SAN FRANCISCO—“Kaposi’s sarcoma (KS) conforms very poorly to conventional notions about cancer,” Donald Ganem, MD, of the University of California, San Francisco, said at a conference on globally emerging viral infections. “It’s properly classified as in the gray zone between proliferative hyperplasia and frank neoplasm.”

Histologically, unlike typical cancers derived from the clonal outgrowth of a single cell, KS tumors have three phases: a proliferative component of mainly spindle-shaped cells probably derived from endothelium, an inflammatory component made up of many different cells of lymphoid and other origins, and a highly aberrant angiogenic component of slit-like neovascular spaces that give the lesion its red-purple color.

“Epidemiology started us thinking that a virus is involved in KS,” Dr. Ganem said. Although KS became prevalent in the West in conjunction with AIDS, it was evidently not caused by HIV, as it existed in Africa in HIV-negative people.

The search for agents or cofactors led to the discovery in 1994 of a virus named KS-associated herpes virus (KSHV), or human herpes virus 8 (HHV8), that is present in KS tumors. Using an immunofluorescence test for one KSHV component, latency-associated nuclear antigen (LANA), the prevalence of KSHV was found to be 1% to 2% in healthy US blood donors, 30% to 35% in HIV-positive homosexual men, 3% to 4% in HIV-positive women, and 2% to 3% in HIV-positive hemophiliacs.

A more sensitive commercial test that used immunofluorescence on the lytic cycle showed seropositivity of 5% to 6% among Western blood donors, 50% to 60% in HIV-positive gay men, and 95% to 96% in KS patients.

Asymmetry in Sexual Transmission

The asymmetry in sexual transmission of KSHV, shown by the much higher seropositivity in gay men than in women, is unique among sexually transmitted diseases, he said. Because the viral genome is recovered from semen in less than 1% of HIV-positive men with KSHV but is highly prevalent in saliva, oral-anal contact, which is more prevalent in gay men, may explain this disparity.

In Africa and the Mediterranean, KS is endemic, and HIV-negative KS is 10 to 20 times more common than in the US. Seroprevalence of KSHV in Africa is 60% to 80% of the general population: Prepubertal children are often seropositive, and prevalence increases with increasing age. That a very important component of infection in Africa occurs before the onset of sexual activity indicates nonsexual modes of transmission in the first decade of life and thereafter, as well as sexual modes later on.

This difference from the largely sexual transmission in the United States is not understood. Nonetheless, he said, “in every culture around the globe, there is a very strong linkage between seropositivity for KSHV and the incidence of KS.”

At the individual level, seropositivity for KSHV increases the risk of KS in those with HIV. Studies of young healthy gay men showed that all of the KS cases occurred in HIV-positive men and none in HIV-negative individuals. Among those seropositive for both HIV and KSHV at entry to the studies, after 10 years, 40% to 50% developed KS, a three- to fourfold higher rate than among men initially seronegative for KSHV.

“If you are infected with both viruses and get no effective treatment for either virus,” Dr. Ganem said, “after 10 years you have a 50% risk of developing KS. That is strong evidence that infection with KSHV both antedates the neoplasm and is predictive of an increased risk of KS.”

The weight of this evidence, Dr. Ganem concluded, “supports the notion that KSHV is necessary but not sufficient for KS development; you also need HIV or some other predisposing condition.”

The contribution of HIV may be only to create an immunodeficiency state, allowing progression to tumor, a view supported by the fact that cyclosporine use in renal transplants is also linked with the development of KS.

It is also possible that HIV makes some specific genetic contribution to the lesion, since in vitro studies have shown that products from HIV-infected cells can act as a growth factor in cell cultures derived from the KS tumor. “At this moment, school is out on exactly what is the contribution of HIV to KS,” Dr. Ganem said, “but in my view HIV-induced immunodeficiency may cause not the decline of immune surveillance against tumor cells but rather the loss of control of the lytic infection by KSHV.”

There are homologs between KSHV and other herpesviruses, but there are also genes unique to KSHV, many of which have distant homologs in host cell genomes that regulate some aspects of immune function. These homologs may play a role in the disease.

“Pathogenesis is going to be more complicated than in other tumor viruses because the neoplasm is not a traditional tumor,” Dr. Ganem said, “but if we break down the viral gene expression program into its proliferative, inflammatory, and angiogenic components, we may discover what each contributes to the lesion.”