Two cellular medicines for multiple myeloma were associated with an increased risk of early death in clinical testing, raising questions about whether they should be earlier in a patient’s disease, Food and Drug Administration scientists wrote in a briefing document published on Wednesday. 

On Friday, the FDA will convene a panel of experts to discuss whether the two cell therapies — Legend Biotech and Johnson & Johnson’s Carvykti, and Bristol Myers Squibb’s Abecma — should be used as early as after a patient’s first relapse. 

The drugs are already cleared to treat patients following at least three lines of treatment, a setting in which both have proven powerfully effective. A clearance earlier would open up use more broadly in the U.S. European regulators have already endorsed such a decision. The requests are supported by large trials of Carvykti and Abecma that each enrolled hundreds of participants.

Analyzing the data, however, FDA reviewers identified better survival for patients on standard of care regimens, which included such medications as Darzalex and Pomalyst.

That elevated risk lasted through the first 11 months of the Carvykti trial and 15 months in the Abecma study. In the first three months after treatment, people receiving Carvykti had a six-times-higher relative risk of death than people taking comparator regimens containing such drugs as Darzalex and Pomalyst, while in the same time period, patients who got Abecma were more than twice as likely to die than the comparison drugs.

FDA reviewers said that, for Legend and J&J’s drug, the improved overall survival data over the long term was affected by incomplete information about the entire study population, and “the pattern of early deaths attributed to adverse events in the [Carvykti] arm is concerning.”

For both drugs, FDA staff concluded that “it is unclear whether the overall benefit-risk assessment is favorable.” They also critiqued Abecma for a lack of demonstrated survival benefit.

Carvykti and Abecma are so-called CAR-T therapies, which involve re-engineering patients’ own immune cells to attack tumors. The treatment is lengthy, requiring patients to have their blood drawn and to undergo an immune-suppressing chemotherapy regimen before the cells can be reinfused.

They’re also associated with low blood-cell counts and an immune-related side effect called cytokine release syndrome.

Seven of the 20 deaths attributed to adverse events in the Carvykti arm were from COVID-19 pneumonia, while only one patient in the standard therapy group died from COVID-19. Bristol Myers didn’t break out deaths due to COVID-19 for Abecma, but 14 of the 24 who died from adverse events in its trial were categorized as infections, compared to eight of the 12 who died from adverse events in the comparison group.

In a note to clients, RBC Capital Markets analyst Leonid Timashev wrote that the outlook may not be bad for Carvykti because the higher early risk is attributable to seven deaths, six of which were patients who were randomized to get the medicine but didn’t ultimately receive it.

Timashev, who covers Legend, added that Carvykti’s survival benefit grows with time. Those two factors, he argues, suggest “a high likelihood of a positive vote on Friday and an ultimate approval that can put the drug” on a trajectory toward $10 billion in annual sales.