abstract LB3). Transmission was seen at all plasma HIV levels, except under 1,000 copies/ml, and although transmitting mothers had slightly higher viral loads than nontransmitting ones at delivery, this difference was not statistically significant.
Rhoda Sperling, M.D., (abstract LB1) summarized the long-awaited virologic results from ACTG 076, a study of reducing mother-to-child HIV transmission by administering AZT to pregnant women and, starting immediately after birth, to their babies. Use of AZT in this double-blind, placebo-controlled trial reduced the rate of vertical transmission by about 70 percent (see Treatment Issues, March 1994, pages 15-16). In both the study group receiving AZT and that receiving placebo, HIV transmission took place at all detectable viral loads (but not in women with undetectable levels of HIV). The pregnant women receiving AZT saw their viral load drop by a median of only 0.2 log (37 percent), and differences in the magnitude of this reduction did not correlate with any difference in transmission rate.
The 076 team concluded that the fall in HIV levels with AZT could not explain the entire decline in mother-to-child transmission of the virus. Dr. Sperling conjectured that the AZT given in 076 to the newborn babies, as a kind of immediate post-exposure prophylaxis for those babies contracting HIV during delivery, played a major role in lowering transmission rates.
One limitation of these last three studies is that the observed viral loads in the women they followed were remarkably low (the means or medians were below 10,000 copies/ml). Any differences in HIV levels were relatively small, as was any effect AZT could have on these levels.
But the measured viral loads may not have been accurate. They could have been reduced by the way blood samples were processed: the three late breaker studies all used heparin as an anticoagulant in the blood samples collected. Dr. Bryson contends that heparin results in an immediate 30 to 38 percent loss in viral load values. (Dr. Bryson's group at UCLA processed their samples mostly with EDTA.) Storage of samples could also be a problem: a delay in separating plasma from whole blood can lead to lower viral loads as can a postponement in freezing the processed samples. (The Ariel Project shipped its samples to New York for overnight processing whereas the 076 and WITS studies batch-processed their samples well after the studies were terminated.)
Finally, researchers are attempting to divide up mother-to-child transmission by stage -- during pregnancy (in utero) or during labor (intrapartum). In utero HIV transmission is considered to have taken place if the infant is PCR- or culture-positive for HIV within 48 hours after birth, according to Dr. Bryson's working definition, for example. Such early transmission seemed to be promoted by high maternal viral load in her study, as it did in the WITS. (Note that a low viral load early in pregnancy is not sufficient to establish low transmission risk since viral load can rise abruptly.)
Intrapartum transmission, a much more common occurrence than in utero, is not necessarily tied to high maternal viral load during delivery. Events during delivery that expose the baby to maternal blood or cervicovaginal secretions might promote transmission. For example, the UCLA group detected a tendency for intrapartum transmission to occur in women whose uterine membranes ruptured well in advance of delivery. This trend did not reach statistical significance, though, and the ACTG 076 investigators did not notice any correlation between vertical transmission and such "obstetrical factors." The WITS researchers are continuing to look at maternal cofactors that promote vertical transmission, and the 076 researchers are now analyzing whether the particular infectivity of a woman's HIV strain influences the risk of transmission during her pregnancy.
- Dickover RE et al., Journal of the American Medical Association. Feb. 28, 1996; 275(8):599-605.
MAC Prophylaxis Revisited
by Gabriel Torres, M.D.
MAC (mycobacterium avium complex) infection occurs in eighteen to 40 percent of people with AIDS. It is most common in those with CD4 counts under 50. The symptoms and signs are quite nonspecific and protean: they can mimic those of many other infections and tumors in patients with advanced AIDS.
Several lines of evidence seem to indicate that a few individuals develop MAC from reactivation of an old infection. The preponderance of evidence, however, indicates that most individuals develop MAC following recent acquisition of a virulent MAC strain. Epidemiological studies have found evidence that many patients who contract MAC had new exposure to the bacterium from environmental sources such as tap water or soil.
1The sum of this evidence is inconclusive, and there is no recommendation as to how an individual can avoid exposure to MAC from environmental sources.
The MAC bacteria infects the respiratory system and the gastrointestinal tract, initially forming colonies without causing symptoms. This asymptomatic colonization can lead to dissemination of MAC through the bloodstream (MAC bacteremia) and more aggressive disease. Many experts support treatment of patients who are MAC-colonized but without signs of dissemination in order to prevent spread of the infection via the bloodstream. Virulent MAC strains seem to bind tightly to cells in the gastrointestinal walls or respiratory tree, invading macrophages and triggering the generation and/or release of such inflammatory cytokines as interferon gamma, IL-2, IL-6, IL-12, GM-CSF and TNF, which may promote HIV replication.
Most of the anti-MAC research effort is focused on prophylaxis with drugs that directly kill the organism. The Food and Drug Administration has approved both rifabutin and clarithromycin (the latter just at the end of last year) for the primary prevention of MAC, as placebo-controlled studies have shown a reduction in the rates of MAC bacteremia and symptoms related to MAC disease.
Clarithromycin vs. Rifabutin
At the Third Conference on Retroviruses and Opportunistic Infections this winter, the results of the first MAC prophylaxis study comparing clarithromycin, rifabutin and the combination of the two were presented by Connie Benson, M.D., from Rush Medical College in Chicago (abstract 205). In this phase III double-blind study, 1,216 volunteers were randomized to clarithromycin (500 mg twice daily), rifabutin (450 mg/day) or a combination of the two. The dose of rifabutin was subsequently lowered to 300 mg/day, due to an unacceptably high incidence (three percent) of uveitis (inflammation of the eye's iris and associated tissue). A second study sponsored by the National Institute of Allergy and Infectious Diseases showed that there was a drug-drug interaction resulting in higher plasma levels of rifabutin when clarithromycin was administered concomitantly.
The results of the comparative study found rates of MAC disease to be nine percent in the clarithromycin arm, fifteen percent in the rifabutin arm and seven percent in the combination arm, a highly significant difference in favor of the clarithromycin-containing arms. The death rates were similar in all three arms. The percentage of patients requiring treatment discontinuation was higher in the combination arm (31 percent) than in either monotherapy arm (sixteen percent for clarithromycin and nineteen percent for rifabutin). The most common side effects were nausea and vomiting, diarrhea, abdominal pain, anemia, neutropenia, thrombocytopenia and elevation of alkaline phosphatase blood levels (indicative of liver inflammation).
Preliminary drug susceptibility analysis for isolates recovered from patients who developed MAC bacteremia during the study indicates that 27 percent, zero percent and 24 percent of isolates from patients assigned to clarithromycin, rifabutin or the combination, respectively, were definitely resistant to clarithromycin. In the same three arms, only four percent, two percent and none of the isolates, respectively, were resistant to rifabutin.
The researchers concluded that clarithromycin alone was more effective than rifabutin alone in reducing the incidence and delaying the time to development of MAC disease. Clarithromycin alone also was as effective as the combination but less toxic. Finally, the use of rifabutin alone did not appear to select for rifabutin resistance while the use of clarithromycin did appear to promote clarithromycin resistance. Combining both drugs did not significantly reduce the rate of clarithromycin resistance.
The true impact of clarithromycin resistance is not clearly understood. Clarithromycin's manufacturer Abbott Laboratories found in a previous prophylaxis trial that those patients who developed clarithromycin-resistant MAC while on clarithromycin had the same survival time as those who developed MAC on placebo. But resistance to clarithromycin creates azithromycin resistance, too. This cross-resistance severely limits a patient's treatment options. A number of studies have shown that effective treatment of disseminated MAC requires a combination of drugs that includes one of these two "macrolide" antibiotics.
Azithromycin Once a Week
A study involving prophylactic azithromycin was also presented at the Third Conference on Retroviruses and Opportunistic Infections (abstract 204). It was conducted by Diane Havlir, M.D., and her colleagues from the University of California San Diego and sponsored by Pfizer, maker of azithromycin. In this study, 669 persons with CD4 counts less than 100 were randomized to receive either azithromycin (1,200 mg once weekly), rifabutin (300 mg/day) or the combination (using the same doses). The rates of MAC disease were 15.3 percent in the rifabutin arm, 7.6 percent in the azithromycin arm and 2.7 percent in the combination arm -- a significant difference favoring the combination. There were no differences in survival or dose-limiting toxicities, although the incidence of side effects (mostly gastrointestinal) was highest in the combination arm (22.7 percent). Half of the volunteers receiving azithromycin had diarrhea at least once, but only 3.8 percent of those taking the drug had
to discontinue it.
Only two of eighteen (eleven percent) of the MAC isolates obtained from patients in the azithromycin arm were resistant to both azithromycin and clarithromycin. None of the five isolates from the combination arm had resistance to macrolide antibiotics.
This study found that the azithromycin/rifabutin regimen was associated with the lowest rate of MAC breakthrough. The rates of other bacterial infections in this study were similar among the three groups. In contrast, another Pfizer-sponsored, placebo-controlled study of once weekly azithromycin (1,200 mg/week) in patients with CD4 counts below 100 found that azithromycin also protected against other bacterial infections (abstract 203). MAC breakthrough rates in this study were 8.2 percent with azithromycin and 23.3 percent with placebo. Eighty-three percent of the volunteers on azithromycin reported gastrointestinal side effects compared to 44 percent of the patients on placebo. Only 8.2 percent discontinued azithromycin treatment due to toxicity, compared to 2.3 percent of the placebo group.
Clarithromycin and azithromycin are closely related drugs, with similar anti-MAC activity and toxicity. What sets the two apart is their durability in the blood and cells and hence their dosing regimens. Once weekly azithromycin may be both cheaper and more convenient than twice daily clarithromycin. Without a head-to-head clinical trial, though, it is hard to determine whether one regimen is more effective or less toxic than the other, although it is unlikely that this trial will occur. Pfizer anticipates FDA approval in mid-May for azithromycin at the 1,200 mg once weekly dose. The price is expected to be about $1,000 a year.
Deciding on Prophylaxis
The overall conclusion from these MAC prophylaxis trials is that either clarithromycin (500 mg twice daily) or azithromycin (1,200 mg once a week) are the most effective regimens, with a potential survival benefit reported for clarithromycin in an earlier study.
2Yet those with breakthrough MAC while on either of these macrolide antibiotics have a relatively high rate of MAC resistant to both drugs.
Rifabutin is less effective in preventing MAC, but breakthrough isolates have lower rates of resistance. Rifabutin also interacts with many other common drugs used in HIV disease, including protease inhibitors (saquinavir and ritonavir). It can lower other drugs' blood levels, thus decreasing their effectiveness.
MAC prophylaxis is now recommended for patients with CD4 counts below 75. Many clinicians keep patients on prophylaxis even if their counts climb above this threshold due to protease inhibitors or other drugs. Whether effective anti-HIV therapy can make MAC prophylaxis superfluous is one of the great unanswered questions.
- von Reyn CF et al., Lancet. May 7, 1994; 343(8906): 1137-41.
- Pierce M et al. Tenth International Conference on AIDS. Aug. 7-12, 1994; 10(2):70(abstract 558B).
OAR Puts AIDS Research under Scrutiny
by Derek Link
On March 14, a federal scientific advisory panel released its final report recommending sweeping changes in the nation's AIDS research effort. Under the leadership of Princeton University virologist Arnold Levine, Ph.D., the AIDS Research Program Evaluation Working Group spent fifteen months examining the $1.4 billion AIDS research program at the National Institutes of Health (NIH). Its evaluation of the entire NIH AIDS research program was the first by an expert, non-government body.
The NIH AIDS effort, which incorporates funding for outside investigations as well as for projects within the Institutes' walls, accounts for more than 85 percent of the world's publicly funded AIDS research. Twenty-four separate programs, conducted by each of the NIH's independent research institutes, make up the overall undertaking. The AIDS program at the National Institute of Allergy and Infectious Diseases (NIAID) is the largest and most prominent, but NIAID's endeavors account for only 40 percent of the entire NIH AIDS research budget.
The report assessed how the 24 AIDS research programs fit together and determined whether NIH AIDS research as a whole is moving effectively and efficiently toward an AIDS cure and vaccine.
The Working Group found that the NIH has made substantial progress over the last fifteen years, helping to identify the causative agent of AIDS, establish AIDS research as a scientific discipline, and improve treatment of HIV and associated infections. But a cure and vaccine remain elusive. The panel concluded that the federal government must restructure its AIDS research effort to meet the scientific challenges posed by AIDS.
The evaluation panel was convened by the new NIH Office of AIDS Research (OAR) in late 1994. In 1993, the Congress commissioned the OAR to review how prudently the NIH spent its AIDS research money and strengthened the OAR's authority to coordinate and plan this spending. Since the 1994 Congressional elections, when Republicans won majorities in both houses of Congress, the OAR's authority has been under constant assault. Ironically, just as the Working Group was releasing its findings, Congressional leaders were trying to strip the OAR of its authority to implement the report's recommendations. As Treatment Issues goes to press, the OAR's authority remains an unresolved issue.
Promoting the Nongovernmental Sector
Six themes emerged from the OAR review. First and foremost is the need for stronger oversight and participation by non-government scientists (such as the academic researchers constituting most of the OAR Working Group). Further concerns were better integration and coordination of the NIH's individual AIDS programs; more collaboration between NIH and the drug industry; an increased emphasis on basic research initiatives; and more research on the prevention of HIV transmission, including both biomedical and behavioral approaches. The Working Group's final overarching point was expressed by a strong statement advocating greater involvement of affected communities in the research process.
From these general themes came fourteen specific recommendations:
Increase support for and improve peer-review of investigator-initiated research.
"Investigator-initiated research" refers to research that is developed and conducted by independent scientists. It stands in contrast to "directed research," which is controlled and designed by staff scientists of the NIH. Investigator-initiated research is advanced in the report as the best way to enhance the diversity and productivity of scientific investigations. The report calls for doubling the level of investigator-initiated AIDS research over five years. It notes that only twenty percent of NIH AIDS research is investigator-initiated, far less than the approximately 50 percent that occurs other fields.
Restructure the nation's AIDS vaccine research.
Calling an AIDS vaccine a crucial public health priority, the report calls for a national task force to guide all government AIDS vaccine research programs and urges that distinguished non-government scientists be placed in control of the program. The report sharply criticizes HIVNET, an NIH-funded network of research sites working on an AIDS vaccine, and calls for a comprehensive plan for its organization, governance, research priorities and funding.
Develop an HIV prevention science agenda.
The report suggests the creation of a prevention science agenda that is practical and evidence-based. It says an "ideal" strategy would include behavioral and social interventions and biomedical technologies (topical microbicides, anti-addiction treatments, etc.) in addition to vaccines.
Augment research efforts to better understand the human immune system.
More research into the human immune system is critical for the development of an AIDS vaccine and better therapies. The report recommends more research on human and primate immunity, noting a heavy emphasis to date on mouse immunity.
Integrate all adult clinical trial programs into a single network.
The NIH maintains twelve adult trials networks to test therapies for HIV and related conditions. Among the most prominent are the ACTG and CPCRA at NIAID; the Study of the Ocular Complications of AIDS, or SOCA, at the National Eye Institute; and the AIDS Malignancy Consortium, or AMC, at the National Cancer Institute. Currently, there is no mechanism for cooperation and collaboration between them. The report observes that "there has been overlap and, in some instances, unnecessary competition between these diverse programs that are funded by different institutes with different self-defined missions. The scientific productivity of these independent clinical trials efforts has been quite variable." A single clinical trials network should be run by NIAID with support from other institutes, and the network's scientific leadership should be provided by non-government scientists.
Refocus and restructure the drug discovery research effort.
The NIH drug discovery effort for AIDS includes many different programs in different scientific areas. Some programs, such the National Cooperative Drug Discovery Groups and the HIV structural biology program, are praised by the Working Group. The panel singles out the National Cancer Institute's Developmental Therapeutics Program (DTP) for particular criticism. The DTP annually screens thousands of compounds, both natural and synthetic, for activity against HIV, but the report finds that this vast effort is hobbled by antiquated techniques and a lack of follow through on the potential antiviral drugs it uncovers. The DTP's AIDS resources and enormous collection of compounds should be restructured, serving as a resource for the entire NIH AIDS program.
Augment basic research on AIDS-associated infections and facilitate transfer of new findings for early clinical evaluation.
Although opportunistic infection treatment and prevention have achieved impressive gains over the last decade, few new drugs have been developed specifically for AIDS OIs. (Most OI drugs were first developed for other more common maladies.) The panel concluded that an increased emphasis on the fundamental biology of these pathogens is necessary, given that drug companies have limited incentive to investigate these normally uncommon diseases. Its report specifically recommends greater use of Small Business Innovation Research grants to facilitate transfer of NIH basic science research to the pharmaceutical industry.
Strengthen the scientific base for the assessment of alternative therapies for HIV disease.
The panel found that alternative therapies are widely used by people with HIV and that alternative therapies have the potential to both benefit and harm recipients. To collect more information on the effects of such therapies, the report urges greater collaboration between the AIDS research effort and the NIH Office on Alternative Medicine.
Reorganize the Regional Primate Research Centers to ensure they are responsive to outside scientists.
The RPRCs are a crucial national resource, especially for research into primate models of HIV infection. The panel found the RPRCs do not always support the most meritorious research, and recommends that their resources be refocused on the most promising research ideas. The RPRCs should develop a competitive process to ensure that the best science is supported and allow outside scientists greater access to their primate resources.
Strengthen AIDS Research Centers to foster multidisciplinary research on the disease.
The sixteen NIH-supported AIDS research centers are multidisciplinary organizations that bring together basic and clinical researchers from many medical sub-specialties. The centers "provide a central pool of resources, capable of a flexible and coordinated response to scientific opportunities." The report advocates doubling the funding for AIDS Research Centers.
Ensure that central repositories of biomedical specimens and databases are of the highest quality and accessible to qualified investigators.
Over the last decade, the NIH has generated enormous repositories of specimens and large databases, which represent a "potential national treasure" for AIDS research. But, the report says that these repositories and databases are poorly coordinated, and many investigators have little or no access to them. It recommends that databases use a common, user-friendly information system, and that access to them be peer-reviewed and open to all scientists with meritorious research ideas.
Upgrade the NIH AIDS Research Information System and increase the information base.
The panel said its review was hampered by the lack of an adequate information system at the NIH for AIDS research. Available databases do not accurately and thoroughly track all AIDS research expenditures at the Institutes, producing many gaps in grasping the nature of AIDS research there.
Develop and implement a clear definition of AIDS and AIDS-related research through an evolving process.
The panel found that each NIH institute has its own definition of what constitutes "AIDS research." A clear NIH-wide definition of AIDS research is urgently needed, because, in some institutes, inappropriate classification has led to allocation of a significant amount of AIDS funding toward activities with little or no relevance to AIDS (see Treatment Issues, October, 1995, page 9 on the National Cancer Institute's misallocation of tens of millions of dollars). The panel urges that these funds be redirected to more appropriate projects.
Preserve a strong Office of AIDS Research to provide leadership and coordination to the entire NIH AIDS research program.
The panel "unanimously agreed" that the NIH AIDS research program needs central leadership and oversight. In particular, the panel emphasized that the OAR will be crucial for implementing the recommendations in its report.
Abbott Laboratories claims that the nausea commonly associated with its protease inhibitor ritonavir (brand name: Norvir) goes away after the first few weeks in most people. The digestive upset can be minimized by progressively increasing the initial dose. But the initiation regimen contained in the package insert is too quick, Abbott officials now indicate.
The following schedule is now thought to be the best way to begin ritonavir (the drug is supplied in 100 mg capsules):
First and second day: Take 300 mg twice a day.
Third and fourth day: 400 mg twice a day.
Fifth and sixth day: 500 mg twice a day.
If you are having a problem, stay at this dose. If not, go up to 600 mg twice a day by the end of two weeks.
These instructions will not necessarily do away entirely with the gastrointestinal upset. What you eat with ritonavir is important as well. Take ritonavir with a full meal, particularly a fat-rich meal. Ingest the capsules one at a time with a generous quantity of milk or whatever you find most useful. At least one person with AIDS washes down his ritonavir with yogurt rather than a fluid drink.
Matching up Ritonavir with Saquinavir
A trial to study the much heralded protease inhibitor combination of ritonavir and saquinavir is now screening applicants. Because of its liver-inhibiting properties, full dose ritonavir will raise blood levels of saquinavir about ten-fold and make this drug much more effective. Conversely, increasing saquinavir's effectiveness may allow a reduction in the dosage of ritonavir, thus reducing side effects. The combination also might be less prone to trigger the emergence of treatment-resistant HIV than either drug alone.
At first, the volunteers will receive saquinavir at 400 mg twice daily plus ritonavir at 400 or 600 mg twice daily. Once safety is established, the doses will increase to saquinavir at 800 mg twice daily plus 400 mg ritonavir twice daily or saquinavir at 600 mg twice daily plus 600 mg ritonavir twice daily. Everyone will receive both drugs in this open-label, 48-week trial.
This 120-person trial (in Annandale (Virginia), Boston, Los Angeles, New York, Ottawa, Pittsburgh and San Francisco) is open to volunteers with CD4 counts between 100 and 500, any viral load and no previous protease inhibitor experience. One or two concomitant nucleoside analogs will be allowed if an individual's viral load rebounds to within one log (ten percent) of the pre-trial value. Otherwise, no other anti-HIV drugs will be allowed during the trial.
"This might be a powerful combination. If people are on nucleoside analogs and failing, then this is a great study for them," commented Marty Markowitz, M.D., the trial's New York City investigator.
Further information on this trial may be obtained from the Abbott Laboratories consumer hotline, 800/441-4987.
Nevirapine and Delavirdine Plus Protease Inhibitors
Expanded access programs for the two nonnucleoside reverse transcriptase inhibitors (NNRTIs) nevirapine and delavirdine are now up and running (call 800/595-5494 for nevirapine and 800/779-0070 for delavirdine), and an FDA advisory committee hearing is scheduled for June 7 to review nevirapine's marketing application. People may be tempted to combine the NNRTIs with protease inhibitors to achieve greater, more durable HIV suppression. Like the protease inhibitors, nevirapine and delavirdine are metabolized by the liver, though, and so such combinations may alter the normal blood levels of each component drug.
Delavirdine inhibits the way the liver breaks down certain drugs and may raise blood levels of protease inhibitors. Nevirapine, an inducer of liver activity, could have the opposite effect. Changing protease inhibitor blood levels can sharply alter their activity and side effects, thus, necessitating dose adjustments (as with the ritonavir/saquinavir combination above).
Pharmacia & Upjohn, the maker of delavirdine, and Boehringer Ingelheim, maker of nevirapine, both predict that their drugs will most likely affect the blood levels of saquinavir and are now conducting pharmacokinetic studies of these combinations. More information will be available by this summer.
Co-administration of delavirdine and ritonavir may increase blood levels of either drug. Given the toxicity of ritonavir, it is advisable to wait before trying the combination until safety and the necessary dose adjustments are established. The pharmacokinetic study to accomplish this should be underway in May, Pharmacia & Upjohn says.
Liposomal KS Drugs Approved
On April 8, 1996, the FDA granted NeXstar's DaunoXome full approval as first line therapy in people with advanced HIV-associated Kaposi's Sarcoma. DaunoXome is a liposomal (lipid-encapsulated) formulation of the anticancer compound daunorubicin and is similar to Doxil (liposomal doxorubicin). Doxil (from Sequus) received FDA "accelerated" approval last November for use only in people with advanced KS who could not tolerate or did not respond to standard chemotherapy. Doxil could be used "off-label" as first line therapy, too, but the comparative effectiveness of Doxil and DaunoXome is unknown.
Liposomal drugs tend to be safer and more stable in the body than standard formulations. Doxil is administered by infusion only once every three weeks, and DaunoXome is infused every other week. The drugs' main side effect is bone marrow suppression, which leads especially to low neutrophil counts.
The wholesale price of four month's worth of DaunoXome is $4,300, close to the $3,500 cost of ABV, the old standard chemotherapy combination for KS. Doxil costs about $6,400, but the longer interval between infusions considerably reduces the difference in total treatment costs.
NeXstar has a DaunoXome reimbursement hotline and indigent program that can be reached at 800/226-2056. A similar program exists for Doxil -- call 800/375-1658.
Still Looking for the Real GEM
A persuasive presentation at the 1993 International Conference on AIDS excited many about GEM 91, the first antisense drug targeted against HIV. Antisense compounds are synthetic segments of nucleic acid that bind to viral genetic material and disrupt the viral life cycle at multiple points. In the lab, GEM 91 blocks HIV when the virus infects new cells, when the cell's machinery is commandeered to produce viral protein, and when complete HIV gene sets are transcribed and packaged in new virions.
Clinical trials reported last September (Interscience Conference on Antimicrobial Agents and Chemotherapy abstract I86) dampened the excitement. They found that the body breaks down GEM 91 too quickly for it to reach effective concentrations in people. Despite two-week continuous infusions, volunteers received no benefit.
Hybridon, Inc., GEM 91's developer, now has a new concentrated formulation that is administered as a subcutaneous or intramuscular injection. A French phase I study of the improved GEM 91 is scheduled for this month, and an AZT/GEM 91 drug interaction study will begin soon afterwards in Britain.
Another slowly advancing high-tech antiviral strategy is the "rev mutant" gene therapy investigated by Gary Nabel, M.D., of the University of Michigan. Some mutant versions of the rev protein hobble production of new HIV particles in cells by blocking the transport of necessary material from the nucleus. Dr. Nabel found that cells grown with HIV in the lab survived longer when the gene for this protein had been added to their normal gene set.
In the new study, CD4 cells were extracted from three HIV-infected people. Half of each person's cells were injected with the antiviral mutant gene and half with a neutral mutant gene. Gold microparticle "bullets" successfully implanted the genes in ten percent of the cells. Altered cells were treated with delavirdine to eliminate any HIV present and IL-2 to stimulate rapid proliferation. They were then reinfused into the source patient.
After the first infusion, only very small numbers of the cells could be detected. Patients no. 2 and no. 3 then received a second treatment utilizing ten billion, rather than one billion, cells. In patient 2, the half-life of the cells with the antiviral gene was four days versus one day for the control cells. In patient 3, the half-lives were around fifteen and 3.5 days for the anti-rev and control cells, respectively.
The slowed but relentless drop in the anti-rev cells may be due to instability of the transferred genes, or because even non-infected CD4 cells are subject to elimination in people with HIV. This issue may be resolved as Dr. Nabel's study continues, using different gene transfer methods to achieve a more stable integration of the protective gene in a higher proportion of cells.
Thalidomide for Diarrhea
Celgene is beginning a phase II study of thalidomide as a treatment for chronic diarrhea in people with HIV. Thalidomide, already under testing for AIDS-associated wasting, commonly causes constipation as a side effect. This unwanted activity actually can be advantageous in people with untreatable diarrhea. A recent open label study in London found that thalidomide reduced the number of bowel movements, increased weight and reduced levels of TNF in patients with severe diarrhea (mostly from microsporidiosis) who had failed all other available treatments.
The 28-day trial will randomize 120 people to placebo or 100 mg of thalidomide daily at bedtime. Besides diarrhea, the trial will monitor changes in body weight, quality of life and TNF in small bowel tissue. Enrollment already has begun in London. Other sites are expected to open soon in Mexico and the U.S. (St. Luke's Hospital in New York).