See more of the story

SEATTLE – Ken Shefveland's body was swollen with cancer, treatment after treatment failing until doctors gambled on a radical approach: They removed some of his immune cells, engineered them into cancer assassins and unleashed them into his bloodstream.

Immune therapy is the hottest trend in cancer care and this is its next frontier — creating "living drugs" that grow inside the body into an army that seeks and destroys tumors.

Looking in the mirror, Shefveland saw "the cancer was just melting away." A month later doctors at the Fred Hutchinson Cancer Research Center couldn't find any signs of lymphoma in the Vancouver, Wash., man's body.

"Today I find out I'm in full remission — how wonderful is that?" said Shefveland with a wide grin, giving his physician a quick embrace.

This experimental therapy marks an entirely new way to treat cancer — if scientists can make it work, safely. Early-stage studies are stirring hope as one-time infusions of supercharged immune cells help a remarkable number of patients with intractable leukemia or lymphoma.

"It shows the unbelievable power of your immune system," said Dr. David Maloney, Fred Hutch's medical director for cellular immunotherapy, who treated Shefveland with a type called CAR-T cells.

"We're talking, really, patients who have no other options, and we're seeing tumors and leukemias disappear over weeks," added immunotherapy scientific director Dr. Stanley Riddell. But, "there's still lots to learn."

T cells are key immune system soldiers. But cancer can be hard for them to spot, and can put the brakes on an immune attack. Today's popular immunotherapy drugs called "checkpoint inhibitors" release one brake so nearby T cells can strike. The new cellular immunotherapy approach aims to be more potent: Give patients stronger T cells to begin with.

Currently available only in studies at major cancer centers, the first CAR-T cell therapies for a few blood cancers could hit the market later this year. The Food and Drug Administration is evaluating one version developed by the University of Pennsylvania and licensed to Novartis, and another created by the National Cancer Institute and licensed to Kite Pharma.

CAR-T therapy "feels very much like it's ready for prime time" for advanced blood cancers, said Dr. Nick Haining of the Dana-Farber Cancer Institute and Broad Institute of MIT and Harvard, who isn't involved in the development.

Now scientists are tackling a tougher next step, what Haining calls "the acid test": Making T cells target far more common cancers — solid tumors like lung, breast or brain cancer.

For all the excitement, there are formidable challenges.

Scientists still are unraveling why these living cancer drugs work for some people and not others.

Doctors must learn to manage potentially life-threatening side effects from an overstimulated immune system. Also concerning is a small number of deaths from brain swelling, an unexplained complication that forced another company, Juno Therapeutics, to halt development of one CAR-T in its pipeline; Kite recently reported a death, too.

And, made from scratch for every patient using their own blood, this is one of the most customized therapies ever and could cost hundreds of thousands of dollars.

In Seattle, Fred Hutch is working to tackle those challenges. At a recently opened immunotherapy clinic, scientists are taking newly designed T cells from the lab to the patient and back again to tease out what works best.

"We can essentially make a cell do things it wasn't programmed to do naturally," explained immunology chief Dr. Philip Greenberg. "Your imagination can run wild with how you can engineer cells to function better."

Small, early studies in the U.S. made headlines as 60 percent to 90 percent of patients trying CAR-Ts as a last resort for leukemia or lymphoma saw their cancer rapidly decrease or even become undetectable. Last week, Chinese researchers reported similar early findings as 33 of 35 patients with another blood cancer, multiple myeloma, reached some degree of remission within two months.

Too few people have been studied so far to know how long such responses will last. A recent review reported up to half of leukemia and lymphoma patients may relapse.

There are long-term survivors. Doug Olson in 2010 received the University of Pennsylvania's CAR-T version for leukemia. The researchers were frank — it had worked in mice but they didn't know what would happen to him.

"Sitting here almost seven years later, I can tell you it works," Olson, now 70, told a recent meeting of the Leukemia and Lymphoma Society.

If the FDA approves Novartis' or Kite's versions, eligible leukemia and lymphoma patients would be treated at cancer centers experienced with this tricky therapy. Their T cells would be shipped to company factories, engineered, and shipped back. Gradually, more hospitals could offer it.

"This is the hope of any cancer patient, that if you stay in the game long enough, the next treatment's going to be just around the corner," said Shefveland, the Hutch patient.