This review by Dr. Aboulafia presents aspects of the epidemiology of
acquired immunodeficiency (AIDS)-related lymphomas and their
pathogenesis. The authors main focus is in the molecular area,
and the summary of epidemiology is well known to those interested in
The article contains a few minor mistakes. For example, it was not
involvement of the central nervous system, but rather, the high grade
and distribution to extranodal sites, including the brain, that first
drew attention to lymphomas in the AIDS setting.
Perhaps of more substance, the estimate that 10% to 20% of AIDS
patients ultimately develop lymphomas is exaggerated; 5% to 6% is
more likely, based on current data.[1,2] This figure could change if
large numbers of HIV-infected persons were to live for years with
partial immunodysfunction because of new, highly active therapies to
control human immunodeficiency virus (HIV). Presumably, if therapies
were so successful that the immune system were completely restored,
this population would not be at excess risk of developing lymphoma.
Role of c-myc
The author then discusses selected recent reports about lymphoma
clonality, the roles of Epstein-Barr virus (EBV) and human
herpesvirus-8 (HHV-8), the function of oncogene translocation, and
the place of HIV-related immuno-suppression. The role of c-myc
is a central focus of this discussion.
Several points made in the authors presentation can be
questioned. For example, the translocation that places c-myc
adjacent to the immunoglobin switch gene areas probably occurs before
mutations of c-myc, rather than after these mutations, as the
authors phrasing suggests. Furthermore, although Pelicci et al
find c-myc rearrangements only after lymphoma develops, their finding
conflicts with reports[3,4] not cited in this review. These other
reports indicate that c-myc translocations are not infrequent
in the peripheral blood mononuclear cells of HIV-infected (or even
uninfected) homosexual men, and that the risk of lymphoma is not
significantly higher in individuals with detectable c-myc
(15%) than in those without this gene rearrangement (12%).
A Thesis--Clonality is Central to Understanding Pathogenesis
A review of a field of research is useful for those who seek a
synthesis of recent reports in that area. Accurate facts are
essential, and one role of such a review is to present and critique
the published observations to ensure that the presented results are
valid and to compare the consistency among studies. Facts, even if
accurate, do not provide insights into the nature of the problem,
however. Another role of a review is to synthesize these facts into a
thesis--right or wrong--in order to stimulate responses that can move
the field forward.
I propose such a thesis: Clonality is central to understanding the
pathogenesis of lymphomas and other cancers. Underlying the emergence
of a malignant clone is proliferation. The clonality of lymphomas is
well accepted, even though an occasional tumor can be demonstrated to
have more than a single clone. The contention, described in this
review, that the frequency of clonality in East and West Coast
non-Hodgkins lymphomas differs must be a matter of definition,
assuming that the quality of the work in both areas is adequate. For
me, true polyclonal expansions are proliferations, not cancer, even
when they can kill, as in the X-linked fatal EBV syndrome.
Distinguishing Between Proliferation and Cancer
This distinction between proliferation and cancer is essential
because it permits a differentiation of the mechanisms for each and
probably has practical applications as well. Proliferations are
physiologically mediated expansions of cells that are potentially
reversible. Because proliferation is a "normal" process,
studies of it will lead to reproducible results.
Furthermore, I hypothesize that proliferations increase cancer risk
through introducing genetic errors--a point of view that I think Dr.
Aboulafia would accept. If so, then control of proliferation will
reduce cancer risk. Since some proliferation is essential, it can
never be eliminated entirely, but it can be minimized by the
prevention of avoidable proliferation. Thus, understanding
proliferation may yield crucial insights into practical means to
prevent at least some cancers.
Cancer, in contrast, is beyond physiologic control, except perhaps in
the very earliest phases. Although clonal, cancer is an unstable
entity. In normal cells, there must be constellations of events
controlling the checkpoints of replication, and these will differ by
cell type and even within the same cell type, by the level of
differentiation. In cancer, by definition, these differing
checkpoints must have been bypassed. This can happen when genes that
inhibit replication are mutated or deleted (eg, TP53, bcl-6)
or when genes that enhance replication are turned on (eg,c-myc
translocated to the immunoglobinswitch gene areas) or amplified (eg,
No doubt, we are only just beginning to understand how checkpoints
are bypassed. Even within a cancer from the same clone in a patient,
however, early tumors may have different cellular mechanisms, since
later cancer cells will accumulate genomic changes that permit more
efficient means of replication, losing inhibitors and augmenting
genes that increase growth potential. This hypothesis, for which
there are already some data at both the chromosomal and gene
levels, can be examined in serial samples from cancers in individual
patients. These samples should show progressive simplification, with
genes that inhibit tumor growth being lost and genes that enhance
growth being duplicated (eg, by trisomies).
Thus, cancers, even of the same cell type, should not be expected to
have identical genetic make-up. This hypothesis could be tested by
examining the same tumors (including stratification by stage and
site) to determine whether they have the same basic genetic
construct. Analyzing cancer cells may be of limited value in
developing preventive strategies. This is not to imply that such
research is unimportant. It is essential to gaining an understanding
of the basics of cell replication because inhibition of the
mechanisms will limit replication of already existing cancer cells.
However, the insights gained from this understanding can then be
applied appropriately to cancer chemotherapy, not prevention.
Resolving Some of the Confusion in This Field
Separating proliferation and cancer biology may resolve some of the
confusion that currently attends this field. All viruses associated
with cancers, for example, are proliferative influences. The human
immunodeficiency virus acts indirectly, not being present in the
cancers that occur in AIDS patients. However, even those viruses
implicated as being directly involved appear to be acting as
proliferative stimuli, because all infected cells (and there are
millions if not billions of them) harbor the virus, but only a single
cell emerges as a cancer.
Even c-myc translocations do not, themselves, cause cancer, as
they can be present in persons without cancer and also may not be
present in every cancer in which they are expected. According to this
hypothesis, c-myc translocations would not be expected to be present
in every cell and could be present in cells that have not yet escaped
regulatory control. Furthermore, there is no single point at which
cells become malignant, and even clonal expansions in their earliest
stages can be reversed if the proliferating stimulus is removed.
Is this thesis correct? At least it clarifies why there are divergent
findings in the studies presented by Dr. Aboulafia, and supplements
his report by offering a vision that may challenge readers to either
confirm or refute the concepts. That is what science is about.
1. Biggar RJ, Rosenberg PS, Cote TR: Kaposis sarcoma and
non-Hodgkins lymphoma following the diagnosis of AIDS: The
Multistate AIDS/Cancer Match Study Group. Int J Cancer 68:754-758, 1996.
2. Cote TR, Biggar RJ, Rosenberg PS, et al: Non-Hodgkins
lymphoma among people with AIDS: Incidence, presentation, and public
health burden. Int J Cancer 73:645-650, 1997.
3. Müller JR, Janz S, Goedert JJ, et al: Persistence of
immunoglobulin heavy chain/c-myc recombination-positive lymphocyte
clones in the blood of human immunodeficiency virus-infected
homosexual men. Proc Natl Acad Sci USA 92:6577-6581, 1995.
4. Rabkin CS, Müller J, Goedert JJ: Residual clones in childhood
leukemia (letter). N Engl J Med 33750-33751, 1997.
5. Seemayer TA, Gross TG, Egeler RM, et al: X-linked
lymphoproliferative disease: Twenty five years after the discovery.
Pediatr Res 38:471-478, 1995.
6. Orr-Weaver T, Weinberg RA: A checkpoint on the road to cancer.
Nature 392:223-224, 1998.
7. Biggar RJ, Lee EG, Nkrumah FK, et al: Direct cytogenic studies by
needle aspiration of Burkitts lymphoma in Ghana, West Africa. J
Natl Cancer Inst 67:769-776, 1981.
8. Gutierrez MI, Bhatia K, Cherney B, et al: Intraclonal molecular
heterogeneity suggests a hierarchy of pathogenetic events in
Burkitts lymphoma. Ann Oncol 8:987-994, 1997.