Epilepsy & Behavior 51 (2015) 191–198

Contents lists available at ScienceDirect

Epilepsy & Behavior

journal homepage: www.elsevier.com/locate/yebeh

Comparison of short-term effects of midazolam and lorazepam in the

intra-amygdala kainic acid model of status epilepticus in mice
Mairead Diviney, James P. Reynolds, David C. Henshall ⁎
Department of Physiology and Medical Physics, Royal College of Surgeons in Ireland, Dublin, Ireland

a r t i c l e i n f o a b s t r a c t

Article history: Benzodiazepines remain as the first-line treatment for status epilepticus (SE), but debate continues as to the
Received 26 June 2015 choice and delivery route of pharmacotherapy. Lorazepam is currently the preferred anticonvulsant for clinical
Revised 27 July 2015 use, but midazolam has become a popular alternative, particularly as it can be given by nonintravenous routes.
Accepted 28 July 2015
Anticonvulsants are also commonly used to terminate SE in animal models. Here, we aimed to compare the effi-
Available online 24 August 2015
cacy of midazolam with that of lorazepam in an experimental model of focal-onset SE. Status epilepticus was in-
Keywords:
duced by intra-amygdala microinjection of kainic acid in 8 week old C57Bl/6 mice. Forty minutes later, mice were
Status epilepticus treated with an intraperitoneal injection of either lorazepam or midazolam (8 mg/kg). Electroencephalogram
Anticonvulsants (EEG) activity, histology, and behavioral tests assessing recovery of function were evaluated and compared be-
Midazolam tween groups. Intraperitoneal injection of either lorazepam or midazolam resulted in similar patterns of reduced
Animal welfare EEG epileptiform activity during 1-hour recordings. Damage to the hippocampus and presentation of postinsult
3Rs (Replacement, Refinement, Reduction) anxiety-related behavior did not significantly differ between treatment groups at 72 h. However, return of nor-
Behavior mal behaviors such as grooming, levels of activity, and the evaluation of overall recovery of SE mice were all su-
perior at 24 h in animals given midazolam compared with lorazepam. Our results indicate that midazolam is as
effective as lorazepam as an anticonvulsant in this model while also providing improved animal recovery after SE.
These data suggest that midazolam might be considered by researchers as an anticonvulsant in animal models of
SE, particularly as it appears to satisfy the requirements of refining procedures involving experimental animals at
early time-points after SE.
© 2015 Elsevier Inc. All rights reserved.

1. Introduction the route of administration, where intravenous delivery is not always

Status epilepticus (SE) is a life-threatening neurological emergency
and is regarded as one of the most extreme and severe forms of seizure
activity. It is defined as a prolonged seizure and, operationally, as five
minutes of continuous seizure activity or as two or more seizures with-
out complete recovery between seizure episodes [1]. Status epilepticus
is associated with a risk of significant damage involving brain structures
such as the hippocampus, cognitive impairment, and epileptogenesis
[2,3]. Status epilepticus is also associated with high morbidity and mor-
tality [4], and a critical variable in these outcomes is the duration of the
SE period [5]. Specifically, the longer that SE persists, the more likely it is
to become unresponsive to drug therapy, resulting in poor patient out-
comes [6]. As such, SE requires prompt treatment using anticonvulsant
drugs [7]. However, there remains a debate regarding the most effective
prehospital treatment regimen with a particular consideration being
Abbreviations: SE, status epilepticus; EEG, electroencephalogram.
⁎ Corresponding author at: Department of Physiology and Medical Physics, Royal
College of Surgeons in Ireland, 123 St. Stephen's Green, Dublin 2, Ireland. Tel.: +353 1
402 8629.
E-mail address: dhenshall@rcsi.ie (D.C. Henshall).
practical.
The cellular and molecular mechanisms by which SE develops are in-
completely understood but are thought to be from a combination of ex-
cessive excitatory and impaired inhibitory neuronal mechanisms [8].
Current treatment, therefore, includes administration of benzodiaze-
pines which act by potentiating GABAergic inhibition [9]. The most
widely used agents for first line therapies for SE in the clinical environ-
ment are. lorazepam and diazepam [10]. Lorazepam remains the phar-
macotherapy of choice as it has fewer contraindications, such as lower
incidence of respiratory depression [11]. The mode of delivery of these
drugs is via intravenous (i.v.) administration, and while effective, estab-
lishing i.v. access in a patient having a seizure can be difficult and time-
consuming [12]. These treatments can be administered intramuscularly
(i.m.); however, they are absorbed more slowly [8], and due to their
vehicle (i.e., propylene glycol), irritation at the injection site can occur.
An alternative benzodiazepine, midazolam, can be safely administered
i.m. enabling quick and accurate delivery. It has also been shown to be
highly effective at terminating seizures, both when given as an initial
agent for SE and when given as second line therapy for refractory
SE [13,14]. Recent evidence has also emerged to suggest that the anti-
seizure efficacy of midazolam for SE is equivalent to lorazepam in

http://dx.doi.org/10.1016/j.yebeh.2015.07.038
1525-5050/© 2015 Elsevier Inc. All rights reserved.
192 M. Diviney et al. / Epilepsy & Behavior 51 (2015) 191–198
prehospital settings [12,15,16]. Patients treated with midazolam also
had better short-term outcomes, with a significantly lower number of
patients requiring hospital admittance than those treated with loraze-
pam [12]. Lorazepam, however, has often been chosen over midazolam
in clinical treatment as it has a longer duration of action, with one of the
major disadvantages of midazolam being tachyphylaxis, resulting in the
need for increased dosing to maintain seizure control [17]. Furthermore,
with prolonged infusion, midazolam begins to accumulate, resulting in
a protracted time to consciousness [18]. Therefore, it can be difficult
to reconcile the need for immediacy of treatment with longer term out-
comes, adding further complexity to decisions around anticonvulsant
treatment.
Animal models of SE are widely used to study epilepsy and
epileptogenesis and are crucial in the development of effective thera-
peutics [19,20]. Antiseizure drugs are often administered to reduce sei-
zures during SE and minimize morbidity and mortality. Indeed, the
timing of benzodiazepine administration has time-dependent effects
on attendant brain damage and epilepsy development [21]. The choice
of anticonvulsant is, therefore, important for ensuring satisfactory
interanimal reproducibility in seizure suppression and postinsult recov-
ery. Commonly, either lorazepam [22,23] or diazepam [24] is adminis-
tered via i.v. or intraperitoneal (i.p.) injection in models of chemically
induced SE in rodents. Little is known, however, about the performance
of midazolam in rodent SE models. A small number of studies have
reported that midazolam provides protective effects against pentylene-
tetrazol (PTZ)-induced seizures and also results in a reduction in overall
time spent in seizure [25,26]. Others have focused on the pharmacody-
namics of midazolam in various epilepsy models [27]. Differing behav-
ioral profiles in parameters such as spontaneous motor activity and
muscle relaxation have also been noted in naïve mice following treat-
ment with four benzodiazepines, suggesting that drug type may influ-
ence recovery [28].
For many years, our group has used systemic injection of diazepam
or lorazepam as anticonvulsants in the intra-amygdala kainic acid
(KA) model in mice [22,29]. Here, we aimed to compare the effect of
midazolam with that of lorazepam as an alternative anticonvulsant
in this model, focusing on EEG, hippocampal injury, and recovery of
normal behavior patterns in an experimental mouse model of SE.
2. Materials and methods
2.1. Seizure model
All animal experiments were performed in accordance with the
European Union Directive (2010/63/EU) and were reviewed and ap-
proved by the Research Ethics Committee of the Royal College of Sur-
geons in Ireland, under license from the Health Products Regulatory
Authority, Ireland (AE19127/P001; AE19127/I089). All experiments ob-
served the principles outlined in the ARRIVE guidelines and the Basel
declaration including adherence to the 3Rs (Replacement, Refinement,
Reduction). Adult (20–25 g) male C57BL/6 mice were purchased from
Harlan (UK). Animals were housed in a vivarium on a 12-hour light/
dark cycle with access to food and water ad libitum. Mice were anesthe-
tized using isoflurane (3–5%) in oxygen and placed in a mouse-
adapted stereotaxic frame. Following a midline scalp incision, bregma
was located and three partial craniectomies were performed for the
placement of skull-mounted recording screws (Bilaney Consultants)
to record EEG (see Fig. 1A). A fourth craniectomy was drilled for
the placement of a 26-gauge steel guide cannula (coordinates from
bregma; AP = − 0.95 mm, L = − 2.85 mm) based on a stereotaxic
atlas [30]. The cannula and electrode assembly were fixed in place
using dental cement (Plastics One, Inc.), anesthesia was discontinued,
and mice were placed in an incubator until they had recovered and
were freely moving. For EEG recordings, mice were placed in a clear
Perspex recording chamber which allowed free movement. The EEG
was recorded using a Grass Comet digital EEG (Medivent Ltd, Lucan,
Ireland). Baseline EEG was established over a 10-minute recording
period. Following this, the animal was lightly restrained while a
31-gauge injection cannula was lowered 3.75 mm below the brain sur-
face for injection of KA [0.3 μg in 0.2-microliter vehicle (phosphate-
buffered saline (PBS)), pH adjusted to 7.4; Sigma-Aldrich] into the
basolateral amygdala nucleus to induce SE. After 40 min, mice were
administered with either lorazepam or midazolam (see Section 2.2).
Electroencephalogram was recorded for a further 1 h thereafter before
animals were disconnected and placed in a warmed recovery chamber.
Mice were sacrificed 72 h after anticonvulsant administration by pento-
barbital overdose and perfused with ice-cold PBS to remove intravascu-
lar blood components. Whole brains were flash-frozen in 2-
methylbutane (at −30 °C) and stored at −80 °C for cryostat sectioning
(Fig. 1B).
2.2. Drug treatments
For anticonvulsant treatment, animals were randomized to either
lorazepam (LZ, 8 mg/kg) or midazolam (MDZ, 8 mg/kg; n = 6/group).
Drugs were delivered via a single intraperitoneal injection in 0.25-
milliliter volume of either PBS (lorazepam) or sterile H2O (midazolam),
40 min after intra-amygdala KA injection.
2.3. Recovery and behavior scoring
Following SE, mice were housed individually for behavior assess-
ments. To determine the effect of treatment on recovery, mice were
observed at 4 h, 24 h, 48 h and 72 h postanticonvulsant treatment.
Measurements included weight, with baseline weight measured imme-
diately postsurgery — this was to account for additional head-piece
weight (i.e., EEG electrode, cannula, and cement assembly) — and
were monitored at each time-point post-SE. Behaviors observed fell
into a number of distinct categories including: 1) well-being scoring,
2) activity-related behavior, 3) grooming behavior, and 4) attentive
behavior. A modified welfare score sheet which comprised of four cate-
gories including appearance, locomotion, unprovoked behavior, and
behavior to external stimuli was used (see Table 1). These categories
are modified from Kirsch et al. [31] and Roughan and Flecknell [32].
Subscores for each category were determined, and the total, overall
score was calculated by adding all subscores. Scores were interpreted
as: 0 to 3, normal; 4 to 7, animal requires careful monitoring; 8 to 11, an-
imal appears in distress and requires regular observation; and 12 to 15,
severe distress [31]. Activity, grooming, and attentive behavior were
scored in relation to the number of occurrences within a 10-minute ob-
servation period. Activity-related behaviors included ambulation, head
turns, drinking, eating, and digging, climbing, and biting bars. Grooming
included licking fur, washing face, scratching body, and licking paws.
Attentive behavior included sniffing, high rearing, and low rearing.
The observer scoring all behaviors was blinded to treatment group.
2.4. Exploratory and anxiety-related behaviors
Seventy-two hours after SE, exploratory behavior and anxiety were
assessed in mice from both treatment groups using the open-field task
[33,34]. Each mouse was placed individually in the center of a clear,
plexiglass open-field apparatus (ENV-510; 27.9 × 27.9 cm; Med
Associates) and allowed to freely explore for 10 min. Arenas were
wiped with 70% ethanol before each session and between mice. An
overhead camera captured all animal movements, and video recordings
were later analyzed using EthoVision videotracking (Noldus Informa-
tion Technologies, Wageningen, Netherlands). To assess exploratory be-
havior, distance traveled (cm) and velocity (cm/s) were measured.
Anxiety was determined by measuring time spent in either a demarcat-
ed central zone or peripheral zone of the arena. Prior to behavior testing,
mice were allowed to acclimatize to the testing environment for 1 h
 M. Diviney et al. / Epilepsy & Behavior 51 (2015) 191–198 193

Fig. 1. Electroencephalogram and status epilepticus in mice treated with LZ or MDZ. (A) Schematic showing cannula and electrode implantation site. (B) Experimental design illustrating
time of intra-amygdala KA injection followed by 40-minute recording post-KA and a further 60-minute recording after anticonvulsant treatment. (C) Representative EEG traces depicting
electrographic seizure development following KA and also after treatment with either LZ or MDZ. Below, an EEG heat map depicting typical frequency–amplitude data during status
epilepticus (SE) and after anticonvulsant administration. Electroencephalogram recordings show no differences between mice in LZ (n = 6) or MDZ (n = 6) treated groups during SE
in (D) total power or (E) amplitude during SE. Antiseizure treatment type also does not result in significant alterations in (F) total power or (G) amplitude during the 60 min after admin-
istration of either LZ or MDZ.

prior to experimentation. The observer was blind to the treatment of
the mice.
2.5. Histopathology
Brains (n = 6 per group) were sectioned at 12 μm on a cryostat
(Leica) in the coronal plane and stored at −80 °C until further use. For
detection of neurodegeneration, sections at the level of the dorsal hip-
pocampus (−1.7 mm) [30] were processed for Fluorojade B (FJB) [35]
(Millipore, Cork, Ireland). Flash-frozen tissue sections were postfixed,
dehydrated, immersed in 0.06% potassium permanganate solution
followed by a rinse, and then stained with 0.001% w/v FJB (Millipore
Ireland B.V., Tullagreen, Ireland). Staining was examined using an
epifluorescence microscope (Nikon 2000s) under Ex/Em wavelengths
of 472/520 nm (green), and positive cells within the hippocampus
were counted in a blinded fashion under 20× magnification. Represen-
tative images were obtained using an Orca 285 camera and processed
using Wasabi software (Hamamatsu Photonics Germany GmbH,
Herrsching, Germany). Images were converted to grayscale and
inverted such that degenerated neurons appeared dark on a light back-
ground. Hippocampal FJB counts were carried out on areas CA3, CA1,
and the hilus and taken as the mean of two adjacent sections. The ob-
server quantifying FJB positive cells was blinded to treatment group.

2.6. Quantification of EEG

Table 1
Well-being recovery evaluation scoring.
Electroencephalogram data were exported to Labchart 7 (AD Instru-
Category Score Operational Definition ments Ltd, Oxford, UK) for analysis. Baseline EEG was recorded for
Appearance 0 ○ Normal, haircoat smooth, flat with sheen, eyes clear 10 min for each animal. The effects of drug treatments on EEG were
1 ○ Ruffled fur, lack of grooming, lubricant in eyes determined for the 60-minute recording period after injection. Mean
2 ○ Rough hair, pointy, hunched look, eye and nose discharge
amplitude and total power were calculated, and measurements were
3 ○ Ungroomed, abnormal posture, eyes glazed, animal
appears severely depressed normalized to baseline for each animal. The total power was automati-
Locomotion 0 ○ Walking normally cally measured by the software with the default frequency range of
1 ○ Limping, stiffness 0–500 Hz. Quantification of EEG was performed by an observer blinded
2 ○ Swollen limbs to treatment. Treatment group was revealed after all analyses were
3 ○ Severely restricted mobility
Unprovoked 0 ○ Normal — exploring cage, grooming, feeding, bright, alert
complete and only revealed at the time when statistical comparisons
behavior & responsive were conducted.
1 ○ Altered slightly, huddled
2 ○ Abnormal behavior, reduced mobility, alertness or 2.7. Data analysis
responsiveness, inactive, lame, aggressive or huddled
3 ○ Severe distress, vocalizations, restless/no movement,
unresponsive, twitching, lameness All data are presented as mean ± standard error of the mean (SEM).
Behavior to 0 ○ Normal response to reaching into and tapping cage Two group comparisons were made using independent t-tests, while
external 1 ○ Shows minor exaggerated responses repeated measures analyses were made using analysis of variance
stimuli 2 ○ Moderate abnormalities, more aggressive or docile (ANOVA) followed by appropriate post hoc testing. Significance was
3 ○ Overreaction to stimuli or nonresponsive
accepted at p b 0.05.
194 M. Diviney et al. / Epilepsy & Behavior 51 (2015) 191–198

3. Results
All mice used in this study (n = 12; n = 6 LZ; n = 6 MDZ) survived
SE induced by intra-amygdala KA, and all reached the planned end-
point at 72 h post-SE.
3.1. Effect of anticonvulsant treatment on EEG after SE
Electroencephalogram was recorded for all mice for 10 min prior
to injection of KA and referenced as a baseline (100%) to determine
changes in EEG during and after SE. Following intra-amygdala injection
of KA, all mice entered SE with no differences in EEG total power be-
tween the mice later treated with lorazepam (LZ) (587.7 ± 149.0%) or
midazolam (MDZ) (648.1 ± 129.4%; t(10) = 0.31, p = 0.76) or in am-
plitude between the groups (LZ: 186.6 ± 18.97; MDZ: 214.5 ± 26.66%;
t(10) = 0.85, p = 0.41; Fig. 1C–E). This would indicate that all
mice, prior to random assignment to treatment group, experienced
similar SE. To assess the effect of drug treatment on the reduction of
EEG activity after SE, total power and mean amplitude were analyzed
for a 60-minute period following anticonvulsant drug administra-
tion. No differences were observed in EEG total power between LZ
(398.0 ± 122.9%) and MDZ (328.5 ± 127.2%) treated mice (t(10) =
0.54, p = 0.59). Similarly, analysis of mean amplitude also indicates
that treatment with MDZ (163.9 ± 6.38%) did not alter the pattern
of seizure termination when compared with LZ treatment (173.2 ±
27.46%; t(10) = 0.33, p = 0.75; Fig. 1F, G).
3.2. Effect of anticonvulsant treatment on neurodegeneration after SE
To determine whether there were differences in hippocampal dam-
age between the MDZ and LZ groups, we next examined hippocampal
damage in tissue sections from mice from each treatment group, 72 h
after SE. Semiquantitative counts of FJB positive cells in ipsilateral CA3
indicated no significant difference between LZ-treated (62.75 ± 6.9)
and MDZ-treated mice (76.17 ± 15.92; t(10) = 0.77, p = 0.45;
Fig. 2A). Similarly, no differences in FJB counts were noted between
groups in the ipsilateral CA1 (t(10) = 0.036, p = 0.94) and the hilus
(t(10) = 0.08, p = 0.93; Fig. 2B, C). There were also no differences
observed in any subregion of the contralateral hippocampus (see repre-
sentative image Fig. 2D).
3.3. Midazolam improves speed of recovery after SE
We next examined a number of behavioral parameters to determine
any changes in the recovery and well-being of mice following the termi-
nation of SE. Mice were assessed for weight, alongside a number of
other parameters which would indicate the occurrence of any alter-
ations as a result of anticonvulsant treatment on recovery from SE.
The postsurgical weight of mice was recorded at baseline and change
in weight was measured at 4 h, 24 h, 48 h, and 72 h after SE. A repeated
measures ANOVA revealed an overall reduction in weight for both
groups across time (F(4, 40) = 12.76, p b 0.001), with a significant re-
duction from baseline to 4 h (p b 0.001), 24 h (p b 0.001), and 48 h
(p b 0.002). However, there was no difference in weight between treat-
ment groups overall (F(1, 10) = 0.038, p = 0.85; Fig. 3A).
We next sought to determine if there were changes in recovery
evaluation scores (see Table 1). Mice were observed over a 10-minute
period in their home cages at each of the four time points. A repeated
measures ANOVA indicates an overall improvement in general recovery
scores over time (F(3, 30) = 96.75, p b 0.001). There was no difference
in recovery between groups (F(1, 10) = 2.73, p = 0.13); however, a sig-
nificant interaction effect was found for time × group (F(3, 30) = 5.57,
p b 0.01). Further analysis revealed that MDZ-treated mice displayed
significant improvements at 24 h in comparison to LZ-treated mice
(p b 0.05; Fig. 3B). Grooming, a good indicator of normal behavior
in mice, also improved over time for both groups (F(3, 30) = 14.03,
p b 0.001), and while there was no overall effect for group (F(1,
10) = 3.09, p = 0.109), there was a significant time × group interaction
effect (F(3, 30) = 6.31, p b 0.01). Crucially, further analysis suggests that
MDZ had a significant impact on initial recovery after intra-amygdala
injection of KA, with a significantly higher level of grooming behavior
in MDZ than LZ mice 24 h after SE (p b 0.05; see Fig. 3C). Similarly, the

Fig. 2. Hippocampal damage 72 h after status epilepticus in LZ-treated and MDZ-treated mice. FJB-positive cell counts in the subfields of the ipsilateral hippocampus 72 h after SE
(n = 6/group). Treatment type did not significantly alter the level of neuronal death in ipsilateral (A) CA3, (B) CA1, or in the (C) hilus. (D) Representative photomicrographs of FJB-stained
neurons in the ipsilateral (right) and contralateral (left) hippocampi at 72 h for each group. Note the little damage evident in the contralateral hippocampus with no differences between
groups in the contralateral area CA3 (p = 0.41), CA1 (p = 0.27) and the hilus (p = 0.50). Scale bar: 500 μm.
 M. Diviney et al. / Epilepsy & Behavior 51 (2015) 191–198 195

Fig. 3. Welfare and typical behavior assessment of mice treated with LZ and MDZ. (A) There was an initial reduction in weight 4 h and 24 h post-SE; however, treatment had no effect
on weight change. (B) Improvements in welfare were observed across days in all mice. Midazolam-treated mice display improved typical behaviors and appearance at 24 h in comparison
to LZ-treated mice. (C) Significant improvements in levels of activity across time in all mice were noted. Again, significantly higher active behavior was seen in mice treated with
MDZ. (D) Patterns of grooming behavior indicate increased activity at 24 h for MDZ mice when compared with LZ mice. (E) The ability of mice to attend to environmental cues also sig-
nificantly improved over time. Again, treatment with MDZ resulted in a higher level of occurrence of this behavior when compared with LZ treatment at 24 h (p b 0.001). *p b 0.05,
**p b 0.01, ***p b 0.001.

levels of activity-related behavior increased significantly over the 72 h
after treatment for both groups (F(3, 30) = 14.43, p b 0.01) with a sig-
nificant difference in the level of activity between treatment groups
(F(1, 10) = 5.11, p b 0.05). Further analysis indicates that MDZ mice
were, again, significantly more active than LZ mice at 24 h (p b 0.001;
Fig. 3D), displaying higher occurrences of ambulatory behavior, head-
turns, climbing, eating, and drinking. The occurrence of attentive behav-
iors, such as high and low rearing and sniffing also increased for both
groups at each time point subsequent to KA injection (F(3, 30) =
16.01, p b 0.001). Treatment did not appear to have an overall effect
on the level of attentive behaviors (F(1, 10) = 2.01, p = 0.187); how-
ever, a significant interaction effect was noted between group and time
after SE (F(3, 30) = 3.03, p b 0.05), with MDZ-treated mice showing sig-
nificant improvements over LZ mice at 24 h after treatment (p b 0.001;
Fig. 3E). On all behavior measures, differences between treatment
groups were no longer present after the 24-hour time-point as mice
continued to recover over the subsequent 48 h.
3.4. Exploratory and anxiety-related behaviors
Seventy-two hours after SE, exploratory behavior and anxiety were
assessed in mice using the open-field task to determine if the treat-
ment had any gross effects on animal behavior. To assess exploratory
behavior, distance traveled (cm) was measured with no differences
noted between LZ (2582 ± 545.1 cm) and MDZ (2845 ± 588.6 cm)
treated mice (t(10) = 0.87, p = 0.74; Fig. 4B). Similarly, assessment
of velocity (cm/s) in the arena identified no significant differences
between groups (LZ: 4.23 ± 0.95 cm/s; MDZ: 4.72 ± 0.99 cm/s;
t(10) = 0.36, p = 0.72, Fig. 4C) suggesting that anticonvulsant treat-
ment type did not differentially affect exploratory behaviors in
the open-field task. Anxiety was assessed by measuring time spent in
either a demarcated central zone or peripheral zone of the open-field
arena. Animals normally spend greater time exploring the periphery,
and this was found for both groups (Fig. 4D). There was no significant
difference in total time spent in the peripheral zone between LZ

Fig. 4. Exploratory and anxiety-related behaviors in mice treated with lorazepam and midazolam. (A) Representative image of the open-field and demarcated peripheral and central zones.
The open-field, carried out 72 h after SE, revealed no differences between treatment groups (n = 6/group) in (B) distance traveled (p = 0.74) or (C) velocity (p = 0.72). Assessment of
anxiety-related behavior revealed no differences in percentage of time spent in a (D) peripheral zone (p = 0.40) and (E) central zone (p = 0.43) between LZ or MDZ mice (p N 0.05).
196 M. Diviney et al. / Epilepsy & Behavior 51 (2015) 191–198
(68.46 ± 14.04%) and MDZ (81.21 ± 3.75%) treated mice (t(10) = 0.87,
p = 0.40; Fig. 4D). Similarly, there was no difference between the LZ
(32.96 ± 13.99%) and MDZ (21.01 ± 4.22%) groups in time spent in
the central area of the arena (t(10) = 0.81, p = 0.43; Fig. 4E).
4. Discussion
The main findings of the present study are that midazolam was
noninferior to lorazepam in the treatment of SE in an experimental
model. Our data indicate that midazolam was as effective as lorazepam
in curtailing seizures during SE and did not alter the level of acute neu-
ronal loss as a result of prolonged seizures, indicating that it was neither
significantly more protective nor deleterious. There was also no impact
of treatment on later-measured anxiety-related behavior or in locomo-
tor activity in an open-field task. Surprisingly, we found that midazolam
was superior to lorazepam in initial animal recovery 24 h after SE. Be-
havioral measures including welfare monitoring, grooming, and activity
levels all indicated that midazolam-treated mice displayed better out-
comes than lorazepam-treated mice at the early stages of recovery.
Thus, midazolam offered equivalent seizure-stopping performance
and reliable histopathology while also improving early recovery.
Time is one of the key factors in treating SE, with the overall dura-
tion of seizure activity affecting later outcomes [5,36]; as such, early
treatment is essential. There is disagreement, however, on the best
treatment regimen in clinical settings [17,37]. The factors to be consid-
ered include ease of delivery, access routes, onset, and duration of ac-
tion, as well as recovery time. In the present study, we explored the
efficacy of midazolam in a model of partial-onset SE, which is the
most common form in patients with epilepsy [38]. The intra-
amygdala (i.a.) kainic acid model in mice has been in use since 2001
and is becoming a popular model due to its consistent and reliable
SE response, reproducible hippocampal pathology, being relatively
high-throughput, with minimal mortality and versatility to translate
to other strains and transgenic lines [22,39–41]. Here, we find evidence
to support the use of midazolam as an anticonvulsant treatment in this
experimental model of SE. Our first major finding was that while both
lorazepam- and midazolam-treated animals experienced similar epi-
sodes of SE as measured by EEG, no differences were found between
groups during a 1-hour recording after anticonvulsant administration.
This would indicate that midazolam is as effective as the frontline
treatment lorazepam in terminating SE.
Furthermore, the level of neuronal loss was also similar between
groups. The i.a. KA model is associated with a highly reproducible injury
to the ipsilateral hippocampus, and the damage seen here was very
similar to that reported previously in this model when performed by
another researcher [23]. Here, we found again that damage was mainly
confined to area CA3, with some lesser and more variable injury to CA1.
There was also no difference in damage between lorazepam-treated
and midazolam-treated mice at 72 h. This fits with the EEG data and
with previous studies having found a close association between seizure
duration and resulting damage [22,42]. Since hippocampal pathology is
a necessary feature of the i.a. KA model in order to assess potential
disease-modifying treatments and in longer-term studies of epilepsy,
midazolam could be used in place of lorazepam while displaying the
advantage of improved animal recovery at early time-points.
When looking at the mechanism of action of these drugs, it is
perhaps not surprising that we see similar effects of midazolam and
lorazepam treatment in these parameters. Similar to lorazepam and
other benzodiazepines, midazolam acts primarily through potentiating
the benzodiazepine–GABAA receptor complex [43]. This enhances the
inhibitory action of GABAA by increasing the number of chloride channel
openings resulting in the hyperpolarization of neurons. Status epilepti-
cus is thought to result from a failure of the normal mechanisms for sei-
zure termination, in particular, a decrease in GABAA receptor function
[44] and an increase in glutamate receptor density [1]. In line with this
view, both benzodiazepine treatments resulted in similar reductions
in seizure activity, supporting the utility of midazolam as a treatment
in experimental SE.
While the electrographic and histologic features of the i.a. KA model
are well documented, relatively less is known about post-SE recovery
and behavioral deficits. Work to date has focused on hippocampus-
dependent tasks [40], but here, we show that a number of behaviors
are altered post-SE in mice. Examination of these other outcome param-
eters revealed significant differences between treatment groups, with
midazolam-treated animals displaying a better recovery of function at
24 h after SE. Specifically, midazolam-treated mice displayed higher
levels of overall activity including eating, drinking, and ambulation, as
well as a higher frequency of grooming behavior and ability to attend
to environmental alterations. While no differences in EEG between
groups during SE would indicate that any alterations in behavior are
due to anticonvulsant effects and treatment, there may be, albeit unlike-
ly, alternative targets beyond seizure management resulting in im-
proved outcomes. Although there are many similarities among the
various benzodiazepine compounds, the pharmacokinetics vary and
may account for clinically relevant differences between drugs. The supe-
riority of midazolam at 24 h after administration is potentially a result of
elimination half-life differences between therapies. The elimination
half-life in humans for lorazepam is 12–15 h, significantly longer than
midazolam at approximately 3 h [37]. These are, however, more rapidly
eliminated by rodents with an elimination half-life plasma of 1.5 h for
lorazepam [45] and 0.5 h for midazolam [46] and, as such, may not
fully explain the improvements seen. Midazolam's shorter duration of
action is due to its rapid metabolism by the liver via the cytochrome
P450 enzyme system to active and inactive metabolites [47]. Whether
this is beneficial or not remains contentious. Some argue that due to a
short duration of action, seizure relapse can occur in patients in the ab-
sence of repeated dosing [17]. However, in the current study, the supe-
rior animal recovery over lorazepam treatment at 24 h suggests the
utility of midazolam in treating SE in this experimental model. These
data could also assist researchers in meeting obligations under the
3Rs (replacement, refinement, and reduction), now a mandatory re-
quirement under EU legislation.
Our findings are also in agreement with work comparing the clinical
efficacy of various benzodiazepine therapies. Crucially, our findings sup-
port a randomized control trial that specifically compared the efficacy of
i.m midazolam treatment with i.v. lorazepam in the treatment of SE in a
nonhospital setting [12]. Patients treated with midazolam displayed at
least comparable outcomes as those administered lorazepam, with ad-
ditional parameters, such as percentage requiring admittance to hospi-
tal, indicating better outcomes for midazolam-treated patients [12].
Midazolam was also found to be as effective as lorazepam in treating
SE in a pediatric population [15]. When compared with other benzodi-
azepine treatments, midazolam has also shown effective results, with
a recent meta-analysis of 19 studies concluding that non-i.v. midazolam
is as effective or superior to diazepam in treating SE [48,49]. In experi-
mental models of SE, midazolam has also shown its effectiveness, with
tonic–clonic seizures being terminated within 1.5–2 min of i.m. midazo-
lam administration [26,50].
While the majority of studies report i.m. administration of mid-
azolam, the current study is the first to determine its effectiveness
administered i.p. in an experimental animal model of SE. Electroen-
cephalogram recordings showed an ~ 50% reduction of EEG power
over a 1-hour recording after administration of midazolam. This is
potentially an underestimation of its effect as it is compared with EEG
recordings from the entire 40-minute period prior to midazolam treat-
ment and which takes in a 10- to 15-minute period after KA injection
before seizures begin in the model. Previous observations from our
group indicated that lorazepam administered i.p. results in similar anti-
convulsant properties as i.v. administration [22]. Intramuscular admin-
istration is also not advised in smaller rodents such as mice [51]. Here,
we show that midazolam is as effective in treating SE as lorazepam
when administered i.p. This is likely due to its solubility in water,
 M. Diviney et al. / Epilepsy & Behavior 51 (2015) 191–198
resulting in rapid absorbance from the injection site [52]. A potential
limitation of our findings is that this effect could be unique to mice
or to the i.a. KA model of SE. Furthermore, due to the drug solubil-
ity, there were different vehicles used for each drug, i.e., phosphate-
buffered saline and sterile water. This is potentially a confounding factor
that should be considered and may have indirect effects on resultant
recovery.
A potential limitation of our study is the short duration of EEG re-
cording and relatively short-term behavioral observations. While we
saw similar EEG patterns during the 60-minute posttreatment between
groups, there is the possibility we missed recurrence of seizures over
the subsequent 24 h. In addition, although there is a return to a similar
level of recovery of function by 72 h in both treatment groups and no
differences in the levels of neuronal loss at these early time-points,
it is possible that differences between treatments emerge later, for ex-
ample during epilepsy development, or affect delayed neurodegenera-
tion. Future studies could incorporate prolonged EEG recording, for
example using telemetry, and explore longer-term outcomes between
midazolam-treated and lorazepam-treated animals. Also, our study
did not include a saline-treated group of SE mice, and it may be valuable
to have data on the duration of SE and recovery pattern in animals in the
absence of treatment.
In summary, the present study provides important insights into the
efficacy of midazolam as a pharmacotherapy for SE in an experimental
model of epilepsy. We found evidence supporting the utility of midazo-
lam as an anticonvulsant treatment as compared with the more widely
used lorazepam. Importantly, our data indicated significant benefits of
midazolam at early time-points after drug administration on animal re-
covery without altering baseline EEG and histological features. Midazo-
lam is quickly becoming considered as the drug of choice for persistent
acute seizures and SE [53], and the results reported here support the ef-
ficacy of midazolam administered i.p. in treating acute, early SE in a
chemically induced experimental model of epilepsy.
Ethical publication statement
We confirm that we have read the Journal's position on issues in-
volved in ethical publication and affirm that this report is consistent
with those guidelines.
Acknowledgments
We would like to thank Dr. Edward Mealy for his assistance and ad-
vice in the care and welfare of experimental animals. This research was
funded in part by Science Foundation Ireland award 13/IA/1891 and the
Health Research Board PHD/2007/11.
Conflict of interest
None of the authors has any conflict of interest to disclose.
References
[1] Wasterlain CG, Liu H, Naylor DE, Thompson KW, Suchomelova L, Niquet J, et al.
Molecular basis of self-sustaining seizures and pharmacoresistance during status ep-
ilepticus: the receptor trafficking hypothesis revisited. Epilepsia 2009;50(Suppl. 12):
16–8.
[2] Neligan A, Shorvon SD. Prognostic factors, morbidity and mortality in tonic–clonic
status epilepticus: a review. Epilepsy Res 2011;93:1–10.
[3] Lowenstein DH, Alldredge BK. Status epilepticus. N Engl J Med 1998;338:970–6.
[4] DeLorenzo RJ, Pellock JM, Towne AR, Boggs JG. Epidemiology of status epilepticus.
J Clin Neurophysiol 1995;12:316–25.
[5] Towne AR, Pellock JM, Ko D, DeLorenzo RJ. Determinants of mortality in status
epilepticus. Epilepsia 1994;35:27–34.
[6] Loscher W. Molecular mechanisms of drug resistance in status epilepticus. Epilepsia
2009;50(Suppl. 12):19–21.
[7] Rossetti AO, Lowenstein DH. Management of refractory status epilepticus in adults:
still more questions than answers. Lancet Neurol 2011;10:922–30.
197
[8] Towne AR, DeLorenzo RJ. Use of intramuscular midazolam for status epilepticus.
J Emerg Med 1999;17:323–8.
[9] Prasad M, Krishnan PR, Sequeira R, Al-Roomi K. Anticonvulsant therapy for status
epilepticus. Cochrane Database Syst Rev 2014;9:Cd003723.
[10] Alldredge BK, Gelb AM, Isaacs SM, Corry MD, Allen F, Ulrich S, et al. A comparison of
lorazepam, diazepam, and placebo for the treatment of out-of-hospital status epilep-
ticus. N Engl J Med 2001;345:631–7.
[11] Chiulli DA, Terndrup TE, Kanter RK. The influence of diazepam or lorazepam on the
frequency of endotracheal intubation in childhood status epilepticus. J Emerg Med
1991;9:13–7.
[12] Silbergleit R, Durkalski V, Lowenstein D, Conwit R, Pancioli A, Palesch Y, et al. Intra-
muscular versus intravenous therapy for prehospital status epilepticus. N Engl J Med
2012;366:591–600.
[13] Mandrioli R, Mercolini L, Raggi MA. Benzodiazepine metabolism: an analytical
perspective. Curr Drug Metab 2008;9:827–44.
[14] Portela JL, Garcia PC, Piva JP, Barcelos A, Bruno F, Branco R, et al. Intramuscular
midazolam versus intravenous diazepam for treatment of seizures in the pediat-
ric emergency department: a randomized clinical trial. Med Intensiva 2015;39:
160–6.
[15] Welch RD, Nicholas K, Durkalski-Mauldin VL, Lowenstein DH, Conwit R, Mahajan PV,
et al. Intramuscular midazolam versus intravenous lorazepam for the prehospital
treatment of status epilepticus in the pediatric population. Epilepsia 2015;56:254–62.
[16] Lowenstein DH, Alldredge BK, Allen F, Neuhaus J, Corry M, Gottwald M, et al. The
prehospital treatment of status epilepticus (PHTSE) study: design and methodology.
Control Clin Trials 2001;22:290–309.
[17] Costello DJ, Cole AJ. Treatment of acute seizures and status epilepticus. J Intensive
Care Med 2007;22:319–47.
[18] Naritoku DK, Sinha S. Prolongation of midazolam half-life after sustained infusion for
status epilepticus. Neurology 2000;54:1366–8.
[19] Bialer M, White HS. Key factors in the discovery and development of new antiepilep-
tic drugs. Nat Rev Drug Discov 2010;9:68–82.
[20] Loscher W, Schmidt D. Modern antiepileptic drug development has failed to deliver:
ways out of the current dilemma. Epilepsia 2011;52:657–78.
[21] Löscher W, Brandt C. Prevention or modification of epileptogenesis after brain
insults: experimental approaches and translational research. Pharmacol Rev 2010;
62:668–700.
[22] Tanaka K, Jimenez-Mateos EM, Matsushima S, Taki W, Henshall DC. Hippocampal
damage after intra-amygdala kainic acid-induced status epilepticus and seizure
preconditioning-mediated neuroprotection in SJL mice. Epilepsy Res 2010;88:
151–61.
[23] Engel T, Gomez-Villafuertes R, Tanaka K, Mesuret G, Sanz-Rodriguez A, Garcia-
Huerta P, et al. Seizure suppression and neuroprotection by targeting the purinergic
P2X7 receptor during status epilepticus in mice. FASEB J 2012;26:1616–28.
[24] Kretschmann A, Danis B, Andonovic L, Abnaof K, van Rikxoort M, Siegel F, et al. Dif-
ferent microRNA profiles in chronic epilepsy versus acute seizure mouse models.
J Mol Neurosci 2015;55:466–79.
[25] Dhir A, Zolkowska D, Rogawski MA. Seizure protection by intrapulmonary delivery
of midazolam in mice. Neuropharmacology 2013;73:425–31.
[26] Raines A, Henderson TR, Swinyard EA, Dretchen KL. Comparison of midazolam and
diazepam by the intramuscular route for the control of seizures in a mouse model
of status epilepticus. Epilepsia 1990;31:313–7.
[27] Cleton A, Voskuyl RA, Danhof M. Adaptive changes in the pharmacodynamics of
midazolam in different experimental models of epilepsy: kindling, cortical stimula-
tion and genetic absence epilepsy. Br J Pharmacol 1998;125:615–20.
[28] Bourin M, Hascoet M, Mansouri B, Colombel MC, Bradwejn J. Comparison of behav-
ioral effects after single and repeated administrations of four benzodiazepines in
three mice behavioral models. J Psychiatry Neurosci 1992;17:72–7.
[29] Jimenez-Mateos EM, Bray I, Sanz-Rodriguez A, Engel T, McKiernan RC, Mouri G, et al.
miRNA expression profile after status epilepticus and hippocampal neuroprotection
by targeting miR-132. Am J Pathol 2011;179:2519–32.
[30] Paxinos G, Franklin KBJ. The mouse brain in stereotaxic coordinates. 2nd ed. San
Diego, California: Academic Press; 2001.
[31] Kirsch JH, Klaus JA, Blizzard KK, Hurn PD, Murphy SJ. Pain evaluation and response to
buprenorphine in rats subjected to sham middle cerebral artery occlusion. Contemp
Top Lab Anim Sci 2002;41:9–14.
[32] Roughan JV, Flecknell PA. Effects of surgery and analgesic administration on sponta-
neous behaviour in singly housed rats. Res Vet Sci 2000;69:283–8.
[33] Prut L, Belzung C. The open field as a paradigm to measure the effects of drugs on
anxiety-like behaviors: a review. Eur J Pharmacol 2003;463:3–33.
[34] Engel T, Sanz-Rodgriguez A, Jimenez-Mateos EM, Concannon CG, Jimenez-Pacheco
A, Moran C, et al. CHOP regulates the p53-MDM2 axis and is required for neuronal
survival after seizures. Brain 2013;136:577–92.
[35] Jimenez-Mateos EM, Hatazaki S, Johnson MB, Bellver-Estelles C, Mouri G, Bonner C,
et al. Hippocampal transcriptome after status epilepticus in mice rendered seizure
damage-tolerant by epileptic preconditioning features suppressed calcium and neu-
ronal excitability pathways. Neurobiol Dis 2008;32:442–53.
[36] Kapur J, Macdonald RL. Rapid seizure-induced reduction of benzodiazepine and
Zn2+ sensitivity of hippocampal dentate granule cell GABAA receptors. J Neurosci
1997;17:7532–40.
[37] Loscher W. Single versus combinatorial therapies in status epilepticus: novel data
from preclinical models. Epilepsy Behav 2015. http://dx.doi.org/10.1016/j.yebeh.
2015.02.027 (pii: S1525-5050(15)00083-9, in press).
[38] Wasterlain CG, Treiman DM. Status epilepticus: mechanisms and management.
Cambridge, Mass.: MIT; 2006
[39] Henshall DC, Skradski SL, Bonislawski DP, Lan JQ, Simon RP. Caspase-2 activation is
redundant during seizure-induced neuronal death. J Neurochem 2001;77:886–95.
198 M. Diviney et al. / Epilepsy & Behavior 51 (2015) 191–198
[40] Liu G, Gu B, He XP, Joshi RB, Wackerle HD, Rodriguiz RM, et al. Transient inhibition of
TrkB kinase after status epilepticus prevents development of temporal lobe epilepsy.
Neuron 2013;79:31–8.
[41] Li T, Ren G, Lusardi T, Wilz A, Lan JQ, Iwasato T, et al. Adenosine kinase is a target for
the prediction and prevention of epileptogenesis in mice. J Clin Invest 2008;118:
571–82.
[42] Engel T, Hatazaki S, Tanaka K, Prehn JH, Henshall DC. Deletion of Puma protects hip-
pocampal neurons in a model of severe status epilepticus. Neuroscience 2010;168:
443–50.
[43] Amrein R, Hetzel W. Pharmacology of drugs frequently used in ICUs: midazolam and
flumazenil. Intensive Care Med 1991;17(Suppl. 1):S1–S10.
[44] Naylor DE, Liu H, Wasterlain CG. Trafficking of GABA(A) receptors, loss of inhibition,
and a mechanism for pharmacoresistance in status epilepticus. J Neurosci 2005;25:
7724–33.
[45] Granvil CP, Yu AM, Elizondo G, Akiyama TE, Cheung C, Feigenbaum L, et al.
Expression of the human CYP3A4 gene in the small intestine of transgenic
mice: in vitro metabolism and pharmacokinetics of midazolam. Drug Metab
Dispos 2003;31:548–58.
[46] Mandema JW, Tukker E, Danhof M. Pharmacokinetic–pharmacodynamic modelling
of the EEG effects of midazolam in individual rats: influence of rate and route of
administration. Br J Pharmacol 1991;102:663–8.
[47] Thummel KE, Shen DD, Podoll TD, Kunze KL, Trager WF, Hartwell PS, et al. Use of
midazolam as a human cytochrome P450 3A probe: I. In vitro–in vivo correlations
in liver transplant patients. J Pharmacol Exp Ther 1994;271:549–56.
[48] McMullan J, Sasson C, Pancioli A, Silbergleit R. Midazolam versus diazepam for the
treatment of status epilepticus in children and young adults: a meta-analysis. Acad
Emerg Med 2010;17:575–82.
[49] Brigo F, Nardone R, Tezzon F, Trinka E. Nonintravenous midazolam versus intrave-
nous or rectal diazepam for the treatment of early status epilepticus: a systematic
review with meta-analysis. Epilepsy Behav 2015. http://dx.doi.org/10.1016/j.
yebeh.2015.02.030 (pii: S1525-5050(15)00090-6, in press).
[50] Domino EF. Comparative seizure inducing properties of various cholinesterase
inhibitors: antagonism by diazepam and midazolam. Neurotoxicology 1987;8:
113–22.
[51] Shimizu S. Routes of administration. The laboratory mouse. New York: Elsevier
Academic; 2004 527–42.
[52] Barr J, Zomorodi K, Bertaccini EJ, Shafer SL, Geller E. A double-blind, randomized
comparison of i.v. lorazepam versus midazolam for sedation of ICU patients via a
pharmacologic model. Anesthesiology 2001;95:286–98.
[53] McDonough JH, Van Shura KE, LaMont JC, McMonagle JD, Shih TM. Comparison of the
intramuscular, intranasal or sublingual routes of midazolam administration for the
control of soman-induced seizures. Basic Clin Pharmacol Toxicol 2009;104:27–34.