Nature
375: 167, 18 May 1995
HIV infection provokes hyperactivity of the immune system, but the causes
of that are far from understood.
The clutch of contributions to Scientific Correspondence (page 193)
this week deserves a reading, both for its inherent interest and for what
it says about the present state of AIDS research. It will be recalled this
journal published in January an account of research that showed that the
infection of a person by the virus HIV ordinarily evokes not the previously
suspected quiescence of the immune system, but a rapid turnover both of
the vulnerable lymphocytes and of the virus itself. The then-general opinion
that the first reaction of the human body to infection by HIV is a kind
of indifference was dramatically and directly challenged. Nothing that
has since come to light denies the challenge. But it has also become plain
that too little is yet known of the dynamics of the immune system. That
is a gap to fill.
The second arresting feature of this correspondence is the letter from
Dr. Peter Duesberg and his colleague, Dr. Harvey Bialy, which has been
published without change. Sadly, there seems no way in which the authors
concerned can be persuaded that "free and fair scientific debate"
is ordinarily understood to mean a progressive process, one in which each
of two sides learns from what the other says. A restatement of earlier
and well-known positions is not that at all. On this occasion, Duesberg
and Bialy's citation of Loveday in their cause is especially inappropriate,
given Loveday's name among the authors of a letter supporting Wei et
al. and Ho et al. But no further solicitation of Duesberg's
opinion is called for.
Letter from Peter Duesberg and Harvey Bialy, Nature 375: 197, 18
May 1995
SIR - In an editorial in the 19 January issue of Nature, John
Maddox invited "Duesberg and his associates" to comment on the
"HIV-1 dynamics" papers published the previous week, indicating
that these new results should prove an embarrassment to us. Although we
do not think that a scientist should be embarrassed for pointing out inconsistencies
and paradoxes in a hypothesis that have only been reportedly resolved 10
years later, we nonetheless prepared a fully referenced, approximately
2,000-word critique of the Ho et al. (2) and Wei et al. (3)
papers that we believed met the criteria of "not being longer than
it needs to be, and pertaining to the papers at hand" that Maddox
set out in his widely read challenge.
Unfortunately, he did not share our view and agreed to publish only
a radically shortened version, and only after he had personally "gone
over it with a fine-tooth comb" to remove our perceived misrepresentations
of the issues. We found these new conditions so totally at variance with
the spirit of free and fair scientific debate that we could not agree to
them.
Readers of Nature who are interested in these questions, and
feel that they do not need to be protected by Maddox from our ill-conceived
logic, can find the complete text of our commentary in the monograph supplement
to the most recent issue of Genetica (4). Here we would point out
only that the central claim of the Ho et al. (2) and Wei et al.
(3) papers-that 105 HIV virions per ml plasma can be detected in AIDS patients
with various nucleic-acid amplification assays is misleading. The senior
author of the Wei et al. paper has previously claimed that the PCR
method they used overestimates by at least 60,000 times the real titer
of infectious HIV (5): 100,000/60,000 is 1.7 infectious HIVs per ml, hardly
the "virological mayhem" alluded to by Wain-Hobson.(6) Further,
Ho and a different group of collaborators have just shown (7) that more
than 10,000 "plasma virions," detected by the branched-DNA amplification
assay used in their Nature paper, correspond to less than one (!)
infectious virus per ml. And infectious units, after all, are the only
clinically relevant criteria for a viral pathogen.
Finally, in view of Wain-Hobson's statement (6) that "the concordance
of their [Wei and Ho's] data is remarkable," note that Loveday et
al. (8) report the use of a PCR-based assay and find only 200 HIV "virion
RNAs" per ml of serum of AIDS patients-1,000 times less than Ho and
Wei. So much for the "remarkable concordance."
Peter Duesberg
Department of Molecular and Cellular Biology, University
of California, Berkeley, California 94720, USA
Harvey Bialy
Bio/Technology, New York, New York 10010, USA
Notes and References
1. Maddox, J. Nature 373, 189 (1995).
2. Ho, D. D. et al. Nature 373, 123-126 (1995).
3. Wei. X. et al. Nature 373, 117-122 (1995).
4. Duesberg, P. & Bialy, H. Genetica Suppl.
(in the press).
5. Piatak, M. et al. Science 259. 1749-1754 (1993).
6. Wain-Hobson, S. Nature 373, 102 (1995).
7. Cao, Y. et al. New Engl. J. Med. 332, 201-208
(1995).
8. Loveday, C. et al. Lancet 345, 820-824 (1995).