Chimeric antigen receptor (CAR) T cells are genetically engineered to recognize tumour-associated antigens and have potent cytolytic activity against tumours. Adoptive therapy with CAR T cells has been highly successful in haematological cancers. However in solid cancers, CAR T cells face a particularly hostile tumour microenvironment and clinical trials of solid tumour antigen-targeted CAR T cells have shown limited efficacy. Hence, it is clear that CAR T-cell technology requires further optimization before it becomes a viable treatment for non-haematological malignancies. Our research team is currently conducting the CARPETS study, a phase I clinical trial of third generation GD2-specific CAR T cells in metastatic melanoma patients at the Royal Adelaide Hospital. This trial was a world-first in using CAR T-cell therapy for melanoma, and in combining CAR T cells with standard kinase inhibitor-targeted therapy. So far, six patients have been treated safely with a further six to be recruited. Dose escalation has been completed and the treatment has been shown to be safe with no serious adverse events. However CAR T cells failed to persist beyond four weeks in five of six patients; the only patient to demonstrate long-term persistence was pre-treated with lympho-depleting chemotherapy. Our research suggests that GD2-specific CAR T cells have a predominately effector memory phenotype and a subset of CAR T cells upregulate and maintain PD-1 expression following antigen encounter. CAR T cells also demonstrate loss of function and increased susceptibility to activation-induced cell death with repeated stimulation. To counter this we have adjusted the CAR T cell manufacturing conditions, and investigated a combination therapy approach with anti-PD1 monoclonal antibody, with the aim of promoting a central memory phenotype and improved persistence and function of CAR T cells. Our hope is that this will improve CAR T cell efficacy for solid cancer patients.