GMHC Treatment Issues
April/May 1997

Now There are Eleven,
but So What?

Reading delavirdine's package insert is a scary experience that will put off most people from trying the compound. In the first place, resistant HIV emerges within a few weeks of taking the drug, due to a single mutation. That mutation most often occurs at reverse transcriptase's amino acid 103. This same mutation also reduces HIV's sensitivity to Du Pont Merck's experimental and much more active NNRTI DMP 266. It would be a shame to lose some of the advantages of DMP 266 through previous use of delavirdine. Largely because of its resistance problem, delavirdine provided no statistically significant benefit in two of the three clinical trials that led to its approval. One of these, ACTG 261, involved participants who had no prior therapy or less than six months of AZT. Preliminary analyses of the viral load and CD4 count data in ACTG 261 indicate that AZT/ddI is virtually equal to AZT/ddI/delavirdine or ddI/delavirdine and decidedly better than AZT/delavirdine. In another study, the AZT/delavirdine combination did do better than AZT alone in a cohort with six months or less of prior AZT therapy. This improvement was not apparent when evaluating CD4 count changes, only in terms of viral load. (A further reduction in HIV levels of only about 0.1 log -- 20% -- after week eight could be attributed to delavirdine. At week four, this difference briefly reached 0.5 logs or 67%.)

This is not a very attractive drug. Except, except... One of the pieces of information that led the FDA to OK delavirdine was some data from Paul Bellman, M.D., a New York City AIDS specialist in private practice. Taking advantage of the preapproval delavirdine expanded access program, Dr. Bellman added the compound to the regimens of 60 of his patients who were experiencing rising viral loads despite being on AZT/3TC/indinavir. (In some patients, the AZT was switched to d4T at the same time). Lo and behold, 15 of the 27 patients whose records have been made available experienced a rapid viral load decline amounting to more than one log (90%) by the second or third month. Dr. Bellman has no way of checking, but he thinks that the people failing the indinavir regimens were not achieving high enough blood levels of this protease inhibitor, possibly due to poor absorption in the digestive tract. Delavirdine conceivably could have raised the patients' average indinavir levels by two-fold or more, due to its effect on the liver, where 80% of indinavir is broken down.

Dr. Bellman observes, "Delavirdine added to nucleoside analogs does not have these benefits. But adding delavirdine, a potent drug, to indinavir makes it potent, too. So now there are two potent new drugs in the patient's antiviral combination."

Dr. Bellman combines delavirdine with the standard indinavir dose of 800 milligrams three times a day. Upjohn, though, recommends that indinavir should be reduced to 600 mg three times a day when co-administered with delavirdine. The worry with high indinavir levels in the body is that some of the compound will end up in the kidneys as an extremely painful precipitate or sediment. Dr. Bellman says he has noticed no increase in this kidney stone problem, which strengthens his feeling that his failing patients had abnormally low indinavir levels to begin with.

The doctor's whole strategy rests on his personal intuition. He has not actually monitored the indinavir levels in any of his patients since such a test is not commercially available. His experience points out a gap in current research -- why people fail treatment regimens is not examined systematically. In particular, the drug levels achieved in patients are rarely looked at.

Merck, which developed indinavir, is now collecting blood samples from people who have not responded well to the compound. The reasons why people fail seem varied, according to Merck researcher John Condra. He says, "We don't have any evidence that failures have low blood levels. This is one possibility. Among the other explanations are drug resistance and inadequate compliance." If the problem is failure to take indinavir on schedule or to take it with food rather than on an empty stomach, this should be corrected first before adding delavirdine. If the problem is indinavir-resistant HIV and not low indinavir levels, the benefits and dangers of adding delavirdine are uncertain.

Adding to the mystery, Dr. Bellman notes that he has seen patients continue to fail when delavirdine was added to their indinavir regimen and then succeed on indinavir plus nevirapine. But nevirapine reduces indinavir levels rather than raising them.

Like the reasons for protease inhibitor failure, protease inhibitor/NNRTI combinations have suffered from lack of attention. Manufacturers in the past have tried to position their NNRTIs as substitutes for the more expensive protease inhibitors in combination regimens with nucleoside analogs. It nonetheless makes more sense to combine NNRTIs with protease inhibitors rather than nucleoside analogs since then two different HIV enzymes and points on the virus lifecycle are attacked rather than just one. In one small Vancouver trial, for people with AIDS who had failed all nucleoside analogs, the combination of 3TC, indinavir and nevirapine was surprisingly successful (see Treatment Issues, February 1997). Also, the two-drug combination of DMP 266 and indinavir in trials has suppressed HIV as effectively as most three drug combinations.

Dr. Bellman's results in his first patients are drawing increasing attention in New York City and elsewhere. The complete results are not in, though. For most of these initial patients, adding delavirdine to indinavir did not drive HIV down to undetectable levels (below 400 HIV RNA copies/ml of plasma). This leads to fears that long-term stability will not be maintained, given HIV's propensity to develop resistance to either of these two compounds when residual viral replication is possible during treatment. The clock may be set back, but usually it is still ticking; successful ways to recover from treatment failure need further elaboration.

Part of that solution appeared to be Upjohn's "spawn of delavirdine," a series of altered delavirdine compounds that the company tested in the lab and found to be active against delavirdine-resistant HIV. These compounds were described in a series of four papers over the past year, but Upjohn now says it has no attention of developing them further because they exhibited poor stability in animal studies. More recently, Upjohn has come up with a new set of NNRTIs, the "pryrimidine thioethers," which also are potent against delavirdine-resistant HIV as well as nonresistant virus. Their absorption and stability reportedly are excellent in rats, but Upjohn is still evaluating their future development. There is considerable speculation in the AIDS community that Upjohn is pulling back from HIV research. It will be interesting to see what the company does.

Ganciclovir Implants:
One Year Later

The Implant and Retinal Detachment

After having performed the operation on hundreds of patients, Dr. Martin now feels that the problem was somewhat overstated. Although he maintains there is an increased risk in the range of 5% to 10% in the first two months after the implant surgery, he estimates that the one-year risk is about the same as for those patients on systemic treatment. Similarly, Murk-Hein Heinemann, M.D., of Cornell University Medical College, presented a study in May at the Association for Research in Vision and Ophthalmology (ARVO) that found a total 4.8% detachment rate following implantation (abstract 3415-B16). The Cornell study concluded that detachment is an uncommon complication with the implant.

Dr. Martin stated that CMV activity itself is probably the single most important risk factor in retinal detachment, followed by location and extent of disease. "Every time you get a [CMV] reactivation, you get areas of atrophy and small holes can develop in the retina. It makes sense that an implanted eye over time is going to have lower risk because you don't get the multiple reactivations that you do with IV meds."

Joseph Eviatar, M.D., an ophthalmologist in New York City, also voiced early worries about the risk of retinal detachment. He now concurs with Dr. Martin's observation that the rate of detachments appears to be about the same with the implant as with systemic treatment. Dr. Eviatar said that his experiences have been overwhelmingly positive. "Patients have had their CMV controlled for six to eight months with a single implant and have been able to preserve vision. Most patients have a couple of weeks of slight blurring of vision immediately after the surgery, but that usually resolves."

There have been problems in the past obtaining insurance reimbursement for the implants, which cost about $4,000 plus surgery expenses. These seem to have been largely resolved. Anyone having difficulty with reimbursement or lack of insurance coverage may call Chiron Vision's patient assistance program at 800/EYE-IMPL.

Concurrent Systemic Treatment

Documenting the value of systemic protection is critical. Oral, not to mention IV, ganciclovir is a hard drug to take. Therapy requires 12 to 18 pills a day, at a cost of at least $15,000 a year. Many patients on oral, as well as IV, ganciclovir experience bone marrow suppression, evidenced by reduced numbers of white blood cells (29% of patients on the oral formulation) and red blood cells (19%). Intravenous Neupogen (G-CSF) or Epogen (erythropoietin) can counteract the effect on white or red blood cells, respectively. On the other hand, all patients not on systemic medication must be monitored closely so that treatment can begin at the first sign of CMV progression to new tissue.

Replacing Vitrasert: Immediate or Deferred?

With the advent of highly active antiretroviral therapy (HAART), some doctors hoped that improvements in CD4 counts would translate into lower risk for reactivation of CMV disease, calling into question the need for pre-planned implant replacement. However, increasing a patient's CD4 count through HAART does not always prevent CMV reactivation. Dr. Martin stated, "I stopped doing scheduled exchanges for awhile and was disappointed. I had some patients with very high CD4 counts and their CMV reactivated at about the same time I would have expected."

Dr. Martin advises those who have Zone 1 (central) disease to have scheduled replacements because they cannot afford to take a chance on loss of vision. For those with Zone 2 (peripheral) disease, Dr. Eviatar stated, "If the patient is doing better systemically and the lesion is not site-threatening, I think it's reasonable to wait and see how they respond. It may be that with the protease inhibitors, they are better able to control the CMV on their own or with oral ganciclovir."

The Effect of Anti-HIV Treatment on CMV

Caution should be used when interpreting this small number of case reports. The first author of the letter and clinical director of the National Eye Institute, Scott Whitcup, M.D., stated, "We will need more information before recommending that all people with CMV retinitis and elevated CD4 counts stop their anti-CMV medications." Dr. Whitcup would not recommend coming off anti-CMV medication outside of a carefully picked patient population with extremely close supervision.

Mark Jacobson, M.D., and colleagues at the University of California San Francisco presented a study at the Fourth Conference on Retroviruses and Opportunistic Infections (abstract 353) representing the other end of the spectrum. The Jacobson study looked at five patients who had CMV retinitis diagnosed despite a CD4 count of 200 or more. Four to 24 weeks before the diagnosis, all patients had CD4 counts below 85, and four to eight weeks before the CMV diagnosis, all patients had initiated HAART. The researchers then examined data from 76 consecutive patients enrolled in a treatment trial for CMV retinitis. During the period from July, 1995 through February, 1996, only one patient out of 27 had CD4 counts above 50 and zero out of 27 patients had CD4 counts above 100. From March, 1996 (when HAART became widely available) through August, 1996, 14 out of 49 patients had CD4 counts above 50 and seven out of 49 had CD4 counts above 100. The investigators concluded that the immunological benefits of HAART may not fully protect against CMV retinitis, which may occur more commonly now in patients with CD4 counts above 50.

The authors of the JAMA letter suggest that such cases of CMV may be due to subclinical (asymptomatic) CMV retinitis existing prior to the onset of HAART. Dr. Whitcup notes that retinitis developed in patients in the Jacobson study shortly after initiation of HAART and feels that CMV would have eventually emerged in these patients in any case. Dr. Whitcup has seen his own patients develop such symptoms of CMV retinitis as vitreous floaters shortly after starting on HAART. It is possible that HAART cannot immediately protect a patient who already has a subclinical case of retinitis, but it may be protective against CMV progression over the longer run no matter what the initial status of the disease.

Dr. Whitcup said, "I have yet to see someone develop CMV retinitis who has been on HAART for over three months and whose CD4 counts are over 50 -- although I think there may be cases out there where that occurs." Of course those who have been on combination therapy for extended periods may experience "HAART failure" and subsequent CD4 count declines. Once CD4 counts fall below 50, there will be renewed risk for CMV progression.

The National Eye Institute is seeking to recruit AIDS patients with non sight-theatening CMV retinitis for further study at the NIH. A small 12-person trial will attempt to provide further insight into when it is appropriate to discontinue CMV treatment in patients on HAART. Anyone interested in this trial should contact Cheryl Perry at 301/435-4559.

Improvements in the Pipeline

There also is room for improvement in oral ganciclovir for maintenance therapy to prevent CMV reactivation. The current standard three-gram dose is not as effective as IV ganciclovir. Raising the daily dose of oral ganciclovir to 4.5 grams works better (see Treatment Issues, September 1996), but requires even more pills and expense. Roche's ganciclovir prodrug (RS-79070) is much more absorbable in the gut than the present version of ganciclovir and breaks down into the active form once in the body. The prodrug's oral bioavailability is about 61%, compared to about 8% for the current oral formulation. Data presented by Roche this past winter at the Fourth Conference on Retroviruses and Opportunistic Infections (abstract LB19) demonstrated that the same levels of serum ganciclovir can be achieved with the prodrug as with IV ganciclovir.

Providing that the pharmaceutical companies move forward, the future standard of care might employ Vitrasert II to treat the local manifestation of CMV retinitis for two years and one pill of ganciclovir prodrug per day to control extra-ocular disease. This treatment strategy would avoid what Dr. Martin calls "the emotional roller coaster of going back and forth from induction to maintenance therapy," which is the current experience of most people with CMV retinitis.

As this article was being readied for press, Hoffmann-La Roche announced that it was deferring development of the ganciclovir prodrug -- see box.

Happiness Reigns in the Drug Marketplace

Once lured to court AIDS and gay community groups because of their wide consumer reach, drug companies are resorting to more traditional methods of direct-consumer marketing as competition among the protease inhibitors heats up. The market's dominant player, Merck, has felt this pressure the most.

Its newest and fiercest rival, Agouron, launched an aggressive campaign to publicize Viracept's low toxicity and ease of use upon its approval in March. Only one month later Viracept represented 23% of new protease inhibitor prescriptions, according to one analyst report.

With an estimated 10,000 new prescriptions being filled each month, drug companies are becoming more and more anxious to get the message out that theirs is the superior product by whatever argument they can advance. The irony is that the messages conveyed about HIV through the media continue to reduce the disease to mere imagery, even as treating it grows more medically complex. Such imagery includes the notion of a war being waged in the blood where billions of T-cells are destroyed every day, says Joseph Sonnabend, M.D., and that "now we have drugs that can stop it all." In Merck's ad, phrases like "battle against HIV" and the tagline "focus on the rest of your life" are typical of ideas feeding the popular perception that AIDS is a manageable, even curable disease. Roche's tagline "consider a protease inhibitor you can live with," which accompanies pictures of happy, smiling people, strikes a particularly ironic chord: though perhaps tolerable, the product in its present formulation and dose does little to reduce viral load.

All ads include the highly technical patient package insert as required by the FDA. The duality of advertisements, where self-promotional imagery juxtaposes objective information, is reflective of the larger issue that is at stake with this approach to disseminating information on the treatment options available to people with AIDS. "Maybe these companies are reaching people who aren't getting this information elsewhere, and maybe they're giving them very useful treatment information," says Spencer Cox of the Treatment Action Group, "On the other hand, these people may be getting information that is not accurate or which they can't evaluate within the appropriate informational context."

Indeed, if raising awareness about HIV and combination therapy among the public-at-large can be construed as one advantage of mass-advertising, then the consequences of increased brand-name recognition should be construed as a disadvantage. Given the highly complicated regimens that characterize triple-combination therapy, using potent imagery and slogans to describe a protease inhibitor can seem oddly inappropriate. "This is not Rogaine or an antihistamine," says Jeanne Bergman, a policy analyst at Housing Works, "I find it peculiar to see ads for such heavy, heavy drugs."

With this comes the fact that combination therapy is still a matter of choice best addressed through active dialogue between a patient and his or her physician. Brand-name recognition should carry no clout in a decision-making process that involves individual considerations for physician and patient alike, including patient history and ease of use. As Spencer Cox put it, "I'm not sure that engendering consumer confidence is the most effective way to ensure the best healthcare." Rather, says Mark Snyder, M.D., an infectious disease specialist with the Mid-Atlantic Permanente Medical Group in Washington, D.C., the advertising can do more harm than good by "creating the belief that one drug is superior to the others when, in fact, it is not."

But creating that impression seems to be what the pharmaceutical industry is focusing more of its efforts on doing. With analysts predicting that the protease inhibitor market will exceed $2 billion by 1999, protease inhibitors are becoming a promising commercial enterprise indeed (Alex Brown Report, April 29, 1997). Concerns about cross-resistance are further supporting the ambitions of companies like Merck, Abbott, Roche, and Agouron to make their protease inhibitor first in line for inclusion in triple combination therapy. Drug resistance is a completely unsettled field that lends itself to widely different spins by the various protease inhibitor manufacturers.

The AIDS community has been insisting that drug companies proactively communicate information about their products since the beginning of the epidemic. "PWAs revolutionized the way medicine is practiced in our country," says Joy Schmitt of Agouron, "They are our most educated consumers, and the industry has responded to that." Yet information about just how these new drugs will affect the future of people with AIDS or HIV remains scarce. HIV therapy has come a long way since the introduction of AZT and ddI. Today, the only thing that is definite about HIV treatment is the dissenting medical opinion on when to initiate therapy, and the plethora of possible combination protocols. Observes Harold Grossman, M.D., "The truth is always somewhere between what all the companies tell you."

As Abbott, Agouron, Merck and Roche ponder their response to each other's latest ads, the bottom line is that clinical data remain the sole basis for judging the value of their products. Generating solid, clear information on potency and durability of effect is the best marketing technique there could be, whether that information is communicated through patient education material, community groups -- or ads.

Roche Brings New Formulation
of Saquinavir To FDA

The recommended dose of the new version is 1,200 mg three times a day (TID), compared to 600 mg TID for the hard capsules. Taking this much will require 18 saquinavir soft gel capsules per day and provide eight to ten times the drug exposure of the current formulation. The eight-week dose-ranging study (NV15107) that identified 1,200 mg TID as the preferred dose recorded a respectable, though hardly record-breaking, 1.43 log (96.3%) drop in viral load among 22 volunteers on saquinavir monotherapy.

The safety study (NV15182) showed few adverse side effects with the new formulation, but caution should be advised for those with a history of liver disease. In addition to the safety data, the FDA has recently requested a head-to-head activity study comparing the two formulations to prove that the soft gel is an enhancement over the hard capsule. Sixteen-week activity data from this study (NV15355) should be available by mid-summer and will be submitted to the FDA during the review process. Roche expects to receive the FDA's go-ahead to market the improved saquinavir by the end of this year.

Community activists have raised concerns about the continuation of Roche's aggressive marketing campaign which is advertising saquinavir as "a protease inhibitor you can live with." The hard capsules are administered at 600 mg TID, a suboptimal dose that may breed resistance to both the new formulation and other protease inhibitors by failing to have much impact on HIV replication. At a community meeting in May, activists asked Roche to pull the campaign until the soft gel capsules are available. Tammy Lewis, Product Director for saquinavir, conceded that the best way to use the current formulation is in combination with the protease inhibitor ritonavir, which raises blood levels of saquinavir ten-fold by blocking the drug's metabolism in the liver. But the saquinavir/ritonavir combination has not been reviewed, much less approved, by the FDA, and Roche is prohibited from promoting it.

Roche is now pursuing approval of an expanded indication of saquinavir for use in combination with other protease inhibitors. Approval is not likely within the next 12 months and would probably be limited to specific drug combinations initially (in particular, 400 mg saquinavir plus 400 mg ritonavir), rather than a broad saquinavir/other protease inhibitor indication. Discussions have already taken place between Roche, Abbott and the FDA regarding registration trials in Europe for saquinavir/ritonavir.

Another long-standing community complaint is the current price of saquinavir. At $5800 per year wholesale, the hard capsule saquinavir is at least as expensive as the other protease inhibitors even if it is the weakest. Ms. Lewis says it is "premature" to comment on the pricing of the soft gel capsule. Especially at the preferred 1,200 mg TID dose, the soft gel could prove much more costly than the present version.

No Attrition in Research
on Wasting Therapies

It may be that once the conditions for wasting exist, the syndrome persists even when HIV levels decline. Several small surveys presented at April's Nutrition and HIV Infection conference in Cannes, France found that about a fifth to a quarter of those on aggressive antiviral therapies continue to lose weight or lean body mass. One of the studies observed that there was little correlation between viral load response to therapy and changes in weight. Serono Laboratories has made similar findings in an ongoing survey it is conducting. And the initial results from a study at Tufts University indicate that over the first seven to nine months on HAART, patients gain significant amounts of weight as they eat more, but it is all fat. (presented by Sherwood Gorbach, M.D., at the NIH Conference on AIDS Wasting Syndrome, May 20-21, 1997).

An inability to restore lean body mass (muscle in particular) after acute or chronic weight loss has been the hallmark of wasting syndrome. That inability remains even when a person recovers from an opportunistic infection that triggered the weight loss episode. Another on-going study at Tufts University is observing that, at least in people with high HIV levels, even strenuous exercise programs do not improve body composition. Trial participants did increase their weight by an average 3.8% after eight weeks, but the percent increase in fat and lean tended to be the same (presented by Ronenn Roubenoff, M.D., at the NIH Conference on AIDS Wasting Syndrome, May 20-21, 1997).

(A more definitive word on recovery after successfully suppressing viral load may come from the weight gain substudy that is part of the recently concluded ACTG 320, a trial of AZT (or d4T) plus 3TC with or without indinavir. The substudy will try to determine whether adding the protease inhibitor and obtaining greater reductions in viral load result in greater increases in lean body mass, and whether increased lean body mass is associated with changes in appetite and the body's pro-inflammatory agents (such as TNFa, see below). Until now, drug companies have either not collected or refused to release much information on weight changes in their protease inhibitor trials.)

Growth Hormone Hits the Market

One of the oddest things about the Serostim approval is that the duration of therapy and a precise definition of what population will benefit remain elusive. Controlled trials of growth hormone lasted only 12 weeks. Average weight gain in these trials was about 2.5 pounds. In the most exhaustive trial, a 178-person affair that included body composition analysis, the weight gain was found to consist of an average lean body mass increase of 6.8 pounds and a 4.3 pound loss in body fat. But, 22% of the participants dropped out or were noncompliant, and of the rest, 26% of those on growth hormone continued to lose weight. Participants in the open-label continuation phases of the trials maintained their weight gains while on growth hormone, but it was difficult to say what would have happened had they stopped. There were many trial dropouts here too, and these people may or may not have resumed their weight loss.

The FDA-approved package insert that comes with Serostim limits itself to recommending that treatment be initiated when a patient's unintentional weight loss reaches 10% of normal body weight. If weight loss continues after two weeks of treatment, the patient's condition should be reevaluated. The insert notes that any acute opportunistic infection may need to be cured, or at least suppressed, before restoration of weight or body composition is possible.

Serostim, injected subcutaneously once a day, is hideously expensive, costing $210 or $252 retail for a daily dose of five or six milligrams. Serono has established a yearly price cap of $36,000 for those who must keep taking Serostim indefinitely to prevent a recurrence of loss in weight and lean body mass. A patient assistance program will also provide free drug to a limited number of people with restricted financial means. For more information on Serostim distribution and financial assistance call the "Serostim Access Line," 800/714-2437.

The enormous expense is not the only reason to use Serostim judiciously. The doses of growth hormone administered for AIDS wasting syndrome (about 0.1 mg/kg of body weight) are very high compared to what is used for other conditions. In the brief AIDS wasting trials, rates of muscle pain (53.7% of those on growth hormone), tissue swelling and stiffness (27.3%), and carpal tunnel syndrome (relatively infrequent) were markedly elevated.

Much of the confusion around growth hormone stems from the fact that AIDS wasting is really a collection of syndromes. The variety of factors at work range from lack of appetite to malabsorption in the gut to metabolic derangement triggered by generalized inflammation, and the precise mix differs from individual to individual. Many of these factors must be treated in tandem. Growth hormone cannot work without sufficient nutrient intake, for example.

The confusion concerning Serostim's therapeutic role may be resolved by an FDA-imposed post-marketing trial. That trial will enroll over 700 volunteers to check on changes in lean body mass, body weight, physical function and quality of life after 12 to 48 weeks on therapy. A series of changes in study design have caused this study to be postponed several times, unfortunately. Some of the existing uncertainties in the protocol may be cleared up by an FDA conference on evaluating wasting syndrome described at the end of this article.

Testosterone Replacement Therapy

Testosterone has a strong stimulating effect on muscle tissue, and replacing lost testosterone might help reverse the attrition of lean body mass. In fact, the Alza Corporation has been implying that its testosterone skin patch (brand name: Testoderm) is useful for AIDS wasting since it was approved last year as testosterone replacement therapy for those with low levels of the hormone. Results from two trials for men and one open-label follow-up are now in, and they show no benefit for the Alza patch as far as weight and body composition are concerned. The patch did substantially increase the male volunteers' testosterone levels, which may have other benefits in terms of libido and overall energy. Alza officials and outside researchers are interpreting these results to mean that testosterone levels were not raised sufficiently to have a detectable anti-wasting effect.

Judith Rabkin, Ph.D., of the New York Psychiatric Institute, argues, "With injectable testosterone, you get a bigger bang for your buck. The patches are hard to stick, so delivery may be a problem. Also compliance could be an issue -- you have to stick a new Alza patch on your scrotum each day." Dr. Rabkin's trial of injectable testosterone is now being written up. Also, Dr. Steven Grinspoon's group at Massachusetts General Hospital (Boston) should soon finish analyzing a trial of intramuscularly injected testosterone in 50 men with AIDS wasting syndrome. Like the Alza trials, this is a six-month testosterone replacement study -- it seeks to restore testosterone to normal in those with low untreated levels. Dr. Grinspoon comments, "Our study will definitively answer the question" of whether mere testosterone replacement is sufficient to reverse AIDS-related weight loss. Finally, the University of Southern California is nearing completion of a trial that specifically compares the Alza patch to injected testosterone.

A competing testosterone patch can be placed on any part of the body, not just the scrotum. It was developed by the TheraTech Corporation and is marketed by SmithKline Beecham under the brand name Androderm. A Los Angeles trial with this patch in men with intermediate HIV infection recently ended. Of special interest, TheraTech has developed a skin patch geared specifically to replace the small amounts of testosterone normally produced by women. A trial of the women's patch is taking place at Massachusetts General Hospital in Boston, and results should be available this summer.

The Grinspoon paper cited above also presented findings that men with AIDS-related wasting actually had higher than normal levels of natural growth hormone, but their bodies did not seem to be responding to it. (In particular, there were insufficient levels of insulin-like growth factor 1 (IGF-1), which is produced by liver cells in response to growth hormone.) "Our data support the case for using very high levels of growth hormone," says Dr. Grinspoon. The recommended dose for Serostim is indeed very high, and the cost is a certain number of generally tolerable side effects. With testosterone, safety may be more of an issue, in part because of its masculinizing side effects. Possible adverse events include acne, increased irritability and, over time, cardiovascular and prostate problems.

A recent paper, however, found achieving supernormal testosterone levels was safe over the short-term as well as highly effective in stimulating muscle development among HIV-negative volunteers (see Shalender Bhasin et al., New England Journal of Medicine, July 4, 1996; 335(1):1-7,52). The New England Journal paper found that adding weight training to testosterone was even more effective.

Dr. Bhasin, who is from Charles R. Drew University in Los Angeles, argues that testosterone might be preferable to growth hormone. Testosterone acts to raise IGF-1 locally, within muscle tissue, whereas the IGF-1 produced by the liver in response to growth hormone spreads throughout the body. He is currently investigating the use of testosterone with and without resistance weight training in men with HIV-associated weight loss, but the testosterone dose administered, 100 mg/week, is only one-sixth that used in the HIV-negative trial. Dr. Bhasin estimates that the cost of testosterone therapy could amount to a mere $150 per year.

Aside from adding exercise to testosterone, Dr. Grinspoon and others are considering combining human growth hormone and testosterone. Other possible combinations include the use of testosterone plus Megace, an appetite stimulant that on its own tends to increase fat stores, with little effect on lean tissue. Megace is a derivative of the female hormone progesterone and actually suppresses testosterone production. ACTG trial 313 is testing whether testosterone replacement in both men (200 mg every two weeks) and women (100 mg every two weeks) can overcome Megace's drawbacks. And in New York, Dr. Rabkin is testing testosterone therapy along with a high amino acid nutritional supplement.

Exploring Anabolic Steroids

The most tested anabolic steroid so far has been oxandrolone, which is produced by Bio-Technology General under the brand name Oxandrin. Oxandrolone is an oral drug approved by the FDA in 1962 as a general remedy for weight loss in a variety of conditions including chronic infection. The agent is now being touted in advertisements that strongly imply its efficacy for AIDS wasting syndrome. Oxandrolone's usefulness in AIDS has yet to be established, and an early trial utilizing a 5 mg/day dose found that oxandrolone had no appreciable benefit in AIDS.

Recent trial data suggest that higher doses will be more effective, though. In a paper published last December describing a trial in AIDS wasting syndrome, administration of 15 mg of oxandrolone per day resulted in statistically significant weight gain over those taking placebo through 14 weeks. At week 16, however, this difference (a gain of 1.32 pounds versus a loss of 2.42 pounds in the placebo group) failed to reach statistical significance. Body composition was not measured in this trial. Two studies of oxandrolone were reported last winter at the Fourth Conference on Retroviruses and Opportunistic Infections, and both looked at the effect of oxandrolone at 20 mg a day on body composition (see abstracts 692 and 695). The initial results of each of these studies (after an average of about a month) found about a 6.6 pound increase in weight. Body cell mass (a measure of increased protein stores) averaged 2.6 and 3.3 pounds higher in the two studies. Both these small, open-label trials are continuing, with further improvements noted in the trial participants who have already been on treatment for more than a month.

Two major ongoing oxandrolone trials are following a blinded, placebo-controlled protocol with higher doses. One study is trying 20, 40 and 80 mg per day in men with AIDS-related weight loss. Another is administering 20 and 40 mg per day to women. These trials will enroll about 300 men and 200 women, observing them for 12 weeks to compare oxandrolone versus placebo. In a second 12-week period, everyone will receive active treatment.

At 20 mg/day, it would cost more than $12,000 for a year's supply of oxandrolone at today's prices. A definitely cheaper substitute is the injectable anabolic steroid nandrolone decanoate. Nandrolone is made by Organon Teknica under the brand name Deca Durabolin, and there are also several generic versions. On a per milligram basis, nandrolone costs a sixth to a twelfth of what oxandrolone costs. Unfortunately, no one is pursuing a development strategy for nandrolone the way Bio-Technology General is for oxandrolone.

Some data are available, though. At the Eleventh International Conference on AIDS, Gary Bucher, M.D., and colleagues at the Center for Special Immunology in Chicago described gains in weight and fat-free mass in HIV-positive male volunteers receiving 100 mg/week of nandrolone for 12 weeks (abstract Mo.B.423). These volunteers had not necessarily experienced previous weight loss, however. In another report (see AIDS, June 1996, 10(7):745-52), a group from Sydney, Australia noted gains averaging 8.8 pounds over 16 weeks in 24 men who did have HIV-associated wasting. These men had not gained weight when enrolled in a program consisting of resistance exercise and dietary counseling. (Note that these 24 were from a cohort of 220 men, of whom the rest did succeed in gaining weight from the exercise and dietary counseling alone). The volunteers received 100 mg of nandrolone injected every other week in four months on/one month off cycles. The investigators estimated that the total yearly cost for such treatment would be just $600. A new San Francisco study of nandrolone using 100 or 200 mg/wk has found equivalent results -- a gain of 8.8 pounds at 12 weeks, of which two-thirds was ascribed to increased body cell mass (presented by Marc Hellerstein, M.D., at the NIH Conference on AIDS Wasting Syndrome, May 20-21, 1997).

Summing up his experience with nandrolone, Dr. Bucher said, "The anabolic effect seems equal to growth hormone, but there may be fewer side effects with deca [nandrolone] in terms of joint pain and tissue swelling." The AIDS Clinical Trials Group is at present sponsoring a small trial of nandrolone in women with greater than 5% HIV-associated weight loss. The women will receive 100 mg of nandrolone or placebo every other week for 12 weeks and then everyone will receive nandrolone for another 12 weeks. A second study at the University of Southern California is testing nandrolone with and without resistance weight exercise in men with HIV-associated weight loss. In this trial, the high dose of 600 mg/week will be administered. A third trial is being sponsored by the American Foundation for AIDS research for both men and women with moderate weight loss. This 120-person, 16-week trial will test biweekly nandrolone doses of 200 and 400 mg in men and 100 and 200 mg in women.

Thalidomide: New Results

In April, a preliminary analysis of this trial was unveiled at a conference on TNFa blockers in Santa Fe. Commenting on the state of data, the study's principal investigator, Morris Schambelan, M.D., of San Francisco General Hospital, said, "The primary efficacy data are pretty solid, but the rest needs careful scrutiny." The results are a little curious in that the higher dose had less effect than the lower one: During the first eight weeks, participants gained on average 4.5 pounds (3.3% of their mean weight) on 100 mg/day of thalidomide, compared 2.2 pounds (1.2%) on 200 mg/day and 0.7 pounds (0.5%) on placebo. Participants continued to gain weight in the eight-week continuation phase. Dr. Schambelan estimates that about half the total weight gained was in lean body mass, although the complete body composition analysis has not been done.

One surprise is that HIV viral load went up slightly (an average of 0.3 logs, or two-fold) in the course of the trial. Reducing TNFa levels should have slowed HIV replication by lessening CD4 cell activation. But then, does thalidomide really reduce TNFa? TNFa levels are still to be assessed in this trial, but Dr. Schambelan noted that in a newly released study of thalidomide in oral aphthous ulcer treatment, the level of both TNFa and its cellular receptors rose somewhat. Thalidomide also failed to reduce IL-2 induced increases in TNFa levels or relieve side effects reputedly associated with the TNFa increases during a trial of IL-2 plus thalidomide (reported at the Fourth Conference on Retroviruses and Opportunistic Infections, abstract 36).

It may be that total TNFa is hard to quantify because blood levels are not necessarily indicative of cells' exposure to the substance in lymph tissue or that some sort of rebound effect occurs in response to thalidomide. Alternatively, thalidomide's primary anti-inflammatory property, the key to its effectiveness, may not really be related to TNFa.

Another surprise in this trial was how poorly tolerated was the 200 mg per day dose. Seventeen of the 31 people receiving the 200 mg dose dropped out, most complaining of somnolence or skin rash. Otherwise the results are in line with, though more modest than, two recently published articles: one describing a small six-month trial in Mexico City (Gustavo Reyes-Tran et al., AIDS, December, 1996 10(14):1501-7), which administered 400 mg/day and a three-week, 300 mg/day Thai study in 32 men with HIV, half of whom also had tuberculosis (Jeffrey Klausner et al., Journal of Acquired Immune Deficiency Syndromes, March, 1996; 11(3):247-57). In the Mexican study, no effect of thalidomide on plasma HIV or TNFa was observed, but the Thai study did find reductions in both these values among the study participants who were co-infected with HIV and TB.

Among the Thai volunteers, baseline TNFa plasma levels seemed to correspond with HIV viral load, with the TB infected group tending to have higher levels of each. Other studies have observed that tuberculosis hastens HIV disease progression, and the enhancement of TNFa production may be related to this acceleration.

Celgene has filed a New Drug Application with the FDA for use of thalidomide for treating a type of aggressive leprosy. Based on the latest results concerning weight gain, it plans to do the same for AIDS wasting syndrome. Further investigation is also taking place concerning thalidomide's ability to reduce oral aphthous ulcers and diarrhea, both of which can interfere with food intake and lead to weight loss. Until the drug is on the market, it is available through a compassionate use protocol for those with uncontrollable AIDS-related weight loss. For more information, call Celgene at 800/801-8328.

Of course, thalidomide is notorious for causing birth defects, and distribution of the substance will have to take place within the context of a rigorous education program to prevent its use during pregnancy. A program that could serve as a model is the one involving Hoffmann-La Roche's widely used acne medication Accutane, which also causes severe birth defects. Aside from the literature it distributes, the Accutane program requires a signed consent form and a negative pregnancy test for women starting Accutane. Roche also recommends that doctors obtain a pregnancy test from their female patients before they renew the Accutane prescription each month. Although Accutane is administered for just five months, pregnancies are still known to occur in women taking it, at a rate of 3.4 per 1,000 female patients according to a Roche survey. Eight percent of these pregnancies were carried to term and resulted in live births.

Evaluating Effectiveness

One difficulty in combining approaches is that the evaluation of wasting treatments is still in its infancy. Until last summer, the FDA was insisting that the prime standard for approving an effective anti-wasting therapy was proof that it prolonged patients' lives or at least reduced the incidence of opportunistic infections. Even weight gain was not necessarily sufficient for the FDA: We know that losing weight predicts further disease progression and death, but does gaining weight improve survival? Measures such as improved quality of life or physical performance are extremely subjective or inconclusive, although one could argue that a therapy that makes people function better is worthy of approval. The accuracy of methods for measuring body composition also remain controversial, with many fearing that growth hormone and other anabolic agents largely increase intracellular water or blood proteins rather than the protein content of muscle tissue.

In the case of Serostim, the FDA was faced with considerable pressure from AIDS activists and physicians who insisted that indications of increased lean body mass should be sufficient to allow sales of an anti-wasting treatment. The begrudging approval, granted without evidence of a positive effect on survival or disease progression and conflicting data on weight gain and improved functioning, is evidence of the agency's increasing confusion in this therapeutic area. Last year, the researcher in charge of oxandrolone's clinical trials, Rudolfo Ferraresi, M.D., complained, "When we tried to pin down the FDA on what data we should collect, they didn't come up with anything. They just indicated that if we come to them with compelling data, they might approve the drug. They said wait for the public meeting we are going to have to develop a consensus."

That public meeting finally took place on May 22 and 23, at Treatment Issues' press time. Cosponsored by the Federation of American Societies for Experimental Biology, it featured workshops on wasting syndrome in AIDS and other chronic diseases. At the end, small discussions groups focusing on specific diseases were to formulate guidelines for measuring wasting in patients as well as for evaluating wasting therapies during clinical trials. This meeting is a sequel to the two-day NIH conference referred to above that was devoted to reviewing current knowledge on the causes and treatment of AIDS wasting syndrome.

Bill Thorne, who has kept track of AIDS wasting therapies for ACT UP/Golden Gate, summed up his expectations for the FDA meeting by saying "Similar to the exploration of HIV viral load as a tool for clinical trials, the introduction of body composition measurements will lead to faster approval of therapies and a significant modification of treatment for AIDS-related wasting."

TREATMENT BRIEFS

Glaxo's Parsimonious Compassion

Discussions between Glaxo and community activists have taken place sporadically for over a year concerning some sort of expanded access program to supply such people while 1592 is under development. On April 28, the company announced the impending commencement of a limited compassionate use program. But this program is so restricted that it has touched off a tidal wave of community criticism.

Glaxo's proposal consists of three programs: a pediatric compassionate use for children who have failed one nucleoside analog, a program for people with "severe" AIDS dementia, and a general adult program. Entry to the general adult program would be restricted to people with viral loads over 50,000 and CD4 counts under 100 after having failed at least two nucleoside analogs and one protease inhibitor. The total worldwide availability will have a ceiling of 2,500 people, which by itself has been cause for a considerable outcry: "We will lose another five to six thousand people in the next year who could be saved by 1592," commented Jeff Getty of ACT UP/Golden Gate.

The enrollment criteria generated a whole slew of objections. Requiring adults to fail two nucleoside analogs and one protease inhibitor before receiving 1592 means that there will be little if anything left to combine the new drug with, given the implications of cross-resistance. To ensure adequate viral suppression and durability of treatment effect, people should take 1592 in combination with other drugs still active against the HIV in their bodies. Preferably that combination would include a protease inhibitor, which 1592 cannot replace. At the very least, the requirement to fail a protease inhibitor should be eliminated. (An alternative would be to undertake a joint expanded access program with Glaxo's experimental protease inhibitor 141W94, which lab studies indicate could be active against HIV resistant to the protease inhibitors on the market.)

Glaxo in addition is forcing both adults and children to wait until they have very high viral loads and low CD4 counts before they can receive 1592. Waiting until that point allows irreparable damage to be done as patients' CD4 cells are decimated and HIV replicates to levels that are hard to bring down. It would have been better to define treatment failure as either an upward trend in viral load or stable, but high HIV levels.

A similar concern arises in relation to the dementia program: Why require that dementia be "severe"? Full recovery may no longer be possible at that point.

Glaxo has responded to these objections by saying that the current supply of 1592 is low and that there is a need to collect extra safety and efficacy data before more broadly releasing the drug. To facilitate this data collection, the general adult compassionate use program will take place at "geographically dispersed centers" rather than the usual central mail order operation. This raises the question as to how people in remote areas will obtain the drug. According to a company announcement, instituting a network of separate sites is part of Glaxo's attempt to build within compassionate use an "open label study which allows for collection of safety and efficacy data... in a broad population of people in addition to that which is being collected through our clinical trials."

It is true that testing has been slow, but the supply problems that supposedly held up trials have been resolved. If Glaxo expects to market 1592 in a year, it must now have the capacity to make commercial-size batches of the substance for the traditional 12-month shelf-life test that new drugs go through. It could use that capacity for a true expanded access program, too.

The company does promise to institute a larger expanded access program next winter, but that will be only a few months before Glaxo anticipates receiving FDA approval for 1592. In the meantime, Glaxo's AZT and 3TC have been the best-selling HIV drugs. Although 1592 will steal some of that market , it can be expected to sell exceedingly well, too. The present demand for immediate access can only bolster its future sales. There is little reason for the miserly expedience that would divert compassionate use from an emergency distribution program into a stop-gap trial.

Some Competition for 1592

A Welcome and an Apology

You also may have noticed that Treatment Issues has been very late this spring. We are very sorry. TI editor David Gilden has been working two jobs lately -- in addition to his regular duties, he has been acting director of GMHC's Treatment Education and Advocacy Department. His temporary tenure in that position is now over, and with him back on the job full-time and Jill Cadman settling in, TI should resume its normal schedule -- or maybe not. All GMHC is moving next month, and who knows what chaos will ensue.

Treatment Issues is published twelve times yearly by GMHC, Inc. All rights reserved. Noncommercial reproduction is encouraged. Subscription lists are kept confidential.