Epilepsy Research (2013) 103, 254261

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

Bioavailability of Intranasal vs. Rectal Diazepam

Vijay Ivaturi a,b,, Robert Kriel a,b, Richard Brundage a,b, Gordon Loewen c,1,
Hank Mansbach c,2, James Cloyd a,b
a

Center for Orphan Drug Research, United States

Department of Experimental and Clinical Pharmacology, University of Minnesota, United States
c
Valeant Pharmaceuticals, United States
b

Received 1 June 2012; received in revised form 25 July 2012; accepted 30 July 2012
Available online 13 September 2012

KEYWORDS
Intranasal;
Diazepam;
Seizure emergencies

Summary There remains an unmet medical need in the out-of-hospital management of seizure
emergencies because older children and adults often refuse treatment with diazepam rectal gel due to social objections. We have previously reported that intranasal diazepam (DZP)
administration is feasible, with maximum plasma concentrations (Cmax ) and time to maximum
concentration (Tmax ) that are comparable to rectal DZP; but tolerability was poor. In the present
study, the tolerability and pharmacokinetics of two investigational nasal formulations were
compared with DZP rectal gel.
Twelve healthy volunteers were enrolled into an active-control, double-blind, four-period,
crossover pharmacokinetic and tolerability study. Three intranasal treatments (Nas-A 10 mg,
Nas-B 10 mg and Nas-B 13.4 mg) were compared to a 10 mg dose of the rectal gel. A single
dose of each formulation was administered followed by at least a 14 day washout period. Blood
samples for plasma DZP concentrationtime characterization were collected pre-dose and at
regular intervals to 240 h post-dose. Tolerability and sedation were assessed using visual analog
scales.
Mean DZP Cmax (SD) was 181.8 84.16, 151.3 108.1 and 180.7 82.1 ng/mL for Nas-A
10 mg, Nas-B 10 and Nas-B 13.4 mg respectively; in comparison the Cmax for the rectal gel was
160.9 109.4 ng/mL. Median Tmax was 0.75 h for all treatments.

Corresponding author at: Department of Experimental and Clinical Pharmacology, University of Minnesota, United States.
Tel.: +1 46 736282714; fax: +1 612 626 9985.
E-mail address: ivatu001@umn.edu (V. Ivaturi).
1 Present address: EnVivo Pharmaceutical, Watertown, MAS, United States.
2 Present address: Medivation, Inc., San Francisco, CA, United States.

0920-1211/$ see front matter 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.eplepsyres.2012.07.018

Bioavailability of Intranasal vs. Rectal Diazepam

255

Both intranasal formulations were well tolerated and exhibited relatively rapid, but variable,
absorption with bioavailability of 7090% relative to DZP rectal gel. This study shows that the
development of a well-tolerated nasal formulation is possible and that the rate and extent of
absorption approximates that of DZP rectal gel. We conclude that intranasal DZP offers a viable
alternative to rectal administration, but enhancement of formulations is needed to improve the
extent and consistency of absorption.
2012 Elsevier B.V. All rights reserved.

Introduction
Diazepam (Valium ) possesses properties that make it a
particularly good candidate for intranasal administration as
compared to other benzodiazepines. Its lipid solubility and
potency are comparable to midazolam and it has a substantially longer elimination half-life, which may provide
a greater duration of effect as compared to midazolam
(Cloyd, 2007). Based on these considerations, our group
has undertaken a series of studies investigating the potential of intranasal diazepam as an out-of-hospital treatment
for seizure emergencies. We recently reported the pharmacokinetics of an intranasal diazepam formulation in 2
pilot studies (Ivaturi et al., 2009a,b). The glycofurol-based
formulation used in these 2 studies resulted in 75% and
74% bioavailability after administration of 5 and 10 mg DZP
doses, respectively. The formulations were absorbed rapidly
and concentrations were sustained above a target therapeutic level of 200 ng/mL for 24 h after drug administration.
Our preliminary studies demonstrated that nasal administration of diazepam is feasible, resulting in Cmax and Tmax
values that are comparable to diazepam rectal gel (Diastat), an FDA-approved product for out-of-hospital treatment
of seizure emergencies. However, the tolerability of our initial formulations was poor, which made them unsuitable for
further development.
Several intranasal diazepam formulations have subsequently been developed that may be better tolerated than
our original formulation. In the present study, we compared the tolerability and pharmacokinetics of two new
investigational nasal diazepam formulations with diazepam
rectal gel. An additional objective of this pilot study was to
determine which of the two nasal formulations was better
tolerated and had superior pharmacokinetics.

Methods
Subjects and study design
This study compared the pharmacokinetics, safety, and tolerability of 2 intranasal formulations of diazepam with
an FDA-approved rectal diazepam formulation (Diastat) in
healthy volunteers. This was a 4-period, 4-way crossover
study conducted in 12 healthy subjects. Prior to participating in the study all subjected provided signed, informed
consent. The study was approved by the University of Minnesota Institutional Review Board and conducted under
amended IND # 64,933.
Subjects were admitted to a clinical research facility
the evening before drug administration and remained at
the facility for 24 h following dosing. They then returned
as outpatients for subsequent blood sampling.

All subjects received a 10 mg rectal diazepam dose (as

Diastat) in the rst period. Thereafter, subjects received
each of the following treatments in random order over the
next three periods:
10 mg diazepam nasal formulation (Nas-A).
10 mg diazepam nasal formulation (Nas-B1).
13.4 mg diazepam nasal formulation (Nas-B2).
There was at least a 14-day washout period between
dosing. For the initial dosing period with rectal diazepam,
subjects were conned to the research unit from 1 day prior
to dosing until 2 days after dosing for pharmacokinetic, tolerability, and safety assessments. In order to minimize the
number of subjects rst exposed to any of the two investigational nasal formulations given as three treatments in
Period 2, a lead sub-set of six subjects was admitted to the
research unit and given one of the three nasal treatments
(two subjects each for formulation Nas-A, Nas-B1 and NasB2), with the remaining subjects in Period 2 to be dosed only
if acceptable tolerability was observed in the lead sub-set.
Stop criteria were established to terminate dosing under the
following conditions (a) if one of the investigational treatments caused a pain score of 8 or greater for more than
5 min or resulted in continuous nasal bleeding for more than
5 min in more than one subject, that treatment would not
be administered to any other subjects; or (b) If any one subject has pain scores of 8 or more or continuous nasal bleeding
for more than 5 min following exposure to the rst two nasal
treatments, that subject would not receive the third nasal
treatment. As the rst stop criteria were not triggered in the
six lead subjects in Period 2, thereafter in periods 3 and 4
all subjects (4 subjects each for the 3 treatments) received
their designated treatments on the same day.
Prior to each of the three treatments, the subjects eligibility was reviewed. Subjects were instructed to abstain
from prescription or over-the-counter medications beginning 24 h prior to each admission through each 48 h blood
draw. They were also instructed to not consume alcoholic
beverages 24 h before and after drug administration study
days.
On the evening before the dose of diazepam was administered, subjects entered the clinical research unit for
baseline assessments. Subjects were conned to the clinical research unit for 24 h following dosing for blood draws
and tolerability assessments. Each subject had an indwelling
catheter placed in her/his arm. All intranasal doses were
administered while the subjects were in the supine position. The rectal formulation was administered according
to instructions provided in the package insert. All subjects
were placed in a bed in supine position and continued to
remain so till 4 h post dose.

256

Study drugs
All study medications were dispensed by a licensed pharmacist at the clinical research unit to the study nurse. Two
investigational intranasal diazepam formulations were supplied by DPT Laboratories (Lakewood, NJ). The formulations
contained diazepam 5 mg/0.1 mL (NAS-A and NAS-B2) or
diazepam 6.7 mg/0.1 mL (NAS-B2). The intranasal diazepam
formulations included ingredients that are generally recognized as safe (GRAS) and used in other injectable and
ophthalmic products, or had been shown to be safe in animal studies. Each intranasal formulation was supplied in a
single-use, pre-lled, bi-dose Peffer nasal spray device. A
single spray of the device delivered approximately 0.100 mL
of the intranasal formulations containing either 5 or 6.7%
diazepam. Each subject was administered two sprays (one
in each nostril), thus, the total intranasal dose received by
the subject at each administration was 10.0 or 13.4 mg of
diazepam.
The rectal diazepam formulation (Diastat AcuDial,
Valeant Pharmaceuticals North America, Aliso Viejo, CA) was
provided by the manufacturer. A 10 mg dose of the rectal gel
was given to each subject.

Drug assay
Five-milliliter blood samples were collected for determination of plasma diazepam and the active metabolite,
desmethyldiazepam (DMD) concentrations at the following
times: predose (time 0) and 5 1, 10 1, 15 2, 20 2,
30 3, and 45 5 min and 1, 1.5, 2, 3, 4, 6, 9, and 12, 24, 48,
72, 96, 144, 192, 240 h postdose. A validated assay employing high performance liquid chromatography was used to
simultaneously measure DZP and DMD. The analytical work
was performed by MDS Pharma Services.

Pharmacokinetic analysis
DZP concentrationtime data were analyzed using a noncompartmental pharmacokinetic approach with WinNonLin
software (version 5.2; Pharsight Corporation, Mountain View,
CA, USA). The terminal elimination rate constant (z ) was
determined from the slope of the terminal log-linear portion
of the plasma-concentrationtime curve, and the terminal
elimination half-life (t1/2 ) was calculated as ln 2/(z ). Maximum plasma concentrations (Cmax ) and time to maximum
concentration (Tmax ) were determined by direct observation
of the data. The area under the concentrationtime curve to
the last non-zero plasma concentration (Clast ) that was above
the lower limit of quantication was calculated as AUClast .
The area under the concentrationtime curve extrapolated
to innity (AUC0 ) was calculated as AUClast + (Clast /z ).
To assess initial exposure after intranasal administration,
partial area under the concentrationtime curve was determined at various time points in the rst few hours (AUC0t h ).

Safety and tolerability assessment

Safety assessments included adverse events, clinical laboratory tests (serum chemistry, hematology, and urinalysis),

V. Ivaturi et al.
vital signs, 12-lead electrocardiograms, pulse oximetry and
physical examinations.
Tolerability was evaluated after both nasal and rectal
administration of each dose of diazepam. The evaluation
consisted of three components:
Sedation
A sedation score was used to assess the degree of drowsiness
of the subjects after administration of both the rectal and
nasal diazepam formulations. Sedation scores were reported
by the subject (if awake) as well as by a trained observer,
using the same rating scale, just prior to (baseline) and at
5, 15, 30, 60 min and 2, 3, 4, 6 and 8 h post dose. Subjects
were also questioned by the trained observer regarding their
degree of drowsiness. Scores were recorded on 15 scale
as follows: 0 alert, not drowsy; normal conversation; 1
awake, talking; but somewhat drowsy; 2 napping or
sleeping, but easily awakened; 3 sleeping, awakened only
with loud voice or shaking; 4 sleeping, very difcult to
awaken; promptly returns to sleep; 5 sleeping, cannot
awaken.
Pain scale
This score was used to assess the subjects overall feeling
of pain after the administration of diazepam by either route
using a visual analog scale (VAS) that consistsed of a 10 cm
(100 mm) horizontal straight line with markings from 1 to
10 (Scott and Huskisson, 1976). The ends of the scale are
dened as extremes limits of pain sensation: 0 no pain, 10
extreme pain. The subjects marked a point on the scale
which best described their intensity of pain and discomfort
just prior to and at 5, 15, 30, 60 min and 2, 3, 4, 6 and 8 h
post dose.
Nasal irritation
Nasal irritation was evaluated after administration of the
intranasal formulations. The scoring was done by a trained
observer based on an assessment of the nasal mucosa prior
to and at 0.5, 1, 2, 4, 8, 24, 48, 72, 96, 144, 192 and 240 h
post dose. Irritation was assessed by evaluating the degree
of mucosal inammation and bleeding. Subjects were also
required to report any incident of bleeding or inammation
in-between the actual evaluation time points. Scores were
recorded on 15 scale as follows: 0 normal appearing
mucosa, no bleeding; 1 inamed mucosa, no bleeding; 2
minor bleeding which stops within 1 min; 3 minor bleeding, taking 15 min to stop; 4 substantial bleeding for
460 min, does not require medical intervention; 5 ulcerated lesions, bleeding which requires medical intervention
(such as ER trip).

Statistical analysis
Descriptive statistics were used to summarize the data. The
study was not designed to have sufcient power to establish
bioequivalence of the intranasal formulations compared to
the diazepam rectal formulation. However, relative bioavailability of the nasal formulations compared to the rectal
formulation was computed and used to obtain an estimate of
bioequivalence. For each variable, the data were analyzed
using a mixed effects modeling approach in SAS v.9.2 (Cary,

Bioavailability of Intranasal vs. Rectal Diazepam

Table 1

257

Mean SD of DZP pharmacokinetic parameters following rectal and intranasal administration.

PK parameter

10 mg rectal
mean SD

10 mg Nas-A
mean SD

10 mg Nas-B1
mean SD

13.4 mg Nas-B2
mean SD

Cmax (ng/mL)
Tmax (h) (median)
AUC01 (ng h/mL)
AUC04 (ng h/mL)
AUC0 (ng h/mL)
F (AUC0 IN/D)/(AUC0 R/D)

160.9 109.4
0.75 (0.36.0)
112.66 80.35
386.51 253.0
5051.0 3722.7

181.8 84.16
0.75 (0.251.5)
122.61 52.63
387.58 153.42
4450.0 1992.4
88%

151.3 108.1
0.75 (0.253.0)
113.46 60.2
400.95 187.54
3494.7 2179.9
70%

180.7 82.1
0.75 (0.254.0)
100.38 69.02
283.99 167.57
6079.6 4055.6
89%

NC). AUC0t , AUC0 , AUC0240 h , and Cmax analyses were

based on natural log (ln)-transformed values. The ratios
between 2 treatments (each nasal formulation versus the
rectal formulation) for these parameters (point estimates)
and the corresponding 90% condence intervals (CIs) for the
ratios were obtained by exponentiating the differences and
the corresponding 90% CIs in logarithms. These data were
then used to assess the bioequivalence of the nasal formulations relative to rectal diazepam. Bioequivalence was
concluded if the 90% CIs for the ratio of the geometric means
fall completely within the 80125% bioequivalence interval.
Identication of outliers was done by visual inspection of
the concentration time proles. After the identication of
outliers, two datasets were prepared, one with intent-totreat population which included all subjects and the second
was the pharmacokinetic population which had the outliers
removed. The partial AUCs were then compared between
the two populations. Statistical tests of comparison between
the two populations were however not performed.

Results
Pharmacokinetics
12 subjects, 9 male and 3 females with an age range of
1865, were enrolled and assigned to receive all four treatments. Mean diazepam pharmacokinetic parameters are
summarized in Table 1 and the mean proles are presented
graphically in Fig. 1. In general, the nasal spray formulations
had an absorption and elimination prole similar to that of
the rectal gel formulation. However, there was considerable
subject to subject variability in the proles of both nasal
spray and rectal gel formulations. There was no consistent
pattern of observations that would suggest a failure of any
one specic formulation, and in general, there was no indication of an alteration of the elimination prole of drug in
any of the groups.
Median Tmax values were 0.75 h for all treatments suggesting that all formulations had a similar rate of absorption.
Mean Tmax values were similar for all three nasal spray doses,
ranging from 0.83 to 1.05 h versus 1.3 h for the rectal dose
group. The longer rectal gel Tmax was due to two subjects
with Tmax values of 3 and 6 h; the remaining subjects (N = 10)
had Tmax values of 1.5 h. Thus, the nasal spray formulations
did not appear to be more rapidly absorbed than that rectal
gel.
As can be seen from the Table 1, mean observed Cmax
values with the 10 mg nasal spray formulations were close

to those observed with the 10 mg rectal gel dose. However,

median Cmax values with the 10 mg nasal spray formulations
were approximately 1525% less than the median Cmax for
the rectal gel. In contrast, the mean Cmax value for the
13.4 mg Nas-B1 group was approximately 12% greater than
that of the rectal gel, with the median Cmax values for these
two groups being essentially the same. The ratios of Cmax
and AUC0 for the nasal spray groups relative to the rectal gel group are presented in Table 2. The point estimates
and the condence interval for all formulations suggested
lack of bioequivalence to the rectal gel in this exploratory
analysis.
Considering the primary purpose of the study was to
determine the relative potential for the nasal spray formulations to provide an acceptable DZP pharmacokinetic
prole, a pharmacokinetic population dataset was identied representing those pharmacokinetic proles that were
determined to be more representative of the acceptable use
of the tested treatments. An analysis of individual plots as
shown in Fig. 2 identied 10 proles from 5 of the 12 subjects
as potential outliers from this pharmacokinetic population
dataset.
An analysis of partial AUC values up to 12 h post-dose
for the intent-to treat and the pharmacokinetic population
was performed to compare diazepam exposure over initial
time periods following dosing (Fig. 3a and b). In general, the

Figure 1 Mean plasma DZP-concentration time proles

after rectal and intranasal administration in twelve subjects
(012 h). Inset shows the complete prole (0240 h).

258

V. Ivaturi et al.

Table 2 Mean ratios (%) and 90% condence intervals of ln-normalized DZP nasal spray pharmacokinetic parameters relative
to rectal gel parameters.
End point

10 mg Nas-A

10 mg Nas-B1

13.4 mg Nas-B2

BSV (%CV)

Residual (%CV)

AUC0
Cmax

0.99 (0.63, 1.55)

1.43 (0.77, 2.66)

0.68 (0.44, 1.00)

0.86 (0.47, 1.60)

0.87 (0.55, 1.36)

0.98 (0.53, 1.81)

53.60
48.21

55.33
80.23

BSV, between subject variability.

in vital signs or ECG measurements and no clinically relevant

changes in laboratory parameters during the study.
Fig. 4 shows the pain scores for the rectal and nasal
doses. Rectal DZP was well tolerated by all the subjects
with a mean maximum pain score of 0.3. In contrast, the
mean maximum scores for the three nasal treatments were
2.6, 1.6 and 1.4. The discomfort was noted immediately
after dosing and lasted for up to 5 min. Sedation scores
for the intranasal treatments were 1.0, 1.4 and 1.1 2 h
after administration. Sedation score for rectal diazepam
was 0.8. Subjects noted increasing sedation which reached
a maximum effect about 2 h after dosing (Fig. 5). However, sedation scores did not correlate with either dose or
maximum plasma concentrations. Physician observed nasal
irritation scores did not change from baseline for any of the
three intranasal treatments.

shorter the time interval post-dose, the higher the bioavailability of the nasal spray doses relative to the rectal gel
dose. For example, the ratio of AUC04 h values for the nasal
spray doses relative to the rectal gel dose was between 10
and 30% greater than respective AUC0 ratio values. However, this effect was primarily attributable to 12 subjects
within each comparison, and not reective of a consistent
formulation effect.

Safety and tolerability

No unanticipated adverse events were reported by subjects
following nasal administration of diazepam. All subjects,
however, reported swallowing a portion of the nasal dose.
There were no observed clinically signicant abnormalities

NASA 10 mg

NasB 10 mg

100

204
209

211

208

DZP Concentration (ng/mL)

10
211

NASB 13.4 mg

Rectal Gel 10 mg

100

210
211
209

10

207

211

10

12

10

12

Time after dose (hours)

Figure 2 Plots showing outliers in each treatment group. The concentrations are plotted on a log scale over the rst 12 h of
dosing. The dashed gray lines represent individual subjects (with their subject IDs) who were agged as outliers in each treatment
group when compared to the rest of the subjects represented in solid gray.

Bioavailability of Intranasal vs. Rectal Diazepam

259

Figure 5 Comparison of sedation scores (1 minimum; 5

maximum) after rectal and IN administration (n = 12).

Figure 3 Comparative partial AUC plots for (a) intent-to-treat

population and (b) pharmacokinetic population.

Discussion
This is the rst study which directly compares the pharmacokinetics of intranasal and rectal administration diazepam.
The data suggest that either Nas-A or Nas-B have the theoretical potential to provide an absorption prole very similar

Figure 4 Comparison of mean global pain scores (1 no pain;

10 extreme pain) after rectal and IN administration (n = 12).
Baseline scores were zero (0) for all subjects.

to that of DZP rectal gel, including attainment of a Tmax

within approximately 0.51 h post-dose, and maintenance
of diazepam concentrations above a target threshold up to
various time periods within the rst 12 h of dosing depending
on the formulation used.
Diazepam, regardless of the route of administration,
exhibits wide intra-individual variability in exposure (Kaplan
et al., 1973; Schmidt, 1995). For example, several studies have reported 3040% variability (SD of AUC/mean
AUC) in diazepam exposure following an intravenous dose.
We observed similar variability with the nasal and rectal
formulations both within and across subjects. No clear pharmacokinetic advantage was observed for either Nas-A or
Nas-B. Although the single Nas-A dose appeared to have AUC
parameters more similar to the rectal gel than the Nas-B
formulation, it should be noted that there was considerable variability among all the nasal treatments and that the
condence intervals for all parameters are quite large. The
variability was typically associated with a general reduction in relative exposure to a specic formulation within
a subject, suggestive of incomplete dosing and/or absorption of the intended dose, or loss of drug through nasal or
rectal leakage. Another possible contributor to the lower
AUCs, was apparent incomplete nasal dosing in some subjects where we found residual solution left in the chamber
in 5 out of 36 devices while checking the nasal spray device
after drug administration. As expected given DZPs inherent
variability, exclusion of the data from these subjects did not
reduce the variability nor change our conclusions.
When compared to our previous investigations using a
supersaturated formulation of diazepam (Ivaturi et al.,
2009a,b), the intranasal formulations used in this study
were also rapidly absorbed with good, relative bioavailability. The absolute bioavailability of the nasal formulations
is unknown, but based on our previous studies which
compared a supersaturated nasal formulation with intravenously administered diazepam, absorption appears to be
incomplete. However, unlike the supersaturated formulation, which caused considerable discomfort immediately
following administration and was poorly tolerated, the formulations used in this study were well tolerated with

260
minimal discomfort and/or irritation. The sedation scores
for the subjects after intranasal administration were no different than those after the rectal gel administration. In
response to a subjective questionnaire regarding preference
of intranasal or rectal route of administration outside a hospital setting for the given indication of seizure emergencies,
all subjects preferred the intranasal route.
Diazepam, as well as other benzodiazepines used to treat
seizure emergencies, is poorly water soluble. In order to
solublize the drug so as to make it suitable for parenteral,
rectal or nasal administration, manufacturers have resorted
to the use of organic solvents, or in the case of injectable
midazolam, adjustment of the pH to 34. The solubilizing
agents in the present investigational nasal formulations are
used in marketed ophthalmic preparation with a considerably higher aqueous component, thus we did not expect any
lasting irritation or side effects. This change was reected in
low pain scores as shown in Fig. 4. Three subjects reported
scores above 7 immediately after a nasal dose, their scores
quickly returning to average score of the study cohort. All
subjects reported a disagreeable after-taste minutes after
nasal administration which was consistent with reports of
swallowing of some part of dose after administration. This
phenomenon commonly occurs with many intranasal products, but is a formulation issue which is potentially xable.
The aftertaste was partially relieved with sips of water.
Although not extensively discussed in the literature,
there is considerable variability in diazepam pharmacokinetics following rectal administration. This is consistent
with the inherent variability in diazepam pharmacokinetics
when administered by other routes (Kaplan et al., 1973;
Schmidt, 1995) and compounded by factors uniquely associated with rectal administration such as the effect of fecal
material and defecation. The controlled clinical safety and
efcacy trials of diazepam rectal gel did not include a
pharmacokinetic component. However, in a Phase I study
in healthy volunteers (Cloyd et al., 1998) and unpublished
study from a sponsor (Xcel Pharmaceuticals), which more
rigorously characterized rectal diazepam pharmacokinetics;
rectal diazepam exhibited similar variability to the results of
the present study. A population analysis of rectal diazepam
gel (Diastat ) in 75 healthy volunteers pooled across 5
different studies (manuscript under preparation) shows considerable between-subject variability ranging from 30 to 70%
in different pharmacokinetic parameters. Considering that
these estimates reect variability in controlled pharmacokinetic studies in healthy adult volunteers, variability could
be expected to be even greater when being administered
to patients (children and adults) in an out-of-hospital setting by a care-giver in a stressful condition at the time of a
seizure.
It is reasonable to conclude that rectal diazepam is effective despite considerable variability in exposure. Further, as
treatment is intended as rescue therapy where drug effect
should have a rapid onset and persist for at least few hours
after the onset of a seizure emergency, exposure in the
early time period after seizure onset (04 h) is the primary contributor to efcacy. One could hypothesize that the
main considerations in the development of a nasal benzodiazepine product for treatment of a seizure emergency are
the time to attain a therapeutic concentration, i.e. the time
to minimum effective concentration, and maintenance of

V. Ivaturi et al.
that concentration for a sufcient duration to ensure seizure
control (Wermeling, 2009). Absolute bioavailability, Tmax and
Cmax of a potential nasal product may be less important
factors in determining safety and efcacy. Future clinical
studies with rescue therapies should potentially explore the
relationship of response to drug concentrations attained
within the 04 h period after dosing.
The intranasal doses in our study, 10 and 13.4 mg,
provided maximum concentrations in the range of
150190 ng/mL. Further optimization could result in
higher concentrations to rapidly attain and maintain the
proposed minimum target therapeutic plasma concentration of 200 ng/mL (Milligan et al., 1982). Furthermore,
based on the characteristics of the intranasal formulations
evaluated in this study, target concentrations could be
attained at the same time or earlier than via the rectal
route (0.75 h). It is anticipated that it would be possible
to attain therapeutic diazepam concentrations with the
current intranasal formulations by giving a second nasal
dose 5 or 10 min after the rst. This would not be expected
to pose any additional safety issues as diazepam is known
to have a wide therapeutic and safety window as shown
with rectal diazepam (Brown et al., 2001; Mitchell et al.,
1999; Pellock, 2004). Practically, it is easier to repeat a
nasal dose than a rectal dose to achieve therapeutic levels
if required.
The bioavailability of the two nasal formulations relative to rectal diazepam was in the range of 7090%, albeit
with high variability. Review of the data and comments
by study staff in the case report forms, raised concerns
regarding the correct dispensing of the nasal spray formulation during at least the rst day of Period 2, and possibly
other days. In addition, the rectal gel appeared to be poorly
absorbed or incompletely administered in some cases. Furthermore, plasma diazepam concentration proles after
nasal administration in periods 3 and 4 suggested incomplete absorption, nasal leakage and/or technical difculties
with administration. For these reasons, the intent-to-treat
data were considered too variable to accurately assess the
potential utility of the nasal spray formulations relative
to diazepam rectal gel. Considering the primary purpose
of the study was to determine the relative potential for
the nasal spray formulations to provide an acceptable
diazepam pharmacokinetic prole, a pharmacokinetic population was identied representing those pharmacokinetic
proles that were determined to be more representative
of the acceptable use of the tested treatments. Moreover,
as initial exposure levels after treatment are more critical
than looking at bioavailability estimated using the complete
area under the curve from 0-inf (Wermeling, 2009), the
partial AUCs shown in Fig. 3 indicate that the rate and
extent of absorption as similar during early time periods
for both the intranasal and rectal gel formulations. This
observation was more consistent across the pharmacokinetic than the intent-to-treat population with the former
showing relatively higher exposure levels. Our intranasal
formulations also fulll most of the requirements of an
ideal intranasal benzodiazepine formulation outlined by
Wermeling (2009).
Neither of the two nasal formulations was clearly superior in either bioavailability or tolerability. What our study
does show is that the absorption is relatively rapid and early

Bioavailability of Intranasal vs. Rectal Diazepam

exposures are comparable to rectal diazepam combined
with a good tolerability prole indicating that intranasal
diazepam is a viable alternative to rectal diazepam.

References
Brown, L., Bergen, D.C., Kotagal, P., Groves, L., Carson, D., 2001.
Safety of Diastat when given at larger-than-recommended doses
for acute repetitive seizures. Neurology 56, 1112.
Cloyd, J., 2007. Pharmacologic considerations in the treatment of
repetitive or prolonged seizures. J. Child Neurol. 22, 47S52S.
Cloyd, J.C., Lalonde, R.L., Beniak, T.E., Novack, G.D., 1998. A
single-blind, crossover comparison of the pharmacokinetics and
cognitive effects of a new diazepam rectal gel with intravenous
diazepam. Epilepsia 39, 520526.
Ivaturi, V.D., Riss, J.R., Kriel, R.L., Cloyd, J.C., 2009a. Pharmacokinetics and tolerability of intranasal diazepam and midazolam
in healthy adult volunteers. Acta Neurol. Scand. 120, 353357.
Ivaturi, V.D., Riss, J.R., Kriel, R.L., Siegel, R.A., Cloyd, J.C., 2009b.
Bioavailability and tolerability of intranasal diazepam in healthy
adult volunteers. Epilepsy Res. 84, 120126.

261
Kaplan, S.A., Jack, M.L., Alexander, K., Weinfeld, R.E., 1973.
Pharmacokinetic prole of diazepam in man following single
intravenous and oral and chronic oral administrations. J. Pharm.
Sci. 62, 17891796.
Milligan, N., Dhillon, S., Oxley, J., Richens, A., 1982. Absorption of
diazepam from the rectum and its effect on interictal spikes in
the EEG. Epilepsia 23, 323331.
Mitchell, W.G., Conry, J.A., Crumrine, P.K., Kriel, R.L., Cereghino,
J.J., Groves, L., Rosenfeld, W.E., 1999. An open-label study
of repeated use of diazepam rectal gel (Diastat) for episodes
of acute breakthrough seizures and clusters: safety, efcacy,
and tolerance. North American Diastat Group. Epilepsia 40,
16101617.
Pellock, J.M., 2004. Safety of Diastat, a rectal gel formulation of
diazepam for acute seizure treatment. Drug Saf. 27, 383392.
Schmidt, D., 1995. Benzodiazepines: diazepam. In: Levy, R.H.,
Mattson, R.H., Meldrum, B.S. (Eds.), Antiepileptic Drugs. Raven
Press, New York, pp. 705724.
Scott, J., Huskisson, E.C., 1976. Graphic representation of pain.
Pain 2, 175184.
Wermeling, D.P., 2009. Intranasal delivery of antiepileptic medications for treatment of seizures. Neurotherapeutics 6, 352358.