THE DRUG-TRIALS DEBACLE, PART II: WHAT TO DO NOW

"If you ask researchers the question, as I have several of them, 'If you did not need to deal with the FDA regulations, if you did not need to face all the questions of marketing and licensing, etc., and just get an answer, could you determine
within six months whether a drug is going to be useful in the fight against AIDS?' If you just ask the question in that simple way, all of them I've asked say, 'Well, of course we could.'

"That being the case, let's get some of this baggage out of the way, and get these answers more quickly, and act like this really is the emergency that everyone says it is."

-- Martin Delaney, co-founder and director of Project Inform, on KQED public radio "Forum" program, May 3, 1989.



Part I of this article (AIDS TREATMENT NEWS #77, April 21, 1989), presented a simple mathematical model which anyone can use to calculate the loss of human life caused by delays in AIDS research and treatment access. We showed that because of the geometric progression of the epidemic, a delay of 18 months to three years** would cause half of the total of all the deaths due to AIDS -- deaths of people whose lives can be saved if the delay can be avoided. This model brings home the cost of the widespread unspoken attitude that everyone infected with HIV is going to die anyway, and therefore we can write them off and ignore research and access delays, in favor of other things such as services for the dying short of saving their lives, or grief support. If we assume that the epidemic will someday be controlled, then it is certain that we can save many thousands of lives by eliminating some of the unnecessary delays now built into the research and treatment access and delivery systems.

In Part I, we also showed that if a cure were found, there would be no mechanism to release it to people quickly -- that in fact the bureaucratic incentives are to avoid risks and therefore to conceal treatment advances rather than release them. We pointed out that delays in the new-drug research and regulatory "pipeline", delays long enough to cause tens of thousands of deaths, serve the interests of major corporations by rationalizing and protecting their drug-development investments -- and that perhaps as a result, the "reforms" allowed to take effect have been only those which could not change the outcome, even if they worked perfectly as intended.

Part I showed what is wrong. Here we suggest what we believe can and should be done toward correcting the problems. We will show what kinds of studies could quickly and inexpensively produce information that would assist the treatment decisions physicians and patients must make now -- and why the research establishment largely rejects such studies. We will show that the mainstream research tradition in the United States has become skewed toward producing the kinds of information which corporations and regulators need to make their decisions -- not the information which patients and physicians need to make theirs.

We will show that much of the needed research could be done legally in the United States today, using community funding to bypass the financial control and ineffectual bureaucracy of the research establishment -- and that some of the studies which could not be conducted in the United States could be done elsewhere.

Politically, we will suggest that a key, doable first step is to develop clearer statements of consensus within the coalition of individuals and groups already committed to saving lives, then use this consensus to get our friends on board.

The Mainstream View Of Research

One school of thought has come to dominate government funding and permissions, and therefore research careers, in the United States. This establishment approach is not, of course, all bad. The problem is that it has achieved such dominance that it can insist on applying its own ways of doing things to all situations, no matter how inappropriate the result.

Mainstream treatment research is based on the following mindset:

* The first goal of clinical research is to prove, to a statistically stated degree of confidence, that a drug does work better than nothing, or better than some existing treatment.

* There must be a control group, as otherwise there would be no way to justify the statement that the drug to be tested is better than something else.

* Above all, clinical trials must guard against the danger that a drug which is useless and perhaps harmful could become accepted and generally used in medicine, as has happened many times in the past.

* Since most drugs being tested will show only a small benefit, trials must be designed to distinguish a small benefit
from none at all.

* If patients use other treatments during a trial, their effects could interfere with the results. Therefore subjects must refrain from other treatments -- even if the drug being tested will in fact be used with other treatments after it is approved.

* The fact that a trial imposes an unrealistic environment which may never occur in practical use of the new drug does not matter. What is important is to learn about a drug in isolation, not a therapy in practical use.

* If for whatever reason (such as lack of national political will) it is impossible to arrange a trial which meets these and all other standards of pure research, then it is better to do nothing until such future time as trials may be done, instead of doing any other kind of a study, which could lead to error.

* There is no need to design trials in such a way that it is feasible for any particular patients to volunteer, or for physicians to recommend their patients. If the trial fails to recruit subjects, that is not the fault of the researcher, whose job concerns pure science, not practical medicine.

* It is ethical to deny access to treatments until the trials are complete or at least well along -- either to force patients into trials, or to maintain a stockpile of untreated patients available for future studies which may occur when someone gets around to paying for them.

* If the necessary trials do not get done, no one is responsible, since no one has the job of expediting trials, or untangling the snafus which block them.

* No one is responsible for the tens of thousands of unnecessary deaths which will result from this approach. All involved can make the case that they have done their jobs. The final outcome is no one's responsibility.

This system developed to serve the needs of the powerful players: drug companies and Federal agencies. The companies want above all to get their "NDA" (new drug approval), allowing them to market a drug which they have exclusive rights to and have chosen to push. The Food and Drug Administration, supported by consumer protectionists in Congress and elsewhere, wants above all to protect the public from unsafe or unproven drugs. The National Institutes of Health wants to pursue studies which are scientifically interesting. No one has the mission of making sure that trials which are critically important for saving lives get done quickly, or making sure that patients have access to treatments which are clearly beneficial but which for any of a multitude of reasons have not gone through all the steps necessary for full marketing approval.

Prospective Monitoring Studies: Another Kind of Trial

The main problem with controlled, randomized trials, the kind the U.S. research establishment has insisted on, is that they are very difficult to get going. They are difficult because the procedures which take place in these trials are so different from those in the normal practice of medicine. There are major ethical and practical difficulties in giving patients a placebo -- or in asking them to submit to any randomized study, in which they do not know which of two or more medicines they will receive. Only large, well-financed institutions can manage such trials, and the red tape involved usually creates months or years of delay.

It may seem that these trials, however cumbersome, are the only way to get credible information. After all, researchers can show with statistics that using fewer patients, or deviating from their rules for running trials, may cause wrong answers to be found. It is well known that many inadvertent biases in the design or conduct of clinical trials can cause a drug to appear appear effective when really it is not, or vice versa.

But these arguments assume that the goal of the trial is to prove (or fail to prove) that a drug is effective. This is the information which drug companies and regulators need for their decisions, and of course this information is useful to patients and physicians too. However, there are other kinds of studies which do not even try to prove whether or not a drug is effective, which can produce information useful to patients and physicians (but much less useful to corporations and regulators).

One of the problems we face today is that if the large, cumbersome trials have not been done, then the only alternative has been anecdotal information, which is notoriously unreliable, and often under the control of self-interested parties -- true believers or promoters with a product to sell. The obvious problems of such reports have discredited any information not confirmed by big-money, big-bureaucracy trials.

If you are considering a new treatment now, you have several alternatives, all of them unsatisfactory:

* Wait several years or more until official trials have been done and the drug is approved.

* Try to get into a trial -- but it may not start for a year, the nearest site may be hundreds of miles away, you may not qualify, you may have to stop other drugs (or have never used them), and then you may get a placebo.

* Ask around. Maybe a friend knows two people who used the treatment and whose T-cells went up. Maybe they forgot to say that the before and after tests were at different labs, that they began other treatments at the same time, and that other blood values deteriorated.

* Read articles about the treatment. Unfortunately, the writers may have an interest in promoting it. And even if not, they will usually have had to base their articles on anecdotal reports, since good information is not available.

There is another alternative, however -- very well managed collection, handling, and presentation of data about a particular treatment, in the environment and context in which that treatment is actually used. Community-based research organizations can take treatment information out of the hands of the true believers and promoters, and have it controlled instead by professionally guided research teams which serve no interests except those of the patient community.

Besides waiting for years for randomized trials, or using anecdotal reports to make treatment decisions, there should be another choice. Prospective monitoring studies could provide another source of treatment information. Here is how they can work:

When persons with AIDS or HIV start using a new treatment (for example, hypericin), a research organization could offer to monitor perhaps 20 to 50 persons, paying for blood work and physical examinations. The study would be designed in advance (that is why it is called "prospective") and approved by a scientific advisory committee, so all the important data would be collected for every patient, in a uniform way; for example, all blood work would be done by the same lab, to avoid inter-lab variations. All physical exams and medical histories would be conducted uniformly. Identical patient diaries can be used. All patients would be accounted for. Data handling would be audited and would meet the same standards as in any other clinical trial.

A purely monitoring study cannot ask patients to change what they were doing for the sake of the research. Therefore patients can use whatever other treatments they want during the study, as long as they tell the researchers what they are doing.

As the study proceeds, data is statistically summarized and given to one or more leading HIV physicians for their interpretation. Is anything happening which is dramatically different, either better or worse, from what would have been expected without the treatment being tested? Or is it unclear whether or not the treatment has helped -- meaning that the benefits, if any, are less than dramatic? These evaluations by the physicians, along with the statistical summaries, would be published as the report of the study. Because there is no control group and no randomization, this study is not designed to "prove" the drug safe or effective. The treatment group is in effect being compared with the expectations of the expert physicians chosen to evaluate the data obtained -- a method not as statistically precise as using a placebo control, but certainly able to pick out a decisively effective treatment, which is what these studies will be looking for.

Instead of asking for statistical proof, the important question for judging a monitoring study is whether it provides infor-
mation useful for making treatment decisions. Primary-car physicians will make this determination, when they decide what studies are credible. Our own expectation, after reporting on AIDS treatments for three years, is that for many unapproved therapies, a single such study, scientifically designed and professionally managed by an unbiased research organization and collecting complete data from several dozen patients, could produce better information on the use of the treatment for HIV than all of the world's anecdotes and rumors put together, even for substances which had already been widely used for months or years.

Perhaps most importantly, such monitoring studies could be used as a quick screening for the most promising treatments now entering human trials (such as compound Q, DDI, or D4T). The goal would be to look for very dramatic benefits, in order to bypass years of ineffectual bureaucracy for any treatment found to work so well that there could be no dispute about its value.

Monitoring studies have several advantages of flexibility and ease of use:

* No FDA permission is required, because the study does not give any drug to people -- it only collects data. Major delays are therefore avoided.

* The cost is low. Less than a thousand dollars per patient will pay for physical examinations and for six to eight months of blood work more complete than that of many official "phase II" studies. Therefore this research can be supported directly by contributions from the public, bypassing government agencies which often take more than a year to award money -- a year after completion of the ponderous applications, which can run to hundreds of pages. Since there is no major overhead cost, monitoring studies can start with whatever funding is available and add more patients later as additional money comes in, or let patients pay for their own blood work until funding can be found.

* Reports can be compiled and published at any time, not just after the study is done, as is usually the case with randomized trials. Mainstream medical journals may reject these papers because there is no control group. But the results can be distributed immediately by community organizations to patients and physicians, without being restricted for months by pre-publication secrecy.

* The fact that patients can use other treatments during the study will make the results more difficult for the physician(s) engaged for that purpose to interpret. But in return there are two advantages of not restricting other therapies. First, the treatment is studied in the actual context of its use, not in an artificial context of a single drug tested in isolation. And second, in the traditional trials which kick people out for using additional treatments not in the protocol, patients whose lives are at stake often use other drugs anyway, and conceal what they are doing. The difference is not whether other drugs are used, but whether the researchers know about it.

* Since these studies do not require any medical sacrifice of the patient -- they simply offer free blood work -- recruitment can be much easier. Randomized trials are often delayed for months and sometimes cancelled entirely because they cannot recruit patients.

* Since no big institution is needed to run monitoring studies, this research can be more responsive to community needs than the official, randomized drug trials.

* Monitoring studies can legally be done in the United States, provided that patients can obtain and use the drug without the help of the researchers. If patients cannot obtain a drug here, the study might be conducted abroad.

Monitoring studies are already occurring -- for example, an antabuse project of the Community Research Initiative in New York, and monitoring of AZT, and of alternative therapies in general, by the County Community Consortium in San Francisco. Although the idea of community-based research is to conduct trials through physicians' offices, monitoring studies might work better if the blood tests, physicals, and patient interviews were handled at central locations when possible, so that staff can be trained to do these consistently. The research organization must coordinate with primary-care physicians, of course, and give them copies of laboratory reports and other information. But most primary-care physicians are too busy to go out of their way to collect data in a specified, uniform manner. This job can be done by medical staff trained by the project and following written guidelines.

Treatment Politics: Challenging the Death Consensus

The fast, inexpensive kind of study suggested above is only one example of how AIDS treatment research could be improved. The political task is more basic: how to overcome the widespread fatalism which makes even friends of the AIDS community unwilling to deal with treatment issues, as they have already given up on saving the lives of persons now ill or infected, and written them off as dead. How to we respond to the widespread, often silent assumption in professional and institutional circles that saving the lives of those now infected or ill is either impossible or not worth doing?

Congress, for example, is today largely a wasteland on the issues addressed here. The usual attitude toward those who bring the subject up has been described by one treatment advocate as, "You are the doomed or advocates for the doomed, and the doomed always want more drugs." End of conversation -- and of any effort or interest in dealing with the issue.

To change this attitude, which today forms a consensus even among many friends of persons with AIDS in Congress, we need to start in our own community.

Recently a leading gay rights lobbyist, describing his commitment to AIDS work, was quoted in a major newspaper as saying, "I feel compelled to use my professional skills to make it easier for those who will die, and to prevent others from getting sick." Too many organizations have written off much of their constituency as dead and left out any involvement in saving the lives of the tens of thousands who will die unnecessarily as a result of current policies. No wonder Congress and the research community have failed to examine their own fatalism, their unwillingness to lift a finger to change policies which make thousands of unnecessary deaths inevitable, since even the AIDS community's organizations and advocates have not done so. How can we expect others to speak for us when we will not speak for ourselves?

Why have most AIDS organizations been so reluctant to work on treatment issues? There seem to be many reasons. One is that they fear differing with their political allies, usually liberals, who for years have been fighting for consumer protection. Consumer protectionists want to see the strongest possible FDA, strong regulations, and the most exhaustive testing of new drugs before they are released. They are afraid that AIDS will allow the pharmaceutical industry to weaken the regulatory system they have worked so hard to build, and that flexibility in treatment access will facilitate quackery and unscrupulous exploitation of persons who are desperate. (We too support consumer protection -- but not at the cost of human lives. AIDS must be treated as an emergency, as it would have been if it had not first been perceived as a gay disease. It is not enough to simply apply old battle lines in utter disregard of the existence of this emergency, and of the effects of its eight years of its malign neglect under the Reagan White House.)

A second reason is that AID S service organizations are usually publicly funded, sometimes with Federal funds, so they may be fearful of questioning Federal agencies.

A third reason is the emotional issue of HIV testing. Of hundreds of AIDS organizations in the country, only a handful, mostly in San Francisco, are now willing to recommend that persons at risk of AIDS seek voluntary, anonymous testing. The others may be deterred from becoming involved in treatment issues, because if they did, they would face the contradiction that people cannot obtain early treatment (such as aerosol pentamidine before a first attack of pneumocystis) unless they seek testing to find out whether they need it. Organizations must re-evaluate strongly held positions in light of the fact that early, voluntary testing has now become a medical issue, as there are many patients who clearly should receive preventive treatment even though they feel fine and have no outward sign of illness.

The AIDS community needs to tell its advocates what kind of representation it wants. Are we satisfied to accept projections of tens or hundreds of thousands of deaths, without making any effort to change a system which keeps new treatments in the drug development and regulatory "pipeline" for years longer than necessary? Are we willing to accept a consensus which keeps designing trials which are so unworkable and inhumane that it is widely believed that patients must be denied access to treatments outside of trials, or else nobody would volunteer and the trials could not be conducted?

The death consensus is so entrenched that it is hard to know where to begin to change it. One Washington, DC-based PWA organ izer made what seems to be an excellent suggestion. The way to start, he suggested, is to develop a coalition of those who already agree, then use that coalition to force other friends of the AIDS community to face the issue. He only saw three groups already mobilized for saving the lives of those now infected or ill: persons with AIDS or HIV, "treatment physicians", and some AIDS activists.

While today the picture is bleak, there are great pressures for change. Treatment will inevitably become a central issue in AIDS, as more and more people see that it affects them. Meanwhile, the first steps are clear. We need to develop explicit consensus among those already committed to saving the lives of persons with AIDS or HIV, and then talk with those among our friends who have so far refused to become involved.

Footnote:

** The 18-month figure for the doubling time of the AIDS death rate, used in part I of this article, is approximately consistent with the projection of the U.S. Public Health Service of 179,000 deaths at the end of 1991. In San Francisco, however, the pro- jections fortunately indicate a longer doubling time, between two to three years. There are many possible reasons for this differ- ence, among them different statistical methods used, the later stage of the epidemic in San Francisco, and much earlier safer- sex education in San Francisco than nationally.


STATEMENT OF PURPOSE

AIDS TREATMENT NEWS reports on experimental and complementary treatments, especially those available now. It collects information from medical journals, and from interviews with scientists physicians, and other health practitioners, and persons with AIDS or ARC.

Long-term survivors have usually tried many different treatments, and found combinations which work for them. AIDS TREATMENT NEWS does not recommend particular therapies, but seeks to increase the options available.

We also examine the ethical and public-policy issues around AIDS treatment research.