November 10, 2019
In 2010, Emily Whitehead was diagnosed with Acute Lymphoblastic Leukemia, a cancer of certain cells in the immune system.
This is the most common form of childhood cancer, her parents were told, and Emily had a good chance to beat it with chemotherapy. Remission rates for the most common variety were around 85 percent.
It would be 20 months before they’d understand the shadow behind that sunny statistic, and the chilling prospect of volunteering their daughter as patient zero for the world’s first living drug.
Emily started on the 26-month chemotherapy regimen. She lost her hair and most of her kid energy, and the curative poison seemed to be doing its job, sickening her body as it killed the disease. But her cancer, like all cancers, was alive, a constellation of mutant cells that continued to mutate into new variations. Some of these new mutants were immune to the chemotherapy and continued to thrive.
By October 2011, Emily had relapsed; in the language of immunotherapists, her cancer had “escaped.” Her physicians at Pennsylvania’s Hershey Medical Center could only offer more chemo, more aggressively. In February 2012 she relapsed again.
Now it was painfully obvious that Emily was one of the 15 percent of kids with leukemia for whom chemotherapy did not work. The cancer was doubling daily in her bloodstream, and it was too late for a bone marrow transplant—she was too sick. Oncologists now referred to Emily’s cancer as “terminal.” She was 6 years old.
Cancer is shitty and unfair, but that shitty unfairness reaches a whole other level when it happens to a kid. Tom and Kari Whitehead were told that they needed to consider hospice for their daughter. Or, if they wanted, she could die at home. Traditional medicine had nothing else to offer her. But a researcher at the Children’s Hospital of Philadelphia might, if Emily’s parents were willing to take the risk.
The Whiteheads learned of this possibility on a Sunday. By Monday, they were in Philadelphia. Emily Whitehead would be the world’s first kid to try an experimental cancer therapy, called CAR-T. Researchers were offering to reprogram her immune cells into a clone army of cancer-targeting serial killers.
A CAR-T cell is a reengineered T cell that has been removed from the cancer patient, tweaked in the lab to recognize that patient’s cancer, and then injected back into the patient. Because each of these reengineered cells is a monstrous robocop-like assemblage of immune cell parts, researchers had given their invention the equally monstrous name of “Chimeric Antigen Receptor T cell” (in Greek mythology, the chimera is a patchwork monster combining aspects of a lion, goat, and serpent), but “CAR-T” sounds much better.
CAR-T is often called the “most complex drug ever created,” but it is not really a drug in the traditional sense. Unlike an inert molecule introduced to the body for some temporary effect, CAR-T is alive. If it worked as designed, this “living drug” would go on living in Emily’s bloodstream like a cancer-killing superpower, providing her with a sort of immunity against her disease. And in the process, it would give humanity a revolutionary new weapon in the war on cancer.
And if it didn’t work? If the unleashed cellular serial killers turned on the girl’s healthy cells instead of the cancer? Well, the Whiteheads decided, best not to think about that. At this point, their only daughter had nothing to lose.
The hundreds of millions of T cells that patrol our bloodstreams and lymph nodes are expert at recognizing sick body cells and killing them. And, although the idea was dismissed by most scientists for the past 100 years, a handful of these T cells are predisposed to recognizing and killing cancer, too.
So, why doesn’t our immune system do that job? You always know when you have a cold or the flu, but cancer arrives without so much as a sniffle. Why does it usually require a test to know that we have this deadly disease?
The answer to that question came in a series of breakthrough discoveries of how cancer uses tricks to turn off, hide from, and overwhelm our immune response. Cancer shuts down T cells before they get a chance to call for reinforcements, reproduce into an overwhelming clone army, and do their job. But what if there was a way to overwhelm cancer instead, barraging it with huge numbers of immune cells capable of recognizing and killing it?
The group of researchers considering this possibility were called cancer immunotherapists, and by the time Emily Whitehead showed up at the hospital, they had already spent decades on the problem.
But before they could hope to make that clone army, they needed to comb through the hundreds of millions of cells in a patient’s immune system and identify the one or two T cells that happened to be perfectly tuned to recognizing that patient’s personal cancer.
Not surprisingly, Mr. Perfect was hard to find. In fact, until the 1980s, even cancer immunotherapists weren’t entirely certain that Mr. Perfect existed.Identifying, extracting, fertilizing, growing, cloning, and then activating the perfect T cell against cancer—this was largely trial and error work, done with little funding and little grasp of the overwhelming biological complexities of cancer or the immune system. The science was all impossibly new; T cells had only been discovered in the late 1960s.
Cancer immunotherapists floundered for decades, the laughing stock of the research community, unable to prove their theory that the immune system could be helped recognize and kill cancer cells, and largely unable to help real cancer patients.
Meanwhile, another group of cancer immunotherapists had started to consider a different approach: Rather than hoping to somehow locate the perfect cancer killing T cells in a patient’s body, they’d make their own Mr. Perfect, engineering a Frankenstein T cell stitched together from various parts in the lab. The Weird Science T cell would be designed specifically to seek and destroy a patient’s specific cancer.
The engineering is complex, but the concept is simple. An individual T cell recognizes only the distinct sick cell protein (called an antigen) that they are born to “see,” as determined by a random assignment process. The business end of that “seeing” is called the T cell receptor, or TCR.
Change the TCR, and you might be able to change what that T cell targets. Change it to the right one, and you might even be able to get it to target a specific disease. That was exactly what occurred to a charismatic Israeli researcher named Zelig Eshhar.
In the early ’80s this beekeeping PhD started thinking about the business end of the TCR—the part that extends out through the surface of the T cell like a grabby protein antenna and “sees” specific antigen targets.
To Eshhar, that looked a lot like the grabby protein claws of an antibody. It seemed to work the same way too. These Y-shaped immune structures come in lots of flavors (hundreds of millions), each sticky to a different disease-specific protein. Each was a key in search of its lock.
Eshhar could imagine popping off the end of the TCR and popping on a new antibody like a vacuum attachment; change the antibody, and you might change what the T cell targets. In theory, you could have a near infinite number of new attachments, each specific to recognize and bind with a different antigen, and thus target a different disease. Such a technology would create a whole new class of medicines.
Turning Eshhar’s theory to reality required a fancy bit of bioengineering, but somehow, in 1985, he managed to produce a simple proof of concept.
He called his primitive CAR a T-body. It was a T cell retooled to recognize a relatively obvious antigen target that he had selected, a telltale protein worn by the fungus Trichophyton mentagrophytes, better known as athlete’s foot. This humble experiment cloaked mind-blowing possibilities.
And it caught the attention of those who’d spent their lives laboring in the trenches of cancer immunotherapy, including a pioneering immunotherapist Steve Rosenberg. Rosenberg had first become convinced of the immune system’s potential to kill cancer in the 1960s, after examining a former stage IV cancer patient whose immune system had spontaneously cured his own disease. Rosenberg had wondered whether the man’s supercharged immune cells could help other cancer patients, too.
In experiments unthinkable today, Rosenberg had tried just that, transfusing the cured man’s blood into the veins of a terminal cancer patient in the next bed. It didn’t work, but the promise of cellular transfer therapy stuck with him.
For the next five decades, Rosenbergs’s National Institutes of Health laboratory (and that of Philip Greenberg at the Fred Hutchinson Cancer Research Center in Seattle) would serve as a sort of hive and haven for immunotherapy talent.
In 1989 Eshhar was persuaded to spend a sabbatical there, joining another brilliant young NIH researcher named Patrick Hwu to create an updated take on what would eventually be known as “adoptive cell therapy.”
Examining a patient’s tumors under the microscope revealed that, even when the larger immune attack had failed, a few T cells still managed to successfully recognize the tumor antigens and nose their way in. These robust infiltrators would be their Mr. Perfect T cells and, hopefully, seeds for their clone army of targeted cancer killers.
Hwu’s focus was to try to weaponize this subset of successful “tumor-infiltrating lymphocytes,” or TILs, by packing them with an additional payload of powerful tumor-killing hormones. “Zelig had shown that an antibody and a T cell could be combined to target something,” says Hwu, who serves as the head of the division of cancer medicine at Anderson Cancer Center in Houston, Texas. “Now the question was, could we get it to target cancer cells?”
In order to work as little guided missiles, they needed a guidance system, one that researchers could choose and customize to target various types of cancers. Starting with a batch of T cells they’d found to be Mr. Perfect TILs active against melanoma, Hwu and Eshhar Frankensteined them with new TCRs to instead target ovarian, colon, and breast cancers. “Zelig made the receptor, I put it into T cells,” Hwu remembers. “It was really hard to do that in the 1990s.”
Without the benefit of retroviral vectors or Crispr, the task required sticking a little needle into a T cell and micro injecting the new TCR genes one cell at a time. “We spent a lot of time together,” Hwe says with a laugh. “A lot of all-nighters in the lab.”
None of the results was perfect, but the TILs they had retargeted to ovarian cancer worked best of the three, and the team was able to publish on the result, heralding the new CAR-T name and the enticing implications of the technology.
They hadn’t cured any cancer, but they’d advanced the science. They had successfully replaced the T cell steering wheel and that knew how to find a specific cancer. “The first time I got that to work I was so elated,” Hwu remembers. But it would take more than retargeting to engineer a cancer-killing machine.
To be effective, these new cells also needed to thrive and replicate themselves, like normal T cells do. Their first-generation cars didn’t do that. It was as if some vital essence had been lost during the retrofit, resulting in lemon CARs that didn’t run long enough to replicate or kill. Their Frankenstein would rise from the table, only to keel over.
It would be up to researcher Michel Sadelain to provide the clever workaround for this and several other engineering problems, creating a truly “living drug,” as Sadelain called it, a second generation CAR that could recognize a target, expand clonally, and retain its other T cell functionality, with a life span as long as that of the patient’s.
Working in his lab, Sadelain (a laconic scientific intellectual who is the founding director of the Memorial Sloan Kettering Cancer Center of Cell Engineering, among other things) also gave his new CAR an important new target—a protein called CD19 found uniquely on the surface of certain blood cancer cells.
CD-19 seemed like a good CAR choice. It was found in abundance on the surface of certain cancers. It was also expressed by some normal B cells, but that was acceptable. If the CAR attacked healthy cells as well the cancer, the collateral damage was survivable.
In a healthy human, B cells are essential aspects of the normal immune system. But in patients like Emily, those B cells had mutated and become cancerous. To survive, she would need to lose them.
Luckily, physicians had long ago learned to keep patients alive without B cells. “If you’re facing terminal cancer,” Sadelain says, “losing your B cells isn’t so bad.”Sadelain now had a sleek, stylish, and self-replicating second-generation CAR with plenty of fuel and a realistic cancer target. His group shared the sequence of their new CAR with Rosenberg’s group at the National Cancer Institute, as well as the lab of University of Pennsylvania researcher and physician Carl June. (June in turn also based aspects of his CAR design on a sample borrowed from Dario Campagna of St. Jude’s Children’s Research Hospital.)
These three groups—all pushing for human trials of this complex and powerful new cancer therapy—were now competitors. At the same time, they worked together, borrowing and improving upon each other’s ideas.
Sadelain’s group had been first to start CAR-19 T cell clinical trials, Rosenberg’s first to publish; their successful CAR-T trial shrank tumors in a lymphoma patient. But it would be Carl June’s trial with Emily Whitehead that would take the spotlight and determine whether there was a future for CAR-T.
June was well aware of the stakes. If his CAR was too aggressive for a pediatric patient, if his powerful Franken-drug proved to be a killer too powerful to control, Emily would die. And any hope of saving hundreds of other children with this technology would likely die with her.
Though June is trained as an oncologist specializing in leukemia, his work on the AIDS crisis had convinced him of the cancer-killing potential of the immune system. Several cancer immunologists had gained their faith that way. Witnessing the prevalence of previously rare cancers in immune compromised patients seemed proof of a connection between the immune system and cancer, even if the scientific concensus was that no such connection existed.
But if the little girl died from the experiment, if his powerful Franken-drug attacked her body instead of the cancer, he was equally certain that the result would be horrifying and tragic. And that any possibility of CAR-T ever curing cancer in the hundreds of other children dying from ALL would likely die with her.