Stem cells, melanoma and cancer stem cells: the good, the bad and the evil?

Most cancers contain morphologically heterogeneous populations of cells. While this observation may partly be explained by the coexistence of multiple genetic sub-clones arising through independent somatic mutations and/or as a result of differentiation processes in the tumor microenvironment, it also implies that the tumor may be formed from undifferentiated ''stem cell-like'' cells called ''cancer stem cells'' or ''cancer-initiating cells''. These cells are thought to constitute one or several rare subpopulations in a given tumor and would be strongly responsible for initiation of tumor development and growth as well as for metastasis and recurrence after cytoreductive therapy. However, while the concept of cancer stem cells has been first established for human myeloid leukemia in the 1960s, it has only much later been extended to other solid tumors such as breast or brain cancers and most recently to melanoma. Thus, it is presently unclear which role a sufficiently characterized population of melanoma stem cells plays in cancer promotion and progression. Here, we review the emerging melanoma stem cell model and discuss the biological and therapeutic implications of the model.

Decreased density of epidermal dendritic cells in melanocytic naevi: the possible role of in vivo sun exposure.

Melanocytic naevi are benign skin tumours that originate in the epidermis. The pathogenesis of naevi and cutaneous malignant melanoma has been linked to sun exposure. This study evaluates alterations in the density of immunologically active epidermal dendritic cells (EDCs) in naevi in response to sun exposure. Immunohistologically stained sections of 266 naevi from patients from Israel (n=135) and Germany (n=131) were evaluated. The proportion of naevi with decreased density of HLA-DR+ (dDR+) and CD1a+ (dCD1a+) EDCs was analysed according to country, last exposure to sunlight, anatomical location and histological subtype. The risk of dDR+ was found to be linked to residence in Israel compared with Germany (odds ratio [OR] = 4.2; 95% confidence interval [CI] = 2.0-8.9), suggesting a latitude-dependent effect. Naevi removed in summer had a higher risk of dCD1a+ (OR = 4.7; 95% CI = 2.3-9.8) compared with those removed in winter. The most conspicuous dDR+ among the German cases, and dCD1a+ among the Israelis, occurred in naevi located on commonly exposed skin. The similar densities of EDCs in the lesional and perilesional skin of the majority of the naevi indicates that the underlying naevus cells have no effect on EDC density. It is not unlikely that an altered immune response due to dDR+ and dCD1a+ in sun-exposed skin in the vicinity of naevi contributes to the subsequent melanoma risk in highly susceptible individuals.

Intact survival with transfusion-associated graft-versus-host disease proved by human leukocyte antigen typing of lymphocytes in skin biopsy specimens.

A transient transfusion-associated graft-versus-host disease occurred in a premature infant of 30 weeks of gestation. We demonstrated donor lymphocytes in a skin biopsy specimen with a two-step immunoperoxidase technique using monoclonal antibodies against human leukocyte antigen determinants specific for the donor. The girl survived and is immunocompetent.

MHC antigens in human melanomas.

The distribution of MHC antigens in human melanocytic lesions, i.e. HLA class I and HLA class II antigens is reviewed. HLA class I antigens have a broad distribution, but may be lost during tumor progression. In contrast, HLA class II antigen expression appears with neoplastic transformation. The mode of regulation of HLA antigens in melanoma lesions is complex. Immunohistochemical demonstration of HLA antigen expression in primary melanoma lesions and in locoregional metastases has prognostic relevance. Expression of HLA-DR in primary melanoma lesions is associated with an unfavorable prognosis, as is a decreased expression of HLA-A,B,C antigens in locoregional metastases.

Level of HLA antigens in locoregional metastases and clinical course of the disease in patients with melanoma.

Immunohistochemical staining with monoclonal antibodies showed marked variations in the percentage of melanoma cells stained by anti-HLA Class I and anti-HLA Class II monoclonal antibodies among 48 locoregional metastases removed from 39 patients with malignant melanoma. On the other hand there was limited variation in the percentage of melanoma cells stained by anti-HLA antibodies in autologous locoregional metastases removed from 8 of 9 patients. In the remaining patient marked differences were found in the percentage of melanoma cells stained by anti-HLA Class I antibodies in the two parts of the lymph node metastasis analyzed. Therefore this patient was not included in additional analyses to correlate the level of expression of HLA antigens with the clinical course of the disease. In all the lesions tested the percentage of melanoma cells stained by anti-HLA Class II antibodies was lower than or equal to but never higher than that stained by anti-HLA Class I antibodies. According to the level of expression of HLA Class I and Class II antigens the 38 patients could be divided into three groups: Pattern A included lesions with more than 50% of tumor cells stained by anti-HLA Class I antibodies (mean, 86.1; median, 85) and 50% or less by anti-HLA Class II antibodies (mean, 10.5; median, 5); Pattern B included lesions with 50% or less tumor cells stained by anti-HLA Class I antibodies (mean, 14.9; median, 5) and by anti-HLA Class II antibodies (mean, 4.1; median, 1); Pattern C included lesions with more than 50% tumor cells stained by anti-HLA Class I antibodies (mean, 88.8; median, 92) and by anti-HLA Class II antibodies (mean, 70.0; median, 70). The survival of 21 patients with Pattern A was significantly longer than those of 13 and 4 patients with Patterns B and C, respectively. No difference in the survival of patients in the latter two groups was found. These results suggest that HLA antigens play a role in the biology of melanoma and that analysis of the level of HLA antigens in locoregional metastases of patients with melanoma may provide clinically useful information.