So close, yet so far: Zogenix LGS trial succeeds, but results underwhelm
Merely crossing the finish line isn’t enough — the win must be decisive. Ask Zogenix, whose experimental seizure drug met the main goal in a pivotal study in patients with Lennox-Gastaut Syndrome (LGS) on Thursday, but saw its shares plummet after the magnitude of the therapy’s effect fell short.
The drug, fenfluramine (to be branded Fintepla) — one-half of the notorious axed prescription weight-loss cocktail fen-phen — was tested against a placebo in 263 patients with the rare, treatment-resistant form of childhood epilepsy. The data comes months after Zogenix’s revamped application was accepted for review in patients with Dravet syndrome, another rare seizure disorder.
In the LGS study, patients on the higher dose (0.7 mg) of the drug saw a statistically significant median reduction of 26.5% in seizure frequency, versus 7.8% in patients taking placebo (p=0.0012). Other secondary endpoints including patients with ≥50% reduction in monthly drop seizures; the proportion of patients improved; and the proportion of patients much improved or very much improved were all in favor of the drug.
“The difference (18.7%) is substantially smaller than the ~60% reduction in seizures over placebo seen for the high-dose cohort in the Dravet fenfluramine trial,” Stifel’s Paul Matteis wrote in a note.
Matteis raised his approval probability in LGS to 80% but said that the poor magnitude of efficacy prompted a reduction in his market share estimate to 15% from 20%.
Last April, Zogenix said the FDA had rebuffed reviewing Fintepla’s marketing application for Dravet patients, citing the lack of certain non-clinical studies key for the assessment of chronic administration of the drug as well as an incorrect version of a clinical dataset. In November, the FDA accepted the company’s resubmitted application, and the agency is expected to make its final decision by March 25.
“Epidiolex absolute seizure reduction in LGS was greater than seen here, however, the placebo effect in GW’s studies is consistently much bigger, which may be due to the expectation bias that consistently hovers around CBD. This dynamic was first observed in Dravet, where the placebo arm in GW’s study showed a 13% reduction in convulsive seizures, vs 1.1% and 1.2% in ZGNX’s two Dravet ph3’s. Similarly, in LGS, while Epidiolex showed 42% and 44% reductions in drop seizures, the placebo arms in the two studies showed reductions in seizure frequency of 17% and 22%,” Matteis said.
“We’re therefore unsurprised to see a significantly smaller placebo effect in fenfluramine’s LGS study (just 7.8%), and any analysis of the absolute seizure reduction on Fintepla needs to be interpreted in the context of this dynamic. Patients in both Zogenix and GW’s LGS studies look fairly severe, and Zogenix made the case that the Fintepla improvement in seizure reduction, when looked at as a multiple of placebo, was robust.”
Overall, the efficacy of Fintepla and Epidiolex is likely more similar than different in LGS (compared to the large difference in Dravet), Matteis concluded. “Therefore if one models Fintepla as the more expensive option, which we do, it probably makes sense to assume that it’s used later in the LGS treatment paradigm.”