One of the biggest challenges for the companies now racing to develop T-cell therapies for cancer is figuring out how to make personalized, living products on an assembly-line scale.

Each patient's own T-cells - the soldiers of the immune system - must be siphoned from the blood, coaxed to multiply, genetically engineered to recognize and attack cancer cells, then returned to the patient.

Now, Cellectis, a French biotechnology company partnering with Pfizer, says it has used gene-editing technology to achieve a major advance: a "universal" T-cell product, made with healthy donor cells and used "off-the-shelf." The therapy successfully eliminated terminal leukemia in two British babies treated at Great Ormond Street Hospital in London, the company says.

The company disclosed the first case in November. That girl, now 23 months old, has been cancer free for almost a year.

The second case, a 21-month-old girl who remains in remission five months after treatment, was announced earlier this month at a scientific conference.

The universal T-cells "are 'off-the-shelf' products, whose production can be industrialized and standardized . . . over time and from batch to batch," Cellectis said in a news release.

Robert M. Frederickson, editor of the journal Molecular Therapy, hailed the announcement of the first case in an editorial titled "A new era of innovation" for T-cell therapy.

"The intervention . . . has induced remission where all other treatments had failed," Frederickson wrote.

A more skeptical view came from University of Pennsylvania pathologist Bruce Levine, who is part of the team that was the first to report stunning blood cancer remissions with T-cell therapy five years ago.

Levine pointed out that both babies received intensive chemotherapy, then got Cellectis' T-cells, and then underwent stem-cell transplants, which completely replace the patients' blood-making systems.

"It's very hard to know what was the effect of those T-cells unless you have more data and very discrete measurements," Levine said. "I think the consensus [in the field] is that we don't know that the T-cells did - I'm using a scientific word - bupkes."

In response, Cellectis vice president Julianne Smith said the two babies "had already gone through previous stem-cell transplants and relapsed." They received the T-cell therapy under United Kingdom rules for compassionate use of unapproved therapies in dire situations.

"The chance of survival is only 10 to 15 percent after a second stem-cell transplant," Smith said. "Time will tell [how good the therapy is]. More patients have to be treated. But for now, it's already a pretty good sign."

The only personalized cellular therapy for cancer to reach the market so far was a flop. Dendreon's Provenge was approved by the U.S. Food and Drug Administration in 2010, but it was costly and barely effective against prostate cancer. In 2012, Dendreon sold its production plant to Novartis, which is commercializing Penn's T-cell therapy. Dendreon declared bankruptcy in 2014.

"Dendreon's example shows that cell therapy products can be brought to market, but that cost and efficacy . . . will be important factors determining their success or failure," Frederickson wrote in his editorial.

The biotech industry, he noted, has "seen several boom-and-bust cycles" and the latest rush toward commercialization "might end in disappointment."

Penn's T-cell pioneer, Carl June, tallied the scope of that rush last year in the journal Science Translational Medicine. He listed nearly three dozen pharmaceutical and biotech companies trying to develop T-cell therapies. Even though experimental versions have been effective only against certain blood cancers - and the effectiveness has varied - developers are banking on extending the technology to treat solid tumor malignancies.

The vast majority of these firms are working on "autologous" products, which are tailor-made with the patients' own T-cells.

While this approach is inefficient, the resulting product has the advantage of being compatible with the patient's immune system. If not compatible, the patient's body would reject the engineered T-cells as invaders, while the T-cells could attack the patient's healthy tissue.

Still, many researchers are trying to do what Cellectis says it has achieved - create T-cells that can work in anyone.

"We are pursuing it," Levine said of Penn's team. "We plan to be into the clinic by the end of year" to test a version of universal T-cells.

Cellectis says it overcame the immune incompatibility problem by using a gene-editing technique to cut out two bits of DNA from the T-cells. One snip rendered the T-cells unable to target healthy cells. The other snip made the T-cells resistant to a drug that was then used to temporarily suppress each baby's immune system.

Making the immune system too weak to kill invaders gave the universal T-cells time to attack the leukemia cells - at least, until the babies' entire blood-making systems were destroyed and regenerated through stem-cell transplants.

Ultimately, Smith said, Cellectis intends to enhance the T-cell therapy so "it will not be a bridge to transplant, but the treatment in itself." Cellectis recently signed a deal with Servier, which has an agreement with Pfizer to co-develop UCART19 and expects to begin clinical testing later this year.

"There will be more efficient techniques" to overcome the immune incompatibility problem, he said. "But we're not there yet."