[Corrected, 11/1/17, 5:20 p.m. ET. See below.] The annual American Society of Hematology meeting is next month. But previews of the big data presentations are out today, yielding more insight into what promises to be the first major competition in CAR-T cell therapy: Treatments for people with non-Hodgkin lymphoma (NHL) who have run out of other options and likely have only a few months to live.
Yesterday, Novartis said it has asked the FDA to approve its CAR-T cell treatment tisagenlecleucel (Kymriah) for a certain kind of NHL. If approved, it would compete with axicabtagene ciloleucel (Yescarta) from Gilead Sciences (NASDAQ: [[ticker:GILD]]), which the FDA approved two weeks ago.
Today, Novartis and academic partners released new data from their key study meant to lead to approval, showing how a subset of patients have remained cancer-free for six months. In the abstract, Novartis (NYSE: [[ticker:NVS]]) and partners said that 46 patients could be assessed at the 6-month post-infusion mark, out of 99 who received a dose of their own modified T cells in the study, dubbed JULIET. Of those 46, 30 percent had a complete response—disappearance of cancer—and 7 percent showed a partial response, or some tumor shrinkage.
Those rates are about the same as what Novartis saw at the three-month mark: among the 81 patients who could be evaluated at that milestone, the complete response rate was 32 percent. Another 6 percent showed a partial response.
Why is that important? A big question about CAR-T is durability. Will the cells circulate in the patient for a long time, sniffing out and killing tumor cells? If not, cancers could come roaring back. The durability question is also a health economic issue: Insurers might balk at paying hundreds of thousands of dollars per treatment if the benefits for some patients are ephemeral.
Gilead is charging $373,000 for Yescarta. Novartis is charging $475,000 for Kymriah, which has been approved for kids and young adults with acute lymphoblastic leukemia. It was the first-ever CAR-T approval. That price tag comes with two caveats: Novartis promised to waive the charge for patients who don’t respond to treatment after one month, and it also said it would likely charge a different price for NHL, if approved.
Overall, Novartis enrolled 147 patients in JULIET and as of March gave Kymriah to 99 of them. Novartis has previously stated that the gap is due mainly to patients growing too sick to receive an infusion. As of June, nine patients could not be infused because of cell manufacturing problems.
It’s not clear how much the JULIET abstract data will be updated at next month’s conference. Novartis representatives did not respond by press time. But the more details about long-term effectiveness of Kymriah, the better Novartis’s chances of competing with Gilead’s Yescarta. (Gilead bought Kite Pharma in August to gain ownership of Yescarta and other Kite work.)
The data used for the approval of Yescarta hinted that cells were more effective cancer killers the longer they lived in a patient’s body. The complete response rate in 101 patients was 51 percent nearly nine months after their dose, on average. That’s an uptick from the roughly 35 to 40 percent complete response rate after three and six months.
Doctors involved in Gilead’s major study of Yescarta, called ZUMA-1, will present 12-month data at ASH next month. Those data were not included in the abstract released today.
Another big question about CAR-T is safety. In its ASH abstract, Novartis reported that a large majority of the patients, 86 percent, had serious side effects, but none died because of them. Three patients died soon after infusion, Novartis reported; all because their cancer progressed too rapidly. (All the data were as of March 8. For the ASH presentation in December, the outcomes could be updated beyond the March cutoff.)
NHL accounts for about 4 percent of all cancers in the United States. Only patients with a certain kind of NHL, diffuse large B-cell lymphoma, who have failed to benefit from other treatments would be eligible for CAR-T treatment, at least in this first wave. With this form of CAR-T, a patients’ T cells are extracted, shipped to a lab to become genetically modified cancer killers, then sent back and infused into the patient.
Another set of CAR-T NHL data released today, from Juno Therapeutics (NASDAQ: [[ticker:JUNO]]) of Seattle, adds to the competitive intrigue. In a trial dubbed TRANSCEND, Juno said a higher dose of its experimental JCAR017 prompted 11 complete responses (73 percent) and one partial response out of 15 “core” patients after three months. There are many other participants in the TRANSCEND study, but these core patients, the company says, are the true target of the therapy, because they have the same profile as the patients in the main Novartis and Kite studies. These higher-dose core patients received 100 million cells, while others in TRANSCEND received 50 million. [This paragraph has been changed to clarify the relationship between doses and core patients.]
Juno also reported data for more patients beyond the high-dose core group. Those data lowered the TRANSCEND three-month complete response rates to 50 percent. It’s not immediately clear if the FDA will also take into account the broader data if Juno asks for an approval decision.
For patients across all groups, the safety outlook has been positive. JCAR017 has produced relatively few safety problems: 44 of 69 have had no CRS, cerebral edema, or other neurological side effects. Only 1 of the 69 has suffered severe CRS, and 10 have had severe neurotoxicity, Juno reported.
Juno has a lot riding on TRANSCEND. JCAR017 is now its top product candidate after a high-profile blow-up of its leukemia CAR-T JCAR015 last year. Investors cheered the small data sample today. Juno shares were up 7.6 percent to $48.31 at Tuesday’s close.
Juno spokesperson Chris Williams said that the ASH presentation in December won’t have a significant update to these data, although some six-month results are likely. Those will be crucial: Juno has said its goal for approval-worthy data is to follow about 75 patients for at least six months. [This paragraph has been changed to correct the potential content of the ASH presentation.]
The upshot is that for NHL, Juno has excellent three-month durability data in a very small number of patients, while Novartis is starting to fill in the durability picture at six months. Meanwhile, Gilead and Kite are keeping a few laps ahead of their would-be competitors. They will unveil a year’s worth of durability data for Yescarta at ASH.
Image: Human lymphoma cells courtesy of the National Cancer Institute.
