Kymriah’s shocking failure in earlier lymphoma treatment has raised doubts about the Novartis drug’s potency in light of impressive victories from rival CAR-Ts by Gilead Sciences and Bristol Myers Squibb. With detailed data now coming in, the Swiss pharma believes it has a pretty good idea what went wrong.
Compared with standard of care, Kymriah showed no additional benefit in shrinking tumors or staving off disease progression or death in patients with aggressive B-cell non-Hodgkin lymphoma after one line of treatment. It wasn’t just a narrow miss by statistical significance: Patients in both groups who either got Kymriah or standard of care—which involves chemotherapy followed by stem cell transplant—lived a median three months without a negative event that indicates treatment failure.
Kymriah’s complete lack of showing came in stark contrast to the more than 60% reduction on that event-free survival marker posted by Gilead’s Yescarta and BMS’ Breyanzi in the same second-line lymphoma setting. The data were presented at the American Society of Hematology (ASH) annual meeting.
When Novartis first unveiled the flop in August, Cowen analysts said it “may be taken as evidence to support some physician views that Kymriah is less potent than competitors.” But Novartis doesn’t believe that Kymriah is inferior in any way, Jeff Legos, Ph.D., Novartis’ global head of oncology and hematology development, said in an interview.
As Legos sees it, Kymriah’s phase 3 BELINDA trial in second-line lymphoma was failure by a thousand cuts. Among the factors that might have contributed to the final outcome, Legos pointed to the long time to infusion with Kymriah as the “most relevant factor.”
For existing CAR-T cell therapies, a qualified medical center draws a patient’s T cells and ships them to a lab to be engineered to specifically fight cancer. In the BELINDA trial, the median time from the withdrawal of T cells to reinfusion of the final product was 52 days, Legos said.
By comparison, the median time from patient randomization to infusion was 29 days for Gilead’s Yescarta in its ZUMA-7 second-line trial, including five days to prepare a patient for the blood withdrawal, Frederick Locke, M.D., of the Moffitt Cancer Center and primary investigator of the study, told reporters during a press briefing.
The time window to treatment is important, because patients in the Kymriah trial had very aggressive disease, with most refractory to first-line treatment, Legos explained. As a result of having a long wait, at the time of what Legos called “true randomization,” which was the actual treatment, the patient populations were already different between the two arms. That’s because the Kymriah group had more patients whose disease was already progressing or had a higher volume of disease, the Novartis exec contends.
What’s more, the BELINDA trial’s definition of event-free survival counts failure to achieve a response at the Week 12 assessment as a negative incident. But because of the long gap to treatment, some patients may not have adequately responded to Kymriah by that point, Legos said. Some patients responded to Kymriah after 12 weeks without any additional therapy, the trial investigators noted.
Further, Legos noted that patients in the Kymriah arm had some of the poorest risk factors that weren’t captured at the time of randomization, leading to some imbalances in patient prognostic characteristics between the two arms of the BELINDA trial.
But even if Kymriah’s trial failure can be explained by the long time to infusion, product turnaround time is a key factor in CAR-T therapy. That 52-day number from the BELINDA trial doesn’t present the Novartis product in a favorable light.
For his part, Novartis’ Legos says the number doesn’t reflect what Novartis has experienced for Kymriah in the real world. The time to infusion was roughly evenly split between manufacturing and logistics in the study, suggesting that COVID-19 might have contributed to the extended timeline on the logistics front.
Manufacturing hiccups previously tripped up Kymriah’s launch in third-line diffuse large B-cell lymphoma (DLBCL) back in 2018. Industry experts Cowen recently spoke with said that the early manufacturing problems with Kymriah have made them more inclined to reach for Yescarta and that they perceive the Gilead drug as having better efficacy despite some neurotoxicity concerns, the analysts wrote in a report late September.
Now, the BELINDA trial results will likely solidify those opinions. Still, Jason Westin, M.D., leader of DLBCL research at the MD Anderson Cancer Center, said he wouldn’t rule on whether Kymriah is inferior to the other CD19-targeted CAR-T therapies given the flaws with the BELINDA trial. Westin is an investigator in both BELINDA and Yescarta’s ZUMA-7.
In its approved third-line DLBCL indication, Kymriah still has convincing clinical data and positive real-world experience, Westin said. But with Yescarta and Breyanzi now looking set to move up the treatment line, there will be less patients left for Kymriah in the third line, the expert said.
For Novartis, despite an understanding of the BELINDA trial’s problems, the company doesn’t plan to repeat a second-line trial for Kymriah, Legos said. That’s because its focus is shifting.
With a next-generation CAR-T platform called T-Charge, the company aims to reduce the manufacturing timeline for immune cells to less than two days and thus cut the overall timeline “from vein to door” by at least half, Legos said.
Besides a shortened timeline, T cells made from the platform are more potent and have better ability to self-proliferate inside the body, Legos said. Therefore, a therapy would need fewer cells. Novartis is also reporting at ASH 2021 positive first-in-human data for two products coming out of the platform targeting CD19 and BCMA, including a candidate dubbed YTB323, which has a similar construct as Kymriah.
“Instead of repeating the same experiment with our first-generation Kymriah, we’re really focusing on conducting our next experiment,” Legos said. “And we will move YTB323 into definitive pivotal registration trials in 2022, looking at a similar patient population of patients who’ve failed first-line chemoimmunotherapy.”