A recent front-page article in The Scientist (F. Hoke, "National AIDS Task Force Expected To Accelerate Drug Development," Feb. 7, 1994, page 1) reported that a newly formed, United States government-backed, 15-member panel intends, among other things, to improve communication between pharmaceutical and biotech companies and thus speed development of AIDS-combating antiretroviral drugs and vaccines.

One wonders, on one hand, what is wrong with the U.S. drug industry that such facilitation should be necessary and, on the other, whether an AIDS task force can, in fact, do anything that the industry is not already doing. The task force, it seems to me, has better things to do.

In trying to imagine a more appropriate research agenda for the panel, I find myself making three idealistic and naive (and, possibly, incorrect) assumptions: (1) that it has been formed to push AIDS research-not politics-forward; (2) that among its members are the equivalents of J. Robert Oppenheimer (to protect its mission, as he did with the Manhattan Project, against external influences) and Richard Feynman (to say, as he did in the Challenger inquiries, what needs to be said in moments when stark reality confronts us); and (3) that it is willing to put the lives of people with AIDS ahead of political correctness, patronage, and economic advantage.

I hope that my assumptions are correct, but I wonder. The fact that we still can neither treat AIDS effectively nor cure it strongly suggests that we do not understand it. I maintain that we have not yet asked all of the right questions about AIDS, and that our ignorance is therefore profound. Identifying the sources and types of our medical ignorance thus becomes the highest research priority.

A model program for accomplishing this has been developed by Marlys and Charles Witte, two AIDS researchers in the department of surgery at the University of Arizona Medical School, and a colleague of theirs, medical philosopher Ann Kerwin. According to their model, ignorance comes in four forms, each of which must be addressed as we pursue our research on AIDS:

Finally, there are the things we think we do not know but we really do. This is hidden knowledge masquerading as ignorance.Examples of each can readily be found in our current approach to studying AIDS, and, in my opinion, it is the job of the AIDS task force to identify and remedy as many of them as possible.

False Assumptions

An example of something we thought we knew, but did not, is that the human immunodeficiency virus (HIV) is the direct cause of T-cell killing in AIDS. Even such formerly stalwart proponents of this notion as Anthony Fauci and Robert Gallo now admit that this is not the case. Virtually all HIV research is now focused on finding "indirect" mechanisms by which HIV may cause immune suppression.

We also thought we knew that HIV alone is sufficient to cause AIDS. But such researchers as Luc Montagnier, Shyh-Ching Lo, Joseph Sonnabend, and many others-including me-now believe that cofactors are necessary and, therefore, that HIV by itself cannot cause AIDS.

We used to think we knew that everyone is at equal risk for HIV and AIDS, and that a heterosexual epidemic was inevitable. But the epidemiology of AIDS has yet to prove consistent with that view. There is no record of tertiary (non-risk group to non-risk group) sexual transmission of AIDS in any Western country. Indeed, every study of female prostitutes in Western countries has led to the conclusion that those among them who do not use drugs intravenously have almost no risk of HIV infection and that evidence of female prostitutes acting as vectors for spreading HIV into the heterosexual population is, at best, inconclusive.

We thought we knew that people in all AIDS risk groups proceed to AIDS at the same rate following HIV infection, but this also has turned out to be untrue. For young hemophiliacs, for example, the average time is more than 15 years, while for older hemophiliacs and homosexual men, the time to AIDS is 10 years. People who have acquired HIV through blood transfusions have a rate nearly double that of gay men (average onset at six years), but people who become HIV-infected during an organ transplant or cancer chemotherapy develop AIDS, on average, in only two or three years.

We thought we knew that HIV always precedes immune suppression in people who develop AIDS. But many studies show that lymphocyte counts are as low in some HIV-negative gay men, intravenous drug users, and hemophiliacs as they are in nonsymptomatic HIV-positive people-and sometimes lower.

We thought we knew that public health measures to combat AIDS-"safe sex," clean needles for addicts, and so forth-work because they interrupt HIV transmission. But epidemiologic studies have shown that transmission of all suspected infections that may act as cofactors in AIDS is also interrupted. We do not know, therefore, why public health measures work.

Other aspects of AIDS dogma have also been challenged-many of them in the last year. For example, we now have evidence raising fears that current vaccines, designed to spur antibody production, may be useless or even detrimental; we have been told that significant percentages of hemophiliacs have beaten HIV and seroreverted without resorting to antiretroviral therapies; and, most recently, the very poor efficacy of AZT has been revealed to us.

If such important, and previously obvious, "facts" are now called into question, we must seriously consider how many of our remaining notions about AIDS are similarly biased by our preconceptions and are, therefore, not trustworthy.

A basic role that the AIDS task force must perform, then, is to make sure that, even to the point of discomfort, we are constantly skeptical and inquiring about the things we think we know-but really do not.

Hard Questions

The things we know we do not know are much more obvious than the things we think we know but do not. For example, we know that we do not know how HIV causes immune suppression. We are not even sure that HIV is sufficient to cause AIDS without other immunosuppressive cofactors, since it is a documentable fact that no one who gets AIDS has HIV as his or her sole immunosuppressive risk.

Although every person with AIDS has a variety of autoimmune complications, even including antibodies against his or her own T-cells, we know that we do not know what role these play in AIDS. Indeed, we do not know what triggers any human autoimmune disease, whether in AIDS or as a factor in other syndromes. We do not know, in consequence, whether treating HIV will be sufficient to cure AIDS, or whether AIDS may continue to destroy the body through autoimmune mechanisms even after the virus is eliminated.

We know that alcoholism, crack cocaine use, and non-IV use of heroin greatly increase the risk of contracting HIV. But there have been no telling studies of the lifestyle or immunologic function of alcoholics, crack cocaine users, or non-IV heroin addicts at risk for AIDS, nor of the mechanisms by which these people acquire their HIV (and other) infections.

Do crack users, for example, share the very high rates of sexually transmitted diseases, malnutrition, and bacterial and viral infections that characterize IV-drug users and cause immune suppression in them? Do sores in the mouth associated with crack cocaine use and oral sex with HIV-positive partners facilitate HIV transmission? Do people who use non-IV drugs participate more often in unprotected anal intercourse (the most efficient way to transmit AIDS sexually) than other people? Does intercourse during menstruation increase the probability of heterosexual transmission of HIV? Do disease conditions that are associated with immune suppression-such as diabetes, dialysis, anabolic steroid use, anorexia, and bulemia-pose additional risks for contracting HIV? We do not know the answers to any of these fundamental questions.

We must address all of these matters-the things that we know we do not know-and many others if we are to succeed in defeating AIDS. The AIDS task force should therefore promote, as one of its highest priorities, the investigation of every anomalous and unexpected observation that threatens the sanctity of the current AIDS dogma.

'Crazy Hypotheses'

The things we do not know that we do not know are, of course, the most difficult forms of ignorance to identify. And to go about identifying them we must invent crazy hypotheses and do unthinkable experiments. I advocate this approach not because we can expect these hypotheses or experiments to work, but because the history of science shows that employing them is the most successful research strategy for addressing the unknown.

The notebooks of the great biomedical scientists-Jenner, Pasteur, Fleming, Blumberg, and so forth-reveal a record of failed theories and botched experiments that ultimately led to unexpected results. In disproving an incorrect theory or in running an experiment for which there was no sound, establishment-validated rationale, these clever scientists encouraged serendipity, the wellspring of scientific insight. On accepting his 1976 Nobel Prize in physiology or medicine, Baruch Blumberg said: "I could not have planned the investigation at its beginning to find the cause of hepatitis B. This experience does not encourage the approach to research which is based exclusively on goal-oriented programs."

The programmatic, rationalistic converse of the Blumberg approach is much more common, of course. But we must bear in mind the number of times that supposedly well-founded approaches to disease have turned out to be harmful. Physicians in the 19^th century "knew" that germs did not cause disease; it took a physical chemist named Pasteur, working outside of the medical community, to prove otherwise. It is just these things that make sense-but are wrong-of which we must beware; and we can beware only if we control every theory by testing it against alternative theories that we do not expect to be correct. Sometimes we will be surprised, and-by being surprised-we will discover our ignorance about things we did not know we did not know.

To nurture our curiosity and thus broaden our opportunities for serendipitous discovery, I suggest that we cut back funding to those animal models and test-tube studies that do not behave like human AIDS and, instead, put more effort into modeling what really happens in human beings. No one, for example, has ever mimicked in an animal the entire range of immunosuppressive agents that bombard a blood-transfusion patient: anesthetics; surgery; multiple blood transfusions contaminated not only with HIV but also with cytomegalovirus, Epstein-Barr virus, and hepatitis C virus; opiate analgesics; and high-dose antibiotics. No one has yet modeled such IV-drug-user risks as multiple, concurrent infections with sexually transmitted diseases; bacterial infections from unclean needles; constant re-exposure to alloantigens (blood, lymphocytes, tissue) on unclean needles; persistent drug addiction; chronic antibiotic use; and malnutrition. No one has repeatedly reinoculated chimpanzees or macaques rectally with semen containing HIV, herpes viruses, and mycoplasmas, while concurrently exposing them to inhalant nitrites, antibiotics, and other drugs, as has typified so many men and women who have contracted AIDS sexually.

So another of the functions that the AIDS task force must take on is to make sure that no assumption goes unchallenged, and to provide sufficient freedom for nonconformist research that might well yield serendipitous surprises. Let's start with messy reality instead of assuming that HIV is the entire answer.

Hidden Knowledge

Finally, there are the things that we do know, but we think we do not. Most important, we think that we do not know of any cure for AIDS at present, but, really, we do. That there is a cure is clearly evident all around us-the hundreds of documented cases of people who have been infected with HIV, who have developed T-cell deficiencies, and who have subsequently returned to a normal immunologic status, lost all signs of HIV, and remain healthy. There are people who have had antibody to HIV for more than 10 years who display no signs of immunologic damage. And there are those rare cases of people who have had full-blown AIDS for more than a decade and are still alive.

These people are living proof that there are things about AIDS that we do not know we know; they are walking data banks, waiting to be tapped, waiting to reveal to us the knowledge that we possess but are unaware of. We need to uncover the hidden knowledge that is within our reach by studying these people to find out whether the strains of HIV that have infected them are nonpathogenic and therefore protective against pathogenic strains; whether they are genetically different from other human beings; or whether they have treated themselves differently and so hold the clues to treatment of other HIV-infected people.

Limited formal studies and much anecdotal evidence suggest HIV have been treated with retroviral drugs; most have drastically altered their lifestyles to eliminate ongoing immunosuppressive risks, including re-exposure to HIV and other sexually transmitted diseases; they have ceased drug use; and most have adopted high-nutrition diet supplements to boost immune function.

These people may be telling us that we know how to treat AIDS successfully through risk elimination and immunologic boosters rather than antiretrovirals or HIV vaccines. One reason we do not know that we know this is because these people have succeeded by ignoring mainstream medical advice, and their knowledge is therefore outside the structure of biomedical science. I suggest that whatever these survivors know must be combined with what the world of official biomedical investigation knows and become a focal point of AIDS research as soon as possible. We must refocus AIDS research on seroreverters, long-term survivors of HIV infection, and long-term survivors of AIDS.

Of course, we must also place whatever knowledge we get in such unorthodox ways back into a standard scientific form:

What we do not know that we know tends to be expressed in terms of anomalies-things that do not fit our expectations-and are therefore generally overlooked or ignored. The AIDS task force should pursue studies involving these survivors and see how much of AIDS can be prevented by focusing on cofactors.

An Open Field

Recognizing the extent of our scientific ignorance of AIDS leads me to conclude that we have narrowed our focus too precisely. AIDS is more complex than just HIV. It includes everything that predisposes people to HIV infection, everything that can synergistically work with HIV once present, and everything that can cause immune suppression in people at risk for AIDS independently of HIV. Recognizing this increased complexity does not displace HIV from its place at the center of AIDS research, but it does provide a wide range of new targets for prevention and treatment.

The useful function that an AIDS task force can play is to make sure that the utmost reaches of our ignorance are quickly and efficiently identified. We will not do that by staying in the rut that has been gouged out during the past decade, but rather by deviating from it.

A diversity of opinion and of research has never hurt science. Dogmatism and politically motivated programs often have. The AIDS task force can foster one or the other, but not both. *

Robert S. Root-Bernstein, an associate professor of physiology at Michigan State University, East Lansing, is the author of Rethinking AIDS: The Tragic Cost of Premature Consensus (New York, Free Press, 1993) and Diversity (Cambridge, Mass., Harvard University Press, 1989). He is a former MacArthur Fellow (1981-1986)