Authors :
Presenting Author: Mohamed Fayed, MS – University of Minnesota
Lisa Coles, MS, PhD – University of Minnesota
Usha Mishra, MS – University of Minnesota
Marcia Terluk, PhD – University of Minnesota
Ned Patterson, DVM, PhD, DACVIM – University of Minnesota
Su-youne Chang, Ph.D. – Mayo Clinic Rochester
Filip Mivalt, PhD – Mayo Clinic Rochester
Gregory Worrell, MD,PhD – Mayo Clinics
James Cloyd, PharmD – University of Minnesota
Ronald Siegel, ScD – University of Minnesota
Rationale:
Timely intervention in seizure emergencies is critical in preventing neurological injury. Benzodiazepines, including diazepam (DZP) (given IV and IN), are first-line rescue therapies for seizure clusters. Avizafone (AVF), a water-soluble DZP prodrug, when co-administered with aminopeptidase B (APB), is converted to supersaturated DZP in the nasal cavity, enabling rapid absorption and onset of EEG changes similar to IV dosing. The present study characterizes DZP pharmacokinetics (PK) following IN AVF/APB and IV DZP, models the PK-pharmacodynamic (PD) relationship using intracranial EEG (iEEG) as the measure of response, and assesses safety in a translational canine model.
Methods:
Two dogs with implanted wireless iEEG electrodes received IN AVF/APB at 1.0 and 1.5 mg/kg DZP equivalents, and a 0.5 mg/kg bolus IV DZP in a three-way cross-over design with at least 2-week washout periods. Serial blood samples were collected up to 8 hours post-dose. Plasma DZP concentrations were quantified by HPLC-MS/MS. Noncompartmental and compartmental models were used to estimate PK parameters. The iEEG gamma-beta power in band (PIB) was analyzed and correlated with plasma concentrations to identify onset of PD effects using a sequential PK-PD modeling approach. Emax models with and without an effect compartment were evaluated.
Results:
IN AVF/APB resulted in rapid DZP absorption (Tmax = 2–5 minutes) with Cmax values of 470 - 1140 ng/ml, which were dose-dependent. IN bioavailability was approximately 40%; However, one dog exhibited notably lower bioavailability (13%) due to substantial loss of drug at the time of administration of the 1.0 mg/kg dose. Marked increases (50% of baseline) in PIB were observed within 3 minutes following IN administration and 1 minute following IV dosing. This effect remained above 50% of baseline for approximately 20 mins. A temporal delay between plasma concentration and PD effect was observed. First-order absorption best described the PK/PD relationship, along with a two-compartment PK model indirectly linked to a sigmoidal Emax PD model via an effect compartment. The concentration corresponding to 50% of maximum increase in PIB ranged from 270 to 380 ng/ml. No significant adverse effects were observed.
Conclusions:
IN DZP prodrug+converting enzyme administration produces a rapid CNS effect approaching the response following IV delivery, an ideal characteristic for rapid, needle-free seizure rescue therapy. Further optimization is warranted to improve bioavailability and reduce variability in systemic exposure. Results support continued development toward clinical translation of this approach.
Funding:
This study was supported by the Office of Discovery and Translation (ODAT), University of Minnesota, and the Mayo Medical Center.