Reappraising AIDS Vol. 3, no. 4

VOLUME 3, NUMBER 4, PAGE 3

REAPPRAISING AIDS

APRIL 1995

VOLUME 3, NUMBER 4, PAGE 3 REAP PRAISING AIDS APRIL 1995

added the physicist. A good deal of Cohen's section on
drugs as a cause of AIDS was devoted to defending AZT
(which "can be severely toxic"), against Duesberg's
charge that it is "AIDS by prescription."
Perhaps Cohen's most important section dealt with
hemophiliacs. Duesberg and his critics think they are the
best "test case" of the HIV theory, Cohen said, because
they are "not riddled with what Duesberg thinks are confounding variables." In fact, Duesberg thinks hemophiliacs are not a good cohort, precisely because they are
confounded with a variable: contaminants of factor VIII,
the blood-clotting factor available since the late 1960s.
Those who have received a high lifetime dose are much
more likely to be HIV positive, because HIV-until it was
eliminated from the blood supply in the mid-1980s- was
a rare contaminant of clotting factor. In 1994, Aids News,
published by the Hemophilia Council of California, granted foreign proteins transfused with factor VIII the role of a
cofactor in AIDS. "Factor Concentrate is a Co-Factor"
was their headline.
The association between clotting-factor dosage and
HIV has been reported in several studies. Nonetheless,
Cohen's principal witness against Duesberg was the
National Cancer Institute's James Goedert, who said
that dosage of factor VIII has "absolutely nothing to do"
with a person's "risk of dying." If so, HIV itself must have
"nothing to do" with risk of dying, because the correlation between dosage and HIV is well established. See for
example Sullivan et al. who reported in the Journal of
Pediatrics that seropositivity to HIV "varied directly with
the amount of factor VIII received." [5]
Goedert may have missed this association because of
gaps in one of his own studies. In a 1989 report on AIDS in
hemophiliacs, Goedert et al. examined potential AIDS
cofactors, and found that the disease was "not related
demonstrably to the cumulative dose of plasma concentrate." [6] But the accompanying table showed that
he had come perilously close to establishing just such a
correlation, perhaps failing to do so only because the relevant data was missing. Out of 24 cases of AIDS where
the dosage was known, 21 had received a "high" or
"medium" dose. In 26 additional AIDS cases, the dosage
was listed as "missing." But if absent data in his own study
caused Goedert to overlook a well attested correlation,
he might have looked at others; for example Kreiss et al:
"Seropositive hemophiliac subjects, on average, had
been exposed to twice as much concentrate" as
seronegatives. [7]
For Cohen's benefit, Goedert extracted additional
data out of his Multicenter Hemophilia Cohort data
base. Goedert, who is principal investigator of the
cohort, divided hemophiliacs of similar age into HIV-positive and -negative groups, and divided into hic~h, moderate or low cumulative dosage groups. The results confirmed that HIV is a high-dosage marker: no "high dose",
HIV-negatives could be found. Out of 0 high-dose HIVnegatives, Cohen duly reported, 0 had died. The num
bers in the "moderate dose" category were also very
skewed, with 264 HIV-positives and only 17 HIV-negatives.
[1] It seems curious that the principal investigator of a

major hemophilia cohort would doubt the immunosuppressive properties of factor VIII concentrate, collected
from thousands of donors. HIV-positive hemophiliacs
treated with very high purity factor VIII showed no T-cell
decline over 96 weeks. Others who received regular
concentrate showed a sharp decline over the same
period. [8]
In a letter to Science, Duesberg suggested a much
better group to study [10]. The problem with almost all
HIV cohorts since 1984 is that the "positive" arm has consisted largely of people who are already immunocompromised. On the other hand, HIV-positive but presumably healthy subjects are found by the U.S. Army, which
tests more than 2.5 million recruits per year. Blood banks
test more than 12 million blood donors a year.
Cohorts of 3650 men (HIV positive and -negative)
would make the test mathematically simple, Duesberg
wrote [9]. If the 3650-day latent period of AIDS is correct,
every two days one of the HIV-positives, on average,
should develop AIDS; but not the HIV-negatives.
Duesberg predicts the incidence of AIDS-defining diseases in the two groups would remain the same. The test
has not been done. In the fiscal year soon ending, the
Department of Defense will have spent $127 million on
AIDS programs [11 ]-enough, one would have thought,
to do an experiment that would resolve this controversy
one way or the other.
References
[1] J. Cohen, Science, 9 Dec 1994
[2] J. Cohen, Science, 16 Apr 1993
[3] R. Strohman, Science, 13 Jan 1995
[4] HIV/AIDS Surveillance Report v 6 No 1
[5] J. Sullivan et al, J. Pediatr, 1986, 504-510
[6] J. Goedert et al, NEJM Oct 26 1989, 1141-1148
[7] Kreiss et al, Am J. Med 1986, 345-350
[8] de Biasi et al, Blood, 1991, 1919-1922
[9] P Duesberg, Science, 20 Jan 1995
[10] RH. Duesberg, Science, 20 Jan 1995
[11 ] White House Office on AIDS, 6 Feb 1995
Simon Wain-Hobson on
Montagnier, AIDS and HIV
By Tom Bethell
In the vast poster hall of the Sheraton Hotel, literally hundreds of retrovirologists were milling about on the opening day of the American Society of Microbiology
Conference. Simon Wain-Hobson, director of the molecular retrovirology lab at the Pasteur Institute in Paris stood
in the midst of a cluster of listeners. As the author of a
commentary for Nature ("Virological Mayhem") about
the Ho and Show papers. he had been much quoted in
the press.
Identifying myself as a journalist. I asked him if Dr. Luc
Montagnier still believes that HIV needs a co-factor to

cause AIDS. Wain-Hobson, who looked to be in his thirties, is British. He told me I should probably ask
Montagnier himself. "I get the impression he's sort of...