Chimeric antigen receptor (CAR) T cell therapy, which edits a cancer patient's T cells to recognize their tumors, has successfully helped patients with aggressive blood cancers but has yet to show the ability to treat solid tumors. To overcome this hurdle, researchers genetically engineered human T cells to produce a CAR protein that recognizes a glycopeptide found on various cancer cells but not normal cells, and then demonstrated its effectiveness in mice with leukemia and pancreatic cancer. Their proof-of-concept study appears June 21 in Immunity.

"This is the first approach using a patient's own immune cells that can specifically target this class of cancer-specific glycoantigens, and this has the great advantage of applicability to a broad range of cancers," says first author Avery Posey, an instructor at the Perelman School of Medicine of the University of Pennsylvania. "Future cancer immunotherapies combining the targeting of cancer-specific carbohydrates and cancer proteins may lead to the development of incredibly effective and safe new therapies for patients."

CAR T cell therapy involves engineering patients' own immune cells to recognize and attack their tumors. T cells are collected from the patient's blood and genetically engineered to express cell-surface proteins called CARs, which recognize specific molecules found on the surface of cancer cells. The modified T cells are then infused into the patient's bloodstream, where they target and kill cancer cells.

In recent clinical trials, CAR T cell therapy has dramatically improved the outcomes of blood cancer patients with advanced, otherwise untreatable forms of leukemia and lymphoma. But the full potential of CARs for treating solid tumors has not been reached because they have targeted molecules found on the surface of both normal cells and cancer cells, resulting in serious side effects.

Posey, along with co-senior authors Laura Johnson - Director, Solid Tumor Immunotherapy Laboratory, Center for Cellular Immunotherapies - and Carl June - Richard W. Vague Professor in Immunotherapy - both at the Perelman School of Medicine at Penn., were motivated to find a solution quickly, for very personal reasons. One of their colleagues, who was well known for her scientific discoveries and lifelong contributions in the field of cancer genetics, had been diagnosed with end-stage cancer. "She knew of our work and asked if there were any promising treatments we had that might be able to help her," Johnson says. "This really polarized our team, in a worldwide collaboration, to find and fast-track a potential treatment for her cancer to the clinic."

As far as targeted immunotherapy goes, the patient's tumor presented a significant challenge: It had none of the markers that are present on several of the other cancers Johnson and her team had worked on. "That was truly what drove the work that resulted in the CAR in this study," Johnson says. "It was the only marker we could find on her tumor; and it turns out, on just about every other tumor we tested, too."

The cancer cell marker that Johnson and her team identified was a specific change in protein glycosylation, that is, a unique pattern of sugars decorating a protein found on the cell surface. In collaboration with investigators from the University of Copenhagen and University of Chicago, the researchers developed novel CAR T cells that express a monoclonal antibody called 5E5, which specifically recognizes a sugar modification--the Tn glycan on the mucin 1 (MUC1) protein--that is absent on normal cells but abundant specifically on cancer cells.

The 5E5 antibody recognized multiple types of cancer cells, including leukemia, ovarian, breast, and pancreatic cancer cells, but not normal tissues. "This is really the first description of a CAR that can target multiple different solid or liquid tumors, without apparent toxicity to normal cells," Johnson says. "While it may not be a universal CAR, it is currently the closest thing we have."

Moreover, injection of 5E5 CAR T cells into mice with leukemia or pancreatic cancer reduced tumor growth and increased survival. All six mice with pancreatic cancer were still alive at the end of the experiment, 113 days after treatment with 5E5 CAR T cells. Meanwhile, only one-third of those treated with CAR T cells that did not target Tn-MUC1 survived until the end of the experiment.

The downside, Johnson cautions, is that this type of therapy is still very new, and there are numerous factors that are involved at the tumor level that may limit treatment. In particular, more work is needed to determine the safety of this therapy in advanced mouse models that can more accurately predict safety in humans, and its efficacy specifically against metastatic cancer, which is the leading cause of cancer-related deaths. "So while we are hopeful, no one ever knows if a cancer treatment is truly going to work, and be safe, until it actually goes to treat patients in the clinic," Johnson says.

If these preclinical studies are successful, the researchers plan to further develop their CAR T cell therapy and test its safety and efficacy for different types of metastatic cancer in upcoming clinical trials. "Unfortunately, our colleague passed away before this could reach a clinical treatment, but she was happy that even if it couldn't help her, this finding might be able to help other patients in the future," Johnson says.

This work was supported by Novartis, the National Institutes of Health, the Danish Research Councils, and the Danish National Research Foundation. Conflicts of interest: The University of Chicago has filed a patent on the 5E5 CAR and an invention disclosure has been filed on these studies. The University of Pennsylvania has entered into a strategic alliance with Novartis for the development of chimeric antigen receptors.

Article: Engineered CAR T Cells Targeting the Cancer-Associated Tn-Glycoform of the Membrane Mucin MUC1 Control Adenocarcinoma, Posey, Jr. et al., Immunity, doi: 10.1016/j.immuni.2016.05.014, published 21 June 2016.