DDI, PARALLEL TRACK NEGOTIATIONS

During the last week a series of meetings and conference calls, arranged by different organizations and individuals, has brought diverse groups concerned with DDI and with proposed "parallel track" treatment access to the discussion table. Part of the urgency of these discussions was the deadline of a meeting on August 17 to advise Assistant Secretary for Health James O. Mason about the parallel track. We have not attended these meet- ings, but have spoken with some of the participants. Some observations, based on what we have heard:

* A major meeting August 8 in Washington brought together most of the groups involved. Ten persons represented Federal agencies -- including Anthony Fauci from NIH, and Ellen Cooper from FDA. Several came from Bristol-Myers, including at least one of the principal writers of the DDI protocol. There were several representatives from patient and advocacy groups, three community physicians, three clinical researchers, and a representative of the Pharmaceutical Manufacturers' Association.

The AIDS patient advocates were Paul Boneberg from Mobilization Against AIDS, Jim Eigo from ACT UP/New York's Treatment and Data Committee, Jay Lipner (who has worked closely with Lambda Legal in New York, and with Martin Delaney of Project Inform), and Earl Thomas of NAPWA (National Association of People with AIDS). The community physicians were Bernard Bihari of New York, Neil Schram of Los Angeles, and Melanie Thompson of Atlanta.

The information below came not only from that meeting, but also from a New York meeting on the following day called by Jay Lipner and attended by physicians from NYPHR (New York Physicians for Human Rights), and by Ellen Cooper, M. D., of the FDA. Later, a conference call coordinated by the American Foundation for AIDS Research (AmFAR) included additional participants in these discussions.

* Despite major unresolved issues (see below), and specific concerns about process in these discussions, the fact that they occurred at all is a major advance. One FDA representative was heard to comment on the advantage of talking to physicians who see many AIDS patients, instead of only to those who are experts in trial design but see few patients.

It appears that until now, protocols for multicenter AIDS trials have not even been seen by the major private-practice physicians (much less by the patients who will put their lives on the line for them) until after the protocols have hardened into concrete -- been submitted to Institutional Review Boards across the country, which makes even the slightest change impossible, as it would require re-approval of the study by all these boards. Apparently for the first time, the physicians who must refer patients if the trials will be successful, and representatives of patients themselves, have had a chance to look at a study while changes are still possible, and point out the kinds of practical problems that could make the trial unworkable but often could easily be corrected by the trial designers if they knew about them in time.

* Some examples of the kinds of problems that can be worked out (or at least addressed) by bringing the different parties together:

- PWAs are concerned about what would happen to persons randomized to receive either DDI or AZT, probably for two years, and who then did poorly. Would they be left to get a major opportunistic infection or die, or would they be allowed to try the other drug?

This issue is still unresolved, and apparently was not discussed in the meetings except for one trial -- the "AZT worrisome", those who responded well to AZT and can still tolerate the drug, but are showing early signs that AZT is beginning to fail for them. This group is not included in the three trials already planned, apparently because the idea for it came later; but this trial could be very important, because it could obtain statistical proof of efficacy of DDI faster than any of the other trials. And the issue of treatment failures is most important here, because everybody going into such a study would already be starting to fail on AZT, and half of them would be randomized to continue the same drug. There seems to be a developing consensus that criteria could be devised under which the code would be broken for patients who continued to deteriorate, and if they had been receiving AZT, they could change to DDI.

But for the three trials, planned to start in September, this issue of great importance to the PWA community is still unresolved. The question should also be important to the trial designers, because if volunteers who do poorly on the study medication are not allowed to try other options within the trial, they are likely to obtain treatments outside of the trial instead, perhaps without telling the investigators. And patients will be reluctant to volunteer -- and physicians reluctant to recommend that they enter trials -- if there is no provision for changing their treatment if they do poorly.

- AIDS community representatives insisted on allowing compassionate provisions for persons who clearly need an antiviral but could not use AZT because they are already using ganciclovir, which like AZT has hematologic toxicity. (These persons will not be eligible for any of the formal trials.) Since DDI does not have the same toxicity, there is no reason to think that persons using ganciclovir could not also take DDI. But because no one has yet tried the combination, there is no proof that it is safe.

If ACT UP and Jay Lipner had not raised the issue, these patients would have been told to wait for DDI until a safety study of the combination had been done. But from the discussion which developed in the meeting, it quickly became clear that it would in practice be a long time before this study could be finished. So an alternative, apparently proposed by the FDA and acceptable to everyone, was to allow these people compassionate access to DDI, and also to put a small number (perhaps eight to 15) of such patients into a new trial at one of the ACTG sites which has previously done the phase I study of DDI, to find out quickly about any unknown dangers in the combination. If problems are found, then the compassionate access for persons also using ganciclovir could be re-evaluated.

- Community physicians suggested that useful data could be collected from parallel-track access by community-based clinical trials organizations working together with Bristol-Myers. Bristol-Myers has agreed to supply DDI for at least five thousand people within a year (the number could go higher if justified), in addition to those in the trials; but it has not yet decided how to obtain data from these patients. Community- based research organizations could collect data from about two thousand or more of them; the others (for example, those too far from any such organization) would receive the drug without data collection. The cost would be small, and the data would be collected by organizations with a track record in research, rather than by individual physicians who would usually not be interested. Then later, in case the formal studies were slow in recruiting, data would be available; if, perhaps a year from now, the FDA is willing to accept surrogate markers (blood tests, etc., instead of deaths or OIs) as proof of efficacy, the drug could be licensed much earlier than otherwise.

- One physician was concerned that DDI may be blamed for cases of pancreatitis which have been found in some of the people using it, when the drug might not be responsible. He pointed out that New York physicians have seen increasing numbers of cases of pancreatitis recently, in patients who have never used DDI.

- AIDS activists expressed concern about recent moves to limit parallel track to those who cannot get into any trial. There are a number of moribund trials, which fail to recruit patients because they are poorly designed, or because they are testing drugs which do not look good but made it into trials because somebody had the money to pay for them. There is fear that restrictions on access to experimental treatments will be used to blackmail the AIDS community into filling these dead-end trials.

However, Dr. Bernard Bihari of the Community Research Initiative, who attended the meetings, told us that the idea of limiting access in this way was brought up but was clearly rejected, and that it will not happen, at least not for DDI.

* Specifics of the trials -- according to the memory of one participant:

- The three DDI trials to start soon -- hopefully by late September -- will be named ACTG 116, ACTG 117, and ACTG 118. It is hoped that they will enroll 1800 volunteers total.

- The trial considered most important will be DDI vs AZT, for persons with AIDS or advanced ARC. In some ways this trial will try to replicate the early phase II trial which led to approval of AZT. It hopes to have 400 patients in each arm. Some prior use of AZT (up to two months) is OK.

- Another trial is for persons who have taken AZT for a year or more, and are tolerating it well. This study is to find out whether these people do better staying on AZT, or switching to DDI. It is hoped that this trial will produce the earliest conclusive evidence.

- The third trial is for those who have had to stop AZT due to hematologic toxicity. Since these people cannot be randomized to AZT, and a placebo would not be acceptable, they will be randomized to receive one of two different doses of DDI. The low dose will be the lowest which showed efficacy in phase I (apparently 100 mg twice a day), and the high dose will be the standard DDI dose for all three of these trials, (apparently 375 mg twice a day). There is concern that these doses are similar enough that it will take a long time for this trial to prove that there is a statistically significant difference between them.

- Later, there are plans to organize a trial for the "AZT worrisome," those who are on AZT and not intolerant, but show signs that the drug is beginning to fail, short of a major opportunistic infection. There will also be a pediatric trial.

- Two different existing programs for early access -- "compassionate use" and "treatment IND" -- for those who clearly need an antiviral and cannot use AZT or enter the trials, should begin at the same time as the trials. The competing "parallel track" concept will not be used for DDI at this time -- partly because parallel track is not yet fully defined, and partly because in the turf wars between FDA and NIH, the FDA is in the position to make this decision. Parallel track may be the more far-reaching proposal, returning more choice to patients and their physicians when patients cannot enter trials, whereas compassionate access is for those who clearly cannot use AZT effectively and have no other choice.

* As we went to press, we received a two-page consensus statement on the parallel track prepared by 15 organizations: AIDS Action Council, ACT UP/New York, ACT UP/San Francisco, AIDS Project Los Angeles, American Association of Physicians for Human Rights, American Foundation for AIDS Research, Community Research Alliance, Gay Men's Health Crisis, Human Rights Campaign Fund, Lambda Legal Defense and Education Fund, National Association of People with AIDS, National Gay & Lesbian Task Force, National Gay Rights Advocates, Project Inform, and the San Francisco AIDS Foundation. This statement calls for the creation of an indepen- dent panel to make "decisions regarding the definition and implementation of the parallel track." This panel "must include full, voting representation by AIDS primary care physicians, representatives of community-based research groups, and people with AIDS, HIV infection and their advocates" -- as well as representatives from government agencies and the pharmaceutical industry. The statement points to the lack of decision-making representation by the AIDS community as a key factor in impeding previous efforts (compassionate use, and the treatment IND) to make AIDS treatments more available.

The statement also calls on this panel to "consider mechanisms for assuring that any person with HIV infection who has a demonstrable medical need for a parallel track treatment could obtain access regardless of economic circumstances."

* A major concern we have heard about the developing program for DDI involves the responsibility of the AIDS community. Bristol-Myers and other participants have taken risks to handle this drug much better than others have been handled, moving very rapidly at this time to get trials going, consulting with community physicians and patients, and planning to make their drug available free. But DDI is not the last drug, and other companies will be watching to see how well this policy works. If it fails, access to other drugs will be harder in the future.

One fear in the AIDS community is that patients who could qualify for the formal trials, or their physicians, may fudge records to get patients DDI through compassionate access or treatment IND instead, in order to avoid the chance of being randomly assigned to AZT, and to be able to receive the drug directly from one's physician and have the physician know which drug it is. On the other hand, the trials will have the advantage of offering excellent testing for free -- and none of them will give AZT to patients who clearly will not benefit from it.

We hope that fudging records to stay out of trials will not be a serious problem, as it is important for many reasons that the trials be able to recruit effectively -- both to get the efficacy data quickly, and to avoid future restrictions on treatment access.

Another concern, which we have heard from persons in the Community Research Initiative in New York, is that the great attention to DDI is leading to an irrational rush to this particular drug, with people avoiding other trials, or calling off other treatment plans, to get DDI. Not only could a stampede for DDI damage other trials and make it more difficult to get legal access to treatment in the future, but also it does not make sense medically, because the case for this drug is not yet very strong. Only about 90 patients have taken DDI in trials, and there are many questions remaining about risks and benefits. DDI may prove to be more useful and less toxic than AZT -- but at this time the uncertainties are much greater. DDI may be especially valuable to those who cannot use AZT effectively, and it is important that the AIDS community make sure they have the option. But those who can use AZT should weigh the choice carefully, and not overemphasize DDI just because it is new or in the news.

Other new drugs now entering clinical trials, such as D4T and AZDU, may be at least as good as DDI. We hope that DDI will show the way to an effective compromise, in which well- designed clinical trials can be quickly enrolled and conducted, while at the same time patients' options are maximized and there is some provision for everybody.


MAJOR UNRESOLVED ISSUES:

* How will data be collected when treatments are released for compassionate use or parallel track access?

Ideally, data collection from early access could provide valuable experience about how drugs are working in the real world, outside of the artificial, controlled conditions of the trials. Patients want more data collection, and so do community-based clinical-trials organizations, which see a role for themselves in collecting it.

But pharmaceutical companies often fear that in today's regulatory climate, this data could only hurt a drug's licensing, not help it. The concern is that the FDA is unlikely to take seriously results obtained outside of a rigorously controlled clinical trial, meaning that data from parallel track, treatment IND, or compassionate use could not help the drug be licensed. But adverse effects reports are taken seriously. The fear, then, is that results from early release can only hurt -- a fear which makes companies reluctant to allow use their drugs, no matter how medically justified and urgent the case for doing so.

Unfortunately many FDA employees do not believe in any early access -- perhaps because they fear that exceptions to the regular licensing process cast doubt on the rationale of their work. If their standard drug-licensing procedures are sound, then why are exceptions necessary?

Fauci's parallel-track proposal has added a new element to the debate -- an authoritative statement that it is possible to relax some restrictions on treatment access without harming clinical trials. But the FDA has more than research to worry about. Like a circus lion trainer, in a cage with entities more powerful than itself, it must not lose control for a moment. This fight to keep control has led to cruel and irrational outcomes, which are increasingly becoming a mainstream political issue.

This is the basic problem which has stood in the way of earlier access to treatments. The FDA has not wanted to find ways to allow data from pre-licensing use (other than through trials which most patients cannot qualify for) to help a drug get licensed. If early access can only harm their interests, pharmaceutical companies will not provide their drugs for any form of compassionate treatment use. Bristol-Myers may have made an exception for DDI because otherwise it would have faced enormous bad publicity. Instead, by agreeing to provide the drug subject to FDA approval, it put the ball in the FDA's court, where it belongs. But the great majority of drugs do not capture the public's imagination. The only sound solution we see is to insist on a full, ongoing examination of the scientific and public-policy assumptions underlying the drug-approval process.

In addition, pharmaceutical companies are understandably reluctant to have others collecting data about their products outside of their control, because they fear being held responsible for adverse effects which may have nothing to do with their drugs. Most private-practice physicians who are not associated with research organizations do not want to do data collection, either. To them it is just more paperwork.

One possible solution to this problem is for the pharmaceutical company to fund data collection by community-based or other research organizations, under the sponsoring company's control, through negotiated protocols. Data would be collected only for those patients who volunteered (the free blood work would be an incentive), and were geographically convenient to a monitoring organization.

* A major disagreement is what to do for "AZT refuseniks," those who do not want to try AZT but do want access to DDI. Patients and some physicians see it as wrong to force somebody to use a drug they do not want and get sick from it in order to qualify for treatment. But at this time the other side seems to have little sympathy -- as if those who refuse AZT are to be punished for their distrust of the system, or as if the doors are to be closed to "economic refugees" who come to DDI because they cannot afford AZT, or as if patients must do their share of sacrificing themselves for some greater good in the future.

One suggestion is to develop standards for documenting who is a "real" refusenik, for example by chart notes over a period of time, so that there will not be a huge number of people who receive DDI this way.

This issue shows that we have a long way to go to reach consensus that treatment decisions should be made by patients and their physicians, on medical grounds, in the patient's interest. There are excellent arguments for being cautious with DDI until more is known, for not using it if other viable options are available. The case against accepting unknown risks as preferable to known ones must be well stated. But when there are legitimate medical grounds for choosing a drug, the ultimate decisions should be made by patients and their physicians, rather than by scientists, officials, or committees who make rules in advance with no knowledge of each specific case.


PHYSICIAN ACCESS

Bristol-Myers has made plans to distribute DDI rapidly. Physicians will be able to call an 800 number, then receive and return forms by fax, or by express delivery for physicians who do not have fax machines. Then a 30-day supply of the drug can be shipped immediately.


NEED FOR CAUTION

Persons considering using DDI should realize that not everybody in the AIDS community thinks that this drug is beneficial. There are growing questions and concerns. The history of new drugs suggests caution, as serious side effects of AZT and DDC were not noticed in early trials.

The best information we are hearing at this time is that DDI may add a year of life to persons who are burned out on AZT (or who never took AZT), but that it will not replace AZT. Researchers are finding the same pattern of rising T-helper cells, then a plateau, then falling T-cells, so that after about a year or eighteen months, patients will probably be back where they started. DDI is becoming available underground, but it must not be considered harmless, as there are side effects including rash, rise in liver function tests, pancreatitis, and peripheral neuropathy. So far the neuropathy has caused the most concern, as a few patients have had severe pain in the feet after 30 to 40 weeks of the drug at doses about equal to those which will be used in the trials; in extreme cases the pain can be debilitating and not responsive to optiates, and it can take weeks to recover after the drug is stopped. At higher doses, the neuropathy has occurred as soon as ten weeks; the problem may depend on total cumulative dose. DDI might prove most useful when taken in cycles alternating with AZT, instead of taken continuously.

Questions have also been raised about the evidence for efficacy. Rebecca Smith of the AIDS Treatment Registry in New York pointed to the report that patients on four higher doses did better than those on four lower doses (recent article in Science, June 28). But the graphs on page 413 of that article show that patients on the lower doses started with about half the T-cells of those on the higher doses. This difference should be considered when evaluating the fact that all but one of the six opportunistic infections which developed occurred in patients on the lower doses. (Other indications of efficacy are more clear, however, such as the fact that every one of the 13 patients on the four higher doses had T-helper cell increases.)

This article has looked at the the current status of the negotiations concerning the issues and mechanisms of early access to experimental drugs including DDI, but only briefly at the arguments for or against the drug itself. We plan to follow all of these issues as they develop.