Otonomy Provides Update on OTIVIDEX Program

Management will review the OTIVIDEX statistical analysis plan update together with the positive top-line OTO-313 Phase 1/2 trial results during a conference call today at 4:30 p.m. ET

SAN DIEGO, July 06, 2020 (GLOBE NEWSWIRE) -- Otonomy, Inc. (Nasdaq: OTIC), a biopharmaceutical company dedicated to the development of innovative therapeutics for neurotology, today provided an update on the statistical analysis plan related to the ongoing Phase 3 clinical trial of OTIVIDEX in Ménière’s disease. In response to questions received from the U.S. Food and Drug Administration (FDA) regarding use of the Generalized Poisson model to analyze the daily vertigo count data reported by patients, Otonomy submitted a revised statistical analysis plan that uses a statistical test called the Negative Binomial model for the primary analysis of the ongoing trial.

Otonomy selected the Negative Binomial model to address the FDA’s questions because it believes it provides the best fit of the OTIVIDEX clinical data based on the Phase 2b trial (for patients with vertigo enrollment criteria matching the Phase 3 clinical trials), the AVERTS-2 Phase 3 clinical trial, and the integrated dataset from both trials. As the table below indicates, the ad hoc analysis of the Definitive Vertigo Day (DVD) count data reported by patients for Month 3 is statistically significant (p value < 0.05) using the Negative Binomial model for each of these prior trial populations as well as the integrated summary.

p value for Analysis of DVD Count in Month 3	Phase 2b* (n = 97)	AVERTS-2** (n = 111)	Integrated Dataset
Generalized Poisson Model	0.002	0.013	< 0.001
Negative Binomial Model	0.016	0.008	< 0.001

*Patients with baseline DVD count of 4-22 days during the one-month baseline period
**Patients who completed 3-month follow-up period (of which 105 completed daily diaries)

“We have recently completed an extensive review of the statistical analysis model that we believe best fits the OTIVIDEX clinical data, which is characterized by approximately 40% of OTIVIDEX-treated patients having no DVD’s in Month 3,” said David A. Weber, Ph.D., president and CEO of Otonomy. “Based on this review, we determined that the Negative Binomial model provides an improved fit and reduced Type 1 error compared to the Generalized Poisson model. The Negative Binomial model also gives us additional power to detect a treatment benefit enabling us to comfortably reduce the target patient enrollment in the ongoing trial from 160 to 142 patients while maintaining more than 90% power. We look forward to completing the enrollment of this trial during the third quarter of 2020 and announcing results in the first quarter of 2021.”