“It was a Rube Goldberg idea,” says Carl June, recalling the earliest suggestions that the immune system might be retrained to attack tumors. This was the late 1980s, and it seemed unlikely—maybe even crazy—that T cells could be taken from a cancer patient and genetically engineered to become cancer killers. The idea was that the T cells would be reprogrammed to express chimeric antigen receptor (CAR) proteins, tuned to the patient’s unique tumor antigens; the resulting CAR-T cells, when returned to the patient, would thus kill any cells displaying that antigen (see ‘How CAR-T Therapy Works’).

The idea might have been ambitious, but it was based on promising research from the past few decades. Nevertheless, most oncologists remained unconvinced. “No-one ever really thought it would work,” says June, now professor of immunology and director of the Center for Cellular Immunotherapies at the University of Pennsylvania in Philadelphia. “There were only a few fanatics in this country. But those people didn’t give up.”

June was one of the dedicated few, along with Steven Rosenberg at the National Cancer Institute (NCI), Michel Sadelain at Memorial Sloan-Kettering Cancer Center, and Malcolm Brenner at Baylor College of Medicine. Thanks to these researchers’ refusal to give up over the past 20 years, CAR-T cells have been able to move from crazy ideal to clinical reality, with the first major success in around 2010, when small clinical trials produced dramatically positive results in aggressive blood cancers.


It was not an easy journey to those positive results, and it didn’t start with cancer. “I did the first CAR-T cell trials in humans in the mid-1990s,” June says. The treatment safely targeted HIV, but its effects were mild and the effort ended with the emergence of successful HIV drug cocktails—ones not related to CAR-T cells. Significantly, however, “we found that the CAR-T cells survived for more than 17 years in the patients,” he adds.

A decade later, a round of clinical trials for solid tumors failed altogether. But a few years further on, trials for CAR-T drugs that targeted a surface protein called CD19—found only on the immune system’s B cells, which go awry in certain types of leukemia and lymphoma—began to succeed.

During that summer, June’s group was seeing positive results in three patients with heavy loads of leukemia when they got a shock. The CAR-T cell treatment triggered an extreme immune reaction dubbed cytokine release syndrome (CRS), in which the reengineered T cells trigger the release of billions of inflammatory signaling molecules called cytokines. That reaction was unexpected, says June. “It didn’t happen in our experimental mice.”

CRS put the first patient into intensive care. “They thought he was going to die,” June says. “And somehow he didn’t. When he woke, they examined him and couldn’t find the tumor.” Patients in other trials have not been so lucky and a few have died from CRS, which must be controlled.

Still, the CAR-T cells were having an impressive effect on June’s leukemia patients. “It was unbelievable,” he says. “This was the first example of a living drug, where the cells we put in multiplied thousand-fold, and the tumor went away.”

When his team published the results in 2011, “it completely changed everything,” he says. “Something finally worked, and we could prove the mechanism of action.”

The structural similarities between an antibody (left), a chimeric antigen receptor (center) and a T-cell receptor (right). T-cell receptors are similar to one arm of an antibody. Credit: Selvanegra


Until these early trials succeeded, drug companies basically hadn’t invested a dime in CAR-T development. In 2012, however, Novartis inked a deal with University of Pennsylvania to commercialize its technologies. Other companies, large and small, also bought into CAR-T research.

By 2016, larger trials were testing multiple CAR-T-related products at numerous cancer centers. By then, clinicians were well aware of the serious side effects of the treatment, says Caron Jacobson, medical director of the Immune Effector Cell Therapy program at the Dana-Farber Cancer Institute/Brigham and Women’s Hospital Cancer Center in Boston, which was hosting one such trial.

“For the first months, we were doing a dance of trying to figure out how much toxicity a patient could experience before you should treat them for that toxicity,” Jacobson notes, “because we didn’t know if that treatment was going to impact their outcomes.”

Evidence began to accumulate for drugs from Novartis and Kite Pharma—the first companies to have CAR-T drugs approved. In the case of Novartis, “the clinical data was so compelling, the only thing that the FDA really asked Novartis to do was to show that they could duplicate the results we had here in Philadelphia,” June says.

Even in 2017, the U.S. Food and Drug Administration (FDA) still notified companies of drug approvals by fax. In order to be ready to receive the news, Novartis manned its fax machine around the clock. When approval finally arrived, June got a call straight away. “It was a lifetime dream,” he says simply. The FDA approved the exact drug made in June’s lab, started back in 2004 by his postdoc, Michael Milone. Also in 2017, the FDA approved one other CAR-T product, by Kite Pharmaceutical. Today, more than 100 companies around the world are rapidly accelerating engineered T cells in medicine.

The power of these cells continues to amaze. In May 2018, June and his co-workers published a Nature paper that described a patient with chronic lymphocytic leukemia who apparently remained disease-free after five years. Astonishingly, genetic analysis traced the lineage of the triumphant T cells back to a single original cell. “It’s a truly remarkable finding, and essentially tells us that the minimum dose needed for CAR-T cells to do their job is one,” June remarks. “Basically, the patient was a bioreactor.”

To learn more about the next-generation of CAR-T therapies just emerging from the clinic, visit our article.