Commentary: non-HIV hypotheses must be studied more carefully

BMJ 1996; 312 doi: (Published 27 January 1996) Cite this as: BMJ 1996;312:210
  1. Peter Duesberg, professora
  1. a Department of Molecular and Cell Biology, University of California, Berkeley, California 94720-3206, USA

    Patients with haemophilia treated with commercial factor VIII and zidovudine are at risk of developing AIDS defining immunodeficiency diseases like pneumonia, candidiasis, and toxoplasmosis but not Kaposi's sarcoma or dementia. Their risk of these diseases is proportional to their lifetime dosages of factor VIII and cytotoxic DNA chain terminators like zidovudine.*RF 1, 1a*

    Haemophilia specific AIDS diseases have been explained in terms of two hypotheses. A non-infectious hypothesis holds that the long term administration of immunosuppressive foreign proteins contaminating commercial factor VIII and zidovudine causes AIDS. An infectious hypothesis holds that HIV causes AIDS.1

    Still no evidence for infection

    Based on inadequate tests, Sabin et al reject the foreign protein-zidovudine hypothesis.2 However, their data and those of others actually support the hypothesis.

    Sabin et al claim that HIV causes AIDS because all of their haemophilic patients had antibodies against the virus. But HIV cannot be enough for AIDS because six of their 17 HIV positive patients (table 2) remained healthy for 10 years. Likewise 12000 out of 15000 HIV positive American haemophilic patients have remained AIDS free since 1985.1 3

    Sabin et al observe that AIDS diseases occur, if at all, only five to 10 years after the appearance of antibodies to HIV. However, “viral” AIDS should occur within weeks after infection because viruses replicate exponentially. A single infected cell produces over 100 HIV virions within 48 hours,4 generating 1014 viruses within two weeks after infection—enough to infect every cell in the human body. Moreover, viral AIDS should be neutralised—not caused—by the antiviral immunity that is detectable in haemophilia.

    By contrast, long time periods are required to accumulate pathogenic doses of foreign proteins via transfusions.1 Because HIV is a rare contaminant of factor VIII, it seems to be just a surrogate marker for high dosages of foreign proteins. Sabin et al confirm this: “Most patients with severe haemophilia—and therefore the highest users of clotting factor concentrate—were seropositive to HIV….”2 Thus their results support the foreign protein hypothesis.

    An appropriate test of the foreign protein hypothesis would compare HIV positive patients with HIV negative patients matched for the lifetime dosage of factor VIII. By contrast, Sabin et al matched patients for current usage. Plainly, a 60 year old and a 14 year old would not be an appropriate match (table 1).2

    The HIV hypothesis predicts the same pattern of AIDS diseases in haemophilia as in other AIDS patients, but Kaposi's sarcoma and dementia were not observed in the haemophilic patients studied by Sabin et al (table 2). However, the foreign proteinzidovudine hypothesis predicts the restriction to immunodeficiency diseases observed by Sabin et al exactly.

    The HIV hypothesis predicts sexual transmission of AIDS. However, in the United States the wives of 15000 HIV positive haemophilic men have only the normal background prevalence of AIDS defining diseases.1 5

    Strong temporal evidence

    HIV is now said to cause AIDS 10 years after infection, and infections by transfusions were stopped in 1985. Hence AIDS in patients with haemophilia should have peaked long before 1995. Instead, cases declined until 19861 and then sharply increased from 1987 both in the United States and in Britain when zidovudine and other drugs were introduced as treatment for and prophylaxis against HIV infection and AIDS.6 7

    Zidovudine kills all growing cells, particularly fast growing blood cells, and therefore must cause immunodeficiency. Even the manufacturer acknowledges, “It is often difficult to distinguish adverse events possibly associated with zidovudine administration from underlying signs of HIV disease.”8 An epidemiological study cited by Sabin et al directly confirms that zidovudine treated HIV positive haemophilic patients have a 4.5-fold higher annual AIDS rate and a 2.4-fold higher annual death rate than untreated controls.9 This explains why the morbidity of American6 and British7 patients with haemophilia has sharply increased since 1987, when most HIV positive haemophilic patients began zidovudine.*RF 1, 1a* Sabin et al confirm the zidovudine hypothesis, as “all nine HIV positive patients with AIDS received zidovudine” in addition to the “toxic” pentamidine,10 co-trimoxazole, and fluconazole.2

    The foreign protein hypothesis predicts that AIDS in patients with haemophilia can be prevented if not cured by using factor VIII that is free of foreign protein. Sabin et al endorse this: “These concentrates seem to preserve the immune system….” Indeed, not only were the immune systems of HIV positive haemophilic patients treated with pure factor VIII and no zidovudine “preserved” but their T cells increased by up to a quarter over three years—despite the presence of the hypothetical T cell killer HIV.11 12


    The foreign protein-zidovudine hypothesis provides biochemically plausible candidates for pathogenicity—that is, large amounts of foreign proteins transfused over decades and DNA chain terminators prescribed indefinitely such as zidovudine. With a daily dose of 500 mg zidovudine a person receives 2x1020 molecules, or 2x106 for every human cell, every day for the rest of his life. By contrast, after 11 years of research the HIV hypothesis has yet to identify a single molecule that could be pathogenic.13 In view of this and the complete failure of the HIV hypothesis to produce public health benefits, non-HIV hypotheses must be studied more carefully.


    1. 1.
    2. 1a.
    3. 2.
    4. 3.
    5. 4.
    6. 5.
    7. 6.
    8. 7.
    9. 8.
    10. 9.
    11. 10.
    12. 11.
    13. 12.
    14. 13.
    View Abstract