The association of Kaposi's sarcoma (KS) with second primary cancers, especially of the lymphoreticular system, has been frequently noted. To confirm these reports in a systematic way, data on a series of 92 patients with KS treated at Memorial Sloan-Kettering Cancer Center (MSKCC) 2949-1975 were subjected to extensive statistical analysis. Use was made also 4517 double primaries including all sites, diagnosed at MSKCC 1949-1974, and 1959 simultaneous double primaries from the Third National Cancer Survey. Some key results: (1) of KS patients had toher primary malignancies; (2) there was a 20-fold increase in the incidence of lymphoreticular malignancies after diagnosis of KS; (3) in the MSKCC series double primaries, lymphoreticular malignancies were involved in 8% of cases; for KS alone the corresponding figure was 58%. Our findings provide evidence in support of possible etiopathogenic mechanisms that may be involved in the natural course of KS.

DNA-DNA reassociation kinetics, anti-complement immunofluorescence (ACIF) and ACIF-blocking tests were used to search for cytomegalovirus (CMV) gene products in Kaposi's sarcoma (DS) biopsies and early cell cultures deriving from them. Three of eight tumor biopsies were positive for CMV DNA; two of them at 0.35 genome/cell and one at one copy 25% genome/cell. CMV-related antigens, mainly localized in the nucleus, were found in cryostat sections of seven of 31 tumor biopsies and four of 12 KS cell lines at early passage level. It appears that the antigen is present in a high number of tumor cells, like the Epstein-Barr virus nuclear antigen (EBNA) in EBV-transformed cells. Inevitably, the increasing data concerning the oncogenic potential of CMV lead on to consideration of the increasing evidence of its association with Kaposi's sarcoma, a multiple hemorrhagic sarcoma with an epidemiologic distribution in Africa similar to that of Burkitt's lymphoma.

No abstract

The prominent finding of this extended serologic analysis on American and African Kaposi's sarcoma (KS) patients and appropriately matched control groups is the detection of a specific serologic association of cytomegalovirus (CMV) with American KS patients. All American KS sera contained CMV antibodies and their geometric mean titers (GMT) were significantly higher than those in sera of melanoma patients (GMT ratio k = 5.3 to 7.7 by complement fixation [CF], k = 8.9 by indirect hemagglutination [IHA]) or in sera of age- and sex-matched healthy controls (k = 12.6 to 16.0 by CF, k = 12.6 by IHA). The result is strongly reminiscent of the data obtained previously for European KS. Although the GMT to CMV of African KS patients were similar to the GMT of the American KS groups, their significance cannot be demonstrated due to the high background of CMV infections in the control groups. Complex mechanisms are hypothesized, by analogy with the Epstein-Barr virus (EBV) involvement in Burkitt's lymphoma (BL), for a CMV involvement in the development of KS.

Sera from patients with Kaposi's sarcoma (KS) were examined for antibody titres to cytomegalovirus (CMV), Epstein-Barr virus (EBV) and herpes simplex virus (HSV) types 1 and 2 by four techniques: indirect haemagglutination (IHA), complement fixation (CF), virus neutralization (NT) and indirect immunofluorescence (IF). The patients were classified, according to the stage of disease, as progressive and regressive. Control sera were obtained from healthy adults, matched for age, sex, race, socioeconomic status and geographic location, as well as from patients with melanoma, some of whom were receiving chemotherapy similar to that given to the KS patients. All KS sera contained CMV-neutralizing antibodies. Seventy-five percent of the European KS patients, mainly regressors, showed elevated anti-CMV titres by IHA with a significant increase in the geometric mean over the corresponding healthy adult group and the melanoma group. An overrepresentation of high anti-CMV titres, although less marked, was found by CF. There was no significant association with antibodies to EBV, HSV-1 and HSV-2 related antigens. By contrast, the African KS patients, mainly progressors, did not show a serologic association with CMV or with EBV and HSV-1 and 2. The implication of these results is discussed.