by ecancer reporter Clare Sansom

Malignant melanoma is the most deadly form of skin cancer.

Although quite rare, it is responsible for about 75% of skin cancer deaths worldwide, and patients with this condition still have a poor prognosis despite recent significant advances in treatment.

Immunotherapy has shown good results in this tumour type, particularly initially, but patients often relapse.

One such treatment that is often successful in inducing temporary remissions is the transfer of cytotoxic T cells that target specific antigens on melanoma cells, a technique that is known as adoptive cell transfer therapy or ACT.

The mechanisms through which melanomas develop resistance to ACT therapy are as yet poorly understood. Now, however, a team led by Thomas Tüting from the University of Bonn, Germany and Antoni Ribas of the University of California Los Angeles, California, USA has developed an ACT protocol in a mouse model of melanoma and used this to explore these mechanisms.

Tüting, Ribas and their co-workers established a transgenic mouse model in which melanoma development is driven by deregulation of tyrosine kinase signalling pathways. They developed a form of adoptive cell transfer for these mice that is similar to those in clinical use and which involves infiltrating, cytotoxic T cells that recognise the melanoma-specific antigen gp100. Injecting these cells into melanoma-bearing mice caused tumour regression that generally lasted for about two months.

This regression could be extended and survival time increased by vaccinating the mice with an adenoviral vector vaccine expressing gp100 to re-activate these T cells. Comparing the gene expression profile of relapsed melanomas with that of controls showed increased expression levels of pro-inflammatory cytokines in the relapsed tumours.

On the basis of these findings, the researchers proposed that a pro-inflammatory tumour micro-environment might contribution to the acquisition of ACT resistance by these tumours. To confirm this, they first ruled out the possibility that the resistant tumours were newly arising by transplanting single melanomas of known cellular origin into the mice and treating these with ACT. These tumours showed a pattern of regression and re-growth that was similar to that seen in the initial experiments.

The researchers next compared the gene expression profiles of parental and relapsed melanoma cells, and showed that the relapsed cells lacked expression of the gp100 antigen that was the target of the infiltrating T cells.

Expression of genes involved in pigmentation was also down-regulated and expression of genes involved in the immune response up-regulated in cells from relapsed melanomas. These results indicated that the cells had lost some of their specialized phenotype, i.e. that they had dedifferentiated. Similar changes, with a decrease in gp100 expression and increased expression of pro-inflammatory cytokines and of the nerve growth factor receptor NGFR, occurred when melanoma cells were incubated with conditioned medium from the activated T cells.

Further analysis showed these changes to be driven by the expression of TNF-a, which acts as an inflammatory mediator in promoting these changes in gene expression and thus dedifferentiation. These results support a model in which the gene expression of melanoma cells can be altered reversibly by pro-inflammatory signals.

The researchers then tested the validity of the model in human cancer using several different human melanoma cell lines, and showed that TNF-a expression in these cells induced the up-regulation of NGFR and the down-regulation of the melanocyte-specific antigens that are recognized by the ACT cells. Cutaneous metastases taken from patients with melanoma also showed increased expression of NGFR.

This proposed model for the development of ACT resistance in melanoma suggests that melanoma cells escape from the immune therapy through an inflammation-driven, reversible process in which the melanocyte-specific antigen targets of the therapeutic T cells are lost. Tüting and Ribas suggest that the development of ACT protocols that target a wider range of antigens on the tumour cells and the simultaneous blocking of the immune-inhibitory response might help prevent melanoma dedifferentiation and thus relapse after ACT therapy.

Reference

Landsberg, J., Kohlmeyer, J., Renn, M. and 9 others (2012). Melanomas resist T-cell therapy through inflammation-induced reversible dedifferentiation. Nature, published online ahead of print 11 October 2012.  doi:10.1038/nature11538