'A negative trial, not a failed trial': Neurocrine axes in-licensed tremor program on phase 2 data

Two years after licensing the asset for $45 million upfront, Neurocrine Biosciences has pulled the plug on the development of NBI-827104 in essential tremor after the failure of a phase 2a clinical trial to make the case for further investment. 

Neurocrine put down $5 million to secure an option on the T-type calcium channel blocker, then known as ACT-709478, in 2019. The following year, the biotech paid Idorsia $45 million to exercise the option, putting it on the hook for up to $365 million in development and regulatory milestone payments. Neurocrine moved the candidate into a phase 2 essential tremor trial last year.

Now, management has data from the phase 2 study—and the signs are bad. With the phase 2a missing its specified endpoints, including on the TETRAS essential tremor scale, Neurocrine has concluded the totality of the data fail to make the case for more development in essential tremor. No further work is planned in the indication.

“It was not a failed trial, it was a negative trial,” Eiry Roberts, M.D., chief medical officer at Neurocrine, said in a conference call with investors. “We saw some evidence of activity, but the signal was not sufficiently robust to hit statistical significance or to hit clinical relevance, either on the accelerometer or on the TETRAS endpoints that we looked at.”

“We had central assessments as well as local assessments on the TETRAS and there was high concordance between the two sets of assessments,” Roberts added. “So it was a robust data set.” 

The news comes months after Praxis hailed the success of its T-type calcium channel blocker in a phase 2a essential tremor study. Jazz also has a runner in the race, suvecaltamide, that it moved into a phase 2b study late last year.

Neurocrine is pushing ahead with a phase 2 clinical trial of NBI-827104 in epilepsy with continuous spike-and-wave during sleep. Enrollment is complete and top-line data are due by the end of the year. Because of the differences between the two patient populations, Roberts sees little readthrough from the essential tremor results to the epilepsy readout.