The 20-year search for an infectious agent associated with Kaposi's
sarcoma (KS) may be over. Researchers at Columbia-Presbyterian
Medical Center in New York have reported significant evidence
directly linking a probable new herpesvirus to KS in AIDS patients.
Dr. Yuan Chang, Assistant Professor of Pathology at Columbia University
College of Physicians & Surgeons, and Dr. Patrick Moore, Assistant
Professor of Public Health at Columbia University School of Public
Health, co-principal investigators of the research, found unique
DNA sequences from what may be a new type of human herpesvirus
in 93% of KS lesions from 27 deceased AIDS patients. The sequences
generally were not found in the patients' non-KS tissues.
The DNA sequences found in the KS lesions are most similar to
the DNA sequences of a group of herpesviruses known as gamma-herpesviruses.
The new virus, which is thought to be sexually transmitted, is
most closely related to the Epstein-Barr virus.
The Columbia researchers' findings open the way to potential improvements
in the diagnosis and treatment of KS patients, and provide one
more example, among only a handful, of a virus being implicated
in cancer causation.
"Ours is the first evidence, as far as I know, of a reliable
molecular marker, DNA sequences, found specifically in KS tissue,"
says Dr. Chang. The investigators also tested the reliability
of their molecular marker by correctly predicting KS in 11 tissue
samples provided to them blindly from the University of Pittsburgh.
At present, these findings do not change the standard treatment
for KS. It is not yet known whether drugs that are currently effective
against herpesviruses would have any effect on this presumed virus.
In the future, however, a diagnostic test or therapeutic agent
could be developed based on the DNA sequence, or from antibodies
raised against the virus.
"If we can prove that this putative virus actually causes
the cancer, diagnostic tests could be developed that might help
clinicians determine sooner who may be at risk for developing
Kaposi's sarcoma," says Dr. Chang.
For some time, epidemiologists have thought that a sexually transmitted
agent might be responsible for the KS disease pattern seen in
gay and bisexual AIDS patients. Gay and bisexual men with AIDS
are 20 times more likely than hemophiliacs to develop KS (Lancet,
335:125, 1990). Women with AIDS also are more likely to develop
KS if their partners are bisexual men rather than intravenous
"While we have no direct evidence that KS is sexually transmitted,
the higher rates among gay and bisexual men with AIDS, compared
to hemophiliacs with AIDS, suggest that it might be sexually transmitted,"
says Dr. Moore. "For now, it is prudent to say that safe
sex practices, besides preventing the spread of HIV, probably
may prevent the transmission of this presumed herpesvirus."
Dr. Moore adds that there may be multiple modes of transmission
of the new virus, and there is no evidence that it can be transmitted
casually. The virus was not found in tissue from 85 individuals
without AIDS or KS, indicating that it is probably not common
to the general population. The researchers currently are determining
whether the viral DNA sequences they found are present in KS lesions
of non-AIDS patients.
Infectious agents suspected of causing KS over the past 20 years
include viruses, such as cytomegalovirus, hepatitis B virus, human
herpesvirus 6, and HIV; and bacteria, such as Mycoplasma penetrans.
Even noninfectious agents, such as nitrite inhalers, or poppers,
a drug used by the gay community, were thought to be a cofactor
for KS. "Extensive research has not demonstrated a causal
relationship between these agents and KS in AIDS," write
Drs. Chang and Moore used the latest tools in molecular biology
analysis-representational difference analysis and the polymerase
chain reaction-to isolate the extraneous viral DNA sequences.
They have identified portions of three separate genes that appear
to belong to this new virus. The sequences also have been found
in lymphoma tissue and lymph nodes from 3 of 39 other AIDS patients.
The role of the sequences in AIDS-associated lymphomas is being
Although the scientists believe that they have strong evidence
implicating the virus as the causative agent of KS in AIDS patients,
they need to do additional work to prove it. "Although this
virus is not found in most non-KS tissue, it is still possible
that this agent is a common latent virus in humans that preferentially
colonizes KS lesions in immunocompromised patients," the
If this presumed herpesvirus is shown to cause KS, it would be
another addition to the list of viruses known to cause cancer.
These include papillomavirus (cervical cancer), hepatitis B (liver
cancer), and Epstein-Barr virus (Burkitt's lymphoma and nasopharyngeal
Drs. Chang and Moore, a wife and husband team, took a year to
identify the sequences with their collaborators. The other investigators
on the project were Drs. Ethel Cesarman and Daniel Knowles, formerly
of Columbia-Presbyterian Medical Center but now with Cornell Medical
Center in New York City; Dr. Melissa Pessin, a postdoctoral fellow
in Columbia's Department of Pathology; and Drs. Frank Lee and
Janice Culpepper of DNAX Research Institute in Palo Alto, California.
This research was funded by start-up funds from Columbia's Department
of Pathology and the School of Public Health. The full report
is published in Science, Vol 266:1803-1804.