Ephata E. Kaaya

Expression of HHV-8 latency-associated T0.7 RNA in spindle cells and endothelial cells of AIDS-associated, classical and African Kaposi's sarcoma

Analysis by polymerase chain reaction (PCR) and serological studies have demonstrated a close association between the novel human herpes virus, Kaposis sarcoma‐associated herpes virus (KSHV) or human herpes virus‐8 (HHV‐8) and the development of Kaposis sarcoma (KS). To clarify the role of HHV‐8 in KS pathogenesis, we investigated at the cellular level by in situ hybridization the expression of a recently described 0.7‐kb HHV‐8‐encoded mRNA (T0.7 mRNA) in KS tissues of different epidemiological origin (AIDS‐KS, African endemic KS and classical KS). The T0.7 mRNA likely encodes a small membrane protein, supposedly expressed in latently HHV‐8‐infected cells. Indeed, we detected T0.7 mRNA in virtually all cells of the cell line BCBL‐1 established from a body cavity‐based lymphoma (BCBL) and latently infected with HHV‐8. In all KS biopsies examined, independent of their epidemiological type, the late‐stage (nodular) KS tissues showed a high level of T0.7 mRNA expression in typical KS spindle cells but also in endothelial cells lining blood vessels, indicating latent HHV‐8 infection of these cells. The presence of T0.7‐expressing cells was restricted to KS tumor tissue and therefore appears to indicate an important role of latent HHV‐8 infection in KS pathogenesis. Int. J. Cancer 72:68–71, 1997.

Heterogeneity of spindle cells in Kaposi's sarcoma: comparison of cells in lesions and in culture

The immunophenotype of spindle cells in epidemic, endemic, and classic (sporadic) Kaposis sarcoma (KS) lesions was defined by the demonstration of various cell markers and compared with that of KS-derived cell lines. No significant histological or immunophenotypic differences were observed between the three clinical types of KS at comparable stages. The spindle-cell compartment of the different KS types was composed predominantly of a mixture of proliferating CD45+/CD68+ bone-marrow-derived monocytes and TE7+/collagen+ fibroblastic cells with varying expression of EN4/PAL-E/CD31/CD34/CD36 endothelial-associated antigens and/or smooth-muscle-specific alpha-actin (alpha-actin). The latter cells appeared to represent transitional forms of fibroendothelial and fibromyocytic cells. The in vitro cultured KS-derived cell lines (KS-3, KS-6, and KS-8) expressed the fibroblastic antigen TE7 and smooth-muscle-specific alpha-actin but not leukocytic or endothelial-associated antigens consistent with the phenotype of fibromyoid spindle cells of primary lesions. Neither HIV antigen nor provirus DNA was demonstrable in the epidemic KS lesions. The observed heterogeneity of the spindle-cell compartment further substantiates the view that Kaposis sarcoma, irrespective of clinical setting, expresses salient features more compatible with reactive, tumor-like lesion than clonal sarcoma.

Lymphatic and vascular origin of Kaposi's sarcoma spindle cells during tumor development.

The histogenesis of Kaposis sarcoma (KS) tumor spindle cells (SC) remains controversial but several immunohistochemical studies favor a lymphatic origin. Twenty KS surgical biopsies were analyzed for the coexpression of LANA, CD34, LYVE‐1, D2‐40, VEGFR‐2, VEGFR3 by using double or triple immunostaining. Most of the SC in both early and late KS expressed the lymphatic markers LYVE‐1, D2‐40 and VEGFR‐3 and the blood vascular endothelial/endothelial precursor cell markers CD34 and endothelial stem cell marker VEGFR‐2. All the LANA+ SC in early and late KS were LYVE‐1+, but only 75% of these LANA+ cells were CD34+. The CD34+/LANA+ cells increased from early‐ (68.8%) to late‐stage KS (82.2%). However, approximately 18% of the LANA+ SC in early KS were CD34− but were LYVE‐1+, suggesting that resident lymphatic endothelial cells (LEC) are targeted for primary infection by human herpesvirus‐8. This LANA+/LYVE‐1+/CD34− (resident LEC) cell population clearly decreased during the development of KS from early (18.7%) to late KS (2.9%). Thus, in late stages of KS, most SC were LANA+/CD34+/LYVE‐1+. However, in both early‐ and late‐stage KS, approximately 18% of the SC were CD34+/LANA‐/LYVE‐1− and could represent newly recruited endothelial precursor cells, which become infected in the lesion and eventually undergo a phenotype switch expressing LEC markers. Our study apparently indicates that KS represents a unique variant of tumor growth with continues recruitment of tumor precursor cells as well as proliferation and decreased apoptosis of SC.

Clinical features and predictive markers of disease progression in cynomolgus monkeys experimentally infected with simian immunodeficiency virus.

ObjectiveTo study the pathogenicity of simian immunodeficiency virus (SIVsm) in cynomolgus monkeys in order to establish an animal model for human AIDS. MethodsThirty-three cynomolgus monkeys were monitored for more than 2 years following experimental infection with SIVsm. ResultsAll the macaques became SIV-infected, as demonstrated by virus recovery from peripheral blood lymphocytes and by the appearance of viral antibodies. SIVsm was found to be pathogenic, killing 29 out of the 33 monkeys (88%) within 26 months. Clinically, infected monkeys developed lymphadenopathy, splenomegaly, diarrhoea, weight loss, neurological symptoms and a remarkably high incidence (39%) of malignant lymphomas. All lymphomas were high-grade malignant and of B-cell origin. Disease progression was associated with low CD4+ lymphocyte count, involution of initially hyperplastic follicular B-cell areas in lymph nodes, reappearance of viral antigen in serum, loss of anti-Gag antibodies and development of systemic giant cell disease in 55% of the monkeys. ConclusionsThere are many similarities between SIVsm-induced AIDS in cynomolgus monkeys and human AIDS with regard to clinical, virological, immunological and pathological manifestations.

Kaposi's sarcoma before and during a human immunodeficiency virus epidemic in Tanzanian children

Background. With the onset of AIDS increased frequency of Kaposi’s sarcoma (KS) has been reported. However, there is no case-based comparison of childhood (<14 years) KS before and during the HIV pandemic in sub-Saharan Africa. Here we report on the Tanzanian cancer registry data of pediatric KS in Tanzania and implications with regard to pathogenic factors. Methods. One hundred fifty histologically confirmed pediatric KS (PKS) cases registered during 1968 through 1995 (28 years) were analyzed with regard to demographic and clinical characteristics before and during the AIDS epidemic. Statistical analysis was done with the Epi-Info program and chi square test. Results. Of children with PKS 126 (84%) were male and 24 (16%) were female. The gender ratio was 5.1:1 and 5.4:1 during the endemic and epidemic periods, respectively. The highest occurrence of PKS was observed in the 0- to 5-years age group. Overall 73 (4.9/year) of these cases were registered during the pre and 77 (5.9/year) during the AIDS period. Over time a significant increase in anatomically disseminated KS cases was evident during the AIDS epidemic (P = 0.003). Conclusions. These observations indicate that children younger than 5 years are at high risk for developing KS, possibly reflecting low resistance to human herpesvirus (HHV) 8 infection. It is also likely that an increased susceptibility to HHV8 infection and morbidity is related to progressive immunodeficiency. The increase in AIDS PKS incidence appears to reflect a direct or indirect promoting effect of HIV on the development of KS lesions. Recognition of the high KS risk in small children warrants considerations of possible prevention measures including HIV/HHV8 vaccination and therapeutic options.

HHV‐8/KSHV during the development of Kaposi's sarcoma: evaluation by polymerase chain reaction and immunohistochemistry

The human γ‐herpes virus‐8 (HHV‐8) was first described in AIDS‐related Kaposis sarcoma (KS) tumour samples. In this study, we report comparative studies on paraffin‐embedded biopsies of AIDS‐related KS (AKS) and endemic KS (EKS) with regard to HHV‐8 content as evaluated using polymerase chain reaction (PCR) and immunohistochemistry. DNA was extracted either using Chelex‐100 or using Qia‐gene kit and was evaluated with the help of a semiquantitative PCR assay. The PCR detection of HHV‐8 was more sensitive to the Chelex method than to Qia‐gene. The threshold for PCR test sensitivity with the help of serial dilution of DNA was at the level of five plasmid ORF‐26 regions, and DNA from 25 body cavity‐based lymphoma‐1 cells. The results expressed as virus load/actin unit showed progressively higher HHV‐8 levels in late (nodular) cases, compared to those in early (patch/plaque) stages. Evaluation of HHV‐8 DNA levels in tumour tissues, thus, indicates a correlation between virus load and KS stage. Double immunostaining of spindle cells (SC) in KS biopsies for CD34 and HHV‐8/latency‐associated nuclear antigen (LANA) showed an increase in double‐positive SC in the lesions of nodular AKS and EKS cases, compared to that in plaque and patch stages. However, 10–15% of CD34+/LANA– SC cells were observed during the development from patch to nodular cases of AKS and EKS. Our results indicate that PCR analysis is a simple and sensitive diagnostic method for HHV‐8 evaluation in KS tissues, processed for conventional histopathology.

Spindle Cell Ploidy and Proliferation in Endemic and Epidemic African Kaposi's Sarcoma

Comparative studies of ploidy and proliferative activity of spindle cells in sections of 20 (skin, 17; lymph node, 3) biopsy specimens from African patients, 10 with endemic Kaposis sarcoma (EKS) and 10 with AIDS-associated Kaposis sarcoma (AKS) were performed by histopathology, feulgen-based DNA measurement and proliferating cell nuclear antigen (PCNA)/cyclin immunohistochemistry, respectively. All specimens were classified as nodular lesions with basically the same histology. In 17 cases immunostained for cyclin/PCNA, the percentage of proliferating spindle cells range between 2-18, with a higher mean rate in AKS although this was not statistically significant. In situ measurement of DNA showed no significant values greater than the diploid level of control cells indicating that spindle cells in both EKS and AKS have euploid DNA content. Our findings indicate that both EKS and AKS represent the same type of euploid low rate cell proliferations. This corroborates previous suggestions that KS could represent a reactive process to yet undefined stimulus rather than a clonal proliferation, of transformed malignant cells.

Monocyte/macrophage giant cell disease in SIV-infected cynomolgus monkeys.

A non-opportunistic, generalized giant cell disease (GCD) was found in 12 out of 25 (48%) cynomologus monkeys infected with SIVsm. Most organs were affected notably the lymph nodes (LN), spleen, gut, liver, lungs and CNS. The multinucleated GC varied considerably in cell size and in the number and cytoplasmic distribution of the nuclei. Immunohistochemically most GC expressed SIV antigens and markers of mononuclear phagocytes (CD68), CD4 and also occasionally the T-cell markers CD45RO, CD43 and CD2. Monkeys with GCD had more pronounced immunosuppression with lower CD4-cell counts, more often demonstrable SIV antigen in the blood and LN and had been infected for a longer time period, as compared to monkeys without GCD. These findings show that SIV infection in cynomolgus monkeys is frequently associated with extensive formation of multinucleated GC of macrophage origin, which appears to be related to the pathogenesis of the infection and the degree of immunosuppression.

Thymic immunopathology and progression of SIVsm infection in cynomolgus monkeys.

Thymuses from 22 cynomolgus monkeys infected with simian immunodeficiency virus (SIVsm) developed characteristic cortical and medullary changes including formation of B-cell follicles (8/21) and accumulation of virus immune complexes. Advanced thymic histopathology was correlated with more pronounced immunodeficiency. SIVsm provirus was detected by polymerase chain reaction (PCR) in most (16/18) thymuses and spliced viral env mRNA in 3 (3/7) thymuses with advanced histopathologic changes indicative of thymic SIVsm replication. By combined in situ hybridization (ISH) and immunohistochemistry, viral RNA was localized mainly to the follicular dendritic network, macrophages, multinucleated giant cells, and lymphocytes of the medullary regions. Latent infection by an Epstein-Barr-related herpesvirus (HVMF1) was also found by PCR and by ISH in medullary regions of three (3 of 8) thymuses with B-cell follicles, suggestive of an inductive role for B-cell proliferation in these thymuses. In a control group of HIV-2-infected nonimmunosuppressed monkeys, no comparable thymic changes were observed. Our results indicate that SIV, and probably by analogy HIV, can have direct and diverse pathogenic effects on the thymus that are important in the development of simian (human) AIDS.

Human herpesvirus 8/Kaposi sarcoma herpesvirus cell association during evolution of Kaposi sarcoma.

Kaposi sarcoma (KS) is associated with a herpesvirus (HHV-8/KSHV), which expresses a latency-associated nuclear antigen (LANA). The histopathology of KS is characterized by angiogenesis, inflammatory cells, and the development of CD34+ tumor spindle cells (SCs). However, the cellular basis for the recruitment and dissemination of HHV-8 during the development of KS lesions is not clear. Twenty-nine KS biopsies with AIDS (AKS, n = 22) and without HIV infection (endemic KS or EKS, n = 7) were immunostained by a triple antibody method to characterize HHV-8-infected and noninfected (LANA+/−) CD34+ SCs, infiltrating CD3+, CD68+, CD20+, and CD45+ leukocytes as well as proliferating (Ki67+) cells. The CD34+/LANA+ SCs were more frequent in late (nodular) as compared with early (patch/plaque) KS stages. However, in late AKS 36.0% of SCs (median of 11 cases) were CD34+/LANA− compared with 20.7% in early cases (median of 11 cases). Furthermore, both AKS and EKS showed, at all stages, a small (4.1–6.5%) population of LANA+/CD34− cells. Proliferating Ki67+ cells were seen (4.5–11.5%) at all KS stages, and were usually more frequent in early AKS, but no significant difference was observed between nodular AKS and EKS. Most of the proliferating cells in the KS lesions were LANA+/CD34+ but a small fraction was LANA+/CD34−. Lesional CD68+ and CD3+ cells varied between AKS (7.3 and 5.2%, respectively) and EKS (4.9 and 3.1%, respectively) but were not clearly stage related. No LANA+ cells were CD3+, CD20+, or CD45+ and very few (<0.5%) were CD68+. These results indicate that not all CD34+ KS SCs were LANA+, suggesting recruitment of noninfected SCs to the lesions. Cell proliferation in general was much higher in early as compared with the late AKS stages. LANA+ SCs could have a proliferative advantage as suggested by higher frequency of cycling (Ki67+) LANA+ SCs. Few macrophages but no lymphocytes are LANA+.