From: sbharris@ix.netcom.com(Steven B. Harris) Subject: The AIDS Heresies ((all articles, combined)) Date: 22 Dec 1996 Newsgroups: misc.health.aids,uk.gay-lesbian-bi,misc.health.alternative,sci.med THE AIDS HERESIES Introduction: The following paper, _The AIDS Heresies_, was written in response to a large recent increase in unscientific and frankly paranoid theorizing about HIV and AIDS. The recent publication of a book by prominent retrovirus researcher Peter Duesberg, Ph.D., claiming that HIV does not cause AIDS, and the well- publicized statements of a Nobel Prize-winning chemist (Kary Mullis, Ph.D.) also publicly questioning the HIV-AIDS theory, have both contributed to a certain amount of skepticism within the HIV-infected community, if not the general public. It has become clear that some kind of detailed response is needed to Duesberg and his followers, to be used by professionals and laymen alike in the always difficult question of what to do about AIDS. This essay was begun at the end of 1994 with this end in mind. The issue of whether or not HIV causes AIDS is a life-and- death one. Until recently, the treatments for HIV infection were only moderately better than doing nothing, but there was still the question of prevention. It seems likely that the sexual behavior of someone who knows themselves to be infected with a virus which is deadly, versus the behavior of someone who is convinced his or her infection is less serious than a cold, will be quite different. To this end, it is likely that skepticism about the cause of AIDS has already contributed to many HIV infections, and many deaths, as people carrying HIV simply lie to themselves and their partners about carrying the virus. "Sex will not kill you," states Peter Duesberg, who recommends that people worry only about previously known, non-fatal sexually transmitted diseases. But such a low and inappropriate level of concern will be expected to be a disaster in a venereal plague which is, in fact, deadly. Progress against AIDS continues, and in late 1995 and early 1996 a new class of antiviral drugs (HIV protease inhibitors) was introduced. These antivirals have shown the remarkable ability to inhibit HIV replication almost completely, for periods of up to a year and a half, and perhaps far longer. Addition of these new drugs to the medical regimen of tens of thousands of people dying of AIDS has already caused many to regain weight and health, and, to their astonishment, the will to continue with their previous lives. Smaller AIDS wards across the country are literally being shut down for lack of sick patients. Yet in the face of such gains the critics continue to tell HIV-infected people that if they take no drugs of any kind they will surely live, but that if they agree to medical treatment, they will surely die. Never before has the question of belief in one of the more difficult theories of medicine been so important to so many people. This essay tells the story of the discovery of HIV and its lenitvirus relatives. The reader who perseveres will gain an understanding of not only HIV epidemiology and history, but also some understanding of the difficulties of using the inferential scientific method. When does a critic of science become a crank? What makes the establishment so sure about the role of HIV? How much room does the evidence give for alternate theories? All these questions will be addressed. This essay, written as it is for net distribution, is a work- in-progress. The author would be glad for any feedback from any reader. Steven B. Harris, M.D. Salt Lake City, Utah December, 1996 Email: 71450.1773@compuserve.com sbharris@ix.netcom.com N.B. The author is an internist, clinician, and experimental gerontologist who receives no grant money from the government AIDS research establishment, and does not feel himself constrained in his opinion on this subject in any way by his academic status or occupation. This draft information brief may be freely reproduced and distributed in this ASCII form by anyone without license, so long as the ENTIRE contents are reproduced, including author copyright information and notes, and so long as reproductions are not sold for money (this does not include normal connect time charges on computer nets for downloading, but would include any special charges for the article itself on a BBS). Those seeking other arrangements should contact the author. Illustrations and figures for some of the concepts are available, and can be obtained by contacting the author. A shorter edited and illustrated version of this paper appeared in SKEPTIC Magazine in vol. 3 no. 2, 1995, with follow-up letters and an author reply in SKEPTIC vol. 3 no. 3. For SKEPTIC information call (818) 794-3119. THE AIDS HERESIES A Case Study of Skepticism Taken Too Far (c) 1994-1996 Steven B. Harris, M.D. _Felix qui potuit rerum cognoscere causas._ [Fortunate is the man who understands the causes of things.] Virgil "It's the virus, stupid." Dr. David D. Ho AIDS Researcher A Dialogue In Inductive Frustration Let us suppose that you have a bright and iconoclastic friend who smokes three packs of cigarettes a day. You remark one day that you would like to see your friend quit the habit, since he is certainly increasing his chances of lung cancer. "Prove it," says your friend. "Well," you begin, "the Surgeon General and a lot of scientists and doctors say you should quit...." "Come now!" says your friend, "Since when did you become a fan of The Argument From Authority? I can find you scientists who do NOT believe I necessarily should quit, too, as well as a lot of intelligent business executives." "Sure, but all those scientists and all those executives are paid by tobacco companies or grants from the Tobacco Institute," you protest. "Well, what do you expect?" says your friend, lighting up and taking a satisfying drag. "Whenever any scientist takes an anti-establishment, anti-government position like that, all grant funding is cut off. Didn't you know that? The poor scientists then don't have anyone else to support their research _but_ the Tobacco Institute. Do you expect them to starve or drop out of research just because they hold unpopular opinions?" "Okay, let's look at the facts," you say. "What about the fact that 90% of lung cancer occurs in smokers?" "Yes," says your friend, "and that means that 10% of it occurs in non-smokers, doesn't it? Obviously the `cigarettes = lung cancer' hypothesis doesn't explain all lung cancer, and even for smokers there must be `co-factors.' Heck, my grandfather smoked 3 packs a day right up to the day he was hit by a drunk driver at the age of 92. A lot of people smoke for a whole lifetime and never get cancer." "Look, I didn't say the association was statistically perfect!" you protest. "But it is certainly there. Two-pack-a- day people have 13 times the lung cancer risk of non-smokers." "Oh, really?" your friend says, "Now, where do you get that number? I suppose somebody did an experiment where they got together a group of nonsmokers and randomized them to start smoking, or else stay smoke-free, and then made sure each and every person did as told for the next 40 years, so as not to bias the results. I must've missed that study." "You know there is no such study. That experiment would have been impossible, since you can't enforce a random protocol like that. People will start or stop on their own. And besides, any experiment where you try to keep people from quitting would be immoral, since smoking causes cancer." "Sort of circular argument there," sighs your friend, admiring a smoke ring, "since that is what remains to be proved. So you admit you don't have any study where the two groups of smokers and nonsmokers are exactly equivalent, and only differing by chance or random draw? In every study the smokers and the nonsmokers are self-selected for their behavior, aren't they? And bound to be different not only in smoking behavior, but also because of whatever made them smoke or not smoke to begin with! So your experimental `controls' are inadequate, even if I do say so. Not exactly great science, if you ask my layman's opinion." "But darnit, when smokers quit, we know their risk of dying drops!" "You mean with regard to the smokers who don't quit? So what? The people who quit smoking did so for a reason other than chance or the experimental flip of a coin, I'm sure, and again that means they will differ in some way OTHER than their not smoking. Again you don't have a proper "control group" of non- smokers who had that very same reason, but were _prevented_ from quitting. Look here: did you know that for the first year after quitting, the risk of death for a new quitter actually goes UP with regard to his fellow smokers who keep right on smoking?" "I knew you'd bring that up. The mortality goes up for the quitter group for a while after they quit only because those people who quit are quite often sick, and that's WHY they quit." "If so that makes my point about self-selection, doesn't it? These aren't true controlled experiments. You're saying that in that first year of quitting, the higher death rate of quitters is caused by another factor in our study other than smoking-- namely, sickness. Well, so long as we're talking about such third factors, I have a hunch that _stress_ causes cancer, and stressed-out people take up smoking to try to relieve the stress, and that's why there is more cancer in smokers, not because of smoking. Moreover, maybe the act of quitting stresses people out, and that's really why quitters die faster in that first year after quitting. Smoking is just a marker for stress, you see-- what you statistics people call a "proxy variable." "All this is ridiculous! You're just using your intellect to make you believe something you want to believe for other reasons. There is experimental evidence! Smoking causes lung cancer in lab animals! Are THEY stressed?" "Actually, yes-- have you seen what they do to them in a modern lab? Ever seen one of those rabbits with a leather muzzle over its nose, and a cigarette stuck in it which it can't take out? But anyway, I don't even believe you can find me a report of an experiment in which smoking causes lung cancer in animals." Back you go to the scientific literature. And you find---- nothing. There is no such paper.... ----------------------------------------------------- Medical Induction. Because there are many intellectual steps which are not perfectly secure in any generalization, even the most detailed inductive argument only goes so far toward proof, as our fictio- nal dialogue, which is based on facts, demonstrates. It is also unavoidable that the same evidence will mean different things to different people. In particular, it is more difficult to induce a person to follow a complicated inductive-reasoning trail when they dislike, or are threatened by, the conclusion at the end. In the medical sciences, assembling an irrefutable argument for causation is sometimes an impossible task for the same reason it is in astronomy or paleontology: the direct and definitive experiment cannot be done. Scientists cannot travel back in time to watch dinosaurs, nor can they influence the behavior of planets or stars. In medicine, a common difficulty is that the necessary human interventive experiments to perfectly assess "risk factors" for harm may be unethical, and so these risks cannot be studied directly by experiment either [1]. How, then, do we come to "know" (or confidently believe) what things cause lung cancer or AIDS? For that matter, how do we come to know with any confidence that tyrannosaurs ate meat, or what generates the sun's energy? In other words: how _do_ we ever infer causation from looking at events (or records of events) which we cannot influence? However we do it, it does seem that it can (to some extent) be done. Modern science depends on the fact that "correct" causal relationships can often be guessed entirely from logical and indirect observational tests of competing theories, even where direct experimentation is not possible. This is done using help from knowledge of simpler causal mechanisms which we have gained from similar systems in which experimentation _is_ possible. One of the most amazing things about the universe is that this kind of inference is possible at all, as Einstein once commented. Of course, the overall results of this kind of theorizing, like those of any inductive process, are never certain. Still, whenever inferential theories in science finally _do_ become directly testable by some new experimental technique, they often prove to be surprisingly sound (or "robust," as scientists say). Why this should be true remains the deep mystery that it was for Einstein. We are only little more than re-stating the mystery when we note that the universe seems to have a certain degree of uniformity and symmetry at many levels, allowing us to correctly guess at the structure of some levels of uniformity, on the basis of knowledge of others. Human intuition, trained by a core of more secure direct experimental knowledge, does seem often an adequate tool for doing some of this kind of guessing. On the basis of our past success, we expect (on the basis of the symmetry of past and future) that the present inferential methods of science will _continue_ to result in progress toward truth in many fields of knowledge. This progress in understanding will occur even before final experimental "proof beyond all reasonable doubt" arrives in each case. It is because of an inferential process, based on many lines of evidence, that we can be reasonably confident of the tobacco causation of much of lung cancer, even in the absence of a definitive experimental study. In the same way, an examination of a large body of related facts allows us reasonable confidence about the causation of other diseases-- even a disease far more complicated than lung cancer, and with even more money and passion involved on either side of the issues. The HIV-AIDS Hypothesis, and Its Challengers. Recently, several popular lay publications (Reason Magazine, Spin Magazine, the New York Native) have run articles calling into question the theory that the viral agent with the conclusi- on-asserting name, the "human immunodeficiency virus (HIV)," is the cause of the epidemic of human acquired immune deficiency disease, known as AIDS. What do we mean by talking of the "cause" of AIDS? We know that the common cold or the flu-- indeed all infectious diseases- - are in some sense "caused" not only by the pathogenic organism. Also important in the causal chain are host factors (such as immune response), and even simple host-overwhelming factors, such as the infectious dose of organism which enters the body (called the "inoculum"). These additional causal factors, which have nothing to do with the microbe itself, can be extremely important. They may in some cases outweigh everything else. Because (for example) the smallpox virus is necessary for smallpox, if not sufficient, medical science still regards it as "causal" in the sense that if there is no microbe, there is no illness. Eliminate the smallpox virus from the population and one eliminates the disease, as was in fact done in the 1970's. Yet even this kind of "causal" connection between a disease syndrome and infectious agent is what is under attack in recent articles about HIV and AIDS. Some skeptics have claimed not only that HIV is not the only external factor necessary for AIDS, but perhaps even that HIV is not always necessary for AIDS, so that if HIV were eliminated from the Earth, at least some AIDS would still be with us. Still others have gone further and claimed that HIV infection is totally harmless and never even _contribut- es_ to the development of AIDS. These people believe that if HIV were to disappear, AIDS would continue exactly as before. In what follows, we will examine the best evidence behind what most researchers believe is the role of HIV and other factors in AIDS. We will also examine leading skeptical views on the causation of AIDS. Because a great deal of published research is available on this issue, our examination of AIDS will also let us illustrate how science closes in on cause and effect, even when direct experimental "proof" is not available. We will thus be interested in not only AIDS, but also larger questions about science, and scientific debate. What makes a good scientific theory, and what makes a poor one? Are there reasons for hope in looking at the disease of AIDS in particular, and the workings of the biomedical scientific "establishment" in general? Or has science been completely corrupted in working on the cause and cure of AIDS, as many critics suggest, so that with this particular medical problem the scientific method has ceased functioning? Are we making any progress with AIDS, or just wasting billions each year chasing fantasies? This essay will argue that we are not wasting our money, and that when it comes to critics of the HIV/AIDS hypothesis, we have a practical case in which scientific skepticism has been taken too far. Science, we are happy to report, still works, and it is making progress with AIDS. That some critics have failed to recognize this only highlights the fact that science is only partly an empirical enterprise, and that it also has an intuitive and aesthetic side which is subject (to a certain extent) to arguments over taste. This is not a thing which is taught about science in school, but it is a concept key to understanding most scientific controversies. What Is This Thing Called AIDS? Scientific problem-solving begins with definitions, and in choosing a definition for AIDS we run immediately into the HIV/AIDS controversy. Some of the difficulty is that definitions, even in science, are chosen partly on aesthetic grounds. There are utilitarian principles, too, of course. In medical science we rarely know in detail at the molecular or even cellular level what causes most human illness, and so in our ignorance we are often forced to work with "disease syndromes," which are collections of symptoms and sometimes lab tests which seem to "go together." In order to usefully define a "disease syndrome," (which gets promoted to a "disease" after it stands the test of time and we get used to it) we need to pick our defining characteristics so as to include all of the sick people who we are interested in for good clinical reasons, and exclude everyone else. What are "good clinical reasons"? In medicine there isn't much point (other than pleasing some doctor's vanity) in defining a new "disease" which, when present, makes no difference in either prognosis or treatment. Nor is there much point in defining a disease so poorly that it fails to capture all the sick people who seem to have pretty much the same thing wrong with them from the prognosis or treatment viewpoint. If (as always happens) we lack information about what impact certain definitional characteristics _have_ upon treatment or prognosis, then we are forced to _guess_, as best we can, what disease definition will be most useful. It is at this point, in deciding whether two people have "pretty much the same thing wrong with them," that aesthetic and intuitive considerations unavoidably enter into medical science. Utility imposes other constraints, too. A disease definition which is to be used during a hunt for the disease's causation, should not assume any cause which is in question. To be specific, if we choose a definition for "AIDS" which in some cases requires evidence of infection with the HIV virus (the current way it is done in many countries, including the U.S.), then we will have chosen our terms so as be of little help in the question of whether HIV causes AIDS. Obviously, it would be nothing remarkable if we "discovered" that almost all people with AIDS were infected with HIV, if we had previously chosen our AIDS definition to make sure that it was so. In re-opening the question of the cause of AIDS, what we need then is a modified AIDS definition which does not involve HIV, so that the question of whether or not all AIDS cases are infected with HIV is an empiric one, not simply a semantic one. When we have a suitable HIV-free candidate definition for "AIDS," we can then ask two critical questions about it: 1) Have we captured with our definition all of the people with the new medical problem that we historically came up with the "AIDS" label, in order to describe and encompass in the first place? 2) If we test our defined group, are very nearly 100% of the people encompassed by our AIDS definition found to be infected with HIV, an otherwise rare virus in the population? If the answer to both these questions is yes, then HIV is promoted to a good candidate for the cause of the AIDS epidemic. If either answer is no, then our version of the HIV/AIDS hypothesis obviously has severe problems right from the start. Fortunately, it turns out that we can easily construct a workable definition of AIDS which does not include any reference to HIV, but which still describes the new epidemic we are interested in. Such a definition will not be the standard one, of course, but since the standard modern "HIV-containing" AIDS definition is unusable for this purpose, both we and the HIV/AIDS critics are required to construct special AIDS definitions, even to continue to talk about the problem of causation. It's unavoidable. Finally, we should note that in no infectious disease known to man is it possible to possible to define the disease syndrome clinically (i.e., by things observable at the bedside) or by any laboratory means which don't include direct tests for the causal organism, and YET still be able to achieve 100% specificity ("100% specificity" means that in no case before testing for the microbe is the physician ever fooled about whether or not the patient REALLY has the disease). We're not pefectly 100% sure we are seeing ANY infectious disease without a test for the disease causing organism. Since this is true for any infectious disease, we should expect this to be true of AIDS also. Simply put: no definition of AIDS which does not depend on HIV (as the critics insist it must not) can ever be expected to be perfect, and thus NEVER capture anyone who actually does not have HIV. But, as with all infectious diseases, a 99.9% score will be quite good enough for us. _AIDS Is An Acquired Immune FAILURE Syndrome_. What is the best way to define AIDS without reference to HIV? First, the term "AIDS" stands for Acquired Immune Deficiency Syndrome, but it has always been understood (except by certain critics, who we will come to presently) that the amount of immune deficiency we are interested in, is not a trivial one. AIDS is the name historically chosen for a new medical syndrome which is essentially 100% fatal, and thus in defining "AIDS" we are looking for people with an immune deficiency in the range which is life-threatening, and which will continue to grow relentlessly worse until life is impossible. One possible way to define immune deficiency would be to define it by what secondary health problems it causes-- for instance, one could pick people who have gotten so-called "opportunistic infections," which are strange infections which are seldom if ever seen in people whose immune systems are fully functional. In the early days of AIDS, before HIV was discovered, the syndrome was indeed defined using such opportuni- stic diseases (Fig. 1a), and people with these infections are still included in the latest 1993 Centers for Disease Control and Prevention (CDC) clinical surveillance definition of AIDS (but now, in most cases, such people are only included if they are also HIV infected). We will not be able to use this CDC definit- ion of AIDS (Fig. 1c). Not only does it require HIV infection in many (though not quite all) cases, but for historical, political, and technical reasons, it also is constructed in a way which does not assess current immune status in the best way for our purposes. Why is this? The basic problem is that only a limited amount of information about a person's immune system function flows from the bare fact that they have an "opportunistic" infection. Certainly there is a good correlation between immune function and what kind of opportunistic infections occur, but the correlation is far from perfect, since opportunistic infection risk is influenced by not only immune status, but also by the quality of what we may term the infectious "assault" to the system. The infectious assault in turn is influenced by a person's physical location, infection contacts, personal habits, and other exposure factors both known and unknown. In the end, infection assault differences insure that _some_ unlucky, highly-infection-exposed people manage to contract opportunistic infections when only mildly immune compromised (though these are rarely fatal). By contrast, the same assault differences insure that _other_ people who are badly immunologically impaired may escape opportunistic infections for an amazingly long time, simply by missing the microbes which can kill them. When it comes to immune function, then (what we are interested in with AIDS), it is better to have a direct test which is not subject to uncontrolled variables, such as which microbes happen to be in the air or drinking water, and how many. We would like a more direct test. Such a test exists. Quite early in the history of AIDS, it was found that the immune defect in this disease is peculiar, and that it most visibly involves a particular kind of cell in blood and lymphatic tissues (lymph "nodes"), called "T-lymphocytes." In the syndrome of AIDS, certain T-lymphocytes gradually disapp- ear from both blood and lymph tissues, and a simple T-lymphocyte count in the blood can tell how serious the reduction has been in both places (since blood lymphocytes come from the lymphatics). The arm of the immune system which is controlled most directly by T-lymphocytes (the body's defense against viruses and fungi) is what is most defective in AIDS, and viral and fungal infections are (not surprisingly) the main opportunistic infections which appear and cause death in AIDS. AIDS is so specific in its attack that scientists eventually found that only one _subset_ of T-lymphocytes was initially hardest-hit. This was the so-called CD4+ or "helper" T-lymphocy- te, which has the job of stimulating the immune system. The other major type of blood T-lymphocyte, the CD8+ or "suppressor" lymphocyte, is involved in shutting the immune system down; in AIDS, CD8+ lymphocyte blood numbers increase early in the disease, and are not decreased until near the very end of the disease process, when they may also disappear. [See appendix: "What is a CD4+ Lymphocyte"]. CD4+ lymphocyte blood counts tell much of the story in AIDS and other immunodeficiencies involving the T-lymphocyte immune system. A healthy adult might have a CD4+ lymphocyte count of 800 to 1000, with a CD8+ count half of this. These are normal values. Under physical stress, injury, or chronic infection, CD4+ lymphocyte count might drop to 500 (to even less than the CD8+ count), and mild, non-fatal opportunistic infections might be the result. A CD4+ count less than the CD8+ count was once used as a crude marker for AIDS, but today with progress we know that this immune state is non-specific. In AIDS, things eventua- lly become much worse than this, and the worse things get, the fewer possible other causes besides AIDS there are for the problem. In full-blown AIDS, as defined by opportunistic infections and other problems, the CD4+ count is usually below 200. It is at such count levels that Kaposi's sarcoma (a tumor perhaps caused by a virus) and life-threatening infections begin to appear, although approximately 95% of AIDS patients survive beyond this level of decline [2]. Another feature of AIDS, however, is that inevitably the count grows worse over time. Today, in the modern era of antibiotics and more knowledgeable care, 85% of AIDS patients live to see their CD4+ lymphocyte count drop below 50, which is also near the average point at which most AIDS patients in the modern era now suffer their first opportunistic infection [3]. Famous AIDS sufferer Kimberly Bergalis, for instance, had her CD4+ count drop to 41 before her disease was discovered [4]. Many AIDS patients today go all the way to CD4+ counts of zero before the inevitable final infection or other complication. It is because of the implacable and more or less irreversible loss of vital T-cells that AIDS remains a fatal condition, with an average time of less than two years between the first opportunis- tic infection and death. If we wish to define AIDS in terms of immune failure, the essential question is this: where do we draw the line, so as to include almost all people with the new immunodeficiency epidemic, who are going to die from it, but exclude everyone else? If we simply define "immune deficiency" as a CD4+ lymphocyte count of less than 200 (where death begins to become more likely), we will capture about 95% of people who die of what the CDC now defines as "AIDS" (Fig. 1b). Previous to the epidemic of AIDS, of course, people did die of immune failure with low T-lymphocyte counts (including low CD4+ counts) for other reasons, and they continue to do so now. Thus, we must also exclude from our AIDS definition all those people who have one of the classic reasons for a very low T- lymphocyte count-- reasons which were well-known before the AIDS era (cancer, malnutrition, tuberculosis, radiation, chemotherapy, etc). These people don't have AIDS, because the historical epidemic of AIDS consisted of people with no T-lymphocytes for no other known reason. These people had appeared (more or less) newly on the scene in the 1980's with evidence of a fatal kind of immune failure which was _acquired_, meaning that it was an epidemic problem of something "picked up" by previously healthy people. So let us simply collect all the people we can find with sustained CD4+ counts below 200 without known reason, and test them for HIV. When we do, we find that almost all are HIV infected, and any who aren't, don't look at all like typical AIDS patients (as we will see). This, despite the fact that only 0.3% of the general population of the U.S. carries the HIV virus, giving a huge population free of this virus which should still be at risk for any other causes of severe immune problems. Thus, at this point we have no evidence yet to directly contradict the simple theory that HIV causes almost 100% of our conservatively defined "AIDS." In fact, things are looking quite suspicious for HIV. AIDS Heresies, Part 2 _Enter the Critics_ Not everyone will agree to use the above definition of AIDS, of course. Before we go further, we will introduce two major critics of the HIV/AIDS hypothesis. Both are scientists who hold Ph.Ds. Neither one is a physician. [See Appendix: "Two Critics: Duesberg and Root-Bernstein"] The view that HIV plays no role in AIDS has been most notably put forth by Peter H. Duesberg, a molecular biologist spe- cializing in a group of viruses which are related to HIV. Duesberg's view, promulgated since 1987, is that HIV is harmless and has no causal role in AIDS, and that AIDS instead is caused by drug use and by immunosuppressive blood products (page references to Duesberg's views will be from his major 1992 paper on the subject, and to a recently published book by Ellison and Duesberg [5]). At the other end of the skeptic spectrum are hybrid arguments raised by Robert Root-Bernstein, an associate professor of physiology and author of _Rethinking AIDS_ [6], the most carefully-documented work to yet assail the prevailing medical views on HIV and AIDS (Root-Bernstein references will be given in the form of a page reference from this book). Root-Bernstein is less radical than Duesberg, arguing for a somewhat less central role for HIV in AIDS than is generally given it, but still allowing for the virus to have some part in the etiology of the disease. Since Duesberg's original challenge, which has been the cause of much formal debate in the literature [7], a number of scientists, physicians, and lay persons have taken up the cause for a "re-appraisal" of the idea that HIV is THE major causal factor, or even _one_ of the major causal factors, in AIDS. Most respectable is an organization called "The Group for the Scienti- fic Reappraisal of the HIV/AIDS Hypothesis," which has collected over 300 signatures of physicians and scientists, including those of Nobelists Walter Gilbert and Kary Mullis. This group has campaigned to remove the requirement for HIV infection from any medical definition of AIDS, feeling that using this criterion is at best premature, and prejudices any hunt for alternative explanations for the disease. Almost all critics of the AIDS/HIV hypothesis have one thing in common: they insist on using a much broader definition of AIDS than we have proposed, a definition which virtually guarantees that some people who fit the critics' "AIDS" definition will _not_ be HIV infected. To be fair, there is some historical precedent for using a definition of AIDS which relies solely on the patient developing certain of the most serious and specific opportunistic infectio- ns, since this was the way the disease was diagnosed before HIV testing became available in 1985 (Fig. 1d). Today we know that almost all such people with pre-1985 defined "AIDS" are infected with HIV---indeed this was known in late 1983, before the official announcement of viral cause was made the following year. But today we know this figure would not quite be 100% [13]. As we will see below, there is evidence that the few HIV-negatives in this group will be people with lesser degrees of immune suppression (higher CD4+ counts), who will _not_ progress to worse immune function, or quickly die. It seems reasonable, then, with what we know today, to simply exclude them-- since we know that this is not the characteristic picture of AIDS. Again, it is most reasonable for our purposes to diagnose AIDS on the basis of immune function (CD4+ levels) only, since it is immune function, not infection status, which correlates with prognosis in CD4+ immunosuppressed people. The critics, however, will have none of this, and in their definitions are seemingly less interested in clinical utility than they are in collecting ammunition for an argument. The more broadly AIDS is defined, the more "HIV-free AIDS" cases critics can assemble, and these can in turn be used as evidence to the lay public that HIV cannot be the cause of "AIDS." Duesberg, for instance, has insisted upon retaining the early 1980's observation that a CD4+/CD8+ lymphocyte count ratio of less than 1.0 is often seen in AIDS, and he has decided that such a ratio, even in the absence of opportunistic infection, is synonymous with AIDS (Duesberg, p. 260). Duesberg now calls this ratio an "AIDS-defining immunodeficiency," and counts people with this lab result as part of "HIV-negative AIDS," in his shocking and too-often repeated statistic that there are "3000 documented HIV-free AIDS cases" [8]. Here again, Duesberg's chosen definit- ion of AIDS is less than useful epidemiologically or clinically, because people with such mild immunosuppression as he uses to define "AIDS" are not the people who are dying, or are shortly destined to die (as will be made clear below). AIDS is nothing if not a fatal epidemic, and insisting that mildly compromised persons who may or may not eventually get any worse be labeled as having "AIDS," as Duesberg routinely does, only serves to confuse the issue (Fig 1e). Perhaps following Duesberg, there is a general trend for HIV/AIDS skeptics to overdramatize levels of immune deficiency which are not clinically very significant. For example, Root- Bernstein (p. 262), in characterizing a study of HIV-negative (HIV infection free) men newly infected with CMV virus, notes that for a time, some of the men had CD4+/CD8+ cell ratios of less than 0.4, a figure which he claims "represents extreme immune suppression." It would take a medically sophisticated reader to know that in AIDS this ratio would typically be far less than 0.3, and thus these men would not be mistaken for the current CDC immunological definition of AIDS, even if they were HIV-positive. The reader would also need to know that the level of immunosuppression associated with a ratio of only 0.4 is not associated with significant risk of death by opportunistic infection. Such a reader might wonder how we are justified in calling a ratio of 0.4 "extreme immune suppression," if people rarely die from it, as they are known to do in AIDS. Root- Bernstein does not say-- indeed, does not even raise the issue. An tendency toward overdrawn interpretation of the clinical significance of lab results, is one of the places in which the absence of medical training in the chief HIV/AIDS skeptics shows most clearly [9]. Duesberg's paper [5] and Root-Bernstein's book [6] each contain descriptions of groups of HIV-free people who are somewhat immunosuppressed due to low CD4+ counts, but not severely so, as defined by our straightforward criteria of having a significant risk of infectious death due to T-lymphocyte loss. These immune deficient patients in HIV/AIDS skeptic literature are presented along with the inference that perhaps somewhere there exist people with these immune suppressive factors, or combinations of them, who are _severely_ permanently T-lymphocyte immunosuppressed (as AIDS patients generally are), and yet still without having HIV. Duesberg and Root-Bernstein only have one difficulty in this argument-- neither has been able to actually _find_ more than a handful of such people in a country where AIDS sufferers infected with HIV have passed through this clinical phase by the hundreds of thousands [9]. _HIV-Free AIDS?_ Hypotheses may be disproved by the right data with relative ease, and cases of HIV-free AIDS would disprove the idea that HIV causes AIDS, in proportion to how often these are found (i.e., if 10% of AIDS cases were HIV-free, this would prove that HIV is not the cause of AT LEAST 10% of AIDS). Thus, Duesberg and Root-Bernstein aren't the only ones who have been looking for HIV-free people who are badly CD4+ lymphocyte immunosuppressed without any classic reason (i.e., good candidates for HIV-free AIDS). Very recently the CDC reported that after a massive search it had only been able to find less than 100 people without HIV infection across the country whose CD4+ counts were, at one measurement, less than 300 (not quite in the AIDS-class immunosuppression range of 200, but drawing close). This group of people, for ease of reference, was given a special syndrome name: "ICL" (idiopathic CD4+ lymphocytopenia), meaning "people with low CD4+ lymphocyte counts without a medically-defined disease, or other known immunosuppressive reason." Why wasn't ICL simply called "HIV-free AIDS"? Sometimes it is. The 1987 and 1993 CDC AIDS definitions do allow for AIDS to be diagnosed in the absense of HIV, and even with a negative HIV test. This requires certain infections and a CD4 count less than 400, and these things happen extremely rarely in the absense of HIV (fewer than 1 in 1000 cases of AIDS are tested for HIV and found repeatedly negative). Critics have darkly suggested that the reason "ICL" is not called AIDS, is simply a matter of politics, but in fact there were problems with considering these people as AIDS cases, which had nothing to do with AIDS politics or the HIV theory. One difficulty was that two thirds of people labeled as having "ICL" were found not to come from the AIDS risk groups. They did not use illicit drugs, had not been exposed to blood products, and had no evidence of sexual behavior which would have exposed them to a special infection risk. Thus, as we will see, the most popular alternative AIDS hypotheses did not explain the majority of these people _either_ --- a fact which did not keep them from being mentioned in nearly every skeptical treatment of the HIV/AIDS issue. What the skeptics had forgotten (or hoped their readers would not notice) was that the immune deficiency of people with ICL did not seem to be _acquired_ in any obvious way [10]. What justification was there, then, for calling it "AIDS"? To call a disease "acquired" it is not enough simply to show that it appears in adults without previous problems, for even some genetic diseases do this (Huntington's disease). A bona-fide acquired disease must have evidence of an etiology which is not genetic. The classic epidemic we know as "AIDS" does. Most ICL does not. The two syndromes are evidentally not the same thing, even looking at them from basic epidemiology. Moreover, people with ICL were not only epidemiologically, but often immunologically very much distinguishable from AIDS cases: their CD4+ lymphocyte counts swung widely in response to infections and were often much higher than 300 (in contrast to people with AIDS, whose CD4+ lymphocyte counts tend to stay low, and heading on an ever-downward trend). ICL people also often had low total lymphocytes or low CD8+ lymphocyte counts, again indicating that their type of immune failure did not make much distinction between CD4+ and CD8+ lymphocytes, as AIDS always does. Clearly, these people did not belong to the classic AIDS groups which began suffering with epidemic immune problems about 1980. They are not part of the new phenomenon of AIDS, something underscored also by the fact that, although suffering from opportunistic infections, ICL victims did not even seem to share the implacable death rate of AIDS [10]. Searches for HIV-negative people who have AIDS-type severe immune suppression have also been taken specifically within AIDS risk groups. Vermund reported in the United States Multicenter Cohort Study that of the 2713 persistently HIV-negative homosex- ual men in the study, who had had a total of 22,643 blood tests, only one significantly immunosuppressed man (CD4+ lymphocyte counts persistently less than 300) was found. This man was taking chemotherapy and radiation for cancer, and thus had a very good immunosuppressive reason other than his lifestyle to explain his lab results [11]. A similar review of another cohort of homosexual and bisexual men found no persistently lowered CD4+ T cell counts among 756 HIV-seronegative men who had no other cause of immunosuppression [11]. Finally, in the San Francisco Men's Health Study (a population-based cohort recruited in 1984) it was found that among 206 HIV-seronegative heterosexual and 526 HIV-seronegative homosexual or bisexual men, only one had consistently low CD4+ lymphocyte counts, and this man also had low CD8+ T cell counts, suggesting that he had general lymphopenia rather than the selective loss of CD4+ lymphcytes characteristic of AIDS [11]. If these three studies are indicative-- and there is no reason to think they are not-- then most, if not all, male homosexuals with AIDS-range immune failure are HIV-positive, since it has proved very difficult to find any who are HIV-negative. Much the same seems to be true in IV drug users: in a study of 1246 HIV-negative injecting drug users in New York City from 1984 to 1992, for example, only 4 were found with CD4+ lymphocyte counts less than 300 (if IV drug use per se was a major cause of AIDS, the number should have been far higher). In this small group of 4 people, even though infected with multiple other non- HIV viruses, and with a history of heavy drug use, immune function was stable and without the steady decline in CD4+ lymphocyte counts over a time span of years which is characteris- tic of all unselected HIV-positive cohorts [12]. A second study by others found much the same result [12]. Thus, in these studies also, the few HIV-negative people who could be found with even near-AIDS range immunodepression, were _still_ not behaving medically like people with AIDS. Studies of recipients of blood and blood products, and household contacts and sexual partners of transfusion recipients, also suggest that persistently lowered CD4+ cell counts are very rare or nonexistant in the absence of HIV infection [12a]. So far as we know, then, in the United States more than 99.9% of the people who are a part of this new phenomenon of permanently very low (and declining) CD4+s in high risk groups, have been infected with HIV. This does not prove that HIV causes AIDS, but it is surely an important clue. Why Not Merely Use the CDC Definition For AIDS, With HIV Taken Out? A persistent suggestion by critics is that it would be proper to use as an AIDS definition the current CDC definition (which includes all HIV-infected people with a much expanded list of infections and other problems), but with the HIV criteria removed. The problem with this suggestion is that definitions of diseases are chosen by the CDC for maximum clinical utility, and the HIV infection criteria in the CDC AIDS definition were not put there only to insure that there would be less HIV-free AIDS. Rather, HIV infection in a person with opportunistic infection is known to be (alone among all other viral infections) a very good predictor of whether immune status will continue to decay until the person eventually succumbs to opportunistic infections. In people with mildly compromised immune systems, the prognostic value of an HIV infection is large. Even critics admit that the prognostic value is large, without admitting causation. Thus, we cannot simply remove HIV status from the CDC definition and still have the definition do what it was designed to do, which is _predict_ impending death by immune failure. Critics know that if "AIDS" is defined only in terms of today's much broader list of "AIDS-defining" diseases and infections (which are meant to be used only in conjunction with HIV status), it is sure to be quite true that the definition will then become far too broad to be prognostic. Such opportunistic infections, as critics well know, sometimes happen occasionally even without the most severe CD4+ immunosuppression which is characteristic of people who die with AIDS. A study by Salvato illustrates this point [13]. In the study, medical records covering 6 years for 1500 HIV-positive patients were compared with records for 1000 HIV-negative patients who had Chronic Fatigue Immunodeficiency Syndrome (CFIDS) and evidence of immune suppression. It was found that the CFIDS patients had a syndrome much like that seen with early HIV infection -- fatigue, lymphadenopathy (swollen lymph "nodes") and low grade fevers-- but that over the course of 6 years their problems were not severe. Only one of them developed CD4+ lymphocyte counts less than 300 ("ICL"). Still, two had yeast esophagus infections, a severe opportunistic infection rarely seen other than in AIDS, and other people severely immunosuppressed. Three had active CMV virus disease of various tissues-- another disease very often seen in AIDS. A total of 486 patients had evidence of yeast infection of the mouth on exam, a condition suggestive of mild immune problems but one not limited to AIDS. The average CD4+ lymphocyte count in these patients (not including the single ICL patient) ranged from 500-1400, with an average of 650. This was significantly lower than normal, but much higher than typical for AIDS. In this study, 95% of the HIV-negative patients had previou- sly been infected with the EBV, CMV or HHV-6 viruses, and 48% had evidence for continued viral infection (critics such as Root- Bernstein have suggested that these viruses have roles in AIDS at least as important as that of HIV, but this study provides evidence against this idea). Most interestingly, these immuno- compromised HIV-negative patients were followed from 2 to 6 years, and none experienced progressive CD4+ lymphocyte decline (except for the one patient with ICL, who, with treatment of CMV infection, showed increased CD4+ lymphocyte counts again). Such CD4+ count stability is never seen in any random group of HIV- positive people, where an average CD4+ count decline over time would be inevitable. The study authors' conclusion: "Even after a methodical search in a practice that sees a large number of patients with immune problems, only 1 patient was found to have ICL. However, this study demonstrates that patients with normal CD4 counts can develop AIDS defining opportunistic infections [...] Upon long-term follow-up these patients do not appear to experience progressive CD4 depletion." Most importantly, no HIV-negative person died in the study, which illustrates the extent to which chronically virally infected, immune-suppressed people can _approach_ the clinical picture of AIDS, without crossing into the deadly long term immune failure which is characteristic only of people with HIV infection. The reader who is a bit confused at this point should keep in mind simply that the most important thing about the syndrome of AIDS (by any good definition) is that it inevitably and rapidly destroys the immune system, and then kills people by means of the infectious and other diseases they get as a result of having no working immune function. Thus, mild CD4+ cell depression and opportunistic infections are not always AIDS, for only some of these people (as it turns out, the HIV+ ones) will progress to immune failure, and death. It is immune failure (almost complete CD4+ lymphocyte loss) and death by opportunistic infection which is characteristic of AIDS; and it is only such people who are almost _always_ HIV infected. How and Where AIDS First Appeared The story of the detective hunt for the cause of AIDS is told with wit and clarity by Randy Shilts (an author who in 1994, at the age of 42, ultimately became a casualty of the disease himself), in the best-selling book _And the Band Played On_. Other good histories of the early AIDS epidemic are also availa- ble [14]. Historically, what happened in the U.S. in 1981 was that homosexual men began presenting to physicians in increasing numbers with very, very low CD4+ lymphocyte blood counts (but not lowered counts for other subtypes of lymphocytes), a destroyed immune system with lymphatic tissue destruction, opportunistic infections, and Kaposi's sarcoma. Almost no one who had treated diseases in the male homosexual community could remember having seen anything remotely like what had began happening on an increasingly large scale in the early 1980's. The year 1981 was not (in retrospect) exactly when the problem started, but rather when the problem first grew large enough in the U.S. to be brought to the attention of the federally-run Centers for Disease Control and Prevention (CDC) in Atlanta. It was in the Summer of 1981 that alert physicians brought to the attention of the CDC a mini-epidemic of immunod- eficiency and pneumonia caused by unusual organisms (A fungal organism called _Pneumocystis carinii_, and a virus called CMV) in homosexual men in Los Angeles. Before the epidemic of AIDS, there were only 50 to 60 cases of pneumocystis pneumonia reported to the CDC per year, in the United States. After ten years of the AIDS plague, this number had risen to tens of thousands of cases per year, and almost all of these in adults who had never previously been targets for this organism. Because many of the first people to contract AIDS had had sexual contact with each other, CDC researchers thought they might be looking at an unknown sexually-transmitted infectious disease. They also toyed for a time with the idea that sex- stimulant-chemical use or illicit narcotic use, both very common among the first cases of AIDS, might be somehow causing immunosuppression. Perhaps sexual contact was a red herring-- or merely a marker for a small and fairly tight-knit sub-community of people who shared common interests in non-sexual activities which might be damaging their immune systems. Those physicians treating infectious diseases in homosexual men thought not, however. Dr. Joel Weisman, one of the first doctors to put the AIDS puzzle together, noted that initially, within the male homosexual community, the disease seemed to follow lines of sexual contact more than it did drug or sex habits. Not all homosexual men were so promiscuous as to make contact-tracing impossible; Weisman observed that promiscuous men did not always contract the disease, but on the other hand, that even men with few sexual contacts were coming down with the disease if they had had sexual contact with the wrong person. In fact, men with severe immunodeficiency were eventually found to form sexual contact networks, of the kind that have always been seen by researchers using the classic epidemiologic tools for tracing sexually transmitted disease chains. The difference, however, was that for AIDS the contact networks stretched over years, indicating an infectious agent (if there was one) with a very long latency. Still, investigators found that of the first 19 cases of AIDS reported in Los Angeles, 9 had direct or indirect (one intermediate partner) sexual contact with a single French-Canadian airline steward, a man who was also sick with immunodeficiency. Then, starting in 1982, reports began to come into the CDC of the same CD4+ lymphocyte and lymphatic-tissue-destroying immune failure syndrome occurring this time in U.S. citizens who had received transfusions. Soon also came reports that an identical immune deficiency of a new severe variety was now being seen in men with hemophilia, a genetic disease in which sufferers must be injected with concentrates of protein clotting factors made from donated blood plasma. Reports of the first people with hemophi- lia and AIDS emphasized that, in these people, none of the same drug or male-homosexual behavioral factors were present that had been seen in the first group of AIDS sufferers [15]. Further, the same was true of those with "transfusion-related AIDS," who also did not fit into drug-using or male-homosexual lifestyles, and did not resemble them in sex or age either. Former tennis star Arthur Ashe is a well-known modern example. Ashe, like many of those with transfusion-related AIDS, had never had an intimate connection with anyone else with an immune problem, EXCEPT for a history of a blood transfusions years in the past, during the time in which transfusions were associated with AIDS. In late 1982 all this worried epidemiologists as the reports continued to come in. They knew that another viral disease called hepatitis B ("serum hepatitis") was also transmitted epidemically as a sexually transmitted disease in homosexual men, but much more rarely in homosexual women or heterosexuals in the U.S. Hepatitis B had historically also shown up early in people with hemophilia, who because of their large pooled blood-product exposure have historically seemed to be first to suffer from any new organism infecting the blood supply. Hepatitis B had also been known to be one of the worst disease-causing contaminants in donated blood for general transfusion. Thus, the same 3 groups of people who had historically been infected with hepatitis B in the 1970's, had now started coming down with AIDS. Hepatitis B was also a disease of IV drug users who shared needles, and it was not long before the first reports of IV drug users with AIDS came in. By 1983, the CDC was sure it had a new infectious disease on its hands, similar in epidemiology to hepatitis B, but with a longer latency period. Analysis of the habits of donors of the blood components that were transfused into people who had later developed AIDS, indicated one thing different about the donors: it was found that blood products AIDS patients had received had more often come from people who themselves were at "high-risk" for AIDS due to promiscuous male homosexual behavior. On the other hand, matched case-controls who had been transfused identically from the same blood bank but had _not_ developed AIDS after transfusion, were found to be not nearly as likely to have gotten blood components from anyone in a "high-risk group." This initial study concluded that there was only a 1% chance that the statistical association of transfusion-associated AIDS with the lifestyle of the blood-donor would be as close as it was found to be, if only chance had determined the lifestyles of the donors of blood to people who later became sick. Such a chance association would have been expected if there was no contamin- ation, and instead there was something about normal transfusion blood itself, or perhaps some other factor unrelated to trans- fusion, that was causing AIDS in transfusion recipients [16]. The remarkable fact-- from which there was no escape-- was that AIDS in a transfusion recipient PREDICTED the lifestyle of a blood-donor he or she had never met (a donor which generally turned out to be a promiscuous homosexual man who had thought himself to be perfectly healthy). Nothing but an infectious agent (or more than one) could explain a statistical connection between a blood donor's sexual habits, and risk to the person receiving the blood. As for drugs or immune toxins, it was impossible to believe that any chemical toxin could be present in a relatively small amount of blood component coming from a single nominally healthy person, in sufficient quantities to cause total immune failure in the recipient, and do it years after the transfusion. Eventually, with many cases like Arthur Ashe's on record (but showing up in the early 1980s, earlier than Ashe's did), AIDS looked epidemiologically _very_ much like hepatitis B. The hunt was on for the microbe, or microbes, which caused the new syndrome. When the virus now known as HIV finally hit the world news in the Spring of 1984, there was a great deal of skepticism in the scientific and lay communities alike. With the ability to test for antibodies to HIV in 1985, however, there came a way of powerfully sifting through putative causal factors for AIDS, and comparing them with the factor of past HIV infection. HIV infection has emerged from these tests as the clear champion of competing AIDS-causation theories, convincing at present all but the most die-hard skeptics [14]. [See Appendix: "What is an Antibody? and What Does HIV-positive Mean?"] AIDS Heresies, Part 3 The Skeptics and the History of HIV and AIDS But what if AIDS and immune failure are not really new-- perhaps (as some critics suggest) we just look harder for them now that we recognize them? Could our new theories be warping our views so completely that by now that we have made a new "plague" out of something that was here all the time? Epidemiologically, what can we fairly say about the period before 1980, keeping this possible bias in mind? With the new ability to test old preserved tissue specimens for HIV, the first thing that becomes apparent is that AIDS is indeed older than 1981-- perhaps far older. Deaths from what has since been recognized as HIV infection with immune failure have been seen clinically, without being understood, for at least 35 years, and probably much longer. An HIV-infected British sailor, who had traveled widely, was once thought to have died with severe immune deficiency and HIV infection in 1959, the earliest proven case of modern AIDS. The diagnosis was made by means of preserved autopsy tissue specimen HIV testing, 30 years after the fact, although this case is still not without controversy. [Note added in Jan. 1996-- this case has finally been disproven. Apparently reports of this man's tissues containing HIV were in error, see ref 17]. This man's death alone would have provided good evidence that HIV is not a product of deliberate genetic engineering, for in 1959 biologic science was simply too unsophisticated to work with lymphotropic (lymphocyte-infecting) retroviruses like HIV, let alone engineer them [103]. HIV appears still to be an accidental infection of humans with an African primate virus, if it is anything at all. The genetic material of the most common HIV-1 strain is most similar to that of a virus known to natura- lly infect chimpanzees, and it may be that HIV's ancestors have been present in Africa, perhaps even in humans, for a very long time-- perhaps thousands of years [18, 121]. In West Africa, a close cousin of the U.S. HIV-1 strain, called HIV-2, is almost identical to several indigenous African monkey viruses, and almost certainly has been derived from them quite recently in virus evolutionary time (less than several centuries). In the U.S., AIDS-like deaths have been reported as far back as 1934, as Root-Bernstein notes. The first AIDS or AIDS-like death that we believe could have been associated with HIV infect- ion was that of a 17 year-old possibly homosexual male, who died of strange opportunistic infections in St. Louis in 1968 [19]. This early AIDS-sufferer had never been out of the country, showing that the virus was already active in the Western Hemisph- ere in 1968. Unfortunately this case is the clearest on record, and even it has not resulted in recovery of HIV DNA sequences (which would provide absolute proof). Some reports of early HIV are frankly errors: for instance a 4% fraction of preserved serum samples from IV drug users in this era (1971-2) in the U.S., found to be "HIV-positive," were apparently false-positives as shown by later followup [20]. It is possible that HIV viral infection has been present in small contingents of both drug users and homosexual men for some time in the United States, but the case has yet to be absolutely proven. Let us suppose that preserved specimens finally show that HIV was present in America long before the late 1970's. Why, then, was the U.S. first hit with an AIDS epidemic only in the 1980's, with HIV infection quickly rising to 50% in some risk-groups? The answer may be that if it was not the simple presence of HIV virus in the United States that changed, perhaps it was the social milieu. In the late 1960's drug use became far more widespread in the U.S., and the invention of the disposable plastic injection syringe about 1970 made IV drug abuse possible for the first time on a large scale. Also beginning around 1969 (the date of the New York City "Stonewall riots"), homosexuals in the U.S. began to take open political power, and concomitantly one faction of male homosexuals began to engage in the high-infection risk "bathhouse lifestyle" chronicled by Shilts. In addition, the American homosexual-male community was apparently many times re- infected by many world-traveling disease "vectors" from other countries in the 1970's, including the previously mentioned airline steward (described in Shilts as the CDC "patient zero") who traveled widely in Europe, Canada, and the U.S., died of AIDS, and is known to have had sex with no less than 40 of the first 248 Americans to be diagnosed with AIDS by April, 1982 [14]. What happened in the late 1970's in the U.S. is that when a large enough fraction of the American homosexual-male population became infected with HIV, the U.S. blood supply, maintained with volunteer donations only, finally became contaminated with the virus. (This started in 1978, as we know from later testing of archived serum samples taken from homosexual men originally for hepatitis B studies). Similar archived samples tell us that in 1978 the U.S. plasma supply used to make clotting factor for hemophilia treatment became HIV contaminated, no doubt primarily by IV drug users selling plasma to support a drug habit. The dates are not coincidental-- a crossover between initial HIV infected groups occurred as some homosexual men experimented with IV drugs in the late 1970's, and male IV drug users in large cities turned to homosexual prostitution in order to obtain drugs. The resulting new epidemic of transfusion and hemophilia- associated AIDS, beginning in 1982 and rising sharply in 1984, helped to bring the acquired nature of AIDS into focus. The small incidence of AIDS in the American homosexual-male and IV drug-user communities before the late 1970's in no way subtracts from the reality of the dramatic increase in AIDS which took place in the early 1980's on the heels of exploding HIV infection rates in these groups. Although relatively mild immune suppression has apparently always been widespread in many AIDS risk groups, the more complete and deadly immune failure characteristic of AIDS itself has been sporadic and very rare in young cancer-free people in any of these groups, until the 1980s. It is, to be sure, difficult to retrospectively evaluate the health of male homosexuals before the first prospective studies of gay men's health were done in the 1980's AIDS era, but we can be reasonably sure that an epidemic of deadly immune failure among young American men before 1980 would have been duly noted by epidemiologists. The HIV/AIDS skeptic Root-Bernstein docume- nts a few cases of unexplained opportunistic infection deaths from the medical literature before 1980, but clearly an epidemic of immunosuppressive deaths cannot be seen in the historical record before 1980 by any act of imagination. By contrast, at present "AIDS" (a new epidemic of immunodeficiency deaths) shows a high and rapidly rising incide- nce among young men and women in the U.S., and these deaths cannot be simply a new label for an old problem. The reason is that _total_ mortality and cumulated years of life lost to premature death in young persons are observed to be rising rapidly, with all of the change due to "AIDS" deaths, _at_ _the_ _same_ _time_ other leading categories of mortality remain stable. If mere re-labeling of deaths into different categories was a problem, these "newly recognized" AIDS deaths would come out of other previously defined mortality categories, and this clearly isn't happening [21]. AIDS, the disease, may be old, but AIDS, the epidemic, is indeed something new. [See Appendix: "Hemophilia and Life Expectancy in the 80's"] People with hemophilia, unlike homosexual men, represent a well-defined group with long-term documentable changes in morbidity and mortality, since they had been well-studied as a group before the era of AIDS. This research shows that people with hemophilia began to die of dramatically different things, starting about 1982 [22] See Fig 2. A recent check shows little evidence of a special incidence of opportunistic diseases in people with hemophilia in the U.S from the turn of the century up to 1979, although a low incidence of AIDS could not be ruled out in this study, mostly because some cases of fatal pneumonia had no identified infecting organism [23], and because people with hemophilia as a group are immunosuppressed enough to be somewhat more susceptible than normal to bacterial infections. Signifi- cantly, however, in the years before AIDS, people with hemophilia had never been noted to be particularly susceptible to the more obvious _fungal_ infections, such as candida esophagitis, common to AIDS patients and others with low-lymphocyte type immune deficiency. After 1984, however, this type of AIDS-associated opportunistic infection and immune failure rapidly became the single most common cause of death in people with hemophilia in the U.S. [24]. The rise in total mortality risk in people with hemophilia was sudden: total mortality in this population, which had been stable in 1982 and 1983, suddenly increased by a factor of approximately 900% in the first quarter of 1984 [25]. Such an increase in raw numbers of deaths was consistent with an epidem- ic, or some new very toxic contamination of the clotting factor supply. It is not consistent with slower social changes, slower toxin or immune suppression models, multifactorial causation models, or the idea that people with hemophilia were actually at no greater risk than before (i.e., that again perhaps there had been some kind of "cause of death" re-labeling in response to AIDS hysteria). Mortality figures in hemophilia patients also showed something else important, which was that the new deaths of the late 1980s, by virtue of all being judged "AIDS," demonstra- ted that most or all of them occurred in people with hemophilia who were HIV-positive. Since these deaths accounted more or less for the entire new increase in mortality, it could be inferred that the mortality rate for HIV-negative people with hemophilia did not increase much in the 1980's, if at all. How significant was the increase in death rate for HIV- positives in this group? In one _Journal of the American Medical Association_ study it was found that in a cohort of 111 people with hemophilia infected with HIV in the early 1980's, one third had died by 1992 [26]. The reader is asked to imagine any group of this age (a high school class, perhaps) and imagine an overall 33% mortality rate in less than 10 years. Of the estimated 10,000 people with hemophilia to have been infected with HIV in the early 1980s in the United states, a quarter had been reported to the CDC to have died of AIDS by July of 1993. Such death rates were especially shocking in view of strides in hemophilia treatment which had been made in the years before. Total life expectancy in people with hemophilia had risen as clotting factor treatment became available through the 1970's, until by 1980 it was nearly normal [23]. After 1984, however, life expectancy in this group began a steep decline, and by the early 1990's was at a lower level than at any time since before World War II [24]. In the 1980's, total mortality for hemophilia increased in all age groups above 9 years of age, and age at death shifted markedly to lower ages, decreasing from 57 years of age in 1979-1981 to 40 years of age in 1987-1989 [27]. About 50% of people with hemophilia in the U.S. had been HIV infected by early 1986, when screening and treatment of the clotting factor concentrate stopped HIV spread [28]. Still, the long latency of the virus (as long as 15 years for 50% progress- ion to AIDS in this group) caused death rates to rise for long after the window of new HIV infection closed (they are still rising as of this writing, although clinical AIDS or severe lymphocyte loss (ICL) has yet to be reported in children with hemophilia born after 1986). The fact that there was a massive and silent HIV infection of half of the people with hemophilia in the early 1980's is beyond question. The HIV/AIDS skeptics' quest to divorce this event from the epidemic of deaths by AIDS in this same group over the next decade has resulted in some remarkable and curious stateme- nts about hemophilia mortality. Duesberg, for instance (p. 216) quotes only older statistics for hemophilia patients from the pre-1986 period, before AIDS deaths became very large. His practice of using randomly reported AIDS and mortality data for people with this disease (which is often notoriously unreliable in the best of circumstances [29]) instead of the much more reliable cohort study data, also results in figures which minimize the impact of AIDS. Cohort data shows mortality in hemophilia patients to be far higher than Duesberg acknowledges [30]. Duesberg has not been alone in ignoring major trends in hemophilia mortality in the last decade. The very misleading statement that people with hemophilia are living "longer than ever" today is one of the standards among the HIV/AIDS skeptic community. Root-Bernstein does no better than Duesberg at providing updated information in this area, offering one paper's 1979 pre-AIDS statistics [23], without update and without qualif- ication, to represent _contemporary_ life expectancy in people with hemophilia in 1993 (p. 247). This represents very sloppy scholarship (something which stands out particularly in Root- Bernstein), but the oversight does allow the author to skip discussion of the pronounced and otherwise awkward peak in life expectancy for hemophilia in the middle 1980's. Duesberg, though he seems to believe that people with hemophilia have suffered no mortality increases in the age of AIDS, does suggest that people with hemophilia live longer than ever due to recent factor concentrate development, and thus live long enough to die of immunosuppression caused by longer treatme- nts with clotting factor concentrate, instead of from hemophilia (Duesberg, p. 220). Although clotting factor does indeed appear to be mildly immunosuppressive (albeit in a different way than AIDS-- see appendix hemophilia section), the main problem with the hypothesis that clotting factor itself causes AIDS is that two studies of HIV-positive people with hemophilia have found that HIV infection and not clotting factor use is the critical risk for AIDS. These studies found that once a person is HIV- positive, risk of AIDS is _not_ related to amount of clotting factor used or severity or type of hemophilia---effects that would have been expected if clotting factor carried a significant immune risk independent of its HIV content [31]. Available statistics thus strongly suggest that the known association of clotting factor use and AIDS risk is merely due to the increased risk of being infected with HIV the more clotting factor has been consumed; once HIV infection has occurred, it doesn't matter how much clotting factor is used. Can a Viral Epidemic Explain the Historical Timing and Epidemiology of AIDS?: Attacks on Straw Men. It is an unfortunate fact that a great deal of the debate over AIDS and HIV has been over what rhetoricians call "straw- men." A straw man is an argument or viewpoint set up in a debate only for the purpose of being knocked down, and one which the opposite side never really defended or held; or one which is not very important to the central issue of the debate, even if it _has_ been held. Straw man arguments often result from debaters talking "past each other," without understanding the opposing side's position (straw man arguments may also, of course, be used deliberately solely for rhetorical purposes-- a practice genera- lly considered beneath respect in scientific debates). In the HIV/AIDS debate, straw men set up by heretics have most often been medical hypotheses which have previously been put forth in the context of the HIV theory and which have turned out to be wrong, but which were never important corollaries necessarily deduced from the idea that HIV causes AIDS, or were in other any other way central to it. Other straw men are ideas that the orthodox scientific "establishment" never put forth seriously at all, though they may be attacked vigorously by heretics as though they are current medical dogma. We will presently see samples of both. An example of an epidemiologic straw man is the timing of HIV arrival in the Western hemisphere. Root-Bernstein discusses cases of possible AIDS as far back as 1932, notes documented HIV infection with AIDS as far back as 1968 in the US, and argues that these data are anomalous (p. 2) if the virus was transferred for the first time to the Western hemisphere around 1978, as was originally thought (and which is the "just-so story" told by Shilts). And so they are. But if the HIV virus was transferred much earlier than 1978 to the new world, and remained at low levels in male homosexuals and injecting drug users in America until changing social factors in the 1970's encouraged its spread (exactly as Root-Bernstein himself indirectly suggests), no real damage would be done to a suitably modified HIV/AIDS theory. An example of a bad prediction made by the orthodox medical establishment which is not necessarily derivative of the HIV theory, was (or is) the official idea that AIDS is due to be a heterosexual pandemic in America _any_ _time_ _now_. It is argued by Duesberg (p. 203), that the "viral hypothesis" has failed to predict the course of the AIDS epidemic-- namely that AIDS has (at least so far) shown no clear inclination to spread rapidly by a complete heterosexual-sexual-transmission mechanism in the U.S., even though it apparently does so in Africa. It is also asserted in a related argument by Root-Bernstein that the HIV/AIDS hypothesis does not explain the generally-low measured levels of HIV virus in semen, the low (but not zero) rate of HIV infection in mates of HIV-positive men with hemophilia, or the nearly zero rate of infection in U.S. heterosexual prostitutes (unless they are drug users). If AIDS is an infectious disease, ask the critics, then why doesn't HIV infect very well? All these arguments are against straw men, so far as the cause of AIDS goes. There is nothing in the HIV/AIDS theory which demands that any particular transmission mechanism be the chief cause of the spread of HIV infection in any given place, or which demands that the HIV virus be as infectious in one locality as another. For example, it now seems likely from many studies that efficient sexual transmission of HIV requires mucosal tissue trauma, which is much more likely with anal intercourse, and/or a concomitant inflammation or ulcer from a second sexually transmi- tted disease. Because transmission may be inefficient even so, promiscuity also greatly enhances the chance of HIV spread. These requirement(s) for efficient HIV sexual transfer easily explain the difference between spread of HIV in tropical Africa vs. the developed counties. They also adequately explain why a disease which spreads well sexually only in populations with an extreme level of both promiscuity and rectal mucosal trauma (i.e., one sub-segment of American homosexual men) has not yet become a generally spreading epidemic in the U.S. It isn't that the HIV/AIDS heretics haven't come across such explanations. Root-Bernstein, in a good discussion of the epidemiology of AIDS, admits that there is nothing especially strange about a sexually transmitted disease which spreads easily in homosexual males but not heterosexuals. Both syphilis and hepatitis B in the 1970's have been examples of such a phenomen- on, and the "odd" differential epidemiology of both diseases with regard to sexual-preference groups is easily explained by differential _behavior_ in the homosexual and heterosexual populations in those years. Thus, Duesberg argues that a disease which restricts itself to classes of people in America, but not in Africa, cannot be explained by a micro-organism. But while he is doing so, fellow heretic Root-Bernstein (pp. 281-303) is noting that infectious epidemiology in one group of American homosexual males (who may have ulcerative sexually transmitted diseases and in addition be sexually infected with giardia, parasites, amoebas, hepatitis A, and B, shigella, salmonella, etc.), may resemble far more the disease epidemiology of some African countries than that of heterosexuals living next door (p. 290). In this, an AIDS caused by an infectious agent such as HIV may behave just as AIDS statistics suggest it does, and yet merely follow a pattern already amply demonstrated before AIDS, with many another infectious disease. Root-Bernstein is sometimes too competent a scholar for his own good: his Chapters Eight and Nine-- which address the epidemiologic differences and commonalities of U.S. homosexual men and African heterosexuals due to sexual practices and social changes which appeared newly in the 1970's and 1980s-- not only believably explains and refutes most of Duesberg's epidemiologic problems with AIDS (Duesberg, p. 209), but also does the same with many of Root-Bernstein's own epidemiological problems, raised in Chapter One. Unfortunately, Root-Bernstein is willing to let lifestyle and habit differences explain epidemiologic differences when it suits his argument's needs, but much less willing to consider them when they don't. An illustrative example occurs as Root-Bernstein discusses the rectal traumas and infections which occur during certain male homosexual practices, writing of these (p. 283-4): "It is now accepted that such injuries and infections greatly increase the risk of con-current infections (HIV or otherwise) and of semen gaining access to the immune system following anal intercourse." Yet when Root-Bernstein discusses the statistical association of AIDS with receptive anal intercourse (p. 225) he shows an odd difficulty with the same concept: "One possibility is that it is much easier to transmit HIV to a receptive partner than from a receptive partner. No other sexually transmitted disease behaves this way, however. [...] HIV would be the first disease agent to be able to make the discrimination, unless some other factor is involved." Here, unfortunately, Root-Bernstein is wrong, and wrong for the very reasons that he himself discusses in the quote preceding the last. Much like HIV, hepatitis B infection in homosexual men _also_ correlates with rectal trauma and receptive anal intercourse [32], and there is little reason to believe that the "other factor" is anything other than the fairly straight- forward mechanical injury that Root-Bernstein has already helpfully identified for us (see [33] for statistical development of a "rectal trauma index" which partly predicts risk of HIV infection). It is a characteristic of Root-Bernstein's style of argument that it makes causal mechanisms as mysteriously complic- ated as possible-- very often far more complicated than required to explain the facts. Here is another example of Root-Bernstein's difficulties with simple explanations: The known fact that HIV is difficult for an asymptomatic HIV carrier to transmit to another person by needle stick, or by heterosexual contact between married couples, does _not_ necessarily argue for the need for additional unknown co- factors or immune suppression in the more common cases of HIV transmission. Low infection rates for needle sticks may just as easily be explained by the known fact that virus blood levels in "healthy" HIV-infected people are low. Similarly, the known fact of low HIV infection risks during some kinds of heterosexual intercourse similarly admits to several interpretations, but one which requires no additional hypotheses is surely (and simply) that certain kinds of _homosexual_ behavior are more dangerous from an infection transmission standpoint (as discussed above) by their very _mechanical_ nature, without need to resort to the dubious and insupportable idea that the individuals who practice them must also be grossly immunologically compromised for HIV infection to occur. Root-Bernstein, eager to draw attention to any factor other than HIV in the causation of AIDS, does not take into account the most obvious physical factors. "...what is clear from existing studies," he asserts (p. 45), "is that HIV is extremely difficult to transfer to a healthy individual." In fact, existing studies establish no such thing. Studies quoted by Root-Bernstein never demonstrate that only "unhealthy" people in known risk groups contract HIV, only that certain traumatized risk groups (promiscuous gay men, hemophiliacs, transfusion recipients) are _on average_ somewhat unhealthy to begin with. This, of course, is not the same thing. Indeed, there is evidence that within risk groups, even the healthiest of individuals (immunologically) are capable of contracting HIV. Although men with hemophilia and homosexual men are on average mildly immunosuppressed even in the absence of HIV, it is by no means true that all are. A study of army recruits (surely a carefully screened group for health) shows that those who seroconvert to HIV (demonstrating HIV infection) may initially (by the criterion of CD4+ count) have immunity which is in the normal range. This is true in other groups as well [34]. Perhaps the most bloated straw man assailed by Root-Bernstein (and the one that provides the major theme of his book) is the idea that the causal agent of an infectious disease such as AIDS must be both _necessary_ and _sufficient_ to cause the disease in every sense of the terms; and moreover that since Dept. of Health and Human Services Secretary Margaret Heckler's dramatic announc- ement in 1984, most scientists have considered HIV to play this very role for AIDS. Root-Bernstein spends much time attacking what he calls the "HIV-only" theory of AIDS, an idea which actually has never flown, except possibly in the popular press or (at worst) occasionally when some scientist expresses a rash opinion without considering his formal training. (Dr. Robert Gallo, official co-discoverer of HIV, must by now badly regret his remark about HIV being able to cause AIDS in Clark Kent [35]; but this is hardly the N.I.H. official position). The subtitle warning of Root-Bernstein's book is _The Tragic Cost of Premature Consensus_, and it appears from the book that it is upon the "HIV-only" theory of AIDS that the "premature consensus" of the establishment is in dire danger of settling, if it hasn't already. Fortunately, it can safely be said that no such thing is occurring in the biomedical consensus, or about to. This does not prevent Root-Bernstein (p. 331) from logically blasting the somewhat cartoonish view he attributes to medical science: "Two of the most important implications of the HIV-only theory of AIDS are that all the risk groups should develop AIDS at approximately the same rate following HIV infection and that the symptoms they manifest should, on the whole, be the same." Alas for Root-Bernstein, however, since AIDS has from the beginning involved opportunistic infection organisms which vary in prevalence among populations, and since there has been reason to believe from the first that AIDS risk varies greatly with the biological _age_ of the HIV-infected person, scientists have never, even at the beginning, seriously considered such a theory as Root-Bernstein here lays out. "One logical implication [continues Root-Bernstein] is that the immunological status of an infected person should be irrelevant to susceptibility to contagion or to the progression from infection to disease. Acquisition of the retrovirus should be the sole factor determining whether an individual develops AIDS. Everyone should be at equal risk for AIDS, just as everyone is at equal risk for hepatitis B virus, syphilis, or measles." The most troubling thing about such writing is that an unwary lay reader may leave Root-Bernstein's book with the impression that the author has single-handedly discovered that infectious disease risks depend partly on host immune defenses and host behaviors and environments. The reader might well decide further that the biomedical community today does not in general think in terms of individuals having differing resistances to various diseases, and is accepting such advanced ideas only under duress, due to political pressures resulting from the penetrating logic of popular writers such as Root-Bernstein who are "re-thinking AIDS." The facts are more mundane. Obviously, since no microbe infects 100% of people exposed to it, or even causes disease in 100% of the people it infects (not even HIV has been shown to do this), there must be other factors to explain why some exposed people become ill with ANY infectious agent (viral, bacterial or parasitic), and some do not. Medical science certainly recogni- zes such factors, but does not use them to argue that there is in general something badly wrong with the germ theory of disease. Instead, as discussed earlier, medical scientists regard "caus- ality" in infectious disease in merely the sense of "necessity" (i.e., the "causal" microbe is necessary, but not sufficient). Medicine has not regarded the pathogenesis of any natural infection in terms of a "germ only" theory such as Root-Bernstein describes, since Pasteur, referring to disease, said that "The seed is nothing, and the soil is everything." Thus, Root- Bernstein spends many chapters assailing an idea that physicians have not held since the late 19th century, and certainly have never generally held in the case of AIDS. No infectious agent is usually "sufficient" to cause disease in a natural host, although in a laboratory (or perhaps very occasionally in nature) it may be sometimes true that inoculum (microbe "dose") may be so high as to make host resistance almost irrelevant. Naturally-occurring infectious disease organisms at reasonable doses, however, always rely on a chink of some kind in host immunity, with regard to that particular microbe (this is not to say that we must consider any host that is successfully infected to be "immuno-compromised"-- that would cheapen and overly broaden this useful term). The idea that deficiencies in host defense in some sense "permit" all or most infections is indeed a standard medical teaching [36], although a lay reader of Root-Bernstein might be surprised to learn of it after Root- Bernstein finishes misrepresenting the standard views of modern medicine. AIDS Heresies, Part 4 "Why is there such a huge and medically unprecedented variation [in time between HIV infection and death from AIDS]?" asks Root-Bernstein (p. 89). The answer to this rhetorical question is that such variation is _not_ medically unprecedented. Other infectious diseases, from malaria to syphilis to tuberculo- sis to viral hepatitis, may kill years after initial infection-- or within a much shorter time. In a cohort of newly-infected people, any study of a chronic infectious disease cannot help but produce steady increases in the "average" time between infection and death, as deaths accumulate slowly while the study follows the infected cohort prospectively onward in time. Once again, medical science has long assumed that there must exist host factors and other factors which explain why some people die of (say) tuberculosis or serum hepatitis 3 months after infection, and others not until after 3 decades-- but again these factors have nothing to do with our standard way of speaking of infectious disease "causation." The question of whether eventual liver failure from chronic infectious hepatitis is "caused" by the viral infection, is medically simply another way of asking whether or not it would occur _without_ the virus. It does not mean that other causal elements in the chain are being denied by medicine-- only that for very practical and necessary reasons they are being ignored at present (we will return to these reasons later in discussing polio). "No theory based solely on HIV can explain the phenomenon [of variable times of death]," writes [[ part lost ]] was explained as being merely an early warning of impending immune failure in these groups. (In HIV/AIDS skeptic lore, HIV is a harmless bellwether virus which can tell when certain members of modestly immunosuppressed groups are headed for future immune failure, and rapidly infects them ahead of time). The process of multiplying causal theories in order to minimize HIV responsibility for AIDS has finally culminated in the work of Root-Bernstein, which contains an eclectic "multifac- torial" view of AIDS which is so formless and complicated as to be epidemiologically unfalsifiable, even in Root-Bernstein's view (see his p. 92, quote below-- we will turn to this theory later). At the time of the early drug/toxin theories of AIDS, the leading toxin candidates were the inhaled amyl and butyl nitrite street drugs ("poppers") used heavily and almost universally as sexual-experience enhancers in the 1970's and early 1980's by the same fraction of homosexual men who indulged in high risk, promiscuous sexual practices causing injuries to mucosal tissue, and who also historically were the first U.S. group to develop AIDS as an "epidemic" [14]. Since this group was the one that suffered the first major impact of AIDS, a number of early studies found high statistical correlations between AIDS risk and nearly everything to do with this group's lifestyle. Later, after the HIV virus was identified, the CDC found that HIV was universally present and active in such men who developed AIDS. Almost as prevalent were a number of other chronic viruses, such as CMV (cytomegalovirus), HZV (Herpes Zoster virus), EBV (Epstein-Barr virus), and HSV-1, 2, and 6 (Herpes Simplex viruses 1,2, and 6). Many of these chronic viruses were found to be replicating actively in homos- exuals with AIDS. This state of "viral re-activation" (a product of immune suppression) was less common in AIDS sufferers from other risk groups, mainly because other groups had not been infected with as many chronic viruses in the first place. Sorting through the drug and infection variables among promiscu- ous homosexual men with AIDS was a statistical nightmare, although it became easier to separate out important AIDS risks when AIDS in other groups with different lifestyles was consider- ed. Epidemiologists fought it out in scholarly journals. After the main battle was over, they even tried to decide who hadn't guessed from the beginning on epidemiologic grounds that the problem might be infectious, even before a specific causal virus was proposed--- occasionally lambasting each other's past methods in print with words like "Neanderthal" [37,38]. Before HIV was identified, however, the basic problem for epidemiologists was that statistical methods could not by themselves suggest which lifestyles or practices (if any) were causal for AIDS, and which were merely an associative marker for some other causal factor which (perhaps) had not been measured. _After_ HIV was identif- ied, however, a second statistical appraisal could be taken using HIV status as a statistical factor, in an attempt to see if HIV had a closer associational (and therefore presumably more likely causal) relationship with AIDS, than other previously identified factors [39]. It did [40]. Much the same thing happened with other viruses, especially when statistics were extended across different risk groups. Infectious HIV was finally found to be present in essentially 100% of AIDS cases in ALL risk groups-- a higher proportion than was seen with any other virus [41]. Furthermore, although the prevalence of HIV was high (as much as 50%) in certain risk groups, such as homosexual males in some cities, or in people with hemophilia, the difference in HIV status for a risk group, and those who actually contracted AIDS within the risk group, was by far the largest _change_ associated with any virus, increasing from 50% to 100%. Furthermore, as noted, most of the change in HIV infection status had occurred before 1984 in people with hemophilia [28], long before the worst incidence of immunosuppression and in- creased death rate in this community, proving at least that HIV positivity was not derivative of _severe_ immunosuppression, since it preceded it. As judged by CD4+ lymphocyte counts in people who were followed over time, most of the loss of immune function in _individual_ HIV-positive people with hemophilia, came after HIV infection, as well [42]. Finally, it was found in several studies that while HIV- negative homosexual males might be mildly immunosuppressed, their immune function was never seen to drop as low as AIDS-class- immunosuppression (immune failure), as defined by CD4+ lymphocyte counts below 200. Moreover, when followed over time, HIV- negative homosexual men did not become MORE immunosuppressed, but HIV-positive ones did, and in prospective studies when men were followed by blood tests as they actually contracted HIV, this same slow and steady decay in immune status happened to the newly infected group after contracting HIV infection, starting immedia- tely after infection. HIV, when contracted by men being followed in studies, was generally contracted during a time when immune status (CD4+ lymphocyte count) was reasonably good [43]. Men who were severely immunosuppressed (CD4+ lymphocyte counts below 200; essentially immune failure) with no other explanation (such as cancer), invariably had become HIV positive already, or in other words, had become HIV-positive _first_ [44]. Such tight corre- lations between timing of immune failure and time of infection do not hold for any other known viral infection in immunosuppressed people. Lifestyle factors such as non-injected drug use and exposure to blood products (as in transfusions or hemophilia treatments) did correlate with risk of developing AIDS, but this association could be completely explained in the statistics by the fact that these behaviors (including even perhaps nitrite ("popper") use [45]) also increased risk of contracting HIV. To discover which was most important to risk, HIV or drug use, epidemiologists statistically "controlled" for HIV status (i.e., compared people with each other only within HIV status groups), attempting to discover if drug use or blood product exposure was important to AIDS risk AFTER the HIV virus was contracted, or independently of it. The answer, it turned out, was generally no. By and large, drug use and promiscuity were not independent variables after HIV infection was taken into account (with two exceptions, to be discussed). Important studies finding this included the followi- ng: * With regard to risk of AIDS in homosexual men, Susan Buchbinder followed 588 men in the San Francisco Men's Health Study who were infected with HIV at well documented times in the early 1980's (this is known because their sera was being stored for use in a hepatitis study). Of 588 men, 51% had developed AIDS by 10 years, and 69% had developed AIDS within 14 years of becoming HIV-positive. The San Francisco study also found that HIV- negative controls (both homosexual and heterosexual) had stable immune status, and did not develop AIDS, nor AIDS defining diseases. Of 538 men who became HIV-positive before 1983 in this study, only 8% were "healthy" HIV-positives as of Jan 1, 1993. This group of 1 person in 12, who for reasons still not underst- ood possibly will be long term survivors of HIV infection, was found to be quite time-stable in immune function, though mildly abnormal in immune status with respect to HIV-negative controls. [Italics]: _Significantly, the long term healthy HIV-positive men were as likely to have had sexually transmitted diseases or been users of illegal drugs as the other HIV-positive men in the study who became ill_ [46]. Those who progressed to AIDS and those who remained healthy, reported equal and significant use of marijuana, nitrite inhalants, and amphetamines. Both groups had similar herpes and hepatitis B infection immunity on lab testing- - objective markers for risky sexual behavior [46]. Regarding people already infected with HIV, there is no support from this study for the idea that either of the two possible outcomes (eventual progression to AIDS, or the long term health which may be possible for a small subset of HIV-infected people) has anything to do with non-injected drug use or sexual habits _after_ HIV infection, within a fairly wide range of behavior. A specific re-analysis of non-injected illicit drug use in this study [40] showed that 3 stratifications of drug use groups into low, moderate and heavy use, had NO statistical effect on risk of progression to AIDS in groups (even if done imperfectly, as Duesberg contends, some effect should still have been seen). [Fig. 2]. All 3 HIV-positive drug-use groups over several years- time lost CD4+ lymphocytes with equal rapidity and developed AIDS with equal probability; and all 3 HIV-negative drug use groups did not lose CD4+ lymphocytes at all, and did not develop AIDS defining diseases. Crude death rates in groups agreed with CD4+ loss rates and AIDS diagnosis rates, suggesting no bias problems with any of these markers, as has been suggested by critics. Statistical association of nitrite use with AIDS disappeared when only homosexual men, or only HIV-positive men, were examined, and thus nitrite use was seen to be merely a marker for homosexual behavior or HIV risk in San Francisco men, not an independent AIDS risk. Previous investigators in this study had also examined the statistical effect on AIDS risk of reported recreational non- injected drug use, and certain sexually transmitted infections, while controlling for other variables, and also found no associa- tion [47]. * In the Multicenter AIDS Cohort Study following a group of 1835 HIV-positive homosexual men, it was found that the 59 men who developed AIDS over 15 months during the course of the study were no more likely to have used inhaled nitrite "poppers" in the previous two years than were 295 matched controls picked from the same group, who did not develop AIDS [48]. In another paper from this study entitled "No evidence for a role of alcohol or other psychoactive drugs in accelerating immunodeficiency in HIV-1-pos- itive individuals..." Kaslow and coworkers found that the proportion of HIV-positive men at enrollment who developed AIDS during the following 18 months ranged from 5% to 8%, and that among HIV-positive men with low CD4+ lymphocyte counts, those who continued to use drugs showed no significantly higher 18-month risk of AIDS than non-users. No other manifestations of immunodeficiency were positively associated with substance use prior to enrollment in this study. Prior use of non-injected drugs was not associated with low mean CD4+ lymphocyte counts at enrollment, and continued drug or alcohol use after enrollment was not associated with greater subsequent decline in CD4+ lymphocyte counts. The authors conclude that as used in a large cohort of homosexual men, psychoactive substances did _not_ enhance the progression of HIV infection to lower CD4+ counts or AIDS [49]. * Likewise, the Vancouver Lymphadenopathy-AIDS Study (VLAS) comparing HIV-negative to HIV-positive men found no sexual behavior variables which correlated with CD4+ lymphocyte counts independent of HIV status, and for the 25 men who developed AIDS during the study found no significant differences in sexual behavior or illicit non-injected drug use as compared with 80 controls who were HIV-positive but remained AIDS free [50]. These studies included 78 HIV-negative men who were heavy users of many drugs who did not lose CD4+ lymphocytes or develop AIDS defining diseases, and also 19 HIV-positive men who reported no recreational drug use of any kind (and who were followed, incidentally, before AZT became available, yet lost CD4+ lymphoc- ytes steadily [51]). Again, AIDS risk in all studies correlated highly with the use of both intravenous and also certain non-injected drugs in homosexual men, but just as highly with promiscuity and high-risk sex in such men. If non-injected drug abuse (particularly nitrite use) was simply part of the lifestyle which caused one to contract the HIV virus, we would expect the correlation between non-injected drugs and AIDS to be high, without necessarily being causal. To separate out any independent effects of orally administered, inhaled, or intravenous illicit drug use, it is necessary to look at only men who had been infected with HIV while still showing reasonably good immune function (CD4+ lymphocyte count), and ask the question of whether drug use THEN, after infection, had any effect on the risk of later AIDS (which in most cases, appeared years later). The statistical answer was that after HIV infection, contin- ued injection of narcotics did seem to increase risk of AIDS, but that use of no other drug had any effect at all. Specifically, several early (pre-HIV) studies found connections between nitrite use by male homosexuals and Kaposi's sarcoma; but (in addition to the many studies cited above) later studies found that this association between nitrite use and AIDS disappeared after controlling for HIV status [52]. If recreational nitrite inhalants or oral drugs caused AIDS or Kaposi's sarcoma, and HIV was harmless, there should be some continued statistical associa- tion between AIDS risk and nitrite or other drug use, even when looking only at HIV-positive men, or only at HIV-negative men. No such association has been found in more than half a dozen studies [37-40, 46-52]. Illicit drugs did not cause AIDS, with the qualification that injection of drugs was implicated as a cofactor (see below). In the end, the "drug-only" hypothesis of AIDS pathogenesis failed all careful epidemiologic scrutiny. Even among IV drug users, although short term overdose deaths tend to swamp any necessarily long-term consequences of HIV infection [53], studies have generally shown that HIV infection is an additional risk factor for IV drug users [54-56]. Injected or IV drug use, of course, proved an excellent way to _contract_ HIV, if needles were shared. There was no evidence, however, that injected drugs themselves ever led to severe AIDS- type immunosuppression (CD4+ lymphocyte counts less than 200), in the absence of HIV. There was some evidence that IV narcotic use could be quite immunosuppressive (leading at least in part to fatal infections), and a co-factor for rapid AIDS development in people HIV-positive. Some studies found that continued IV injection of heroin, but not use of other drugs, hastened progression to AIDS in HIV-positive people [57], but other studies have suggested that heroin does not decrease CD4+ counts as AIDS does, so the immunosuppression of heroin users is apparently not due to heroin itself, but possibly drug impurities, or to other factors in drug-injectors [22]. One study [58] suggested that HIV-positive IV drug users who switch to methadone (an oral heroin substitute) may have slower progression to AIDS, but there was no mortality difference between using methadone and quitting narcotics completely, so IV injection per se, rather than narcotic use, was possibly the offending practice. Studies of IV drug users who continued IV drug might be implicating not drugs themselves in the rapid production of AIDS, but rather simply continued needle-sharing leading to acquisition of more virulent strains of HIV. The same was possibly true of extreme promiscuity, which also continued to be a risk factor after HIV-infection, in one study [59]. Acquisition of CMV was also a possible consequence of risky behavior, although the role of this virus as cofactor was limited at best [60]. In summary, the search for controllable cofactors in develop- ing AIDS, once HIV infection is present, was disappointing. Needle-sharing and extreme promiscuity very possibly hastened AIDS in HIV-positive people, but if they did so, it wasn't by much. Nor was there any evidence that avoiding these behaviors led to significant delays in AIDS, as compared with risk groups such as HIV-positive hemophiliacs, who still had a 50% chance of developing AIDS over 15 years. The Critics Don't Give In The above results have not convinced those who champion the drug hypothesis as the cause of much of AIDS. Duesberg, for example, accepts a causal role for drugs in AIDS (and for that matter, also for smoking in the causation of lung cancer: Duesberg, p. 213) on much the same grounds which he rejects for assigning a causal role to HIV-- namely, epidemiol- ogical correlations and suggestive lab experiments. The irony of this position is that the correlations are not nearly as good statistically for drug use and AIDS as they are for HIV infection and AIDS, and lab animal experiments with retroviruses mimic human AIDS immunodeficiency far more precisely than lab animal experiments with drugs have. Nitrite recreatio- nal drugs, for instance, do not cause lymphocyte abnormalities, lymphadenopathy, or the almost complete disappearance of CD4+ lymphocytes at the same time as CD8+ lymphocytes counts rise, as all happen in AIDS. Yet chronic retrovirus infection of experimental animals routinely causes these odd and specific phenomena. Nor have opportunistic infections or lymphomas ever been seen in a nitrite treated animal, and yet these are hallma- rks of lentivirus infections. Apparently, Duesberg's standards of evidence change greatly with the hypothesis he likes. Duesberg actually accused a group of scientists of data fabrication [61] after a paper in _Nature_ reported findings not in line with Duesberg's drug hypothesis [40]. Duesberg's letter to _Nature_ calling some results of the paper from the San Francisco Men's Cohort study less than honest, was refused print by _Nature's_ editor, with an accompanying editorial [62]. An independent institutional review board cleared the researchers of Duesberg's charges, which have been answered in print by the authors [63]. In the study printed in _Nature_, the authors had found in the San Francisco cohort men no connection between the four most commonly reported kinds of illicit drug use and later progression to AIDS, after results were controlled for HIV status (i.e., heavy drug users had the same likelihood to progress to AIDS as light users, if HIV-positive, but HIV-negative men did not progress toward immune failure, no matter what their drug use). [See Fig 3]. Moreover, these results held also for the 1985-1986 period before the drug AZT (the use of which Duesberg has suggested may cause AIDS to develop in HIV-positive people) was available. Duesberg's objections were that the study had not controlled carefully enough for drug use between HIV-positive and HIV- negative groups in the study, but Duesberg did not address the obvious question of why such considerable controlling for drug use as _was_ done, had absolutely no effect on differential AIDS risk seen. Duesberg also complained after seeing the raw data that supposedly "AIDS-defining" diseases in the HIV-negative group had not been counted as "AIDS," despite the author's denial that this had happened. Here apparently much depends on a disagreement between Duesberg and others as to what constitutes clinical AIDS. A recent article in _Science_ suggests that one difficulty is over the question of whether mild opportunistic conditions such as oral candida (thrush) constitute clinical "AIDS." Duesberg, ever ready to define AIDS broadly, argues they should [22]. In any case, the specifics of Duesberg's reanalysis of the Nature paper have never been printed, and death rates in this study again underscore the fact that Duesberg's broadly defined "AIDS," which strikes HIV-negative people, somehow does not kill nearly as well as the standard variety. As for worsening immune failure in groups over time (seen as declining CD4+ counts in the HIV-positive men, independent of drug use, but not in HIV-negative men, no matter how much drug use), Ellison and Duesberg have noted that this phenomenon isn't so clear before the data is reduced to averages. This, however, seems a odd complaint (making group trends clear is _why_ scientists calculate group averages). The bottom line is that, for now, the drug hypothesis of AIDS has no epidemiologic associational evidence behind it which is independent of HIV infection. HIV infection, by contrast, is heavily associated with AIDS risk, independent of drug use. [See AZT Section in Appendix on the question of whether AZT contributes to AIDS] Previously Known Viruses Because Duesberg does not regard any virus as being capable of causing a fatal disease long after the body has generated an antibody immune response to the microbe, he rejects a causal role for any virus in AIDS. The evidence for multiple viral infecti- ons in many of the early victims of AIDS, however, has caused many such "non-HIV virus" theories to be generated and tested. For example, though Root-Bernstein does not regard HIV as always a bystander virus in AIDS, he does regard other viruses in AIDS to be just as important as HIV. Are the other viruses (CMV, EBV, Herpes, etc.), or at least their antibodies, present as often in AIDS as those of HIV? Here statistics can help. According to Root-Bernstein, evidences of replication of the viruses CMV and EBV "are just as frequent concomitants of AIDS as is HIV replication" (p. 260). Unfortunately Root-Bernstein fails to note that this is true only in homosexual men with AIDS (where co-infection with EBV and CMV along with HIV is nearly universal). In science, situations in which several possible causes are all nearly 100% associated with a particular effect do not help us to differentiate causality, a point that Root-Bernstein makes (p. 279-280) without taking the next logical step. What are needed with viral studies and AIDS, obviously, are AIDS cases where some of the putative viral causes are present less frequently than 100% of the time. Such cases are available. In both hemophilia and transfusion-associated AIDS, HIV infection is universal, whereas infection and re- activation with other viruses, such as CMV and EBV, is variable [64]. In short, some people with AIDS in these groups have never been infected with CMV or EBV viruses at all in the past-- but all have been infected with HIV. Much the same is seen to be true when people in AIDS _risk_ groups are examined for antibodies to various viruses. EBV or CMV antibodies alone, in themselves, are not predictive for development of AIDS in risk groups where exposure prevalence of CMV and EBV is much less than 100% [60,65], a fact that eludes Root-Bernstein, who writes with too much generality that CMV and EBV antibodies are "synonymous with AIDS" (p. 103). In fact, as we have noted, CMV and EBV antibody are not unusually common in either HIV-positive people, or AIDS patients, from groups such as transfusion recipients, and people with hemophilia, in some countries [66]. Nor are antibodies to these viruses a clear extra risk factor for development of AIDS, even in HIV positive people [60]. Cases of people multiply infected with many chronic viruses, but not HIV, are also well known, and such people are easily differentiable from AIDS on clinical and laboratory grounds (see the study of Salvato [13] above). Despite Root- Bernstein's suggestions, statistical associations clearly differentiate HIV from other viruses in AIDS. It is necessary for heretics to come to grips with the crucial point (hard to explain if HIV has no causal role in AIDS) that the utility of HIV antibody screening is _exactly_ that a positive HIV screen, found in only 0.3% of the population, is predictive of risk for development of severe immunodeficiency, i.e., 50% risk of developing severe, life-threatening immunodef- iciency within less than 15 years. By contrast, EBV and CMV viral immunity and antibodies are acquired by most (well over 50%) of any normal, healthy population of humans during a lifetime, and thus are not predictive of future severe immunod- eficiency and death, except to the extent that they are markers for membership in particular risk-groups (such as gay men). Like many other factors, the association of viral antibodies with AIDS across risk groups disappeared when people were compared within groups-- except for HIV, where the association persists. The fact that HIV antibody is even 50% predictive of disease or death within 15 years is exactly the burden that those who argue that HIV is not causal must labor under, if they are to face the evidence. Other viral hypotheses fail this most simple test of prognostic value. As any life or medical insurance company knows, HIV infection status (HIV-positivity status) is more surely predictive of future death due to future severe immune failure than any other known piece of medical information related to viral infection. In science, if one has a list of factors which are independently statistically associated with an effect (such as AIDS), this does not prove that any of them are causal. If one of these factors (such as HIV infection) is far more highly associated than any of the others, however, it becomes difficult to argue convincingly that it is no more important than the others are. We will return to this point later. AIDS Heresies, Part 5 Why Did Medical Science Historically Implicate HIV? The standard method of trying to identify a new virus in a new disease suspected of being caused by a new virus, is to attempt to culture a new virus from an infected person, then show that antibodies to this virus are present in all people with the disease, but less often in people who are not ill. It is also helpful to show that persons develop antibodies to the virus during the acute illness. Sometimes viruses can be very difficult to culture in lab glassware. This is especially true of viruses which grow in human T-cells-- cells which could not be grown well without certain growth factors only discovered in the 1970's. In 1980, Dr. Robert Gallo of the NIH formally reported isolating a virus he named "Human T-cell Leukemia Virus," (HTLV) which infected T- cells and which was thought to cause some cases of T-cell leukemia in humans. This virus was a retrovirus (a virus type associated with cancer causation), and it was a distant relative of the "Feline Leukemia Virus" (FeLV) which caused leukemia in housecats. Because the transfusion results had shown that AIDS could be infectious, and because AIDS patients had abnormal-looking T- cells which looked something like those from retrovirus-infected animals, or T-cells in cultures infected with retroviruses, early AIDS researchers began hunting a T-cell retrovirus. In early 1983, a team of French scientists led by Dr. Luc Montagnier isolated a new retrovirus which they reported in May of that year, calling it eventually Lymphadenopathy-Associated Virus (LAV), because it had been isolated from tissues of a French patient with enlarged lymph tissues, or "lymphadenopathy" (this man died of AIDS in 1988). The French had been alerted to the possibility of a retrovirus in AIDS patients by an American team, led by Gallo, which was convinced that the AIDS virus was another variety of HTLV. It wasn't. [See appendix: "Montagnier, Gallo, Slip-Ups and Wrong Paths"] The new virus discovered by the French was a tiny, spherical, membrane-coated, protein-studded virus 1/100th the diameter of a lymphocyte, with an inner protein viral core shaped like a truncated cone, with a dense base. Under the electron microscope it didn't look like the feline FeLV or the human HTLV leukemia viruses (Fig 4), which had no distinct cores. Eventually, the LAV virus was correctly understood that Summer by the French team not to be a leukemia virus as they had thought, but rather to be the first human "lentivirus." This hypothesis was first formula- ted when Montagnier, at a suggestion from a colleague, began reading about "lentiviruses" or "slow viruses" -- a class of animal retroviruses he'd never previously heard of. In one book was an electron micrograph of the "equine infectious anemia virus," a virus which sometimes produced a familiar-sounding immunodeficiency and lymphadenopathy disease, after long latenci- es, in horses. Montagnier found himself looking at a tiny membrane-coated virus shaped like a sphere, containing a protein viral core in the shape of a narrow cone [67]. Antibodies against the horse virus cross-reacted with Montagnier's new virus, but not with Gallo's HTLV. Most importantly, coded AIDS patient serum provided by the CDC contained antibodies to LAV, but not HTLV-III, and Montagnier proved his lab could easily pick out AIDS samples from normal samples in the CDC material, without knowing the codes. Thus, Montagnier had his answer (his "LAV" was our modern HIV), but nobody would believe him for almost a year. Lentiviruses and Latent Diseases Most readers will remember that viruses in some sense are not complete living organisms. Animal viruses when outside cells don't metabolize, and cannot reproduce or grow by themselves. Instead, most viruses are little more than tiny floating packages of genetic material, sometimes without much other equipment. Viruses can reproduce themselves only by entering a living cell and commandeering the cell's synthetic machinery to subvert it into making more virus particles, which are then, in turn, released to infect more cells. A metaphor for a virus would be a truck-load of blueprints which rolls into a completely automated factory, and once there, is somehow able to use the blueprints to control the factory's machinery to cause it to make more sets of blueprints and more trucks to carry them, all of which are then assembled and sent out to take over more factories. [See appendix: "How Do Viruses Hide From the Immune System?" and "What is a Retrovirus?"] Members of one class of viruses use RNA as their genetic material, and are called "retroviruses," because their synthesis of DNA from RNA proceeds retrograde, in the opposite direction to what is "normal" in the rest of biology. Retroviruses avoid the body's immune system by inserting themselves into the DNA of the host cell. Most retroviruses cause no major disease, but not all are harmless. A sub-class of retroviruses, called "lentiviruses" is capable of slow infections resulting in death. Lentiviruses typically spend lengthy waiting periods in hiding in the cell nucleus (music lovers will recognize the Latin root _lento_, meaning "slow"), and even lentiviruses may never cause overt disease in their natural hosts. Sometimes, however, the lentivirus disease produced after a latency period can be devastating, though sometimes difficult to detect epidemiologically, due to the delay between initial infection and death. Lentiviruses were named in 1954 in honor of several very slow- acting brain infections of farm animals. The classic example was a sheep disease with a latency as long as a decade, called "visna" (an Icelandic word for "shivering"). Visna wiped out most of the sheep population of Iceland in 1939 because it had not been realized in 1933 that apparently-well sheep brought to the island from Germany had actually been carrying a latent disease. The ability of visna to cause disease many years after infection has since been demonstrated in a series of controlled experiments with sheep [68]. The visna agent proved eventually to be a retrovirus and a lentivirus, and the ominous pattern of the visna epidemic will become familiar to the reader during the course of this essay, as we discuss other better-known viruses in this unique class. HIV, our main subject, is a lentivirus. It is clearly related to the visna sheep lentivirus in structure and genetics, and more closely to "equine infectious anemia virus" a lentivirus which infects macrophages in horses. It is yet more closely related to FIV and SIV, two other immunosuppressive lymphocyte-infecting (lymphotropic) animal lentiviruses which the reader is shortly to meet. How To Dismiss All Laboratory Animal/Virus Experiments Before we meet these viruses, we should note that molecular biologist Peter H. Duesberg's scientific heresies do not only extend to his views on HIV. Duesberg, for instance, is unable to believe that any disease which confines itself relatively rigidly to classes of susceptible people, can qualify as a genuine communicable disease caused (in any sense) by a treatable microbe. In these diseases, Duesberg tellingly prefers to focus entirely on the host, refusing to consider the role of the microbe at all (as we will see, the issue of "causation" is a difficult one for HIV/AIDS skeptics). Thus, Duesberg and Ellison dismiss (for example) Legionnaire's disease and hepatitis C as fantasy diseases made up by the biomedical establishment for purposes of profit and self-aggrandizement (pp. 35-38, 57-59). Duesberg also does not believe in slow infections in general. The modern physician perusing Ellison and Duesberg's AIDS book will be astounded to read that syphilis microbes do not cause neurosyphilis, and that leprosy is not an infectious disease nor is it caused by the organism the medical textbooks say causes it (Ellison and Duesberg, p. 35). Nor does Duesberg believe in slow viruses at all, in the sense that he does not believe that any virus is capable of causing general infection or fatality years later in an organism which had been healthy prior to the initial viral infection (Duesberg, pp 209, 233). "There are no slow viruses," says Duesberg, "only slow virologists" (Ellison and Duesberg make it clear that they consider one of these to be D. Gajdusek, who won the Nobel Prize in 1976 for slow virus work, see pp 50-54). Nor does Duesberg believe that viruses cause cancer in animals, with a single kind of exception, the partial mechanism for which was co-discovered, as it happens, by Duesberg himself. Rous Sarcoma Virus (RSV), a retrovirus, causes muscle tumors in chickens (this fact won F. P. Rous the Nobel Prize in 1966). In 1970, Duesberg and P. Vogt showed that RSV caused tumors by carrying a cancer gene into the cell. (Thus, Duesberg's conten- tion that cancer is never caused by chronic viruses which have no cancer genes, such as HTLV and HIV, perhaps coincidentally has the effect of magnifying Duesberg's own discovery). A review of the evidence that Legionnaire's disease, leprosy, and hepatitis C are real infectious diseases, or that viruses other than RSV cause cancer, is beyond the scope of this article. With regard to other latent diseases it is worth noting that Duesberg's rejection of all current theories that viruses are causal in any latent cancer or disease of normal animals or humans requires that he reject not only all epidemiological evidence, but also all experimental laboratory animal evidence for viral cancer causation and fatal viral immunosuppression. Thus, cancer and immune failure in virally infected lab animals are uniformly dismissed as being responses of weak, inbred lab animals under "odd conditions" (Ellison and Duesberg, p. 91). The reader should note that, if valid, Duesberg's view about lab animals would have the effect of making almost any claim for a virus-caused cancer or immunodeficiency nearly impossible to prove. Epidemiology, after all, can always be dismissed in the absence of lab data. Moreover, it is difficult to impossible to collect experimental lab infection data on animals which have not been bred in captivity to be free of infections other than the ones being studied. In short, if infection-free lab animals were useless for the very infection experiments in which they are needed and now used, this would leave scientists with no good way to study infectious processes at all. Fortunately, however, enough experimental data is available to reject Duesberg's claims in the matter. With the exception of rodents, most animals used in laboratories are in fact no more highly inbred than are pets or domestic animals (or many societies of humans, for that matter). Modern science works with many tumor-virus models, and there is no reason to regard (for instance) the "specific pathogen free" cats used in feline viral leukemia experiments [69], or the woodchucks used in hepadnavirus liver tumor experiments [70], as being inbred, weak, or in any way intrinsically immunologically abnormal. Moreover, although Ellison and Duesberg (p. 84) say that feline leukemia virus and all other retroviruses only cause disease in young inbred lab animals, experiments have proven them wrong. In the next section we will discuss a retrovirus which causes fatal immunodeficiency and cancer, even when injected into so-called "random-source" domestic cats, such as one might obtain from a pet store or animal shelter [71]. Animal Lentiviruses Suitable For Laboratory Study In this section we will examine two very similar animal lentiviruses called FIV and SIV, and will note something of their effects in different animal hosts. The details about these two virus/host systems are given because all are crucial to a pattern which will be apparent by the end of our survey. The two viruses we are about to describe were actually discovered several years _after_ HIV, but it is more illuminating to tell about them first, for nature seldom provides her good clues in proper order. The reader should again bear in mind that our modern idea of the cause of AIDS is based on induction and inference, and inference depends on recognition of common _patterns_. Some of the crucial information for one of these patterns follows. By the end of the discussion, the reader should have some idea for why the "odd" effects of HIV in hu |