Dermatologic Manifestations of Kaposi Sarcoma

Updated: Feb 14, 2017
  • Author: Robert A Schwartz, MD, MPH; Chief Editor: Dirk M Elston, MD  more...
  • Print


In 1872, Moritz Kaposi (1837-1902) of Kaposvar, Hungary, a dermatology faculty member at the University of Vienna, first described idiopathisches multiples Pigmentsarkom der Haut, which has become known as Kaposi sarcoma (KS). Kaposi sarcoma had brownish red–to–bluish red cutaneous nodules that tended to enlarge into dome-shaped tumors. Kaposi observed similar neoplasms of the mucosa, especially of the larynx, trachea, stomach, liver, and colon. Kaposi's original 1872 description of 5 patients is more similar to the Kaposi sarcoma seen in AIDS (KS-AIDS) than the Kaposi sarcoma expected in elderly men of Italian, Jewish, or Mediterranean linkage, in whom the disease behavior is benign. Kaposi's original 5 patients died within 2-3 years. Kaposi later updated his data, noting that all of his 16 patients were men with a prognosis that remained unfavorable.

In 1882, Tommaso De Amici, Professor and Head, Dermatology, University of Naples, Italy published in monograph form a detailed analysis of 12 patients with Kaposi sarcoma. [1, 2]

For most of the first 3 quarters of the 20th century, Kaposi sarcoma was viewed as an indolent slowly growing cancer, and patients were expected to die with, rather than of, Kaposi sarcoma. The aggressive course originally noted by Kaposi has become part of the devastation of AIDS, especially among men who are homosexual. [3, 4]

In 1981, KS-AIDS in America was identified in 3 reports of Kaposi sarcoma as an original defining element of what later became known as AIDS (plus an important editorial and a Centers for Disease Control and Prevention Morbidity and Mortality Weekly Report bulletin). These original descriptions were by Borkovic and Schwartz in San Francisco, Friedman-Kien from New York City, [5] and Gottlieb and associates in Los Angeles. [6] For some time, Kaposi sarcoma was seen in 30-40% of patients with AIDS, often as the presenting sign. The incidence of Kaposi sarcoma has fallen markedly in recent times, although its prevalence has not. The challenge remained to explain the reason patients who are homosexual and have AIDS exhibited Kaposi sarcoma much more commonly than did patients with AIDS unassociated with homosexuality, with the exception of small foci of homosexuals in isolated midwestern communities.

The breakthrough came in 1994, when the Kaposi sarcoma–associated herpes virus (human herpesvirus type 8 [HHV-8]) was identified using representational difference analysis. HHV-8 has been linked closely with all 4 types of Kaposi sarcoma, ie, classic (traditional), endemic (African), epidemic (AIDS related), and iatrogenic (related to immunosuppression). [7] Since then, much research has shown that HHV-8 appears to be necessary to, but not sufficient for, the development of Kaposi sarcoma. [8]

Nevertheless, two critical questions remain. Is Kaposi sarcoma a hyperplasia or a neoplasm? Is it always multicentric or can it be metastatic as well? The authors favor the latter interpretation of both points.

Kaposi sarcoma and its causative agent, Karposi sarcoma–associated herpesvirus, have distinctive largely unexplained geographic distributions. [9] The "oncoweed" hypothesis states that biologic plants in the environment account for the lytic reactivation of human herpesvirus 8. Quinine and its derivatives might better explain the epidemiology of Kaposi sarcoma in Africa than oncoweeds. Thus, an "oncodrug" hypothesis has been proposed, specifically with regard to quinine and its derivatives for African Kaposi sarcoma.

See the image below.

Elderly American man of Armenian origin with chara Elderly American man of Armenian origin with characteristic violaceous plaques of the legs, a good example of classic Kaposi sarcoma.


HIV transactivating (tat) gene, cytokine, and HHV-8 stories in Kaposi sarcoma (KS) are fascinating. Each begins with a classic study. In 1988, the human immunodeficiency virus type 1 (HIV-1) tat gene was introduced into transgenic mice, inducing nodules that resembled Kaposi sarcoma in 33 of 37 males but in none of 15 females. Therefore, it appeared that HIV could play a direct role in causing Kaposi sarcoma. The medication quinine may be a potentially overlooked triggering factor in millions of Africans with Kaposi sarcoma. [10]

The second saga was a result of efforts to grow Kaposi sarcoma cells in culture, requiring a long-term growth factor. Conditioned medium from T cells infected with human T-cell leukemia virus type II (rather than HIV-1 or human T-cell leukemia virus type I) best supported the growth and long-term culture of Kaposi sarcoma cells derived from KS-AIDS lesions. In 1992, this growth factor proved to be a cytokine previously termed oncostatin M, since it had been identified earlier for its inhibitory effects on a variety of cancer cells. Another cytokine scatter factor was found in large quantities in this medium, inducing endothelial cells to demonstrate a Kaposi sarcoma tumor cell-like phenotype. The importance of oncostatin M, scatter factor, and the tat protein has been shown in the pathogenesis of Kaposi sarcoma.

Other cytokines, including interleukin 1 (IL-1), tumor necrosis factor, interleukin 6 (IL-6), and basic fibroblastic growth factor (bFGF), may work synergistically with the HIV tat gene product. Scatter factor may be involved both in initiation and in maintenance of Kaposi sarcoma. Scatter factor stimulates endothelial cells to migrate nearby and become factor XIIIa–positive c-Met- expressing spindle-shaped Kaposi sarcoma cells. The cells further expand neovascularization by producing cytokines and promoting autocrine-mediated and paracrine-mediated growth of Kaposi sarcoma cells. The scatter factor receptor, c-Met proto-oncogene, is expressed by Kaposi sarcoma cells; the oncogene int-2 of the fibroblast growth factor family also may be evident.

Herpes-type viruses have been linked with Kaposi sarcoma for more than 3 decades. A landmark study showed short DNA sequences of a unique human herpesvirus in Kaposi sarcoma tissues via a new molecular biological technique termed representational difference analysis. They resembled herpesvirus saimiri but proved to be a new type of human herpesvirus now termed HHV-8. This virus appears to interact with the HIV tat protein, excess levels of basic fibroblast growth factor, scatter factor, and IL-6. For example, HHV-8–encoded IL-6 has been found to induce endogenous human IL-6 secretion. An HHV-8 oncogene, Kaposin (ORF K12), has been characterized; however, additional factors remain to be found. For example, a 53% prevalence of HHV-8 subtype E in Brazilian Indians does not appear to be linked with the development of Kaposi sarcoma in this population. [11]

Classic Kaposi sarcoma is seen in Italy with hot spots being in the Po River Valley, Sardinia, and southern Italy. It has been suggested that volcanic soil or birthplace/residency in areas abundant with bloodsucking insects may be a risk factor. [12] A survey evaluated the correlation between HHV-8 infection and classic Kaposi sarcoma incidence in northern Sardinia. [13] It revealed that seroprevalence was 35%, within a range of 15.3-46.3% in the five areas. Age was as an important risk factor. Subjects aged older than 50 years had a higher seroprevalence to HHV-8 as compared with younger individuals. A strong direct correlation between HHV-8 prevalence and classic Kaposi sarcoma incidence was also observed.

Kaposi sarcoma–associated herpesvirus (KSHV), or human herpesvirus 8 (HHV-8), is the most frequent cause of malignancy in patients with AIDS. [14] KSHV and related herpesviruses have pirated cellular cDNAs from the host genome. Many of the viral regulatory homologs encode proteins that directly inhibit host adaptive and innate immunity. Other viral proteins may target retinoblastoma protein and p53 control of tumor suppressor pathways, which play key effector roles in intracellular immune responses. The immune evasion strategies used by KSHV in targeting tumor suppressor pathways activated during immune system signaling, may lead to inadvertent cell proliferation and tumorigenesis in susceptible hosts.

HHV-8 variants have been found heterogeneously distributed in diverse geographic regions, but their pathogenicity in Kaposi sarcoma development is unknown. [15] The frequency of KSHV genotypes has been evaluated in numerous settings. [16, 17] The frequency of these genotypes isolated from Kaposi sarcoma lesions among 50 patients from one center in Brazil was evaluated. The most frequently detected viral genotypes were A (50%) and C (48%); the B genotype was isolated only in one case. [18] A genotype was predominant in those HIV-positive, whereas the C genotype was mostly in those in the HIV-negative group. An association between E genotype and Kaposi sarcoma development was observed in Peru.

The origin of the spindle cell, the hallmark cell of Kaposi sarcoma, is unknown. Most research favors a lymphatic endothelial cell or an endothelial-cell precursor evolving into a lymphatic phenotype, both preferentially targeted by KSHV. [19]




United States

The incidence of Kaposi sarcoma (KS) has been estimated at 0.02-0.06%. Traditional Kaposi sarcoma of middle-aged and older American men of Mediterranean and Eastern European (Ashkenazi) Jewish lineage represents approximately 0.2% of cancer cases in the United States. Iatrogenic Kaposi sarcoma has a 400% increased incidence compared to the American population at large. The incidence of Kaposi sarcoma among American renal transplant recipients is approximately 0.4%. In the AIDS grouping, Kaposi sarcoma originally accounted for as many as 35% of patients, an incidence that has been declining with early detection of AIDS, although Kaposi sarcoma prevalence may remain high. In the United States, a few emigrants arrive from Kaposi sarcoma–endemic regions (primarily Africa). The largest immigrant population in this category may be Haitians; however, recent evidence suggests that Haitian AIDS originated directly in New York rather than Africa.


Four groups are predisposed to Kaposi sarcoma including (1) older men of Mediterranean and Jewish lineage; (2) Africans from areas including Uganda, the Congo Republic, Congo (Brazzaville), and Zambia; (3) persons who are iatrogenically immunosuppressed; and (4) men who are homosexual. Kaposi sarcoma traditionally is an uncommon disease in middle-aged and elderly European men of Mediterranean or Jewish lineage.

A similar focus of Kaposi sarcoma exists in the same age and sex groups in Africa. If a crudely calculated incidence of 28 cases of Kaposi sarcoma per 100,000 in the Arabian population is correct, Kaposi sarcoma may be more common among Arabians than among Mediterranean people. A previously unrecognized genetic predisposition for Kaposi sarcoma among Arabians has been suggested. The incidence of Kaposi sarcoma among renal transplant recipients may be as high as 3.5% or higher in regions endemic for Kaposi sarcoma, which is significantly higher than the 0.4% incidence renal transplant recipients in the United States and Western Europe.

Endemic African Kaposi sarcoma has accounted for 10% of cancers and has been seen in a male-to-female ratio of 15:1. The Kampala Cancer Registry, one of the continent's first and foremost, has shown a significant alteration in the incidence of Kaposi sarcoma in the era of AIDS. In Uganda, Kaposi sarcoma has caused almost one half (48.9%) of cancer cases in men and 17.9% in women. The incidence in men (30.1 cases per 100,000) represents a more than 10-fold increase in men since the 1950s and is approximately 3 times the incidence found in women (11 cases per 100,000). The incidence in boys and girls was approximately the same in childhood (birth to 14 y), with small peaks in girls younger than 5 years and boys aged 5-9 years. Subsequently, a progressive rise in incidence peaked in women aged 25-29 years and in men aged 35-39 years. Lymphadenopathic Kaposi sarcoma affected 12% of total cases; 42% of childhood cases were of this type.

Of 73 Ugandan children with epidemic Kaposi sarcoma from the Uganda Cancer Institute in Kampala (from 2004-2007), 37 were boys and 36 were girls, with a median age of 10.1 years (range 2-18 y). [20]

In neighboring Zambia, the disorder was particularly aggressive among children, more than 80% of whom were HIV seropositive. Kaposi sarcoma was found to represent as much as 25% of childhood cancers. The average male-to-female ratio was 1.76:1, with male predominance higher in children older than 5 years (2.5:1 ratio) than in children younger than 5 years (1.4:1 ratio). The prevalence of HIV-related Kaposi sarcoma seems to be increasing in Nigeria, probably owing to more females having HIV disease. [21]

In Italy, KS-AIDS has produced notable epidemiologic changes. A doubling of Kaposi sarcoma incidence rates was noted in Italian men younger than 50 years from 1976-1984 to 1985-1990; however, no change, or possibly a decline, was observed in older men. The incidence of Kaposi sarcoma was estimated in the region around Venice, Italy, with rates higher in the coast and alpine valleys; in the latter, there was an excess of cases for both sexes combined (SIR = 191.1; CI = 113.2-302.0). [12] The standardized incidence rate for Kaposi sarcoma in Italy declined markedly in the highly active antiretroviral treatment (HAART) period. [22]

Classic Kaposi sarcoma in Greece seems to have an older age of onset; lower male-to-female ratio; endemic clustering; and disseminated skin disease at diagnosis, often accompanied by lymphedema and not unusual visceral or lymph node involvement. [23, 24, 25] A clustering was noted, with a high proportion of the patients being born in Peloponnesos (42.42%) and residing in Athens (51.51%) or in Peloponnesos (24.24%).

The incidence rates of classic Kaposi sarcoma in Italy after the spread of AIDS was evaluated. [13] The rates were 1 case per 100,000 population in men and 0.4 case per 100,000 population in women, varying from 0.3 cases per 100,000 population for men in Umbria and 4.7 cases per 100,000 population for men in Sassari in men and from 0.1 case per 100,000 population for women in Parma and 1.7 cases per 100,000 population for women in Sassari.

The rate of classic Kaposi sarcoma in southern Sardinia (Italy) from 1998-2002 was found to be 2.49 cases per 100 000 population per year (standardized), which was the highest rate recorded in the island. [26]

A Finnish nationwide registry–based analysis of cancer clustering detected a strong familial occurrence of Kaposi sarcoma. This study found Kaposi sarcoma had a very high cluster score, with Kaposi sarcoma patients forming clusters of close relatives. [27] One family was documented with 5 affected individuals in 2 generations. Several other families had 2 first-degree relatives with Kaposi sarcoma.


Kaposi sarcoma is an uncommon disease of middle-aged and elderly American and European men of Mediterranean or Jewish lineage. This propensity also is seen in individuals with iatrogenically induced Kaposi sarcoma but not among persons in the KS-AIDS group. Kaposi sarcoma is rare in American blacks, despite its large foci among blacks in certain regions of Africa.


Classic Kaposi sarcoma has an overwhelming male predominance, with a male-to-female ratio of approximately 10-15:1.

For endemic Kaposi sarcoma in Central Africa, the male-to-female ratio is near unity in childhood Kaposi sarcoma cases but often rises in puberty to 15:1.

In Corsica and Sardinia (where classic Kaposi sarcoma is endemic), with the arrival of AIDS, the ratio of male-to-female cases has dropped from 10:1 to 3:1. Children of women who are HIV-1 seropositive without Kaposi sarcoma have an aggressive form of childhood HIV-associated Kaposi sarcoma. A male-to-female ratio of 1.5:1 was observed.

A male-to-female ratio of 1.5:1 also was evident among renal transplant recipients in Arabia. [28]

In Zambia, Kaposi sarcoma represents up to 25% of childhood cancers and has an average male-to-female ratio of 1.76:1, with male predominance higher in children older than 5 years (2.5:1) than in children younger than 5 years (1.4:1).

Clinical manifestations and therapeutic response in men and women with AIDS-associated KS in Uganda differed between men and women. [29] At presentation, the median CD4 T-cell count was significantly lower in women than men, with women more likely than men to initially develop Kaposi sarcoma of the face and hard palate and more likely to develop lesions on the lower extremity. In addition, women were less likely than men to demonstrate clinical improvement.


Age distribution depends on the type of Kaposi sarcoma.

In the United States and Europe, traditional Kaposi sarcoma has a peak incidence between 40-70 years, with a wide range of up to 89 years.

Young men with KS-AIDS who are homosexual also show a wide age range but tend to be much younger, averaging 20-40 years at age of onset.

For endemic Kaposi sarcoma in Uganda, the incidence in boys and girls was approximately the same in childhood (birth to 14 y), with a small peak in girls younger than 5 years and boys aged 5-9 years. Subsequently, a progressive rise in incidence peaked in women aged 25-29 years and in men aged 35-39 years.



Clinical classification of Kaposi sarcoma may be the best prognosticator, comparing localized nodular disease, locally aggressive disease, and generalized Kaposi sarcoma.

Cutaneous skin testing for anergy may correlate with clinical disease type and prognosis.

Prognosis appears to correlate with the CD4 count.

Localized nodular Kaposi sarcoma has the best prognosis, with few deaths directly attributable to Kaposi sarcoma.

Locally aggressive Kaposi sarcoma has an intermediate prognosis.

The authors do not disagree with an old African estimate of a 3-year survival rate of 64%. Generalized Kaposi sarcoma, the form seen most commonly in patients with KS-AIDS, has a 3-year survival rate closer to 0% without therapy.

The low level of access to antiretroviral drugs in HIV-infected patients explains the morbidity and mortality from AIDS-associated Kaposi sarcoma in much of Africa, including Togo. [30]

Seventy children with Kaposi sarcoma and HIV infection were evaluated in South African hospitals. [31] The mortality was high despite use of antiretrovirals and cytostatics, with 32 patients surviving only 4 months on average; 10 were lost to follow-up, and 28 were alive, with an average follow-up of 16 months.

Significant morbidity and mortality has been seen in those with Kaposi sarcoma IRIS. [32, 33]


Patients with traditional Kaposi sarcoma (KS) tend to die with Kaposi sarcoma rather than of Kaposi sarcoma. Patients with KS-AIDS usually die from associated opportunistic infections or from gastrointestinal Kaposi sarcoma with hemorrhage. The mean survival rate of patients with KS-AIDS has been approximately 15-24 months, although the introduction into the United States of apparent immune system reconstitution using highly active antiretroviral therapy (HAART) has extended survival substantially. Kaposi sarcoma also may be fatal as a result of gut perforation, cardiac tamponade, massive pulmonary obstruction or, rarely, brain metastases.

In Kaposi's original description, death usually ensued within 3 years and was linked to fever, diarrhea, and hemoptysis. Inanition may be an important factor, and death may ensue as a result of bulky tumor obstructing the bronchi or larynx.

Patients with AIDS-related Kaposi sarcoma often have widespread visceral Kaposi sarcoma, although Kaposi sarcoma limited to the skin also is common.

Patients with iatrogenic Kaposi sarcoma tend to have gut bleeding resulting from Kaposi sarcoma, although termination or reduction of immunosuppression often, but not always, results in regression of Kaposi sarcoma.