Research is under way to establish new CAR-T cell targets for a wide range of cancers. A major drawback of CD19-mediated CAR-T cell treatment is relapse in patients with CD19-negative disease. As such, CAR-T cells that target CD22, another B cell target, are now in advanced clinical trials. A potential future approach could be to target both antigens simultaneously, theoretically preventing relapse with antigen loss.
B cells are an attractive choice of target compared with other cells, as it is possible to live without them. This makes the destruction of healthy cells, as by-products of targeting cancer cells, acceptable. The challenge for other cancers is to find specific targets that are present on cancer cells but not on healthy cells. Despite these difficulties, the list of potential targets is extensive.
Acute myeloid leukaemia (AML) is a haematological malignancy for which new treatments are urgently required. There are a number of proposed targets including CD33, CD123 and FLT-3, all of which are now in clinical trials. These targets are present on healthy cells at various stages of development but are often expressed at higher frequency on leukaemia cells. Another CAR-T cell that targets a plasma cell-surface protein called B cell maturation antigen (BCMA), also known as tumour necrosis factor receptor superfamily member 17 (TNFRSF17), has shown efficacy in multiple myeloma and is likely to gain regulatory approval in the very near future.
As well as developing CAR-T cells against single targets, some research groups are developing logic-gated CAR-T cells. These cells recognise a target with one receptor which, instead of activating the T cell to kill, induces the transient expression of a CAR gene that recognises a second target, thus leading to CAR-T cell activation and killing. This increases cancer specificity as it is only the combination of target antigens that needs to be restricted to cancer cells, not the individual antigens themselves. Other CAR-T cell targets, particularly for solid tumours, may not necessarily be the cancer cells themselves; attacking the supportive network such as the vasculature may be an effective approach.