Skip to Main Content

The idea is tantalizing: Reengineer a patient’s own immune cells to attack the cancer that’s killing her. Scientists have proven it can be done, curing patients of otherwise terminal blood cancers.

But faith in this approach, dubbed CAR-T immunotherapy, has been shaken in recent months.


First, Juno Therapeutics, a leader in the field, announced that four patients in its clinical trials had died. Then, just this week, Novartis — one of the prime contenders in the race to commercialize CAR-T therapy — shuttered its cell and gene therapy unit and announced plans to lay off 120 employees. The company says it’s still working on CAR-T, but the move struck many analysts as a retreat.

Meanwhile, companies racing to bring CAR-T to market are coming up against daunting challenges as they try to scale up highly personalized treatments and turn them into oncology blockbusters. They’re also running into trouble using CAR-T to treat solid tumors, as opposed to blood cancers, which limits the potential market. Then there are the severe side effects associated with the experimental treatment, which can send patients into comas and cause their brains to swell.

An analyst from BTIG wrote this week that he’s become “generally skeptical” about the market potential for Juno and Kite, since blood cancers are a niche field crowded with prospective CAR-T therapies.


To be sure, there’s still excitement in the field. This week, Amgen bought the global development and commercial rights for a T-cell therapy from Boehringer Ingelheim that targets multiple myeloma. But it’s more subdued.

“I think the story line today is, CAR-T is really walking a fine line,” said Brad Loncar, founder of a cancer immunotherapy fund. “A year ago, there was definitely exuberance. But now, we’re on the precipice of people pulling back.”

Mass producing a personalized therapy

Manufacturing a CAR-T product is complicated. It involves harvesting a patient’s own immune cells, reengineering them so they have the ability to find and kill a cancer cell, then feeding these redesigned T-cells back into the patient’s body.

“CAR-T therapy is a highly individualized product, even though the process that each patient goes through is very similar,” said Dr. Stephanie Goff, a surgeon at the National Cancer Institute who’s working with Kite on its treatments.

Scaling up is a tricky prospect for CAR-T therapy. Increasing the capacity to manufacture a standard biologic drug might involve bumping up the size of a bioreactor so it can pump out, say, 5,000 liters, said Bruce Levine, a gene therapy professor at University of Pennsylvania who has worked with Juno Therapeutics.

But CAR-T therapy is so highly personalized that manufacturers have to effectively build 5,000 one-liter bioreactors — and then train enough staff to be able to operate so many complex machines.

“I think it’ll take us years, though certainly not decades, to get to the point where we have near-complete automation,” Levine said.

To get around that hurdle, one company — the biotech Cellectis, which is headquartered in France — is developing what’s called an “allogeneic” CAR-T product. It’s basically a universal line of T-cells that, in theory at least, could be deployed in anybody. The goal is to cut out the step of extracting a patient’s own cells and reengineering them before putting them back in the body. If it works, it could save both time and money.

“With allogeneic cell therapy, we could have a manufacturing facility prepare the cells ahead of time, so they can be used immediately for treatment,” said Julianne Smith, vice president of CAR-T development at Cellectis, which is working with Pfizer. “We definitely think there’s a big advantage here, because we could increase accessibility.”

Cellectis’s allogeneic products have been tested — successfully — in two children with otherwise incurable leukemia, under a special compassionate use provision.

But some experts are dubious that it will work on a larger scale. Building universal T-cells that could work in anybody is, in a sense, as difficult as building a functional artificial kidney that could be implanted in any patient without fear of rejection. Our bodies may just be too difficult to trick, Levine said.

“There’s certainly a rationale for allogeneic therapies, but it’s important to recognize that our immune system has evolved over several hundred million years to distinguish self from non-self,” Levine said.

“The idea that there will be this bank of cells, and it’s going to be good for everyone as a permanent solution — that’s very far away, if it’s even possible,” Levine said.

Breaking ground on factories

Other companies are moving ahead on the more traditional model of customized CAR-T. Juno Therapeutics, for instance, has opened up a manufacturing site in Seattle that has the capacity to service the whole country, said Steve Harr, Juno’s chief financial officer. It could even provide CAR-T therapies internationally — provided Juno continues to focus on a relatively small niche of patients with select blood cancers.

“We’ve invested a lot since day one in process development and manufacturing,” Harr said. “With our current processes and our current size, we can treat thousands of patients from this facility — but not tens of thousands.”

Juno’s CAR-T machines have a modular design that can be expanded as demand grows, he said. The company’s vision, however, is to build a single machine that can manufacture customized cell therapies for thousands of patients. It has a prototype in the works, but it’s not yet ready for prime time, Harr said.

Kite Pharma held the groundbreaking for its CAR-T factory, near Los Angeles, in June 2015, but hasn’t yet done any large-scale manufacturing.

“Can we manufacture the cells? We’ve shown we can do it in a clinical study setting,” said Dr. David Chang, chief medical officer of Kite. “Can we scale up and manufacture for a large number of patients? We’ve done it as well as the other [companies] — and probably better, in terms of preparations.”

The company is now “moving towards commercializing and launching the product,” Chang said.

Solid tumors are tricky

CAR-T’s future hinges in part on whether scientists can figure out how to get it to work in the solid tumors — like ovarian and colon cancer — that make up the bulk of oncologic diseases. That has proved extremely challenging.

“With CAR-T cells, we have a Model T Ford, but we need a Ferrari to get into solid tumors,” said Dr. Renier Brentjens, director of cellular therapeutics at Memorial Sloan Kettering. He’s working with Juno on its CAR-T treatment.

CAR-T works by engineering the T-cells so they go after a molecule that’s exclusively found on the surface of a cancer cell — but that can’t be found on the surface of any healthy cell in the body.

Researchers have found the key molecules, or antigens, for blood cancers such as lymphoma and leukemia.

But when the National Cancer Institute tried to treat a patient who had a specific mutation, known as HER2, linked to colorectal cancer, the patient died.

The problem, Goff explained, was that the HER2 protein was present not just on the cancer cells but also on the normal cells that line the heart. So when the CAR-T latched onto any cell that contained the HER2 protein, it not only knocked out the colon cancer cells, but also attacked the tissues around the patient’s heart.

“I think that it’s cautionary, that one case report,” Goff said.

Some of the antigens targeted in CAR-T therapy for blood cancers also appear on some healthy cells. But they’re typically B cells, a component of the immune system — and humans can live without B cells, if they get the right kind of support. After receiving CAR-T therapy, patients get infusions of immunoglobulin, an antibody typically produced by B cells.

“It’s not that we’ve found a magic way to treat leukemia and lymphoma,” Goff said. “It’s just that with solid tumors, the targets we’ve found — well, they’re also on tissues that we need. They’re on your pancreas, on the lining of your heart, or your lungs — which aren’t tissues you can destroy without causing problems.”

Of course, it’s just early days for CAR-T therapy.

Another promising avenue, T-cell receptor therapy, or TCR, is also being developed by Kite, Ziopharm Oncology, and many other immunotherapy players. They hope it could prove to be more effective in tackling solid tumor cancers. TCR therapy works, in theory, by penetrating a cancer cell. It allows attacking immune cells to scope out targets that are on the inside of a cancer cell.

But cell therapies will need both clinical success and commercialization to keep up the momentum, Loncar said.

“What we have to keep in mind: What Novartis and Kite and Juno have in trials right now truly is version 1.1 — and companies are learning every single day from what they’re seeing,” Loncar said. “Over time, new versions of this are going to come out. So as a starting point, it’s extremely promising.”

Correction: An earlier version of this story misstated where the biotech firm Cellectis is based.