Development of a novel dosing paradigm to model diazepam rescue therapy in preclinical seizure and epilepsy models

Abstract

Diazepam is a cornerstone immediate‐use antiseizure rescue therapy that may extend the duration between seizure clusters in people living with epilepsy. However, our mechanistic understanding of intermittent rescue therapy on disease progression is limited by the lack of suitable preclinical models. Specifically, the pharmacokinetics of diazepam varies widely between humans and laboratory animals. Here, we developed a novel repeat rescue therapy dosing paradigm in rats to maintain prolonged therapeutic concentrations seen in humans. Rats received three diazepam doses separated by 1 h (0.75, 1.5, or 3 mg/kg, intraperitoneal); plasma and brains were collected at 10 min and 1, 3, or 6 h following the last dose. Plasma and brain concentrations followed a dose‐dependent increase with peak concentrations following the repeat 3 mg/kg paradigm (180 ng/mL) being equivalent to plasma levels observed in human studies with diazepam nasal spray. Increased brain‐to‐plasma ratios in this paradigm indicate that diazepam accumulation in the brain may be long‐acting at the site of action. Overall, our repeat diazepam dosing paradigm mimics drug concentrations and accumulation seen in humans, offering a preclinical tool to study the impact of benzodiazepine rescue therapy on seizure‐cluster biology in rodent models of epilepsy.

Plain language summary

There is more to learn about how diazepam works in the brains of people who use it only when they have two or more seizures in 24 h (this is called a seizure cluster). Ethical studies in animals can be used to learn more about medicines in the body. In this study, we showed that three doses of diazepam in rats give about the same amount of the drug as one dose for a person. We can now test rats with epilepsy to see how the drug might work in people who take it when needed for seizure clusters.

Key points.

• Seizure cluster modelsare novel for mechanistic study of diazepam rescue therapy.

• A repeat‐dose paradigmin rats results in diazepam accumulation in plasma and brain.

• Plasma levels in ratsare within the therapeutic range of diazepam dosing in humans.

• This novel dosingparadigm can test the action of diazepam in seizures in epilepsy.

1. INTRODUCTION

Diazepam (DZP) is a benzodiazepine with anxiolytic effects that is used to treat many disorders, ¹ , ² which includes immediate‐use antiseizure medication as seizure‐cluster rescue therapy in patients with epilepsy. ³ A previous long‐term safety study of intermittent intranasal DZP reported an increase in the duration between seizure clusters in patients; ⁴ however, a preclinical animal model of DZP dosing for treatment of seizure clusters is needed to undertake mechanistic studies, including those on the long‐term impact of intermittent DZP rescue therapy and the potential effect of early intervention on disease progression.

The half‐life of DZP is substantially shorter in rats after intraperitoneal (IP) administration (~1 h) than intravenous (IV) administration in humans (~56 h). ⁵ , ⁶ Single‐dose pharmacokinetics has been described in rodents; however, repeat DZP dosing, particularly in the context of drug accumulation in the plasma and brain, is less studied. ⁵ , ⁷ Previous work by Walker et al. ⁷ suggests that repeat doses of DZP (20 mg/kg IP) can lead to persistent accumulation of DZP in the cerebrospinal fluid (CSF), an indirect marker of brain drug levels. However, significant tolerability issues are associated with doses of this magnitude, limiting translation to a clinically relevant dosing paradigm. A single 20 mg/kg IP dose can cause complete loss of righting reflex for >80 min. ⁸ Hence, a dosing regimen that maintains therapeutic DZP concentrations in the brain without producing substantial behavioral impairment is needed.

In this proof‐of‐concept study, we developed a repeat DZP dosing paradigm that resulted in peak plasma concentrations within clinical levels and brain concentrations that persisted for 6 h after dosing without signs of significant sedation or behavioral impairment. These findings offer a preclinical tool to study the impact of repeat benzodiazepine rescue therapy on seizure‐cluster biology in a rodent model of seizure clusters.

2. MATERIALS AND METHODS

Animal protocols were approved by the University of Washington Institutional Animal Care and Use Committee and consistent with the National Research Council's Guide for the Care and Use of Laboratory Animals and the Basel Declaration, including the 3R concept, and are reported in accordance with the ARRIVE guidelines. ⁹ , ¹⁰ Methodology is detailed further in Appendix S1.

Naive outbred adult male Sprague Dawley rats (~400 g) were obtained from Charles River Laboratories (Wilmington, MA). Pharmaceutical DZP formulations (Hospira, Inc.) were prepared daily in 40% PEG400 in 0.9% sterile saline and administered at doses of 0.75, 1.5, or 3 g/kg (n = 6 rats per dose) or 6 mg/kg (n = 3 rats). There were no a priori inclusion/exclusion criteria. No potential confounders were identified, and treatment groups were randomly assigned. Three doses of DZP were administered systemically (IP), unblinded, each 1 h apart (Figure 1). A separate group of rats (n = 6) received a single 3‐mg/kg dose to compare single versus repeat dosing. Blood samples were collected from the lateral tail vein 10 min following the final injection. For terminal collections at 1, 3, or 6 h after the final dose, rats were deeply anesthetized with isoflurane, and blood samples were collected via cardiac puncture before transcardial perfusion with phosphate‐buffered saline (PBS) and subsequent extraction of the brain. Plasma was separated from blood samples. Brains were extracted, roughly chopped, and flash‐frozen on dry ice. All samples were stored at −80°C until analytical quantification.

FIGURE 1.

Schematic of repeat DZP dosing paradigm (0.75, 1.5, 3 mg/kg) and sample collection for plasma and brain drug concentration analysis. DZP, diazepam.

Behavioral impairment was assessed to identify the maximum tolerated dose that did not induce substantial sedation. Righting reflex is a reliable indicator for assessing arousal and consciousness in rodents ¹¹ and is widely used for evaluating the effects of sedative compounds, including DZP. Briefly, this binary measure involves placing a rodent on its back and measuring whether it can right itself. A negative result (i.e., no observed righting reflex) implies a significant sedative effect of the drug. In this study, the righting reflex of each animal was measured by two investigators blinded to treatment 10 min after administration of each dose (time to maximum plasma concentration for single‐dose DZP IP = ~10 min ⁷ ). Also, stance and gait were observed but not quantitated.

Sample preparation and liquid chromatography–mass spectrometry (LC–MS) analysis was performed as an adaptation of Kim et al., ¹² with minor modifications, as follows: (1) the current analysis used a Waters' LC–MS System with different instrument settings and mobile phase; and (2) sample deproteination used acetonitrile, not ethyl acetate. Lower limit of quantification (LLOQ) of DZP was determined to be 10 ng/mL and 5 ng/g for plasma and brain samples, respectively.

Statistical analysis was performed using GraphPad Prism (Version 9.0.0). Results were calculated as mean concentrations ± SD. Dose–response relationship for drug concentrations was tested by two‐way analysis of variance (ANOVA) with main effects only for time and dose. Statistical significance was set at p < 0.05.

3. RESULTS

DZP was administered hourly for three IP injections at doses from 0.75 to 6 mg/kg. The highest dose tested, 6 mg/kg, resulted in significant impairment of righting reflex in animals after a single dose (3/3 animals) and was not tested further; no other data were excluded. Righting reflex was not impaired at doses up to the maximum tolerated dose (≤3 mg/kg IP), and the remaining three doses were assessed in the study. Sampling up to 6 h post‐dose revealed a dose–response relationship in plasma (Figure 2A; F[2,34] = 6.115; p = 0.005) but not brain concentrations of DZP (Figure 2B; F[2,17] = 1.844; p = 0.19). For the 3‐mg/kg repeat DZP dosing, mean (SD) plasma drug concentrations at 1 and 3 h were 180 (156) and 176 (56) ng/mL, respectively, whereas mean (SD) brain drug concentrations were 140 (126) and 161 (62) ng/g. Brain‐to‐plasma ratios suggest a modest accumulation of DZP in the brain as long as 6 h after the last dose (Figure 2C). However, DZP concentrations remained at detectable limits in both plasma and brain for up to 6 h following 3‐mg/kg dosing only (Figure 2D).

FIGURE 2.

Mean DZP concentration in (A) plasma and (B) brain in the rat at 10 min and 1, 3, and 6 h following three systemic injections of DZP, n = 3–6 animals/dose/time point. (C) Ratio of brain‐to‐plasma concentrations at 1, 3, and 6 h following the final administration of DZP in a repeat injection paradigm. (D) Mean plasma and brain levels at 6 h following DZP. *Significant dose–response relationship on plasma concentrations as determined by two‐way ANOVA with main effects only (plasma: F[2,34] = 6.115, p = 0.005; brain: F[2,17] = 1.844, p = 0.19). #Below LLOQ for assay. ANOVA, analysis of variance; DZP, diazepam; LLOQ, lower limit of quantitation.

Plasma DZP levels were significantly higher in animals receiving three sequential 3‐mg/kg doses on the repeat schedule than in those receiving a single 3 mg/kg injection (Figure 3A; F[1,24] = 4.945; p = 0.035). Brain concentrations were not statistically significant between the single‐ and repeat‐dose groups (Figure 3B; main effect, F[1,12] = 0.7772; p = 0.40), except at 6 h following injection, where brain levels were significantly higher following repeat dosing (Figure 3D; p = 0.028). Brain‐to‐plasma ratios suggest an increase in brain accumulation with the repeat‐dosing paradigm at 6 h after the last dose (Figure 3C,D).

FIGURE 3.

Mean DZP concentration in (A) plasma and (B) brain in the rat at 10 min and 1, 3, and 6 h following a single or three repeat injections of DZP (3 mg/kg), n = 3–6 animals/dose/time point. *Significant difference between injection paradigms on plasma concentrations as determined by two‐way ANOVA with main effects only (plasma: F[1,24] = 4.945, p = 0.035; brain: F[1,12] = 0.7772, p = 0.40). (C) Ratio of brain‐to‐plasma concentrations at 1, 3, and 6 h following the final administration of DZP in a single or repeat injection schedule. (D) Mean plasma and brain levels at 6 h following DZP. ^#Significantly different from 3 mg/kg (single) brain concentration as determined by Mann–Whitney U‐test, p = 0.03. ANOVA, analysis of variance; DZP, diazepam; LLOQ, lower limit of quantitation.

4. DISCUSSION

Herein, we confirm that repeat dosing with DZP results in plasma concentrations in rats that were within the range of therapeutic levels reported for humans at doses that did not cause significant sedation (ie, 3 mg/kg). Specifically, we found that the mean plasma DZP concentrations at 1 h (176 ng/mL) and 3 h (180 ng/mL) following the 3‐mg/kg dose are comparable to peak plasma concentrations found in clinical trials with DZP nasal spray. ¹³ , ¹⁴ Pharmacokinetic studies reported geometric mean peak plasma concentrations of 186 and 226 ng/mL (for 15‐ and 20‐mg doses, respectively) in healthy volunteers and 135 and 153 ng/mL (5‐ to 20‐mg doses) in patients with epilepsy in seizure and nonseizure conditions, respectively. Mean peak plasma concentrations were 164 to 338 ng/mL with oral and rectal DZP formulations in healthy volunteers. ¹³ , ¹⁴

Notably, limited direct information exists on the minimum DZP blood concentration needed for seizure termination. Preclinical data suggest that plasma concentrations associated with an elevated IV pentylenetetrazol seizure threshold are ~70 ng/mL, and concentrations as low as 30 ng/mL may be sufficient for seizure protection. ¹⁵ Two early clinical studies reported antiseizure effects with DZP plasma levels >130 ng/mL in two pediatric patients (one in each study). ¹⁶ , ¹⁷ Interictal spikes on electroencephalogram were reduced at DZP serum levels of 80 to 410 ng/mL (mean, 210 ng/mL) in adult patients with epilepsy, ¹⁸ whereas a study assessing photoconvulsive responses and spontaneous paroxysmal discharges found loss of suppression with DZP concentrations ranging from 100 to 500 and 500 to 1400 ng/mL, respectively. ¹⁹ Importantly, the concentrations seen in the repeat‐dosing model fall within clinical ranges associated with human antiseizure effects, thereby demonstrating that this repeat low‐dose DZP paradigm represents a clinically relevant dosing paradigm that is devoid of significant side effects and serves as a validated preclinical model for assessing the impact of repeat DZP dosing on seizure‐cluster biology.

These findings support the observation that DZP may accumulate in the brain after repeat injections. Others have shown that repeat dosing with substantially higher doses of DZP (20 mg/kg) result in accumulation in both serum and CSF. ⁷ However, even though not explicitly mentioned, it is likely that using such high doses resulted in significant sedation of the animal; for example, we found that a single 6‐mg/kg dose resulted in substantial behavioral impairment, whereas three repeat doses of 3 mg/kg did not cause significant sedation as measured by righting reflex. Notably, the reported TD50 of diazepam in rats is ~2.5 mg/kg (i.p.), indicating that a comprehensive analysis of behavioral impairment could reveal more subtle effects on motor coordination. ²⁰ However, considering the low protective index of diazepam in rodent seizure models where anticonvulsant doses are associated with tolerability issues, it would be unlikely to discover an effective dosing paradigm without any side effects. Our goal was to develop a dosing paradigm achieving measurable and prolonged concentrations at clinically relevant doses that did not render the animal incapacitated. Together, these data suggest that DZP accumulation in the brain may be long‐acting at its site of action even at low doses that do not cause significant sedation in the animal.

A dosing paradigm leading to drug accumulation in the rat brain offers a preclinical tool for mechanistic studies to assess DZP's role in the treatment of seizure clusters, including its disease‐modifying potential. Of particular interest is a post hoc analysis of data from a long‐term safety study of DZP nasal spray that found a significant increase in time between seizure‐cluster episodes (seizure interval, or SEIVAL). ⁴ To explore possible causes of this observation, preclinical research is underway to characterize long‐term seizure‐cluster presentation after repeated DZP dosing using this novel dosing paradigm in an etiologically relevant rat model of temporal lobe epilepsy (TLE). ²¹ , ²² Notably, seizure clusters are common in post‐status epilepticus models of TLE with estimates of >50% of rodents developing some degree of clustering to their seizure patterns, ²³ , ²⁴ , ²⁵ which highlights a suitable model for studying novel intervention strategies on seizure‐cluster biology.

Study strengths include that dosing led to plasma DZP levels comparable to those observed in humans after intranasal administration and within the effective range for seizure suppression. ¹³ , ¹⁴ Moreover, the dosing paradigm resulted in no functional impairment in test animals, strengthening the clinical translation of a paradigm that may be devoid of side effects. Finally, DZP levels were measured directly in the brain, whereas previous research measured CSF levels as a surrogate for brain accumulation. ⁷ A limitation is that 3‐h post‐dose samples were only collected from the 3‐mg/kg dose group; however, given the strong dose–response relationship noted at other time points, similar dose‐dependent changes are expected in the lower dose groups. Moreover, the LLOQ for DZP was higher in plasma than in brain samples, highlighting the possibility that brain‐to‐plasma ratios could exceed 1 at early time points. However, these DZP concentrations are below levels reported to be required for seizure protection, ¹⁵ suggesting that any increase in brain‐to‐plasma ratio at those time points may be negligible. Finally, we acknowledge the limitation of not considering metabolites, such as N‐desmethyl‐diazepam (nordiazepam), oxazepam, and temazepam, that may contribute to the effects of diazepam. However, we sought to develop a dosing paradigm that mirrors clinical rescue therapy scenarios, where the focus is typically on the parent compound's pharmacokinetics and efficacy. Given that the repeat‐dosing paradigm achieved clinically relevant therapeutic concentrations of diazepam, our findings provide valuable insights into the potential use of diazepam immediate‐use antiseizure medication as a rescue therapy for seizure clusters. However, future work will address the various contributions of rapid diazepam metabolism when attributing pharmacodynamic effects to changes in seizure‐cluster biology in a disease model of spontaneous recurring seizures.

5. CONCLUSION

In this study, a repeat‐dosing paradigm of DZP in rats resulted in sustained and measurable DZP levels in brain and plasma that mimicked plasma levels found in humans after dosing with currently approved products and without significant sedative effects. This dosing paradigm can be used in preclinical in‐life studies designed to assess the disease‐modifying effects of DZP on spontaneous seizure clusters when applied to well‐studied in vivo models of epilepsy.

AUTHOR CONTRIBUTIONS

All authors contributed to conceptualization. MG and HSW contributed to study design, execution, acquisition of data, and analysis. All authors provided interpretation, took part in writing, revising, or critically reviewing the article and gave final approval of the version to be submitted; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

FUNDING INFORMATION

Development of this manuscript was funded by Neurelis, Inc.

CONFLICT OF INTEREST STATEMENT

Dr. Guignet received research funding from Neurelis, Inc. Dr. White has received grant funding from Neurelis, Inc., UCB Pharma, and Eisai Pharmaceuticals; consultant fees from Biogen, GW, Neurelis, Inc., Takeda, Inc., and JAZZ Pharmaceuticals; and speaker honoraria from SK Pharmaceuticals and UCB Pharma. He is also co‐founder of NeuroAdjuvants, Inc. Dr. Misra was, at the time of this study, an employee of and has received stock options from Neurelis, Inc. Dr Rabinowicz is an employee of and has received stock options from Neurelis, Inc. Dr. Carrazana is an employee of and has received stock and stock options from Neurelis, Inc. We confirm that we have read the journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

Supporting information

Appendix S1:

Guignet M, White HS, Misra SN, Carrazana E, Rabinowicz AL. Development of a novel dosing paradigm to model diazepam rescue therapy in preclinical seizure and epilepsy models. Epilepsia Open. 2024;9:1575–1581. 10.1002/epi4.12954

DATA AVAILABILITY STATEMENT

All data generated or analyzed during this study are included in this published article.