Let There Be Drugs
by Dave Gilden
The Third Conference on Retroviruses and Opportunistic Infections took place at the beginning of this month in an atmosphere of unaccustomed optimism. There was a sense that the conference marked a watershed in the history of attempted therapies for HIV infection. Now at last, there seem to be some powerful tools to attack HIV, suppressing virus levels by many orders of magnitude.
Emilio Emini, M.D., head of antiviral research for Merck & Co., set the atmosphere for the rest of the conference by describing these new tools in one of the event's lead-off presentations. Reviewing the public data on protease inhibitors and their combination with one or two nucleoside analogs, Dr. Emini urged a "new paradigm" for AIDS treatment, in which the goal is to administer combinations of drugs that lower viral loads for as long as feasible to below current assays' limit of detection. The idea is to shut down HIV replication as much as possible so as to eliminate or delay the accumulation of mutations leading to drug resistance, the chief nemesis of HIV therapy up to this point. (Considering the source, it is not surprising that Merck's protease inhibitor indinavir had a featured role in this "new paradigm.")
The results presented both before and during the conference concerning such drug combinations are very impressive. Combinations of some of the new protease inhibitors and the older nucleoside analogs caused drops in plasma viral loads of a thousand-fold and more, accompanied by average CD4 cell rebounds of up to two hundred and, in one trial, a sharp decrease in new AIDS-related diseases and death. (For details, see Theo Smart's article starting on page 3.)
Yet many questions persist. The data accumulated so far cover only six months of treatment and/or a handful of trial participants. Already it is apparent that there is wide interpersonal variation in response to treatment, especially to the protease inhibitors. Conference reports on the various clinical trials commonly lacked important information on this score, such as baseline viral load
and other characteristics of the best responders.
The duration of treatment benefit remains unknown and probably also differs greatly between individuals. When patients should switch treatment and to what remains the subject of much speculation. And the extent to which people can tolerate the many, sometimes interacting, side effects of combination therapy also remains to be seen.
New Agents Galore
There is no question that a new era is dawning for therapy of HIV and AIDS. As this edition of Treatment Issues goes to press, the FDA's Antiviral Advisory Committee is poised to take up consideration of two protease inhibitors, Merck's indinavir (brand name: Crixivan) and Abbott's ritonavir (Norvir), plus human growth hormone ("Serostim," from Serono) for AIDS-associated wasting. Meetings for these three are scheduled for February 28 through March 1. Delavirdine (made by Upjohn) and nevirapine (Boehringer Ingelheim), two reverse transcriptase inhibitors that are not nucleoside analogs like AZT, also are expected to go before the FDA in coming months.
The FDA reportedly has urged the manufacturers of indinavir, ritonavir and delavirdine to file as quickly as possible, even if trials are not completed or manufacturing facilities ready. Judging by the examples of saquinavir and 3TC, approved last November, the agency may take only a few weeks to license drugs for market after its advisory committee has recommended approval.
By the end of this year, if not this summer, there should be at least ten anti-HIV therapies on the market, and the FDA is in the process of approving a variety of new medications for opportunistic infections, too. All this activity is creating a considerable amount of confusion about when and how to apply treatment.
At the Third Conference on Retroviruses, many experts' inclination seemed to be, don't hesitate, use everything now. In a review entitled "Antiviral Therapy '96," Australian researcher David Cooper, M.D., told his audience that 1995 saw the "demise of monotherapy and the ascendancy of combination therapy." But he warned that in the absence of hard data, there was a "danger of patients getting anecdotal combinations" based on physicians' informal summaries of their own experience.
Because of the rush to combination therapy and the large number of available or soon-to-be-available drugs, recruitment for future clinical trials may be difficult, and dropouts may be common. Nevertheless, testing of innovative combinations against standard ones needs to take place.
One goal would be to find "the most compact, least toxic combinations with the greatest antiviral activity," to use Dr. Cooper's words. Combining different protease inhibitors, an uncharted field so far, would be a particularly fruitful area to investigate. Another area is the use of maintenance therapy: Once HIV replication is tightly suppressed for some months, and nearly all the infected cells harboring the virus are dead, it may be possible to switch to simpler therapies with fewer drugs to provide a barrier against a resurgence of HIV.
Dr. Cooper also proposed looking at disease progression in people with life-threatening opportunistic illnesses for which there are no adequate treatment options. Here, investigators would measure by how much a new anti-HIV therapy increased the time until diseases such as CMV or lymphoma recurred.
Trials monitoring the recurrence of OIs might give some quick answers, or they might have ambiguous results because unhealable damage inflicted during HIV infection prevents the recovery of effective immune function (see below). Certainly these trials would not tell how to best use antiviral therapy earlier in HIV infection. Efforts to evaluate compact and efficient combinations in a more complete fashion could perhaps take the form of large treatment strategy trials involving changing combination therapies as conditions seem to warrant. Such trials could, and preferably should be, conducted by community physicians in a real world setting.
The use of viral load assays -- PCR and bDNA -- to measure changes in the level of free HIV in blood plasma, and hence a drug's antiviral activity, has been widely proposed as a way to radically shorten trials. The problem has been that no one has established how predictive viral load measurements are of a person's long-term disease course. Even more in question has been the significance of short-term reductions in circulating HIV induced by a specific treatment.
Almost unnoticed among the news about protease inhibitors and combination therapy were the Conference presentations that promised to revolutionize trials of anti-HIV compounds by confirming viral load as a marker for drug effectiveness. The announcements concerning viral load began in mid-January, when representatives of the newly merged corporation Pharmacia & Upjohn first broadcasted the preliminary results from two large trials combining delavirdine with either AZT or ddI (see last month's Treatment Issues, page 9).
Researchers found that delavirdine trial participants who registered a 0.5 log (70 percent) reduction in viral load in their eighth week of treatment experienced a 50 percent reduction in AIDS progression (as measured by opportunistic infections or death). In contrast, a CD4 count increase of 25 or 50 at week eight was not predictive of improved clinical outcome over the next year. (See abstract LB8c.)
But a CD4 count gain of 50 sustained for a year was indeed accompanied by less illness and death. The Upjohn researchers argue from their analysis that early viral load changes are nevertheless more useful for evaluating experimental drugs and managing individual patient's therapy. In their cohort, though, those people with reduced viral load tended to keep that reduction for a year's time, leaving open the question of precisely how long a treatment has to show antiviral activity before that activity translates into symptomatic relief for the patient.
A number of other objections could also be raised concerning this confusing trial analysis, which easily leant itself to rationalizing the approval of delavirdine, a drug with modest antiviral effect when combined with AZT and little effect on CD4 counts. "I'd love to see this in the hands of an objective, outside statistician," commented one skeptical conference attendee.
Be that as it may, Upjohn's so far preliminary conclusions about the value of viral load were reproduced in several other conference reports, particularly the virologic analysis of a subset of the participants in the ACTG 175 trial (abstract S24). ACTG 175, it should be recalled, was a 2,500-person, government-sponsored trial spanning three years that compared AZT alone, ddI alone, AZT plus ddI and AZT plus ddC in HIV-positive individuals with and without prior AZT therapy (see Treatment Issues, October, 1995, pages 2-4).
Although CD4 cell count gains sustained for more than a year seemed to best reflect improved outcome during the course of the trial, baseline plasma viral load and treatment-induced reductions at week 8 were more predictive than CD4 count or change in CD4 count in the first eight weeks of trial observation. A ten-fold drop in viral load correlated with a 70 to 87 percent decreased risk of disease progression as measured by various endpoints -- 50 percent drop in CD4 count, AIDS-defining disease and/or death.
The ACTG 175 analysis further found that initial, pretreatment viral load was a good predictor of future health, as did the Upjohn and several other studies. John Mellors, M.D., reported viral load trends seen in a retrospective analysis of 181 gay men with early HIV infection in the years 1984-85 (abstract S22). (The group was part of the Multicenter AIDS Cohort Study, or MACS.) Based on these people's experience, Dr. Mellors observed, "[Plasma viral load] is an earlier and more accurate predictor than CD4 count Viral load reflects the balance between the immune system and the virus. It is the most important determinant of outcome."
These studies do not establish how well viral load can be used when making treatment decisions for individual patients. They are a further indication, though, that in the future, optimum combination therapies will be formulated for each patient according to how his or her viral load responds. More and more experts are concluding that the advantages gained are worth the added complexity and cost. "I'm a believer in HIV RNA [viral load]," concluded Harvard's Scott Hammer, M.D., when reporting on ACTG 175.
Holes in the Defense
In the great new age of elaborate antiviral combinations that pack a real wallop, there will be plenty of reason to closely monitor how well that wallop is reaching HIV. First of all is the issue of side effects, which may make patients take their drugs erratically or cease altogether. It is worth noting that in Jacques Liebowitch's small but highly successful Paris study of ritonavir/AZT/ddC (abstract 285, see page 2 of this issue), eleven of 32 volunteers dropped out. Mostly it was the nausea and vomiting that drove them out, despite almost 1,000-fold drops in HIV levels and gains in CD4 counts of nearly 300. (The new capsule form of ritonavir may induce fewer of these adverse effects than the older liquid formulation used here.)
Also, ritonavir's particular molecular structure was chosen for its inhibitory effect on the liver, which stabilizes the drug's level in the bloodstream. Ritonavir as a result cannot be taken along with the many drugs that depend on the liver to break them down and prevent any toxic accumulation in the blood. If prescribed ritonavir, patients may sometimes be faced with choosing between their anti-HIV medication and their OI drugs.
The other way these therapies will fail is through the development of mutant HIV whose altered enzymes no longer interact with suppressive compounds like the nucleoside analogs or protease inhibitors. The ritonavir trial for people with CD4 counts less than 100 (abstract LB6a -- see page 3 of this issue) did find a rapid reduction in the rates of death and new opportunistic infection, but the level of HIV went right back up after the second week. At sixteen weeks, average viral loads were approaching their initial values, leading one to wonder how trial participants will fare later on.
These volunteers merely added ritonavir to their existing therapy, leading the presenter John Leonard of Abbott Laboratories to reckon that the trial was "almost sequential monotherapy" since the other therapies had probably outlived their usefulness because of drug resistance. Dr. Leonard remarked that it would be preferable "to start multiple agents simultaneously."
The futility of monotherapy with even a very potent protease inhibitor, at least at less than optimal doses, was underscored further in an analysis of seven participants in an early ritonavir study using 600 to 900 mg per day (the current dosage is 1200 mg per day). An outgrowth of HIV with a single specific mutation known to confer resistance to ritonavir in the test tube led to a rebound in viral load (abstract 201).
Disappointing expectations that mutant virus would be somehow impaired, this ritonavir-resistant HIV had a robust doubling time of two to four days, perhaps partially due to a second, compensatory mutation. Worse yet, the mutant HIV was already present in very small amounts prior to ritonavir treatment. The time to virus load rebound after therapy began was highly associated with the pretreatment virus level. It was only weakly associated with the dose of ritonavir.
This ritonavir study was remarkable in that it confirmed some of the predictions made by various scientists, including John Coffin of Tufts University (abstract L13) and David Ho (abstract S21) in their reports to the conference on HIV dynamics: there is a close link between the detection of certain mutations and the rebound in viral load; mutant HIV is able to establish significant growth rates only slightly below that of nonmutant virus; and the HIV with point mutations contributing to reduced susceptibility to particular drugs exists even before exposure to those drugs.
Attaining the goals Dr. Cooper set out for ideal antiviral therapy may require personal modification of proposed treatment plans based on genetic analysis of the resistance mutations present in a patient's pool of HIV. Such complex testing would be in addition to viral load monitoring and the multitude of laboratory tests needed to check for side effects like liver and bone marrow damage.
Holes in the Repertoire
Time will tell if it is possible with our current technology to root out absolutely all the HIV in a person and thus avoid the rise of mutant varieties resistant to treatment.
There are three distinct "reservoirs" beyond the scope of present treatment that may preserve some HIV indefinitely. One is the brain and central nervous system, which is poorly penetrated by most drugs. The second is chronically infected macrophages, which store HIV for many months. Finally, there are antibody-coated HIV particles adhering to the surface of the follicular dendritic cells, which trap such virus in lymph node fluid.
The follicular dendritic cells (FDCs) hold HIV for an unknown length of time, creating a library of the past HIV variants in someone's body, including any previous drug-resistant variants. A paper in the journal Nature last fall (SL Heath et al., October 26, 1995, pages 740-4) reported that virus sticking to FDCs easily infect the circulating immune cells with which they come in contact despite being covered with what are normally highly neutralizing antibodies. "Time bombs waiting to explode," is the way one expert observer, Cecil Fox, Ph.D., characterized them. (Dr. Fox specializes in the analysis of HIV-infected lymph and other tissue.)
It could well be that people with established HIV infections will need some kind of antiviral therapy for the rest of their lives given the limitations on what current treatments can reach. Besides their antiviral cover, people with AIDS may continue to need medications to treat or prevent various infections because they are unable to recover their complete immune response even with HIV permanently suppressed.
Mark Connors, of the National Institute of Allergy and Infectious Diseases, compared the CD4 cell "repertoire" in ten HIV-positive and seven HIV-negative individuals (abstract 397). The people with HIV had a more uniform CD4 cell population containing cells able to respond to fewer different foreign proteins. In theory, these HIV-induced "holes in the repertoire" will prevent their immune systems from recognizing specific pathogens, making them vulnerable to diseases most people are immune to.
The gaps essentially persisted in patients even though antiviral and IL-2 therapy raised their CD4 cell count from about 200 to 1,000. Presumably they could not rebuild their normal CD4 cell diversity because the damage to the thymus and other lymph organs wrought by HIV infection makes it impossible for fresh CD4 cells to develop and mature.
The NIAID study contrasted with a report from Australia of improved CD4 cell response in a few persons with advanced HIV taking ritonavir in one of the original trials (abstracts 232 and 451 -- see page 3 of this issue). Counts of CD8 cells (which play an important role in eliminating virus-infected and cancerous cells) also exhibited a substantial, sustained rise -- the first time this has been observed in trials of anti-HIV agents.
The clinical implications of either the Australian reports or the contrary NIAID study have not been determined and will require prolonged following of many patients' experience. As reported at the Second Retrovirus Conference last year, the Australians did see some resolution in several patients' physical symptoms.
The hope expressed by one of the Australian observers that under the influence of ritonavir, "[advanced] patients' immune profile becomes similar to that of asymptomatic patients" for now seemed more like a prayer than a realistic assessment to many observers. But still, it is encouraging that immune improvements correlated with the depth and length of viral load decrease. Just getting rid of HIV may turn out to be the linchpin of immune therapy.
Who's Footing the Bill?
Roche's protease inhibitor saquinavir (Invirase) entered the market in December with a retail price tag of about $7,500 per year, and Merck and Abbott, though refusing to comment on pricing issues, can be expected to charge a similar amount. It would be one thing if a cure for HIV could be procured for a cost of $15,000 or $20,000. The benefits of combination drug therapy are finite, though. People with HIV will have to come up with the money for all their drugs and for monitoring their own HIV over an indefinite period.
In the U.S., there are a million people, more or less, infected with HIV. Arguably, they all could profit from the new therapeutic advances (although this is not yet proven). Who will pay is going to become a big question, especially when cuts in Medicaid are in the offing and the states' AIDS Drug Assistance Programs (ADAPs) are already in financial crisis (see page 11). Effective treatment requires social commitment as well as technical finesse.
The experience with tuberculosis, which has some parallels to HIV, is not encouraging. TB requires six to nine months of daily treatment, usually with one inexpensive drug if the disease is latent and with four drugs of moderate price in the case of active, symptomatic TB. When TB elimination seemed within our grasp, a concerted effort to finish off the disease was never undertaken. The result has been incomplete courses of treatment, a resurgence in the number of cases and the emergence of drug-resistant TB requiring novel and costly antibiotics. HIV is even more prone to drug resistance, and its treatment is more complicated and prolonged in the first place.
Then there are all the "developing" countries, whose citizens will have almost no access to the new treatment modalities. Scientists may worry about the protected reservoirs of HIV within the body, but political figures will have to worry about the vast reservoirs of HIV within the planet's human population. Unless the virus can be controlled worldwide, it will keep coming back to haunt us, just as TB does.
Protease Inhibitors Come of Age
by Theo Smart
The big news at the Third Conference on Retroviruses was the extraordinary antiviral and clinical effects achieved by protease inhibitors when administered along with nucleoside analogs. The reports, although limited in their scope, suggested that combinations of anti-HIV drugs now available will be considerably more effective than the mediocre HIV treatment options of the past.
In one study, Abbott's ritonavir (brand name Norvir) was found to dramatically prolong survival and slow disease progression in people with advanced AIDS. Other trials found that Merck's indinavir combined with several nucleoside analogs can hold HIV blood levels for many months below minimum detectable levels (less than 500 copies of HIV genes (RNA) per ml of plasma).
Findings from the Ritonavir Clinical Study
Abbott, which was running behind Merck and Hoffmann-La Roche in the race to develop protease inhibitors, is so far the first and only company to show that its drug extends survival and delays opportunistic infections (abstract LB6a). The company accomplished this feat by counting the clinical events that occurred in 1090 volunteers with advanced HIV disease -- CD4 cell counts of 100 or below -- randomized to take either ritonavir (600 mg) or placebo twice daily. The participants had a history of more than nine months of prior anti-HIV therapy and added ritonavir to their existing regimen of nucleoside analogs (most commonly AZT or d4T -- 3TC was not yet approved).
At baseline, the median CD4 cell count was 18 in the ritonavir arm and 22 in the placebo group. The median HIV plasma viral load was over 250,000 copies per milliliter. After completing four months on the study, trial participants who reached an AIDS-defining event were given open-label ritonavir.
When planning the study, Abbott researchers had calculated that it would take only 191 clinical events (the appearance of new opportunistic infections) or deaths to show definitively that ritonavir had an effect. That figure was reached within one month, although the trial continued while the company confirmed each event. When the results were analyzed, it became clear that the people on ritonavir were doing better than those on placebo within ten days. After one month, only 69 opportunistic infections or deaths had occurred in those receiving ritonavir as opposed to 149 in those on placebo, a statistically significant difference. Forty-four of these events were deaths -- seventeen on ritonavir, and 27 on placebo.
After a median of six months follow-up (see table), 26 people on ritonavir versus 46 people on placebo had died, a 43 percent reduction in mortality. New OIs or death occurred in 85 people on ritonavir and 181 on placebo.
Viral load response was analyzed in 159 patients from the trial. Levels of virus decreased by 96 percent in the ritonavir arm but over time returned back upwards toward baseline. The trial's rapporteur, John Leonard, M.D., of Abbott, argued that virologic responses might have been more sustained if trial participants had begun their concurrent therapies at the same time as they started ritonavir. CD4 and CD8 cell responses (followed in 215 participants) were also significantly better in those taking ritonavir (47 CD4 cells and more than 200 CD8 cells above baseline at week sixteen).
An Australian team led by David Cooper, M.D., performed a more in depth exploration of the immunologic effects of ritonavir in 21 patients from an earlier trial by analyzing the functional responses of the new CD4 and CD8 cells (abstract 232). The team determined that the new CD4s acquired during the first four weeks of therapy were primarily clones of existing memory cells (lymphocytes that respond to previously encountered pathogens), and that new "naive" cells (which might be able to mount an immune response to infections for which there is no pre-existing defense) were noted only after the first four week period.
CD8 cells increased here too (abstract 451). Andrew Carr, M.D., one of the Australian team, noted that ritonavir's ability to increase CD8 cell counts has no parallels in the CD8 responses observed for reverse transcriptase inhibitors, whether alone or in combination. In laboratory tests, the ability of some patient's cells to proliferate after exposure to certain foreign antigens improved, as did the ability of cells to secrete the immune stimulant IL-2. Improvements in a few clinical conditions, such as KS and oral hairy leukoplakia, were also noted and attributed to a restored immune response.
Surrogate Marker Data
Abbott:A group of French researchers addressed whether combinations of ritonavir with other antivirals can force the virus into submission in an open-label study of ritonavir plus AZT/ddC in 21 individuals (abstract 285). The baseline CD4 count was 156 and viral load was 69,780 copies per ml. Early data suggested that over time, viral loads dipped below the limit of detectability in a growing number of patients on the combination (see last October's Treatment Issues, pages 4-5).
The combination's antiviral effect, which really improved only slightly after the first month, may now be dwindling (see table). Viral load was undetectable in seven patients at month five, but only in five at month six. The actual change in median viral load between month five and six is very small, though, (median viral load merely went from 351 to 892) so it is not clear yet whether the increase is a random blip or reflects a gradual viral rebound.
Merck:This company may be seeing more sustained results with a different combination. Roy Gulick, M.D., of New York University reported on a study comparing Merck's protease inhibitor indinavir plus AZT/3TC to AZT/3TC or indinavir monotherapy (abstract LB7). The study enrolled 97 people with at least six months of prior AZT therapy, CD4 cells between 50 to 400 and a viral load higher than 20,000 copies/ml. The median baseline viral load was 40,000 copies, and the median CD4 cell count was 142.
The three drug combination most reduced viral load (by more than 99 percent -- see table on page 2). The antiviral effect appears sustained in the combination arm, maybe dissipating slightly in the indinavir monotherapy arm, but is clearly on the way back to baseline in the AZT/3TC arm (only 75 percent below baseline). The results at week 24 are a little shaky, though, since only eight to ten patients in each arm have been on therapy for that long. Viral loads at week 24 were below detectable limits in six out of seven trial participants receiving the triple combination for that long. CD4 cell responses were greater in the indinavir containing arms.
Indinavir plus AZT/ddI may not sustain its antiviral effect as well (abstract 200), but this is a completed study and not merely an interim analysis. A 24-week study enrolled 78 antiretroviral naive patients. At study entry, the median CD4 count was 150 and median viral load was 100,000. Again, the antiviral effect of three drug combination was greater and more sustained than AZT, ddI or indinavir monotherapy (see table for details), although ten patients dropped out of the study because of ddI's gastrointestinal side effects.
Agouron: The company has now shown a dose effect with its protease inhibitor, nelfinavir, by raising the dose to 2,250 mg a day. During the conference, Agouron presented data on the combination of nelfinavir with d4T in a meeting with community representatives, doctors and stock analysts.
An ongoing trial is comparing d4T to d4T plus nelfinavir (two doses) in people with CD4 cell counts 200 or higher and viral loads above 15,000 copies per ml. The median baseline viral load in the different trial arms ranged from 50,000 to 70,000. The two drug combination achieved dramatic reductions in viral load (more than a 99 percent reduction day 45 -- see table on page 2) and increases in CD4 cell counts, but data are available for only seventeen people at day 28 and six to eight at day 45.
Duration of Effect
The Agouron study illustrates the chief problem with the information available on protease inhibitors -- the effect is profound, but given the development of resistance, will it last for more than a few months? Researchers suggest that the combination therapy approach will postpone resistance, but again for how long?
There is no way to say until studies longer than six months are conducted. Some researchers and doctors are pushing patients to use the combination therapies while they still have low viral loads and high CD4 cell counts because they reason that the effect will last longer and that resistance will take still longer to develop with early treatment. But, again, no one knows how much longer.
What scant long-term data there is comes from studies of indinavir. Data out to sixty weeks were presented from an open-label phase 1 trial of indinavir monotherapy in five volunteers (abstract 147). After 24 weeks, four of the five resumed taking their previous nucleoside analog therapy in addition to indinavir. Mean CD4 count for the five at week 60 remained 140 cells above baseline, and viral load was still 96 percent below baseline -- levels not significantly different from their peak responses. Weight also had increased an average of 3.5 kilograms and remained stable.
Another presentation (abstract 146) had a positive report on indinavir monotherapy. The dose-ranging study was in 70 infected people with a median baseline viral load of 126,000 and a median baseline CD4 count of 250. This baseline CD4 count is about 90 to 100 cells higher than in the combination studies above, and the viral load response seemed greater and more sustained. There was little difference between the study arms, but the dose now used in clinical studies (800 mg every eight hours) at week sixteen reduced plasma HIV levels by a median of 99.5 percent, and the effect was sustained out to week 24.
In a state of the art symposium on protease inhibitors, Emilio Emini, M.D., of Merck (abstract L1) briefly presented follow-up data from this study. The suppression of HIV continued unchanged through week 48 in fifteen people who had at least a 99 percent reduction in viral load at week 24. There clearly is not the rebound in viral load that was seen in Merck's earlier studies with lower doses.
The profound antiviral effect seen with indinavir may even be depleting the viral reservoirs in the lymph nodes. The lymph nodes trap large amounts of free virus which can survive for an unknown length of time and may infect cells passing through the nodes. Dr. Emini presented data of antiviral activity in the lymph nodes of five people who had lymph node biopsies after eight weeks on therapy. The people experienced a 99 percent reduction of infectious virus in the lymph nodes that matched a 99 percent reduction in plasma viral loads. Indinavir may also penetrate the brain, another reservoir for virus, at levels 40 to 50 percent of that achieved in the blood.
Side Effects and Drug Interactions
None of the three potent protease inhibitors have been directly compared to one another. In a few months, though, ritonavir and indinavir will be competing for market share, and nelfinavir may be available, too, by the end of the year. The antiviral activity reported for these drugs may sound very similar at this point, but their side effects and drug interactions are distinct.
Ritonavir appears to be the least well tolerated. In the French surrogate marker study mentioned above, which used the old liquid formulation, three-fourths of the study participants experienced nausea, more than half had diarrhea, and more than a quarter had episodes of vomiting. One-third dropped out. Likewise, 91 (seventeen percent) of those receiving ritonavir in the clinical endpoint study for advanced patients stopped taking the drug because of nausea, vomiting, weakness and diarrhea.
Few patients seem willing to quit indinavir due to side effects. The reported elevated levels of bilirubin do not appear to produce symptoms or discomfort. Nephrolithiasis (accumulation of drug residue in the kidneys similar to kidney stones) seems more serious as kidney stones can be extremely painful. In the studies reported at the conference, though, people with this side effect continued on drug without dose reduction. Some new side effects were reported in the indinavir AZT/ddI study: indinavir therapy caused more cases of dry skin, insomnia, rash, pharyngitis and flank pain.
Meanwhile, nelfinavir causes grade I/II diarrhea, characterized as two to six bowel movements per day in 70 to 100 percent of the patients on the three highest doses. The company claims that this diarrhea is mild, "just loose stools, really" and that only 40 percent of the patients choose to treat it with antidiarrhea medication.
Convenience of dosing is a factor that people take into consideration when deciding whether to take a drug. Here ritonavir wins out, with twice-a-day dosing, although people have to take six capsules each time. Nelfinavir has to be taken three times a day with a little food. Conversely, indinavir has to be taken three times a day on an empty stomach, difficult to do for patients struggling to keep their weight stable.
Finally, ritonavir has a powerful inhibitory effect on the liver's ability to metabolize many drugs, which therefore cannot be taken concurrently (see last December's Treatment Issues, page 8).
The Next Generation
Other companies hoping to cash in on the protease inhibitors will have to find some marketing advantage for their compounds -- such as lower price, better potency, or fewer burdens for its recipients. An important positive feature would be that the drug requires of HIV a unique genetic pattern to protect itself from that drug. HIV that develops resistance to other protease inhibitors to which it is exposed would then not be cross-resistant to the new one even before a patient starts taking it.
A number of drug developers have already dropped out because they could not get ahead of the competition. Bristol-Myers did so after discovering that its compounds were not adequately bioavailable. Since these drugs did not stand out from the other current protease inhibitors, the company chose to scrap the whole program rather than spend more funds on improving them. The reverse transcriptase inhibitor nevirapine (see page 9) has exhausted Boehringer Ingelheim's will to do any further AIDS research, so its protease inhibitors are trapped in the starting gate.
For others, the work on perfecting their protease inhibitors continues. Glaxo Wellcome's 141W94 compound, purchased from Vertex, has the advantage of already being in human trials. It also has the distinction of accumulating in the brain -- cerebrospinal fluid levels of 141W94 reach about 170 percent of the levels in the blood.
Upjohn continues preliminary testing of its protease inhibitors, with the first safety and metabolic trial of the discontinued U-103017 in HIV-negative humans reported on at this conference (abstract 153). Upjohn's compounds are known for the unique resistance pattern HIV develops in response to them. They do not suffer from cross-resistance with the more established protease inhibitors. Gilead Sciences has temporarily delayed clinical studies of a potent class of protease inhibitors because the company's chemists believe they can make a few modifications to the lead compound so as to eliminate cross-resistance.
Parke-Davis has a class of novel compounds (abstract 149) that are currently hampered by binding to blood plasma proteins. The Parke-Davis compounds are similar to Searle's second generation compounds, which were reported to have little cross-resistance. Parke-Davis is currently trying to ascertain whether its drugs are active against virus resistant to the Merck and Abbott drugs.
The only new compound entering clinical trails in the near future is Ciba-Geigy's lead compound CGP 61755. A dose-ranging study and a fourteen day viral load study will begin shortly in the Netherlands.
CGP 61755 is potent in the test tube and synergistic with saquinavir (abstract 150). Animal studies suggest that it is safe and highly bioavailable. Plasma protein binding appears to have little effect on its antiviral activity. Unfortunately, there may be cross-resistance problems between CGP 61755, ritonavir and indinavir, which give rise to similar resistance patterns in HIV. CGP 61755's chief advantage is that it is easy to manufacture. That should translate into lower retail costs, one hopes.
Back to the Nukes
by Theo Smart
Clinical and surrogate data were released at the Third Conference on Retroviruses from various studies of nucleoside analogs, as monotherapies or in combination with each other. In light of the profound activity of the soon to be available protease inhibitors, it may seem as though data on this more modest class of drugs is moot. But given the development of resistance to protease inhibitors, there clearly remains a role for these drugs as a means of maximizing the antiviral response and forestalling viral resistance to protease inhibitors. Besides the question of resistance, there will also be a sizable number of people who fail or cannot tolerate protease inhibitors due to drug-drug interactions and gastrointestinal or liver issues (poor absorption or rapid breakdown of protease inhibitors as well as the drugs' toxic effects) -- and then there is protease inhibitors' high cost.
The Old Guard Combinations
New data from various studies largely paralleled the previously reported results from ACTG 175 and Delta 1 trials (see last October's Treatment Issues, pages 2-4). Robert Schooley, M.D., presented data from a Burroughs Wellcome sponsored study of AZT versus AZT/ddC and AZT/ddI (abstract 202). The trial enrolled people with CD4 cell counts below 300 and less than four weeks of prior AZT. The primary goal was to see whether the addition of ddI (at half dose) or ddC delayed the development of resistance to AZT. They did not, but AZT/ddI was associated with a significantly enhanced suppression of HIV and extension of life as compared to AZT alone.
AZT/ddI also produced a significant survival advantage over AZT monotherapy in people with prior nucleoside analog experience in the Delta 2 study, now that more deaths have occurred during the long-term follow-up (abstract LB5). Most of the people who died had entered the study already diagnosed with AIDS.
The Community Programs for Clinical Research on AIDS (CPCRA) network sponsored the NuCombo study, which also compared AZT against AZT/ddI and AZT/ddC. For the trial cohort as a whole, NuCombo found only a marginal survival benefit for AZT/ddI after two years of observation (abstract LB4). The 1,113 participants in this study had more advanced disease status than those in Delta or ACTG 175. NuCombo participants entered the trial with an average CD4 cell count of 119.
In the subset of trial enrollees who were initially AZT-naive (about 23 percent of the total enrollment), combination therapy had a significant clinical benefit over AZT alone. There was no benefit to combination in the AZT-experienced patients. Combination therapy also was poorly tolerated, calling into question the wisdom of adding ddC or ddI to AZT in people with very advanced disease. For ddI, in particular, tolerance and compliance were lower because of gastrointestinal side effects.
Long-term follow-up of NUCA 3001 and 3002 (AZT/3TC studies in antiretroviral naive and experienced patients, respectively) shows that viral load and CD4 cell responses in the 3TC-containing arms at week 24 were sustained in a high proportion of persons who continued treatment out to 68 weeks in NUCA 3001 and 52 weeks in NUCA 3002 (abstracts 198 and 199). In NUCA 3002, no difference in clinical endpoints was seen between arms. The clinical endpoint differences in NUCA 3001 were not statistically significant.
Two posters addressed the effect or safety of 3TC in people with advanced disease. A French group evaluated two doses of 3TC combined with AZT in seventeen people with less than 50 CD4 cells (abstract 299). Both doses lowered viral load and increased CD4 cell counts. In eight patients at the month one timepoint, viral load decreased by more than 90% or became undetectable. This was sustained at month three in four trial participants. At month six, fifteen out of the seventeen 3TC-treated volunteers were alive, only one had had an opportunistic infection.
A Toronto physician, I. W. Fong, M.D., reported that 3TC is not as well tolerated in advanced HIV disease as in earlier stages of the disease and can cause serious side effects (abstract 134). His data came from 36 of his advanced patients enrolled in the 3TC compassionate use program (mean CD4 count at baseline was 80 cells). Fifteen took 3TC as a monotherapy, and the rest used it in combination with AZT, ddI or ddC. Seventeen patients reported serious side effects, with seven discontinuing 3TC as a result. Among the side effects were five cases of hair loss (13.9 percent), five cases of neuritis (13.9 percent), four cases of anemia (11.1 percent), four cases of leukopenia, or low white blood cell count (11.1 percent), including two cases of severe neutropenia (less than 500 of the infection-fighting neutrophil cells per mm3 of blood), and three cases of nausea and vomiting (8.3 percent). More side effects were reported on monotherapy (12 of the 21 events) than on combination
d4T and d4T/ddI
Christine Katlama, M.D., presented data from a placebo-controlled dose-ranging study of d4T (abstract 196). The trial enrolled 66 people with more than 350 CD4 cells (baseline median 527). Viral load (4.2 log or 15,900 copies/ml of plasma at baseline) decreased by 0.8 log (84 percent) on the high dose (80 mg per day), and 0.6 log (75 percent) on the low dose (40 mg per day). These reductions were sustained for the duration of the twelve week study.
Average CD4 cell counts at week twelve were 35 cells over baseline on the high dose and twenty cells over baseline on the low dose. These increases compare with a decline of 45 cells in the placebo group, but the CD4 cell changes were not considered significant.
Forty patients continued with open-label low-dose d4T. Those who had been on placebo had a seventy percent reduction (one-half log) in viral load. The antiviral responses were sustained through month twelve, although viral load rose slightly in the patients randomized to high dose when crossed over to low dose d4T. During the follow-up period, there were nine cases of peripheral neuropathy, leading one patient to stop the d4T.
Meanwhile, Swiss researchers reported on a study in 23 persons with late-stage disease and a long history of prior nucleoside use (abstract 136). d4T had only a minor effect on viral load (0.2 log or 37.5 percent reduction) after one month, and that effect was lost by month three. As with the NuCombo study, this may be another case of a nucleoside analog not performing well in heavily pretreated patients.
Finally, the results reported on the antiviral activity in an ongoing study of the combination of d4T with ddI (abstract 197) were less robust than reported to activists last November (see the November, 1995 Treatment Issues, page 3). The data from this trial is in a constant flux as more people reach later timepoints.
Ninety-four patients have been randomized into the study; 71 are evaluable at present. Viral load responses were presented only for 54 people who entered the study with plasma viral load of over 1,000. The median CD4 cell count in these 54 patients was 310 (average CD4 count: 327), and the median viral load was 4.6 log (39,800 copies/ml). The median CD4 cell count at entry in all 71 evaluable patients was slightly higher at 325 (the average was 336).
More than half of the study participants have had at least 28 weeks of treatment. The average reduction in viral load was 1.4 log (96 percent) in the 32 on treatment at week 28. Contrary to appearances last November, the average viral load response is not increasing over time, but it was sustained at 1.5 log (97 percent) below baseline in the fourteen patients with measurable viral load that have completed 52 weeks of treatment. So far, there have been no statistically significant differences between the various dose arms of the study. The variation in individual response was great, though, with reductions in viral load ranging between 0.29 log (50 percent) and 2.9 log (99.8 percent).
Overall, 26 patients have completed 52 weeks of treatment and their CD4 counts are up by a mean of 68. At present, there appears to be a CD4 response favoring the high dose. Counts increased by 141 cells in the four on this arm at week 52, as opposed to a loss of more than twenty cells on the low dose. Again, these as yet nonsignificant differences are sure to change as more people complete the study.
d4T and ddI both cause peripheral neuropathy as a side effect, but only one episode has been noted in this trial. Six patients discontinued therapy due a variety of other side effects, including increases in liver function tests, abdominal pain, neutropenia and depression.
A Surprisingly Potent Nucleoside Analogue
by Gabriel Torres, M.D.
1592U89 is a new nucleoside analogue under development by Glaxo-Wellcome (see Treatment Issues, June 1995, page 12). It is known to have potent in vitro activity against HIV and to be synergistic with all other nucleoside analogues and protease inhibitors.
At the Retroviruses Conference, Michael Saag, M.D., of the University of Alabama presented the highly encouraging first in vivo efficacy data for this compound (abstract 195). These were the interim results from the international phase 1/2 trial underway in Europe and the United States. The study is a dose-ranging one in HIV-positive adults with less than twelve weeks prior AZT therapy and CD4 counts between 200 and 500.
After four weeks, the average HIV viral load dropped in continuing trial participants by a median 1.9 logs (98.7 percent), and after a further eight weeks, in which participants received either AZT or placebo in addition to 1592U89, the viral RNA level was down by 2.1 logs -- representing a 99 percent reduction from baseline at twelve weeks. CD4 counts had increased by an average of 122 at week twelve.
This first cohort of nineteen patients received 1592U89 monotherapy at an oral dose of 200 mg thrice daily. The group's average CD4 count at entry was 352, and its average viral load was 5.1 logs. All but one of the patients were AZT-naive.
Two participants withdrew from the study, one after developing fever, rash and tingling thought to be due to the drug. The main side effects of 1592U89 were nausea in 28 percent of volunteers, headache in 17 percent and rash in 11 percent. One person had an elevation in liver function tests.
The exceptional results for this nucleoside analogue have prompted the investigators to continue the trial. At present, they are enrolling three additional cohorts for higher doses. Since the drug penetrates into the central nervous system, trial participants are asked to allow spinal taps in order to measure the concentration of the drug in the cerebrospinal fluid.
Pivotal trials in AIDS-related dementia also are being developed for 1592U89, as are novel trials that will combine the drug with AZT and 3TC. The new trials are due to begin once the last of the present cohorts has finished. For more information, call Glaxo at 800/437-0992.
And also, "Adefovir Dipovixil"
by Theo Smart
Adefovir dipovixil (formerly bis-POM PMEA) is an experimental anti-HIV drug made by Gilead Sciences, which presented data on it at the Retroviruses Conference (abstract 407). It has the advantages of a very long half-life within the body (leading to a once-a-day dosing regimen) as well as broad spectrum antiviral activity against hepatitis B and herpes viruses -- particularly CMV and Epstein Barr -- in addition to HIV.
Adefovir (and PMPA, a new derivative now entering clinical studies -- see December's Treatment Issues, page 3) is a nucleotide analog that is similar to nucleoside analogs such as ddI but does not need to undergo as much processing within cells to become active.
Both nucleotide and nucleoside analogs attack reverse transcriptase, so HIV resistance can cover drugs in both classes at once. In particular, HIV mutations that confer a ten-fold loss of susceptibility to adefovir in the test tube also reduce susceptibility to ddI and ddC by four fold and to 3TC by 40 fold. These laboratory derived mutations have not been observed yet in any patients treated with adefovir.
Data from a twelve week placebo-controlled study of adefovir, at 125 or 250 mg per day, were presented during the Retrovirus Conference. Both doses reduced viral load by 70 percent, (roughly equivalent to the effect of AZT), and viral load rebounded when treatment stopped. The CD4 cell response was greater on the lower dose, a boost of 57 versus 27 for the high dose. CD4 counts decreased by 41 in those on placebo. Side effects, mostly nausea and diarrhea, happened more frequently in the high-dose arm.
Researchers at one study site found that adefovir treatment reduced levels of CMV in the semen by more than 50 percent in four out of seven patients. In test tube studies, adefovir is as active against CMV as ganciclovir, while it is much more bioavailable in people than oral ganciclovir. Gilead Sciences plans to look at adefovir both for HIV and CMV in upcoming large phase 3 studies.
Ironically, broad spectrum antiviral drugs like adefovir are a cause of concern for some researchers. They fear that those who take drugs like adefovir for HIV disease risk developing breakthrough CMV infections cross-resistant to adefovir and other CMV medications as well. As of yet, there is no data on whether CMV can evolve resistance to adefovir, but it would be surprising if it could not, especially should the drug enter widespread use. If CMV-resistance to adefovir confers cross-resistance to the other CMV treatments, it might be best to limit CMV's exposure to the drug by reserving adefovir for those needing CMV prophylaxis.
by Theo Smart
Several poster presentations at the Conference on Retroviruses tried to make a case for the HIV reverse transcriptase inhibitor nevirapine (brand name: Viramune).
Richard Pollard, M.D., reported on a study comparing AZT to nevirapine plus AZT in 60 AZT-experienced volunteers (abstract 325). The nevirapine plus AZT cohort saw viral load reduced by a median of over 1.57 log (98 percent), but viral load was on its way back to baseline by week four and at baseline by week fourteen. CD4 cell counts had increased by a median of 60 at week eight but were back to baseline by week 28.
Nevirapine's developer, Boehringer Ingelheim also presented a paper on the duration of the drug's effect (abstract 141) based on an analysis of three separate studies including ACTG 241 -- a "convergent combination" study of nevirapine/AZT/ddI. The addition of nevirapine to AZT or AZT/ddI regimens did delay the "return of viral load to baseline" by a few months. Viral loads may linger slightly below their pretreatment value for months without being significantly different from that value.
In March, Boehringer Ingelheim, intends to request FDA marketing approval for nevirapine. While awaiting approval, the company has announced plans to proceed with an expanded access program to provide nevirapine to people with CD4 counts below 50. For more information on eligibility, call 800/595-5494.
Due to an error in the trial announcement, last month's Treatment Issues
published an incorrect telephone number for information on the currently enrolling large clinical trial of nevirapine. The correct number is 919/544-3170.
Curcumin Trial Finds No Activity
by Dave Gilden and Theo Smart
Curcumin, a component of the spice turmeric and popular herbal medicine, has no anti-HIV activity in people according to a study presented by the Community Research Initiative of New England (CRINE) at the Retrovirus Conference (abstract 140). The report stands in contrast to an earlier, smaller study by the Search Alliance (now the AIDS ReSearch Alliance) of Los Angeles which observed a reduction in viral load. Interest in curcumin increased when Search Alliance announced that 2,500 mg per day of curcumin produced a "modest" reduction in viral load over twenty weeks.
Researchers at CRINE attempted to reproduce these results by enrolling 40 people in an eight-week trial. Two doses were used, 2,700 mg/day and 4,800 mg/day, and enrollees were allowed to take concurrent antiviral drugs.
The results were nil for both doses. Viral load (an average of 61,000 copies/ml at baseline) did not change significantly over the eight weeks, nor did CD4 count (baseline average of 236).
It is not clear why the two community-based research organizations came to such different conclusions. Ten of the eighteen Search volunteers received flu shots during the trial, which probably distorted the viral load data. The accuracy of the Search trial's viral load testing also has been questioned. (See AIDS Treatment News, May 6, 1994, pages 1-7.)
The reason for curcumin's lack of activity may be that it is very poorly absorbed by the digestive tract, according to animal studies. Levels of curcumin in the blood were not measured in either study. CRINE plans to do this shortly.
Hydroxyurea Continues to Show Promise
by Luis Santiago
Further human trial data on hydroxyurea and compounds with similar effect were presented at the Third Conference on Human Retroviruses and Opportunistic Infections. The new data confirm the positive impression that hydroxyurea, especially, has made in past investigations (see the September, 1995 Treatment Issues, pages 1, 2-4).
Hydroxyurea and the related drugs act by restricting the cellular pathways by which new DNA nucleotides are made. These gene-creating building blocks are taken up by the HIV enzyme reverse transcriptase as HIV infects new cells. The resulting scarcity of natural nucleotides allows nucleoside analog drugs like ddI or AZT, which are defective nucleosides, to play a greater role in disrupting HIV.
The most significant new piece of information on hydroxyurea (abstract 406) was presented by Julio Montaner, M.D., head of a Canadian pilot study of hydroxyurea among a group of patients with CD4 counts between 100 and 350 (average CD4 count of 241) and at least six months prior ddI experience. The objective of the study was to assess the viral load effect of adding hydroxyurea to ddI monotherapy.
The study was set up as follows: all trial participants (26) received ddI monotherapy for four weeks (200 mg twice a day). The ddI alone therapy was followed by four weeks of ddI plus hydroxyurea, with half of the group (13) assigned randomly to a low dose (500 mg once a day) and the other half to a high dose (500 mg twice a day) of hydroxyurea. For the last four weeks of the twelve-week study, everyone was put back on ddI monotherapy. Viral load (plasma levels of free HIV) and CD4 counts were measured on a daily basis.
For study purposes, an individual was defined as a "responder" if he or she registered an HIV viral load reduction greater than one log (90 percent). Based on this definition the investigators found that during the four weeks on ddI/hydroxyurea combination, nine out the 26 patients were responders. Six out of the nine responders were on the hydroxyurea high dose, with the remaining three on the low dose. This difference is only great enough to statistically suggest, but not demonstrate, that a greater viral burden reduction can be achieved with the higher dose.
The average viral load reduction (70 percent and 90 percent at week seven in the low and high dose arms, respectively) was statistically significant, as was the viral load rebound after discontinuation of hydroxyurea therapy at week eight. This return to baseline HIV levels occurred during the first week off hydroxyurea.
In the initial laboratory experiments at the National Cancer Institute, inhibition of HIV-1 was still seen several weeks after stopping drug treatment. The rapidity of the rebound in the present human trial indicates that continued viral suppression after withdrawal of hydroxyurea may not occur, at least in this population with advanced HIV infection and high CD4 turnover. (The CD4 cells arising after treatment cessation would be unaffected by the previous exposure to hydroxyurea.)
Many hydroxyurea researchers argue that the ddI/hydroxyurea combination will be of significant benefit only for a relatively healthy, ddI-naive population since this is where a synergistic effect would be seen and toxic side effects (bone marrow suppression) should be minimal.
Considering the advanced disease status and long treatment history of the trial cohort, the viral load reductions were impressive. Investigators are currently analyzing the prevalence of ddI resistance in study participants' HIV and its association, if any, with the magnitude of the response.
No serious toxicities were reported in the study, although some conference attendees pointed out that the time on hydroxyurea was too short to evaluate any negative effects on blood cell counts. Likewise, no change was observed in CD4 counts probably because of the trial's brevity.
Follow-up to the French Trial
Interest in hydroxyurea had been considerably heightened by the results of a French study involving twelve asymptomatic HIV-positive patients (CD4 count above 250) with no previous anti-HIV therapy. (See the September, 1995 Treatment Issues, page 1.) The group received ddI (200 mg twice a day) and hydroxyurea (500 mg twice a day) for three months.
In a follow-up report (abstract 291), researchers presented data on those twelve patients, six of whom stopped the study regimen at the end of the three-month observation period. They also reported on an additional group of six patients that started treatment with this combination since then and also have completed three months of therapy.
At eight months, HIV viral load increases were reported in five out of the six persons that stopped treatment. The other one was stable. CD4 counts held more or less steady in four of them. Of the six patients who have remained on therapy, two have stable HIV levels, one had a major increase and two had slight increases. (No data were reported for one member of the group.) CD4 counts continued to increase in three members of this group.
Of the six patients that started the combination recently, two had HIV levels below detectable limits at three months, one experienced a two log (99 percent) decrease, and the other three a one log (90 percent) decrease.
This study is not a rigorous clinical trial and leaves a lot of room for argument over its interpretation. Along with the Canadian trial, it does at least provide a basis for more sophisticated testing of hydroxyurea.
Where's the Placebo?
Here in the U.S., ACTG 307, a pivotal ddI/hydroxyurea study, was due to start last fall. According to the trial's principal investigator, Joseph Eron, M.D., the lack of a placebo has been a big stumbling block. Bristol-Myers Squibb, producer of hydroxyurea (brand name: Hydrea) has yet to provide the placebo pill, which has to resemble the real thing. A company in Maryland is currently working on it, but researchers are not ready yet to start recruiting. For the latest information and current trial sites, call 800/TRIALS-A.
Alternative experimental medications utilizing the same strategy as hydroxyurea are also waiting for more extensive testing.
Dr. W-Y Gao of the National Cancer Institute described compounds that block the cellular enzyme known as thymidylate synthase (abstract 5). This enzyme helps process the nucleotides for which AZT and d4T are defective analogs. (See Treatment Issues, September 1995, page 5. Hydroxyurea inhibits a different enzyme, ribonucleotide reductase.) In the test tube, very low concentrations of the anticancer drugs FUdR (floxuridine) and 5-FU (fluorouracil) potentiated both AZT and d4T as predicted. FUdR raised d4T's potency eight times, and AZT's effect was six times greater in its presence.
Dr. Gao's presentation was seconded by a report from the Robert Johnson School of Medicine in New Jersey and the University of Rochester (abstract 344). This group used FUdR to potentiate AZT in the test tube.
Both FUdR and 5-FU are extremely toxic, which could limit their use. These drugs' success as antiviral agents depends on their effectiveness within the body at the very low, presumably nontoxic concentrations used in the laboratory experiments.
ADAP Faces Financial Abyss
by Derek Link
As the good news on protease inhibitors and viral load testing was broadcast from the Third Conference on Retroviruses last week, the ability of low-income people with HIV to access these promising developments was dealt a major blow. The states' AIDS Drug Assistance Programs (ADAPs) are running out of money, and no solution is in sight. Increased caseloads and drug usage, higher drug costs, and more expensive combination therapies coupled with stagnant financial resources have placed ADAPs across the country in jeopardy.
ADAPs are the most heavily utilized AIDS programs in the nation, with over 50,000 people with HIV or AIDS enrolled. The federally funded ADAP system began in 1987 as a way of providing free AIDS drugs to low- or moderate-income people not covered by Medicaid or third-party insurance. Since 1990, ADAPs have been funded by the Ryan White CARE Act, the main source of federal funds for AIDS care. Each state administers its own ADAP, so eligibility criteria and formularies (the list of covered drugs) vary from state to state. Some states, like New York, have run generous programs that cover most AIDS-related medications, while other states have restrictive programs that cover only a few (usually AZT, aerosolized pentamidine and Bactrim).
The Gathering Storm
In the last several months, two state ADAPs (Colorado and Missouri) have run out of money and others have cut services, limited enrollment or canceled formulary expansions in the face of growing budget shortfalls. The budget problems with the ADAPs appear certain to grow in the coming fiscal year as new protease inhibitors gain FDA approval and multi-drug combination therapies are prescribed earlier in disease, increasing the cost and use of antiretroviral drugs.
Use of antiretroviral agents has increased dramatically over the last twelve months, following a period of decline after the Concorde study results in 1992. In New York State, for example, expenditures on antiretroviral drugs from fiscal year 1993 to 1996 are projected to almost double, rising from $5.9 million to $11.2 million. 3TC is one significant case. Twelve hundred people obtained 3TC through New York ADAP in the month following its approval in November, and the state ADAP spent $300,000 on the drug in its first month alone. The demand for 3TC represented a record level of interest in a new therapy. 3TC, which is used mostly in combination with AZT, is also associated with increased AZT expenditures.
The National Association of State and Territorial AIDS Directors (NASTAD), a lobbying group, surveyed state ADAPs in January. Forty-two states responded to the survey. NASTAD found nineteen state ADAPs had or will have budget shortfalls this fiscal year. The budget gaps ranged from $5.9 million in New York to $15,000 for Nebraska. NASTAD calculated that a total of $12 million would be needed just to make up the budget gaps this fiscal year. States have implemented a variety of measures to reduce ADAP expenditures, the most common of which are delaying or canceling formulary expansions (twelve states), removing drugs from the formulary (seven states) and establishing waiting lists (seven states).
New York Cuts Back
On January 1, New York State implemented drastic cost-cutting measures for its ADAP. Over 70 percent of the drugs covered by New York ADAP have been eliminated from the program, including critical medications such as Neupogen, Epogen and many antibiotics. In addition to the cuts, New York ADAP is unable to cover any new therapy. Saquinavir, the first approved protease inhibitor, is not covered, and other new protease inhibitors will not be covered when they receive marketing approval later this year.
Despite the budget problems, interest in protease inhibitors remains high in New York State. New York ADAP administrators say that they have received approximately 120 phone calls a week asking for saquinavir since it was approved late last year. New York State's Clinical Advisory Committee for HIV Uninsured Programs, which advises ADAP on clinical issues, this month recommended protease inhibitors be covered by ADAP but not at the expense of other cuts in the program.
With more cases than any other state, New York is the epicenter of the HIV/AIDS epidemic. New York's ADAP is accordingly the largest in the country. Since its inception, New York ADAP has covered over 32,000 New Yorkers with HIV. Its current enrollment exceeds 10,000. About 70 percent of New York ADAP recipients live in New York City, and over 60 percent earn less than $10,000 per year. Seventy-five percent of NY ADAP participants use the program while their Medicaid applications are being processed, which often takes six months, and the remaining 25 percent are uninsured or underinsured.
New York State estimates it needs $49 million in 1996 to maintain ADAP at its current reduced level. The state will need an additional $9 million to keep the current level of service while adding the protease inhibitors. Seventy-five million dollars is required to restore all cuts and add the protease inhibitors, too.
Lobbying For ADAP
The ADAP funding shortfall has led AIDS organizations to press for more funds at the state and federal level. Nationally, funding for ADAP is part of the debate on reauthorization of the Ryan White CARE Act. The CARE Act is stalled in committee, and many AIDS lobbyists worry that it may remain stalled because the 104th Congress has had a very difficult time passing any legislation. If the CARE Act is not reauthorized soon, an even deeper funding crisis may occur in all CARE Act programs, including the ADAPs.
In any event, NASTAD and other groups have called on the Congress and the Administration to increase the CARE Act appropriation for ADAPs. An "ADAP Future Funding Working Group" has also formed in Washington. This working group is composed mostly of lobbyists who represent the companies that make and sell medications for AIDS and HIV.
The Health Resources Services Administration (HRSA), the federal office that administers Ryan White programs, is also developing a national "core formulary" for ADAPs. The goal is to standardize ADAP coverage in all states so that an even level of service can be provided nationwide, among other things allowing for more accurate planning. HRSA expects to have elaborated the "core formulary" by early summer.
In New York State, a group of AIDS organizations and legislators have joined together in a campaign to save ADAP. The New York State group is pressing the government in Albany to supplement the ADAP budget with state funds -- right now, New York uses none of its own funds for ADAP. A bill that would use state funds to fill any shortfall in the state ADAP budget has 31 sponsors in the Assembly but has yet to be introduced in the Republican-controlled Senate.
State funds do subsidize ADAP in California, where ADAP is running a $4-8 million deficit and is estimated to need an extra $18 million annually just to maintain current service and add 3TC and saquinavir. Activists are trying to gain extra budgetary allocations. At the same time, they fear bringing ADAP to the attention of the new conservative Republican majority in the state Assembly.
ADAP's woes underscore a frightening reality in the American health care system. More and more people with HIV who rely on government health care or managed care plans may find that newer and more expensive drugs remain unreimbursed. Breakthrough drugs that have a major life-prolonging potential for people with HIV, like protease inhibitors, may only be available to the very few who have jobs that offer top-of-line insurance coverage. A concerted effort is still needed for comprehensive health care reform, but in the meantime AIDS groups must focus attention on the growing numbers of persons with low and even moderate incomes who cannot gain access to protease inhibitors and other innovative HIV therapies.