Harnessing a patient’s own infection-fighting T cells to fight untreatable cancer has sparked a great deal of excitement and news reports lately, thanks in no small part to several small clinical trials that demonstrated spectacular results, including complete remission, in some patients.

In these trials, T cells were harvested from the patient’s own blood, genetically modified to target cancer cells, and then re-administered, newly-armed, to seek out and destroy the cancer cells. Some of the most positive clinical data has come from trials using chimeric antigen receptor-engineered T cells (so called CAR T cells) designed to recognize tumor-associated antigens, such as CD19, in the treatment of chronic lymphocytic leukemia.

The T cells are genetically modified with a non-replicating virus (i.e. a safe virus) that encodes an antibody gene and a signaling gene. These genes express a new protein on the cell surface of the patients’ T cells; the extracellular region of this new protein recognizes CD19 while the intracellular region—once activated—induces the T cell to kill the cancer cell.

The action of these genetically engineered T cells is therefore to direct the bodies’ own immune defense mechanisms to selectively target cancer cells and induce cell death. Clearly the success of this strategy is dependent upon the selection of the cancer cell antigen since only those cells expressing the antigen will be targeted by the engineered T cells.

The Toxicity Issue

However, engineered T cell therapy is associated with some expected toxicities. Because the cancer antigen is derived from self-proteins—normal proteins expressed at low levels or during development—it may also be present on non-cancerous cells.  For example, engineered T cells directed against CD19 also destroy the patients’ own B cells—important for the production of antibodies. This is why patients administered anti-CD19 T cell therapies have to take other medications to prevent infections.

The administration and proliferation of T cells, which are required to effectively destroy the cancer, may also result in a violent cytokine storm or cancer lysis syndrome due to the destruction of large amounts of tissue within a short time frame.

Fortunately, these side-effects may be managed with anti-inflammatory drugs, but other risks are harder to assess, particularly the recognition by engineered T cells of unrelated antigens expressed on the surface of normal cells. This problem led to a devastating result in one small trial, where two patients sadly died because of acute cardiac failure after administration of engineered T cells against the cancer antigen MAGE-A3.1

Researchers were initially unsure of the trigger for the fatalities, since MAGE-A3 antigen is only expressed in embryos and on a number of different cancer cell types, and cross reactivity was not expected based upon data from earlier preclinical studies.

So researchers from ImmunocoreAdaptimmune and the University of Pennsylvania involved in the development of MAGE-A3 recognizing T cells  investigated the reactivity of their novel T cell product to understand the  mechanism of cardiac toxicity.2 The team showed that engineered T cells did not have any off-target reactivity against a range of primary cells. However, co-culture of MAGE-A3 engineered T cells with beating cardiomyocytes (derived from induced pluripotent stem cells and therefore considered to be a better model), led to the destruction of the cardiomyocytes.

The team then identified the target antigen on the surface of the beating cardiomyocytes by doing a blast search for common amino acid sequences likely to be recognized by the MAGE-A3 chimeric receptor. Using this approach they identified titin—a large skeletal protein expressed by cardiomyocytes—as a candidate, and showed that the expression of titin in non-expressing cells could lead to cell destruction by MAGE-A3 recognizing T cells.  In other words, the observed cardiac toxicity in the clinical trials seemed to be due to the presence of MAGE-A3 cancer antigen in an unrelated protein—titin.

The Future

Because of their extreme potency, engineered T cells provide much needed hope for many patients with incurable cancer. However, determining cross reactivity against non-related antigens will be key to the widespread application of these novel therapies. Preclinical investigations will have to become more extensive and include reactivity against highly differentiated cells in vitro as well as mapping the three-dimensional structure of the engineered T cell receptor along with three-dimensional searches for amino acid sequences in non-cancer antigens that might bind the chimeric engineered T cell receptor.

Solving these problems is critical for the safe use of these novel and highly effective anti-cancer therapies.


  1. Linette et al., Cardiovascular toxicity and titin cross-reactivity of affinity-enhanced T cells in myeloma and melanoma, Blood, 2013.
  2. Cameron et al., Identification of a Titin-Derived HLA-A1–Presented Peptide as a Cross-Reactive Target for Engineered MAGE A3–Directed T Cells, Sci. Transl. Med. 2013.