ORLANDO — Novartis has shipped CAR-T cell therapies to about 1,800 people with blood cancer, a noteworthy accomplishment for a personalized product made from patients' own immune cells.
But some of those people have received treatments that, while safe and effective, don't meet commercial product specifications set out by the Food and Drug Administration when it approved the Swiss drugmaker's Kymriah for leukemia and lymphoma.
And 10% of the time, Novartis hasn't been able to ship a product at all, either due to out-of-specification issues or from manufacturing failures.
Novartis say it has made progress in addressing the issues, and is working with the FDA to put in place solutions. But that's the same message the company shared a year ago, suggesting a lingering hitch in Kymriah's production.
"We have identified the main factors where we can further strengthen the manufacturing process," said Stefan Hendriks, newly Novartis' global head of cell and gene, in an interview at the American Society of Hematology's annual meeting in Orlando. "We're collaborating with the FDA and implementing that into next year."
The drugmakers' case should be helped by real-world data on commercial use of Kymriah, presented at ASH, that showed out-of-specification CAR-T product equally capable at attacking cancer with no evidence of greater safety risk.
In the meantime, however, Novartis can't charge for out-of-specification treatments, delivering Kymriah in those cases through an expanded access protocol that provides the therapy for free. It's not clear how much this has affected Novartis sales of Kymriah, but at minimum it's another hurdle in selling a medicine that's proved uniquely challenging to commercialize.
Sales of Kymriah totaled $182 million through the first nine months of 2019, higher than what Novartis earned last year but lower than some expectations for CAR-T. Hendricks said Novartis is confident Kymriah, which costs $475,000 for leukemia and $373,000 for lymphoma, will reach $1 billion in annual sales within a few years.
CAR-T cell therapies are bespoke treatments. Patient T cells are extracted and then frozen to ship to a manufacturing facility, where the cells are genetically engineered to seek out cancers expressing certain proteins. The souped-up immune cells are then refrozen, shipped back to a medical facility and reinfused into the patient.
It's a complicated, often manual process that takes Novartis 21 days on average to complete. Patients receiving CAR-T are also very sick, making product turnaround time critical.
As with all drugs, there's a checklist of attributes the engineered cell dose must meet in order for Novartis to release the product for use in patients.
One has proved problematic for Novartis. When Kymriah was approved, the FDA set out a requirement that Kymriah product must consist of at least 80% "viable" T cells, meaning they're alive and active. That was a change from clinical testing when Novartis aimed to clear a 70% viability threshold, and is higher than in the other countries in which Kymriah is approved.
In the real-world data at ASH, 29 of 102 lymphoma patients treated commercially and evaluated for the study received an out-of-specification dose with less than 80% viable cells. Among leukemia patients included in a second analysis, 14 of 92 received a dose below 80% viability.
Importantly, there was no evidence that those receiving doses between 60% and 80% viability fared worse. All 14 of the leukemia patients achieved a complete response, according to the study abstract. And a similar number of lymphoma patients given out-of-specification product responded to treatment as those receiving in-specification Kymriah.
When it approved the cell therapy, the FDA identified non-viable cells as a "potential clinical safety issue." The new results suggest that may not be the case.
"Now we have more and more evidence that above and below 80% has the same safety and efficacy, that also will strengthen our dialogue with the FDA," said Hendricks.
Stephan Grupp, an oncologist at Children's Hospital of Philadelphia and a lead investigator of one of Novartis' Kymriah studies, told BioPharma Dive the difference between 70% viability and 80% viability had not worried him previously. Grupp was lead author on the real-world analysis of commercial Kymriah use in leukemia patients.
"We did [the analysis] not because there was a concern, but more to show there wasn't," said Samantha Jaglowski, an oncologist at The Ohio State University Comprehensive Cancer Center, in an interview. Jaglowski was lead author on the other analysis of Kymriah use in lymphoma patients.
Part of the reason the percentage of viable cells may not matter as much is due to another product release specification that lays out the minimum absolute number of viable T cells expressing the chimeric antigen receptor, or CAR.
Novartis' Hendricks said the company is both working to ensure all of its commercial doses of Kymriah meet the 80% threshold, as well as discussing with the FDA whether the requirement in the drug's label could be lowered.
At the same time, Novartis has worked to broaden its manufacturing network for Kymriah, recently opening a new production facility in Stein, Switzerland. The drugmaker also last year bought the French contract manufacturer CellforCure.
Hendricks said both sites will be able to produce commercial product next year, which will "create relief" at Novartis' principal site in Morris Plains, New Jersey.
"By having a global network, we'll be much more flexible to meet demand whenever it arises," said Hendricks.