The Food and Drug Administration on Wednesday approved a new therapy to treat leukemia in kids and young adults—a decision whose importance is as much symbolic as it is practical.

Kymriah, from the Swiss pharmaceutical company Novartis, is a cancer therapy that represents several things at once: a game-changing way to treat cancer through genetic engineering, a novel paradigm for the biotech business, and the latest turn in the debate over just how astronomically expensive a life-saving therapy can be.

Kymriah is strikingly effective for young patients with acute lymphoblastic leukemia, or ALL, but it is far more involved than taking a pill or getting an infusion. It requires inserting a human-designed gene into a patient’s own T cells so they recognize and ferociously attack cancer cells. Researchers began modifying T cells for patients in the 1990s—and now the technology called CAR T-cell therapy is finally ready for prime time in treating cancer.

Of several dozen ALL patients in a clinical trial for Kymriah, 83 percent were cancer-free after three months. It is a lifeline for patients in which traditional treatments like chemotherapy and bone-marrow transplants had failed. When the FDA’s advisory committee initially voted in favor of approving Kymriah, one member called it “the most exciting thing I’ve seen in my lifetime” for childhood leukemia. Novartis is hardly the only company interested in CAR T. Kymriah is the first approved therapy, but several clinical trials—mostly notably Kite Pharma’s for lymphoma—are right behind it.

(To clear up any possible confusion about terminology: The FDA and others have chosen to call CAR T-cell therapy a form of gene therapy—and thus deemed it the first gene therapy to be approved in the United States. There is a debate over whether CAR T truly is gene therapy, since the genetic engineering is not curing a disease. You can go down the semantic rabbit hole on this.)

For all these promising results, CAR T-cell therapy has its challenges, too. CAR T-cell therapy activates the immune system to attack cancer cells, but sometimes it can activate the immune system too much. It has adverse effects like high fever and life-threatening brain swelling. Last year, another CAR T-cell therapy company, Juno Therapeutics, stopped its clinical trial after five patients died of the latter.

More immediately, for Novartis, CAR T-cell therapy is a manufacturing challenge. It involves taking blood from a leukemia patient, separating out their T cells, using a virus to add a gene to the T cells so they recognize cancer cells, growing more T cells, and finally sending it back to the hospital with the patient. Novartis has said it expects to take 22 days to manufacture each patient’s customized therapy. Losing track of a patient’s cells at any point in the process could be disastrous. You can give the same drug to any patient; you cannot give a patient anyone else’s CAR T cells with Kymriah.

The only precedent for such personalized therapy is the prostate-cancer treatment Provenge, from the biotechnology company Dendreon. Provenge was never successful, says Biren Amin, a biotech analyst, because its benefits were unclear and the cost was high. (The company Amin works for, the Jefferies Group, counts the CAR T-cell therapy company Kite Pharma among its clients.) Fortunately for Novartis, it was able to snap up Dendreon’s manufacturing facilities and staff after the smaller company failed.

There may be a way to side step this cumbersome manufacturing process, however. Another biotechnology company, Cellectis, has pinned its hopes on an off-the-shelf version of CAR T made from donated cells. This requires other strategies to prevent the engineered cells from attacking the patient’s body. Cellectis recently infused its first patient in a clinical trial.

André Choulika, the CEO of Cellectis, told me last month that the off-the-shelf approach just makes more business sense. “The individualized approach is more a business of the hospital than the biotech company,” he said. And initially, at least, the first centers to offer CAR T-cell therapies like Kymriah will likely be specialized cancer centers.

And then there’s the issue of cost. Novartis today announced that it will charge $475,000 for Kymriah, but made clear it thought the therapy’s value was closer to $600,000 to $750,000. “Recognizing our responsibility to bring this innovative treatment to patients, we have set the price for Kymriah below that level at $475,000 for this one-time, single-administration treatment,” the company said in a statement. “We believe this will support sustainability of the health-care system and patient access while allowing a return for Novartis on our investment.”

Novartis had come under pressure by Patients for Affordable Drugs, an advocacy group that recently met with Novartis executives to discuss pricing. It’s not unusual for cancer drugs to cost hundreds of thousands of dollars a year, but the price is still eye-popping as an absolute number.

For now, the era of CAR T-cell therapy for cancer is certainly upon us. But for reasons of logistics and cost, the era in which it is widely available is not.