To: Members of the human retrovirus subcommittee of the Retrovirus Study Group
From: Harold Varmus
I have now received results of our straw poll from all but three members of our group, two of whom have already taken strong
positions in favor of certain names. On the whole, the clearest messages to emerge from the poll are that no single name among
those proposed is entirely satisfactory, and that we are all better at finding reasons to oppose a name than at defending
one (and proselytizing in its favor).
Despite the anticipated difficulty of locating a consensus, I was pleased to find among the responses a strong inclination
to reach some agreement and to implement it through the editors of journals and other means. Since there is considerable skepticism
(which I share) about the fruitfulness of a meeting (the funds for which are not readily available in any case), I have decided
to air my own response to the opinions I have received thus far, in hopes that the proposal I shall make either will be greeted
with general agreement or will prompt a yet better solution to our impasse.
As I see the problem at this stage, the names proposed for the AIDS-associated viruses can be grouped in four categories:
(i) those identical or nearly identical to names originally proposed by investigators who have isolated strains of the viruses
in question (some of these names could also be placed in subsequent categories, but they are distinguished by the political
implications of their use, implications which obviously impede acceptance in certain quarters); (ii) names that make use of
the disease states (AIDS and lymphadenopathy syndromes) with which the viruses are associated; (iii) names that invoke some
aspect of pathophysiology, referring either to cell tropism or the effect of these viruses upon immunocyte function; and (iv)
names intended to neutralize the political and scientific problems posed by the names in other categories by assigning numbers
to each group of human retrovirus in some pre-determined (e.g. historical) sequence.
Category (i). Even without responses from some of the more ardent partisans, it is apparent that it would be most difficult
to achieve a consensus for the names HTLV-III, LAV, and ARV. HTLV-III has raised considerably more opposition than the others
(over half of those polled adamantly oppose its use) because of what most of us judge to be clear and fundamental differences
between the AIDS-associated viruses and HTLV-1 and -2. Thus, despite the wide use of the name HTLV-III in the press and among
clinicians and patients, there seems to be substantial opinion within our group that the name is scientifically inappropriate
and that it is not too late to implement a name that will serve us all better in the future. I concur with these views. Opposition
to LAV and ARV is partly political, partly based upon concerns about use of pathological terms, and partly concerned with
maintaining uniformity in retroviral nomenclature (e.g. the name should begin with a term for the host). In my opinion, neither
of these names is good enough to warrant satisfying one political faction after disappointing another.
Category (ii). Two sorts of serious objection have been voiced against names that incorporate the names of disease states,
in particular AIDS. First, we have received the clinical opinion (not, of course, shared by all) that names that include the
term AIDS, especially those such as human AIDS virus or human AIDS retrovirus, will prove troubling in the contexts of medical
care and public health policy. I am dubious about such contentions for several reasons, among them the many precedents the
public tolerates (e.g. polio and hepatitis B viruses) and the "evanescence of euphemism" noted by John Coffin. However,
the opinions on this matter are of sufficient weight to encourage selection of a name that at least diminishes the implication
that infection with the virus is tantamount to the appearance of AIDS. The second major objection is that no name is likely
to cover the full clinical spectrum induced by the virus, so that associating the virus with AIDS or lymphadenopathy syndromes
will ultimately prove to be misleading and insufficient. Although there is merit in this argument, I do not find it strong
enough to reject candidate names; we live happily with the names of many viruses that are frequently encountered in clinical
situations other than those for which they are named.
Category (iii). Names in this group are potentially satisfying on theoretical grounds, but there is also considerable danger
of generating names that will later prove to be highly inappropriate or insufficiently specific. For example, I am attracted
to some of the proposed names that evoke the immunodepressive functions of the AIDS viruses (e.g. HIDV and HISV), but I am
bothered by the lack of specificity: several other kinds of viruses have suppressive actions on the immune system, though
only one kind has been rigorously associated with AIDS. If we were to break accepted convention and include the term retrovirus
in the name (e.g. human immunodeficiency retrovirus or HIRV), matters would be improved in some respects but still confused
by the immunodepressive properties of HTLV-1 (pointed out by one of our members) or by similar properties of any subsequently
identified human retroviruses. Moreover, the situation could then become very confused if the new virus differed as much from
the AIDS viruses as the AIDS viruses differ from HTLV-1. (At this point, such confusion is unlikely to arise from the use
of names based upon diseases [Category (ii)]; the viruses we recognize as the probable causes of AIDS are sufficiently established
that the isolation of another apparent etiological agent from patients clinically similar to AIDS patients would be an occasion
for defining a new clinical entity. In this vein, it is interesting to note that the most recent issue of the MMWR [copy attached]
provides new guidelines for the diagnosis of AIDS that include serological or virological evidence for infection with "HTLV-III/LAV".)
The use of the terms lymphotropic or T cell lymphotropic within names in Category (iii) raises similar problems of specificity,
as well as the further problem of accuracy. Many different viruses are, of course, lymphotropic and even T cell tropic. Moreover,
these terms tend to imply tropism exclusively for the cell type named, yet recent work indicates that the AIDS viruses (like
the T cell leukemia viruses) are able to infect other cells (e.g., neural cells) and perhaps cause disease at other sites.
I foresee future confusion and regret about the choice of such names.
Category (iv). At least three of our members favor taking refuge in this category, away from the political and theoretical
turmoil of the preceding categories. There is undeniable appeal in this maneuver: a system of sequential numbering of human
retroviruses appears at least superficially rational, and similar systems are alleged to work well for other types of viruses.
There are, however, some serious problems. First, the very neutrality of these names makes them difficult to remember. As
long as we are dealing with only HRV-1,-2, and -3 there will be no problem, but remembering the characteristics of twenty
different viruses named only by number would be a challenge. Second, institution of such a system to name the AIDS-associated
viruses would create problems in other sectors. The widely-promulgated and agreed upon names for human T cell leukemia viruses-1
and -2 would have to be changed, and we would have to consider the proper place for the human foamy viruses. (A possible alternative
would be to provide sequential numbers for human oncoviruses, spumiviruses, and lentiviruses; however, in my opinion, the
definition of a lentivirus is still not sufficiently secure to ensure correct categorization of any member other than visna.
And one of our respondents informs me that the alleged homology of the AIDS viruses to visna is due only to short tracts of
adenylic acid in both genomes.) A third problem concerns abuse of the numbering system by assignment of what is intended to
be a species number to new, but not novel, isolates. This is likely to happen in part (as one committee member notes) because
such numbering systems are commonly used to identify strains within a viral species. Adding new numbers within an existing
convention is much easier than gaining acceptance for an entirely new name, and the result is likely to be chaos unless someone
wishes to establish a clearinghouse for assignment of numbers. (I doubt many of us would volunteer for that task.) In sum,
I am not persuaded that our current difficulty is best resolved by a solution that runs counter to long established and generally
successful traditions in retrovirology -- in large part because I think this solution has many problems of its own.
Where does this generally negative discourse leave us? In my review of the problem, it seems to me that, despite objections
to all categories, the objections to the names in Category (ii) are the least troublesome. As a result, I have become a belated
proponent of the name human AIDS-lymphadenopathy virus (or human AIDS-lymphadenopathy associated virus), both abbreviated
HALV. Although you should note the distinction between this name and that proposed recently in Nature by Flossie Wong-Staal
(hers is human AIDS/lymphotropic virus and hence belongs to both Categories (ii) and (iii)), the shared abbreviation has the
advantages of being novel, pronounceable, and sufficiently similar to existing names for AIDS viruses to be quickly recognized
and politically inoffensive. To those who worry about the inclusion of an acronym within an acronym, I would plead for immunity
to pedantry. To the more serious concern about patient and public response to the name, I would point out that the name only
states what everyone knows (i.e. that the virus is associated with AIDS) and that it emphasizes as well the diversity of associated
syndromes (e.g. lymphadenopathy and, by implication or medical explication, less disease or none). Furthermore, I would conjecture
that if the name were accepted, it would soon be known and referred to as "HALV", without great attention to the full
particulars. I don't foresee much difference in the clinic between describing the consequences of a positive antibody
test for HALV or for HTLV-III.
I invite all of you to respond at length to this memo -- both to the analysis of the problem and to my recommendation. But
to insure a swift sampling of opinion, I would again ask that you (also) fill in a brief poll on the following sheet and return
it to me promptly.