Protease Inhibitors 
and Prevention of Cross Resistance

Merck, Abbott Combination Results

Recently Merck and Abbott released information from their ongoing small studies. Merck is studying the combination of indinavir and AZT; the combination seems to achieve a somewhat better viral load drop than indinavir alone, but -- most importantly -- the antiviral effect lasted longer. Indinavir plus AZT achieved a MAXIMUM drop of 2.6 log (400- fold), vs. 2.3 log (200-fold) for indinavir alone, vs. 0.6 log (4-fold) for AZT alone. But after 24 weeks, the combination was SUSTAINED best, with 2.5 log (315-fold) drop in the amount of virus in the blood, vs. 1.5 log (32-fold) for indinavir alone, and 0.3 log (2-fold) for AZT alone.

Abbott Laboratories released information from their ongoing small open-label French study examining 24 treatment-naive participants using AZT, ddC, and ritonavir. So far they only have data for up to six months, but five study participants' viral load remains undetectable, while the mean viral load level achieved for all study participants at six months is still well sustained at a low level, between 2 and 2.5 logs (100-fold to 320-fold). These studies imply that optimal multi-drug combination therapy (including an effective protease inhibitor) should delay drug resistance, and therefore cross-resistance. No one knows how long this delay could last. The theory is that adequate suppression of the virus delays the development of mutations which could lead to resistance. This promising theory needs further research and confirmation.

Merck is conducting two small three-drug studies: indinavir with AZT plus ddI, and indinavir with AZT plus 3TC. Data may be available by January 1996. These results may further confirm the theory that one can delay drug resistance by achieving adequate viral suppression, with well-selected, effective multi-drug combinations. The implications are promising. For example, one might be able to begin with a particular three-drug combination (including a good protease inhibitor), and keep viral activity low enough that resistance does not develop rapidly. Then, before resistance develops, one could switch to a different combination -- without having cross resistance due to one's previous drug use.

Months ago, when the debate surrounding cross resistance was beginning and many of us were confused and discouraged, David Ho, M.D., of the Aaron Diamond AIDS Research Center, predicted that cross resistance would not be as much of a problem as others were claiming. He predicted that sequential use of protease inhibitors would be a useful tool in treatment strategies. The message to people with HIV or AIDS is, don't be discouraged. These theories will be explored and hopefully confirmed. >From this research, improved treatment strategies are likely to emerge. And there may be other developments which are not yet foreseen.

For example, Glaxo Wellcome is developing two new reverse transcriptase inhibitors, one of which may be effective against virus which is resistant to AZT and to other drugs in its class. Those who are already resistant to AZT may be able to switch to new drugs to use in combination with a protease inhibitor. Another major research area is the combination of two or more protease inhibitors. Roche and Abbott will soon be collaborating on studies exploring the combination of their drugs. These drugs appear to have a unique relationship, since ritonavir (the Abbott drug) greatly increases the blood level of saquinavir (the Roche drug).

(Caution: these drugs must not be combined until they have been carefully studied, since the dose of saquinavir must be greatly reduced, probably ten to 100 times or more -- but nobody yet knows what the proportion will be). Also, Merck and Roche may collaborate on a study combining their protease inhibitors. And other protease inhibitors, including Agouron's AG-1343, and the Glaxo Wellcome/Vertex VX-478, are now in earlier human trials. I am hopeful that useful treatment strategies will emerge from this work.

Caution

In some people being treated with the current protease inhibitors, resistance can emerge, sometimes very quickly. To minimize this problem, it is important that you follow the instructions provided by the doctor or study nurse for administering these medications. There are some precautions anyone being treated with indinavir, ritonavir, saquinavir, or other protease inhibitors can take to help prevent the occurrence of resistance and to maximize the benefit of these drugs to oneself.

There appears to be a relationship between dosing and resistance. You should adhere completely to the dosing instructions. Make sure to take your full dose as instructed, at the appointed time, without delay; waiting 15 minutes may not matter, but a delay of two hours could be relevant. Do not skip taking a dose; if you forget or miss the appointed time, take the dose as soon as possible and then take the next dose at the regular time you were supposed to take it. And follow any other instructions on matters such as food intake, or storage of the drug.

AZT Delta Study: 
Major European/Australian Study Finds 
Combinations Better Than AZT Alone

Delta Results

The results of Delta are by far the most impressive. Strictly speaking, it was two separate trials -- Delta 1 for people who had never taken AZT before, and Delta 2 for those who had already taken AZT monotherapy for at least 3 months. The trial was open both to asymptomatic people and those already diagnosed with AIDS. All participants had to have a CD4 count between 50 and 350. Everyone in the trial received a standard dose of AZT (600 mg/day). They were then randomly assigned to take either:

• ddI (400 mg/day), or
  
• ddC (2.25 mg/day), or
  
• no additional treatment (i.e. a ddC placebo and a ddI placebo).

Delta started in 1992, and participants had been followed for an average of just over two years at the time this analysis was performed. Delta 1 recruited 2131 eligible people, and a further 1083 joined Delta 2. 88% of people in Delta 1 and 83% of those in Delta 2 had not been diagnosed with AIDS when they joined, and their average CD4 cell counts on entry were 212 and 189 respectively.

The results of the Delta 1 part of the trial were highly significant. People who began their anti-HIV treatment with either combination had a substantially reduced risk of clinical progression or death compared to people receiving AZT monotherapy. During the two years of follow-up, 16.5% of people assigned to take AZT alone died. But among combination therapy recipients the death rate over the two years was reduced by 38%, to 9.6% in the AZT plus ddI group, and 11.6% in the AZT plus ddC group. The difference in death rate between the combinations was not statistically significant i.e. it could simply have been due to chance. The combinations were also more effective than AZT monotherapy at preventing disease progression in every measure used. People with no symptoms were less likely to become symptomatic, people with mild symptoms were less likely to develop more serious symptoms, and people with AIDS were less likely to develop a more severe AIDS illness. Overall, 28.4% of the AZT monotherapy recipients developed AIDS or died, compared with 17.6% of the AZT plus ddI group, and 23.3% of the AZT plus ddC group.

Dr. Brian Gazzard, the trial's UK principal investigator, said that the message from this part of the trial is absolutely clear. "If you're thinking of starting treatment -- and these results should encourage people to get tested and begin treatment -- you should start with combination therapy, not AZT alone."

AZT Experienced

The results of Delta 2, the arm of the trial for people who had already taken AZT before, were less encouraging. In effect Delta 2 compared the effects of switching to combination therapy, or continuing to take AZT monotherapy. Participants tended to have quite substantial prior use of AZT -- 63% had been on AZT for at least one year. In this arm, there was no significant difference between the treatments. Combination therapy recipients did not have a significantly lower risk of disease progression or death compared with those continuing on AZT. However, the fact that Delta 2 did not detect a significant difference between the treatments does not necessarily mean there is no difference -- only that the trial was unable to detect one. Combining the results of Delta 1 and Delta 2, people taking combination therapy had a 25% reduced risk of death compared with those taking AZT monotherapy.

Side Effects

Combination therapy was no more likely to cause serious side effects than AZT alone. Predictably, people taking ddI or ddC were at increased risk from the side effects associated with those drugs. The most common ddC-related side effects which led participants to stop taking the trial treatments were oral ulcers (12 discontinuations) and peripheral neuropathy (37 discontinuations). Eight cases of pancreatitis occurred, six among AZT plus ddI recipients and two among people on AZT plus ddC. ddI or the ddI placebo (which contained the same antacid buffer as real ddI, but no active drug) caused about 10% of recipients to withdraw from the trial due to side effects of nausea and vomiting, making this the least well- tolerated treatment.

The risk of side effects may deter some asymptomatic people from beginning treatment. But Dr. Gazzard argues that "To reduce their mortality by nearly 40%, I'd have thought most people would accept some discomfort." [Section on ACTG 175 omitted.]

Doubts

Researchers told AIDS TREATMENT UPDATE that the quality of the data from Delta is much better than that of ACTG 175. A fifth of people who joined ACTG 175 were lost to follow-up, and over half of those who remained stopped taking the trial treatments before they reached one of the set endpoints of CD4 decline, development of AIDS or death. This meant that relatively few people reached trial endpoints.

Professor Tony Pinching of St. Bartholomew's Hospital in London told AIDS TREATMENT UPDATE that the difference between combination therapy and monotherapy in ACTG 175 was "a very small effect, involving only small numbers of people out of a very large study". These factors make the findings of ACTG 175 more vulnerable to error, and so less convincing.

In Delta, participants were less likely to stop taking trial treatments before reaching an endpoint. Delta also recruited twice as many AZT-naive participants as ACTG 175, and it was in this group that most of the benefits of combination therapy were seen. These factors make the results of Delta much more secure.

Los Angeles, New York, 
Stanford University: HBY 097, 
New Drug May Weaken HIV

ICAAC's Small Advances

• Researchers from the University of Miami and the University of California in Los Angeles proposed conditions that enhance HIV penetration of the central nervous system. In the laboratory, they created a model of the blood-brain barrier with human brain endothelial cells and a sublayer of fetal astrocytes, and found that this "barrier" could be disrupted by migrating monocytes, by tumor necrosis factor inflammation, and by exposure to cocaethylene, which simulated drug abuse.
  
  (Earlier research has long implicated tumor necrosis factor with the development of dementia, as well as in wasting and disease progression.) The authors of this work theorize that in people, the blood-brain barrier can also resist HIV unless challenged by one of the above disruptions. [Abstract I71]
  
  
• NIH researchers, with others, reported that NAC (N-acetyl- L-cysteine), a mucolytic agent, could inhibit the germination of spores of several species of fungi that can cause pulmonary infections in people who are immunocompromised. One of the fungi used in the cultures was aspergillus, which can definitely be a problem for people with HIV. NAC has already been in wide use in the HIV community since Stanford researchers described its potential to inhibit HIV replication. In this new work, a concentration of 10% NAC was said to work better than more dilute solutions, or than a 10% solution of L-cysteine. These were laboratory experiments, but since NAC is safe and already FDA-approved to treat certain pulmonary disorders, it could conceivably become an adjunctive treatment for aspergillus infections. [abstract E82]
  
  
• A number of other new antifungals were presented at the conference, including those code-named L-733560, Sch 56592, UK-109,496, UR-9746, UR-9751, ER-30346, and D0870, as well as some new antimycobacterials, such as U100-480, T9, CRL-1018, CRL-1072, KRM-1648, tryptanthrin and azaindoloquinazoline-diones. New drugs from both classes are very much needed, because some evolving strains of candida, cryptococcus, tuberculosis, and MAC have lost susceptibility to the older medicines. Apparently only a few of the new agents are in human trials.
  
  
• Azithromycin, an antimicrobial which is often used in AIDS medicine, is capable of causing ototoxicity, or hearing and balance deficits, according to Toronto researchers. They said that eight of 46 patients on azithromycin reported various otologic complaints, including vertigo, hearing loss, tinnitus and a feeling of plugged ears. When the drug was discontinued, the problems resolved within two to eleven weeks [abstract LM19].
  
  (Amikacin, another antibiotic sometimes used to treat MAC, can be ototoxic as well. Patients should be warned about this danger when they are given these drugs.)
  
  
• Johns Hopkins researchers reviewed the follow-up data on patients who had participated in two controlled clinical studies of rifabutin as MAC prophylaxis and found an association with improved survival on the drug. The trials had already established that rifabutin could reduce MAC bacteremia, but its effect on survival has been in question. The authors said that for patients on rifabutin for one year, the relative hazard of dying was 8% lower than for patients not on prophylaxis. After a year and a half, the risk was 14% lower. [abstract I204]
  
  
• A number of presentations supported the use of cytokines, chemical messengers that travel between cells, in AIDS therapies. Stanford researchers, for example, reported that interleukin-12 significantly enhanced the effectiveness of atovaquone and clindamycin in mice with toxoplasmosis. They suggested that the combination could be useful in human toxoplasmosis as well [abstract G99].
  
  Other San Francisco area researchers reported that GM-CSF (granulocyte macrophage-colony stimulating factor) may improve the immune response to MAC bacteremia when added to MAC treatments [abstract G109].
  
  French researchers likewise reported that GM-CSF appeared to enhance the effect of the standard treatment for toxoplasmosis, which is a combination of sulfadiazine and pyremethamine. They surmised that the benefit was from bolstered host responses. [abstract I225]
  
  
• Physicians from Madrid described some modest benefits for patients with PML (progressive multifocal leukoencephalopathy) who were treated with intravenous cytarabine. In a retrospective study of thirteen biopsy-confirmed infections, three people improved on the drug, and five others did not. The five untreated patients did not improve. The patients who received cytarabine had a median survival of 100 days, compared to 36 days for the untreated. The authors added that the treatment was well-tolerated. [abstract I227]
  
  Except for similar humane attempts to address it, PML is probably the longest-running neglect story in the AIDS epidemic. There are no convincing treatments for it, and no convincing research organized to change that. Cytarabine is the only drug that has ever been seriously considered. Most clinicians simply offer AZT to their patients diagnosed with PML, because AZT crosses the blood-brain barrier and has been reported anecdotally to reverse some cases. (The Spanish authors said their results were not affected by the use of AZT.)
  
  Treatment advocate Peter Brosnan in Los Angeles has been like a lone lighthouse in this situation, compiling the meager research data for the last several years and sending it to concerned patients and physicians. We look forward to the conference that offers more than one small study for this infection.
  
• Researchers in Nigeria and Washington, DC, reported that immune globulin was helpful for preventing a number of infections of the ear, respiratory and urinary tracts in children with HIV. This helped to confirm research from previous years, and in fact, immune globulin is considered by some clinicians to be similarly useful in HIV-infected adults with low CD4 counts. [abstract G45]
  
  
• French researchers studied CD8 cell counts, and found that individuals with higher counts, in this case greater than 600 cells/ml, experienced significantly fewer opportunistic illnesses--4% vs. 27.5%--while followed over the course of one year [abstract G49].
  
  And New York researchers reported that HIV-infected infants who had not yet developed CD8- mediated viral inhibition fared much worse than those who did [abstract I8].
  
  These observations support a long-standing contention of Jay Levy of the University of California in San Francisco. Dr. Levy has asserted that CD8 cells can secrete an anti-HIV factor which, at its most effective, explains why some HIV-infected people appear to have the virus under control for many years, perhaps indefinitely. The French researchers suggest the pursuit of strategies to enhance the activity of CD8 cells.
  
  
• Researchers at Henry Ford Hospital observed that HIV levels increase significantly during a bout with an opportunistic infection. They monitored ten patients diagnosed with an opportunistic infections, and of the eight who had detectable baseline levels of HIV RNA, all showed a "striking" increase in viral burden. All eight also experienced a significant decline in plasma RNA upon their recovery from the infection. Interestingly, CD4 counts and p24 antigen levels did not change much [abstract I236]. This could be useful information to physicians who make antiretrovirals a low priority or even discontinue HIV drugs when their patients face a new infection.
  
  
• ACTG researchers reported that apart from mild anemia, AZT does not adversely affect infants whose mothers took the drug during pregnancy. This observation followed a critical study, ACTG 076, which found that AZT could reduce the transmission of HIV from mother to fetus by almost 70%. AZT-exposed infants were compared to those whose mothers got a placebo, and no important differences were seen in parameters such as weight, CD4 and CD8 counts, head circumference and neurodevelopmental tests. [abstract I2]
  
  
• A distressing pattern of undiagnosed infections was reported by physicians at Howard University Hospital. When they reviewed the autopsy findings on 40 HIV-infected women, they found that fully half of them had diseases that were identified only postmortem. Seven had more than one disease. The authors added that even when the infection or neoplasm was diagnosed before death, the extent of organ involvement was often underestimated. Thirty-eight of the women were African-American. [abstract I62]
  
  
• Additional disturbing facts about access to treatments emerged from a survey of HIV clinicians by the Centers for Disease Control and Dun and Bradstreet HealthCare Information. The survey was drawn from selected providers in Atlanta, Chicago, Long Island, Oakland, Portland, Tampa and Washington, DC. It found that women were less likely than men to receive AZT, ddI, or d4T; injection drug users were less likely to get d4T (the newest approved HIV drug) than were gay or bisexual men; patients on public assistance were less likely than those with private insurance to receive any antiretroviral treatment at all, and, among all patients on antiretrovirals, those on public assistance were much less likely to receive combination therapy than were privately- insured individuals. [abstract I19]

Comment

Not clear is whether those discrepancies were a function of bias on the part of individual providers, or of disparate patient populations observed between providers. But either scenario, in our opinion, would be unsurprising under the current system--or nonsystem--of access to medicine in the U.S.

The delivery of health care has become astonishingly undemocratic: some employed persons are insured and are welcomed to comprehensive care with physicians in private practice (who much prefer the reimbursements of insurance to those of Medicaid), while the underemployed, the unemployed and the indigent are not insured and must settle for patchwork services at urban clinics and hospitals which creak along with too many patients and too few resources.

Which venue would be more likely to foster an attentive physician and an informed, motivated patient? Souring the inequity all the more is the influence that sex, race and class hold over who gets stable employment in the first place, especially the kind with health benefits. Unfortunately, the Republicans in Congress lack the heart-- and the Democrats in the White House the spine--to lead America into the developed world on this issue.

AIDS Information Obstacles

What's Good

Consider some positive models first. At last year's International Conference on AIDS in Yokohama, the most successful reporting of the conference may have been by HIVNET, the AIDS computer organization in the Netherlands. The International Conference encourages speakers at the major meetings to make transcripts of their talks available to the conference press office, which photocopies them and distributes them to the press (but not to other conference participants, nor to the great majority of interested people who cannot fly around the world to go to these meetings). HIVNET had people fax these transcripts to Amsterdam, where they were computerized with a scanner, and then placed online, immediately available everywhere.

ICAAC offers a positive example, in making abstracts of the conference available in advance -- something the International Conference on AIDS has not yet done. Having the abstracts allows participants to plan ahead of time, deciding who they want to meet and what they want to see. But if they do not get the abstracts until registration, they have little time for such planning without missing the conference itself. The abstracts would be even more useful if they were available ahead of time in computer form. Then one could search thousands of abstracts for topics of interest, to locate knowledgeable doctors and researchers outside of one's own circle who will likely attend the meeting and can be contacted there.

These computerized abstracts should be available in advance on the Internet -- along with whatever transcripts, slides, or other supporting material the presenter wishes to provide. Then this information would be accessible throughout the world, by those who could not come to the meetings as well as those who could. People would still attend conferences for face to face interaction -- which is what justifies the time and expense of getting there. They should not have to make long trips to spend several days sitting in lectures, for material that they could much better receive by computer or by video. Perhaps fewer people would come to the meetings if they had good advance access to the information to be presented. If so, this would be for the good.

An important impediment to AIDS research is the amount of time that researchers spend in planes, airports, and ground transportation, often giving the same talks again and again for years; no wonder so little work gets done. It is wrong to withhold information which could easily be available, just to inflate meeting attendance. Instead, people should have the information in advance, and come to the meetings when they have additional reason for doing so. If the conferences have to be planned on a slightly smaller scale, what is the problem? They are not supposed to be cash cows. And having the original material accessible everywhere will greatly advance the work the editors and translators who summarize conference information for their particular audiences.

At every international conference, one can see people from developing countries who were sent there, at great expense, to get AIDS information and bring it back to their communities. Often they seem lost; they have the abstract book (in English, usually not their language), they have to decide quickly what talks to hear (again, in a foreign language). They may be able to buy audio tapes, but they cannot get into the press room to obtain what transcripts exist (and they do not know that the transcripts are there). Often these people have no working relationship with anyone else at the conference, because they are the only one from their area able to go. They could do their job much better if the information which now goes by on the slides and the posters, and then is lost, were available to them in permanent form, before they left for the meeting -- and if they could participate in feedback and discussion by email, building professional relationships in advance with others who will attend.

What's Bad

This year's ICAAC in San Francisco illustrated a number of obstacles to AIDS information flow:

• The conference had a new rule this year against photographing the posters or other presentations; we had no advance notice of this change. (Somehow the abstracts and other press materials did not reach San Francisco press registrants in advance, at least not anyone we know; this may have been due to rumors that some local activists, likely to register as press, would disrupt certain corporate booths -- in fact no disruption took place. Other press may have had advance notice of the rule.) Fortunately we took 150 photos before being stopped by the guards; these helped immensely in our reporting for our previous issue. We have never had any researcher at this or any other conference object to our photographing their presentation.
  
  (There is a totally appropriate objection to flash photography when slides are being shown. This is not a problem, as persons familiar with photography would not use flash when photographing slides on a screen.)
  
  We talked to staff and learned that the reason for the rule was fear that scientific rivals would steal data; since the presentations had not been formally published, it might be hard to prove priority of authorship. Note that this issue would be moot if the information were placed on the Internet or otherwise made available by computer, since proof ofpriority would then be established. Photography is only a poor stopgap; officially computerized information would be far more complete and useful.
  
  
• Another problem at ICAAC is that many sessions are not made available on audiotape merely because one of the presenters did not sign the required release form. This kills the only public record not only of their talk, but of everyone else who happened to speak in that session as well. Again, this issue would be moot if the conference provided facilities for computerization. No one would be forced to make their talk, slides, or other information available online. But those who wanted to do so would not be stopped by the few who did not want a public record of their public talk, or who forgot to send in the release form.
  
  
• This year, for the first time at ICAAC, the conference abstracts were available in searchable form on disk at the meeting itself, courtesy of Merck & Co. But not for press. Fortunately, we got our copy just as the Merck booth opened, before either we or the staff at the booth knew about this rule. But many others were turned away.
  
  
• A minor problem, but one that conference participants should know about, is that due to a glitch in a new system for producing badges for registrants, persons registered as press were unable to pick up messages left for them on the computerized message system at the conference. If you left a message there for any press person (or AIDS activist), they never got it.
  
  
• A much bigger and more fundamental problem is that rejected presentations may never reach the public at all -- or only after great delays. Many rejected abstracts clearly have scientific merit, but are turned down due to lack of meeting space, or as a result of professional rivalries or disagreements. Then they are likely to wait months for the next AIDS conference, where they also may be rejected (and where they often have to be submitted months in advance, creating further publication delays, and also assuring that the abstracts are already out of date when they are published). Or the work goes to a journal, which usually is even worse. What appears in journals is what scientists were discussing two or more years ago. At least at conferences, the speech or the poster can include up-to-date material.

What should be done?

In a future article, we will propose a computerized system of "smart peer review." This will allow scientists to release their work to the public immediately, but still have it peer reviewed and modified. And the users of this online journal (which also could be called a computer conference) will have the ultimate decision; they will be able to accept or reject peer review -- or to choose their own kind of peer review as a custom filter through which to view published material. We will show that this system works economically, since professional journals seldom pay their authors or reviewers. Credit for peer-reviewed publication is their draw -- and this can be provided in a much more flexible way, without the intolerable drawbacks imposed by journals today.