See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/7209112

Comparative Absorption Profiles of Divalproex Sodium Delayed-Release versus

Extended-Release Tablets—Clinical Implications

Article  in  Annals of Pharmacotherapy · May 2006

DOI: 10.1345/aph.1G617 · Source: PubMed

CITATIONS READS

15 1,535

3 authors, including:

Sandeep Dutta Ronald Charles Reed

Amgen Husson University
145 PUBLICATIONS   1,753 CITATIONS    66 PUBLICATIONS   1,184 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Pharmacokinetic Simulation of Plasma Concentrations of Antiepileptic Drugs View project

Triptan in Migraine Research View project

All content following this page was uploaded by Ronald Charles Reed on 11 June 2015.

The user has requested enhancement of the downloaded file.

Neurology

Comparative Absorption Profiles of Divalproex Sodium Delayed-

Release Versus Extended-Release Tablets—Clinical Implications

Sandeep Dutta, Ronald C Reed, and Robert F O’Dea

BACKGROUND: The distinct absorption characteristics of the conventional enteric-coated, delayed-release (DR) and the novel
extended-release (ER) divalproex sodium formulations are not well recognized.
OBJECTIVE: To quantitatively and qualitatively differentiate the absorption characteristics of divalproex-DR and -ER formulations.

METHODS: Healthy volunteers (N = 28) received single 1000 mg doses of divalproex-DR and divalproex-ER tablets in a crossover
fashion. Noncompartmental and compartmental analyses were used to estimate valproic acid (VPA) pharmacokinetics from the
plasma concentration–time profiles determined from intensive blood sampling over 48 hours.
RESULTS: VPA was not absorbed from divalproex-DR in the first 2 hours (absorption lag-time) after dosing. After VPA release in the
intestine, approximately 63% of the dose was absorbed in less than 1 hour, that is, 2.9 hours (mean absorption time) from dosing.
Maximum concentration (Cmax) was achieved approximately 4 hours after dosing. VPA absorption was complete (~93% of dose)
within 3 absorption half-lives (~4.5 h) post-absorption lag-time, that is, 6–7 hours from dosing. In contrast, VPA absorption from
divalproex-ER starts immediately after administration, initially at a modest rate, followed by slow and extended absorption at a
constant rate for more than 20 hours; VPA concentrations at 1 and 2 hours were 28% and 40% of Cmax. Approximately 53% of the
dose was absorbed within 12 hours (mean absorption time); complete absorption occurred over more than 20 hours without any
dose dumping.
CONCLUSIONS: VPA absorption from enteric-coated divalproex-DR is rapid following a lag-time of approximately 2 hours and is
complete within 6–7 hours of dosing. In contrast, VPA absorption from divalproex-ER starts immediately after administration, but
occurs at a slow, approximately constant rate over more than 20 hours.
KEY WORDS: absorption, pharmacokinetics, valproate, valproic acid: controlled release, delayed release, extended release,
sustained release.
Ann Pharmacother 2006;40:619-25.
Published Online, 28 Mar 2006, www.theannals.com, DOI 10.1345/aph.1G617

ivalproex sodium enteric-coated, delayed-release
D tablet (divalproex-DR, Depakote; DR used here to
distinguish from extended release formulation, but is not
used in package insert) and the newer extended-release
tablet formulation (divalproex-ER, Depakote ER) have
distinct absorption characteristics. The enteric coating of
divalproex-DR is intended to bypass tablet disintegration
and dissolution in the stomach and mitigate gastrointestinal
(GI) adverse events,1 while the divalproex-ER formulation
was specifically engineered for slow release over approxi-
mately 22 hours2,3 and to enable once-daily dosing.
Author information provided at the end of the text.
Divalproex is well known in the clinical management of
epilepsy, acute mania, and migraine headache prophylaxis.4
A review of the literature shows that the absorption profile
for divalproex-DR has been largely defined by the maxi-
mum observed concentration (Cmax), a parameter derived
via noncompartmental methods. The parameter Cmax is
widely recognized by clinicians, but nonetheless is not an
optimal parameter for precisely defining absorption rate.
Cmax is a hybrid parameter, affected by drug dose, distribu-
tion volume, and clearance.
Despite several articles published in the MEDLINE
database concerning the pharmacokinetic characteristics of
divalproex-DR tablets since its introduction 2 decades ago,
little information exists on the absorption rate constant (ka)

www.theannals.com The Annals of Pharmacotherapy n 2006 April, Volume 40 n 619

Downloaded from aop.sagepub.com by guest on October 11, 2013
S Dutta et al.

of this formulation. Although ka is a parameter that is not PHARMACOKINETIC ANALYSES

well appreciated to have clinical utility, it does have an im- Noncompartmental analysis was performed to estimate VPA pharma-
portant role in drug development pharmacokinetics for sim- cokinetic parameters following administration of both formulations.
ulations of dosing paradigms that are frequently encoun- Two-stage compartmental modeling was performed using WinNonlin
tered in clinical practice. However, these dosing paradigms (Pharsight Corp., Mountain View, CA).
(eg, numerous possible combinations of loading and main-
tenance dosing paradigms, conceptualization of best dival- DIVALPROEX-DR COMPARTMENTAL MODEL
proex-DR to divalproex-ER strategies, and the impact of
A one-compartment model with absorption lag-time, first-order ab-
nonadherence with replacement of missed doses at various sorption, and first-order elimination optimally described the VPA con-
intervals thereafter)5-7 that cannot be studied prospectively centration–time profiles.
for logistical or ethical reasons can be evaluated using clin-
ical trial simulations. Absorption Model
Divalproex-ER is a once-daily formulation that offers The drug input rate was modeled as a first-order process where ka =
the advantages of potentially improved medication adher- first order absorption rate constant and F = bioavailability (Equation 1).
ence, increased patient convenience, and simpler titration dA
= –ka • A; A = Dose • F when t = 0 Eq. 1
to maximum effective dose while possibly reducing the dt
risk of adverse effects by reducing peak–trough fluctua-
tions in plasma concentrations.8,9 Divalproex-ER is formu- Plasma VPA Concentration–Time Profile of Divalproex-DR
lated using hydrophilic matrix technology, and drug re- The first-order absorption model was coupled with a mono-exponen-
lease is controlled primarily by water-soluble polymer (hy- tial disposition model to describe divalproex-DR plasma VPA concentra-
droxypropylmethyl cellulose) erosion from the matrix.10 tion–time profiles where t = time since dosing, tlag = lag time, CP =
Valproic acid (VPA) absorption from divalproex-ER starts plasma VPA concentration, V/F = apparent volume of distribution, Cl/F
= apparent oral clearance, ka = first-order absorption rate constant, and
immediately after oral administration, is multiphasic, oc-
F = bioavailability (Equation 2).
curs over more than 20 hours, and can be approximated as dCp
a pseudo zero-order absorption rate.2 V/F • = ka • A–Cl/F • Cp; Cp = 0 when t ≤ tlag Eq. 2
dt
While divalproex-DR and divalproex-ER bioavailabili-
ty have been compared using traditional bioequivalence
DIVALPROEX-ER COMPARTMENTAL MODEL
criteria of Cmax and area under the curve (AUC),11-14 this
study represents new information in that the distinctive ab- Absorption Model
sorption profiles of divalproex-DR and divalproex-ER The drug input rate profile, AR(t), for a unit dose was modeled3
have been fully characterized and modeled using optimal (Equation 3) where AR(t) (mg/h) = absorption rate at time “t,” A0 (mg/h)
quantitative measures. = zero-order AR, Amax (mg/h) = maximum AR above A0, GT50 (h) or
gastric emptying time = time taken for AR to increase from zero to 50%
of A0+Amax, SIT50 (h) or small intestinal transit time = time taken for AR
Methods to decrease from A0+Amax to A0+(Amax/2), LIT50 (h) or large intestinal
transit time = time taken for AR to decrease to 50% of A0, and γ, δ, & θ =
STUDY DESIGN
sigmoidicity parameters, for model reduction purpose δ = 2γ and θ = 4γ.
This single-dose, fasting, open-label, randomized, 2-period, cross-
over study was conducted in 28 healthy adult male and female subjects.
Prior to the performance of any screening and study-specific procedures,
applicable written informed consent was obtained from each subject.
The study protocol was approved by the institutional review board of
Victory Memorial Hospital. Subjects were randomly assigned in equal
numbers to receive 2 sequences of divalproex-DR and divalproex-ER
formulations. Each formulation was administered as a 1000 mg dose (2
500-mg tablets) with 180 mL of water after a 9 hour fast.
Eighteen blood samples were collected within 10 minutes prior to
dosing (0 hour) and at 1, 1.5, 2, 2.5, 3, 4, 5, 6, 8, 10, 12, 15, 18, 24, 30,
36, and 48 hours after dosing. The blood samples were centrifuged
within 30 minutes of collection to separate the plasma. The plasma sam-
ples were frozen immediately and stored at –6.7 °C until analysis. Plasma
VPA concentrations were determined using a commercially available fluo-
rescence polarization immunoassay on the Abbott TDx instrument. A maxi-
mum of 7 days elapsed between collection and analysis, which was within
the documented long-term stability for VPA in frozen plasma samples. The
assay was reproducible and unbiased, with less than 4% inter- and intra-day
coefficient of variation and a lower limit of quantification of 0.7 mg/L.

620 n The Annals of Pharmacotherapy n 2006 April, Volume 40 www.theannals.com

 Absorption Profiles of Divalproex-DR vs -ER

PLASMA VPA CONCENTRATION–TIME PROFILE OF

DIVALPROEX-ER

The multiphasic absorption model was coupled with a mono-expo-

nential disposition model to describe divalproex-ER plasma concentra-
tion–time profiles as shown in Equation 4 where V/F = apparent volume
of distribution, Cp = plasma VPA concentration, t = time, AR(t) = ab-
sorption rate at time “t”, and Cl/F = apparent oral clearance.
dCp
V/F • = Dose • AR(t)–Cl/F • Cp; Cp & AR(t) = 0 when t = 0 Eq. 4
dt

Results
The subjects’ average (SD) age and body weight were
35.9 (9.6) years and 76.4 (10.03) kg, respectively. The
single-dose plasma concentration–time profiles for dival-
proex-DR and divalproex-ER are presented in Figure 1
(Panels a and b) and the corresponding VPA pharmacoki-
netic parameters are presented in Table 1. The comparative
mean and individual absorption profiles are presented in
Figure 1 (Panels c and d) and Figure 2, respectively, and
the corresponding absorption characteristics are presented
in Table 2.

Discussion
Our study shows that divalproex-DR and divalproex-ER
formulations exhibit distinct pharmacokinetic profiles. For
divalproex-DR, VPA absorption is immediate and rapid fol-
lowing a lag-time of approximately 2 hours. In contrast, for
divalproex-ER, absorption starts immediately after adminis-
tration, without any lag-time, and displays multiple phases
of absorption rate that extend over 20–24 hours.
The comparative mean observed VPA concentration–
time profiles for divalproex-DR and divalproex-ER, pre-
sented in Figure 1 (Panels a and b), result from a convolu-
tion of the overlapping in vivo processes of absorption,
distribution, and elimination. The estimated Cl/F and V/F
values (Table 1) were similar for the 2 formulations and
consistent with results from other VPA studies.13 The dif-
ferences in the observed VPA concentration–time profiles
between the formulations are primarily due to their distinct
absorption characteristics, rather than to differences in
VPA distribution and elimination processes, as quantitated
by V/F and Cl/F values, respectively. The individual ab-
sorption profiles for the 2 formulations presented in Figure
2 are summarized as mean absorption profiles in Figure 1
(Panels c and d) to illustrate more clearly that the differ-
ences in the observed VPA concentration–time profiles be-
tween the formulations are indeed due to their distinct ab- Figure 1. Mean plasma VPA concentration–time profiles are shown for the
first 8 hours postdose for divalproex-DR (Panel a, dashed line) and for dival-
sorption characteristics. proex-ER (Panel a, solid line); the entire profile is depicted from time 0 to 48
hours (Panel b). Symbols and error bars represent mean ± SD of observed
The enteric coating prevented drug release from the di- VPA concentrations and the lines represent the model fit to the data in Panels
valproex-DR formulations for approximately 2 hours (lag- a and b. For direct comparison with VPA concentration–time profiles, the mean
absorption rate–time profiles for divalproex-DR and divalproex-ER over 8
time) in the acidic gastric environment, but rapid absorption hours (Panel c), and over the entire study period of 48 hours (Panel d) are
occurred (Figure 1, Panels c and d; Figure 2, Panel a) in shown. The mean absorption rate–time profiles are derived from the individ-
ual absorption rate–time profiles presented in Figure 2. DR = delayed release;
the intestine immediately following the dissolution of the ER = extended release; VPA = valproic acid.

www.theannals.com The Annals of Pharmacotherapy n 2006 April, Volume 40 n 621

S Dutta et al.

enteric coating. Approximately 63% (Table 2) of the total
divalproex-DR dose was absorbed in less than 1 hour, and
Cmax was achieved 2 hours after VPA was released in the
intestine; mean absorption time from dosing was 2.9 hours
and the ka was 3.5 h-1. VPA absorption was complete
(~93%) within 3 absorption half-lives (~4.5 h) post absorp-
tion lag-time (ie, within 6 –7 hours of dosing). While the
mean absorption time is a measure of the time taken for ap-
proximately 50% of the dose to be absorbed, ka is a mea-
sure of how fast the drug is absorbed following its release
from the tablet. As with all products that exhibit very rapid
absorption, the absorption rate constant is difficult to esti-
mate precisely because ka values of 1.0 h–1 or more do not
result in significant differences in the shape of the observed
in vivo absorption profile (ie, concentration– time profile
up to Cmax). Divalproex-DR individual absorption rate pro-
files (calculated as ka • dose remaining to be absorbed, Fig-
ure 2, Panel a) illustrate considerable variability in the lag-
time. This variability is possibly due to inherent differences
in gastric retention times,15-18 even across healthy individu-
als in this well-controlled study, where doses were adminis-
tered under fasting conditions. Administration under non-
fasting conditions increases the lag-time, as well as the vari-
ability in lag-time across individuals.19
Absorption from divalproex-ER began immediately af-
ter administration, had multiple phases, and the entire dose
was absorbed over 20 –24 hours. Figure 2, Panel b indi-
cates that initially the rate and extent of absorption were
modest due to controlled drug release, where about 15% of
the dose was absorbed in the first 2 hours, probably in the
upper segments of the small intestine. This was followed
by extended absorption at a slower constant (approximate-
ly zero order) rate as the drug was slowly released from the
tablet formulation, with mean VPA concentrations being
approximately 28% and approximately 40% of Cmax at 1
and 2 hours, respectively. Approximately 53% (Table 2)
of the total divalproex-ER dose was absorbed within 12
hours (mean absorption time) with complete absorption,
without any dose dumping, occurring over more than 20
hours in most subjects. However, in a small fraction of
subjects, absorption was complete within 12–18 hours
without any dose dumping, whereas in some other sub-
jects, absorption occurred over more than 30 hours. This
range of absorption durations probably reflects the vari-
ability in gastric transit times in any sample of healthy vol-
unteers.15-18 It should be noted that, in the general popula-
tion, meal conditions vary from day to day within a patient
and across patients; therefore, GI transit times will vary, ir-
respective of drug or formulation.
Our multiphasic absorption model3 may be considered
as a forcing or interpolating function that describes a dis-
tinctive absorption profile of (1) an initial increase in ab-
sorption rate as the drug moves from the gastric space to
the upper segments of the small intestine where the ab-
sorption rate is at its peak (Amax + A0), followed by (2) at-
tenuation of this peak absorption rate as the drug travels
through the small intestine and enters the large intestine,
followed by (3) an extended zero order absorption in the
large intestine at a slower rate (A0), followed by (4) atten-
uation of the absorption rate as the drug leaves the GI tract.

Table 1. VPA Pharmacokinetic Parameters following Single Dose Administration of Divalproex-DR and Divalproex-ERa
Divalproex-DR Divalproex-ER
Pharmacokinetic Noncompartmental Compartmental Noncompartmental Compartmental
Parameter Analysis Modeling Analysis Modeling

tmax (h) 3.8 ± 1.2 21.0 ± 6.9

Cmax (mg/L) 93.9 ± 11.7 48.5 ± 8.7
AUC∞ (mg•h/L) 1818 ± 345 1866 ± 497
Cl/F (L/h) 0.57 ± 0.11 0.59 ± 0.10 0.58 ± 0.16 0.60 ± 0.083
V/F (L) 12.0 ± 3.0 11.0 ± 1.8 11.9 ± 2.4 10.8 ± 2.46
t1/2 (h) 15.2 ± 5.3 13.3 ± 3.1 15.3 ± 4.8 13.1 ± 4.5
Lag-time (h) 1.8 ± 0.74b 2.1 ± 0.65
A0 (mg/h) 0.040 ± 0.012
Amax (mg/h) 0.18 ± 0.19
GT50 (h) 0.032 ± 0.098
SIT50 (h) 1.2 ± 0.70
LIT50 (h) 22.7 ± 7.67
γ 2.5 ± 0.82

A0 = zero-order absorption rate; Amax = maximum absorption rate above Ac; AUC∞ = area under the concentration–time curve from time zero to in-
finity; Cl/F = apparent oral clearance; Cmax = maximum concentration; DR = delayed release; ER = extended release; γ = sigmoidicity; GT50 = gastric
emptying time; LIT50 = large intestinal transit time; SIT50 = small intestinal transit time; t1/2 = half-life; tmax = time to maximum concentration; V/F = ap-
parent volume of distribution.
a
Mean ± SD.
b
Calculated as time to concentrations >0.7 mg/L (lower limit of quantification) for noncompartmental analysis.

622 n The Annals of Pharmacotherapy n 2006 April, Volume 40 www.theannals.com


The 3 sigmoidicity parameters (γ, δ, θ) could not be esti-
mated independently; hence, the model was simplified.
The gastric emptying time (GT50) was not robustly estimat-
ed (not significantly different from zero). However, this por-
tion of the model that characterizes drug entry into the small
intestine followed by increased absorption rate is consistent
with physiologic expectation and is essential to enforce the
plasma concentrations to rise from zero at time zero.
The multiphasic absorption rate of divalproex-ER can-
not be characterized by several atypical absorption models
(eg, mixed zero- and first-order, Weibull function, time-
and/or GI location-dependent, exponential, and several
other absorption models) that have been proposed for other
drugs or formulations.20 In general, the model-estimated
values of the transit times were consistent with physiologic
expectations and previously published work.15-18 However,
the parameter estimates should be interpreted with caution
since the model describing the absorption phase of VPA
from divalproex-ER has several nonlinear components,
with many parameters being correlated. It should be recog-
nized that some parameters of this model could not be esti-
mated robustly in this study because of the absence of ap-
Figure 2. Comparative individual absorption rate profiles after divalproex-DR
(Panel a) and divalproex-ER (Panel b) administration, displaying the large in-
terindividual variability in the lag-times and peak absorption–rate times for the
divalproex-DR formulation and, likewise, interindividual variability in the total
duration of absorption for the divalproex-ER formulation that is not apparent
from the mean curves in Figure 1. DR = delayed release; ER = extended re-
lease.
www.theannals.com The Annals of Pharmacotherapy
Absorption Profiles of Divalproex-DR vs -ER
propriate intravenous data that can be used to “tease” apart
the absorption rate profile.3 In such situations, no physio-
logic meaning should be assigned to the estimated parame-
ters. Although our multiphasic absorption model for dival-
proex-ER is complex, it is consistent with physiologic ex-
pectations and captures all elements of absorption process
that cannot be optimally characterized using any of the
previously described atypical absorption models.20 Our
unique model has clinical and practical utility as an inter-
polating function and can be used to appropriately describe
the absorption rate profile for the purpose of simulation of
various dosing regimens and clinical situations.
CLINICAL IMPLICATIONS OF THE DISTINCT ABSORPTION
PROFILES FOR DIVALPROEX-DR AND DIVALPROEX-ER
Our study demonstrates the distinctly different absorp-
tion characteristics for divalproex-ER compared with di-
valproex-DR. The specific tablet engineering for the dival-
proex-ER formulation enables near zero-order absorption
over 22 hours,2 which statistically significantly decreases
the degree of fluctuation (defined as [Cmax — Cmin]/Cavg) in
plasma VPA concentration, compared with the conven-
tional, enteric-coated divalproex-DR.11-13 The reduction in
the degree of fluctuation with divalproex-ER significantly
reduces adverse effects in both neurologic and psychiatric
patients (adverse effects thought to be associated with
Cmax), including tremor, weight gain, and GI complaints.21
In a nonrandomized clinical trial comparing divalproex-
DR tablets with generic, immediate-release VPA gelatin
capsules (only 0.18% of patients received divalproex-ER),
Wassef et al.22 claimed that, although divalproex-DR may
result in enhanced tolerability compared with immediate-
release VPA capsules, it may also lead to lower effective-
ness in mania. While no significant difference in Cmax is
expected between identical doses of immediate-release
Table 2. Comparative Absorption Characteristics of
Divalproex-DR and Divalproex-ER
Parameter Divalproex-DR Divalproex-ER
a
MAT (h) 2.89 ± 1.01 11.9 ± 3.46b
ka (1/h) 3.5 ± 3.4c 0.0942 ± 0.0387d
Percentage of drug 63.2e 53.0 ± 3.1f
absorbed up to MAT
DR = delayed release; ER = extended release.
a
Calculated as 1/k a plus compartmental model estimated lag-time
(mean ± SD).
b
Calculated as area under the first moment curve/area under the curve
(AUC) of absorption rate (AR) versus time curve (mean ± SD).
c
Compartmental model estimated (mean ± SD).
d
Calculated as 1/mean absorption time (MAT) (mean ± SD).
e
Theoretical percent absorbed for 1 compartment, first-order absorption
process.
f
Calculated as AUC0-MAT/AUC∞ of AR versus time curve.
n 2006 April, Volume 40 n 623
S Dutta et al.
VPA and divalproex-DR,19 Wassef et al.’s results still raise
2 very intriguing questions: Is the effectiveness of various
formulations of divalproex/VPA linked to different phar-
macokinetic parameters (ie, VPA Cmax, Cmin, and/or expo-
sure [AUC]), and do the various diagnoses among psychi-
atric/neurologic patients require targeting different kinetic
parameters to achieve optimal efficacy? In any event, the
relative importance of Cmax versus AUC for efficacy has
not been compared in any well-controlled, prospectively
designed trial as of March 15, 2006. In balance, when con-
sidering efficacy and adverse event data from both neuro-
logic and psychiatric studies, once-daily dosing is accept-
able with divalproex-ER,11-13,23,24 allowing simplification of
dosing regimens,8 which in turn enhances patient adher-
ence and satisfaction.9
Some less-recognized features that are generally associ-
ated with many extended- or controlled-release products
are also evident for divalproex-ER. As divalproex-ER is
designed to be absorbed over more than 20 hours,2,3 the
extent of absorption (AUC) is probably correlated with GI
transit times. Consequently, it may be speculated that for
patients who have had gastric or intestinal bypass surgery,
shorter GI transit time may limit the extent of absorption,
resulting in lower bioavailability of divalproex-ER com-
pared with divalproex-DR. Irrespective of drug or formu-
lation, GI transit times in the general population vary from
day to day within a patient and across patients due to vari-
ous factors, such as meal conditions (caloric content and
timing of meal relative to dosing), physical activity, and di-
urnal effects. Because the extent of VPA absorption from
divalproex-ER depends on the GI transit time, the intrinsic
variability of GI transit time may contribute to its slightly
lower average bioavailability (89%)14 compared with di-
valproex-DR; patients with shorter GI transit times may
have lower bioavailability. Therefore, an 8–20% upward
titration of the total daily dose is required for conversion
from divalproex-DR to divalproex-ER to maintain an
equivalent exposure (AUC),11-13 especially for patients with
refractory or brittle epilepsy.
The tablet engineering for divalproex-ER requires a spe-
cific amount of hydroxy-propylmethylcellulose polymer to
control divalproex release and VPA absorption, thereby
necessitating a relatively large tablet size for the 250 and
500 mg tablet strengths.10 The increased tablet size may
preclude ingestion by young children and patients with
dysphagia (ie, elderly and intellectually disabled pts.). Di-
valproex-ER tablets should not be split, crushed, or
chewed, since that would destroy the extended-release
characteristics of the formulation.
Divalproex-ER and divalproex-DR are not interchange-
able. Straight mg/mg substitution of divalproex-ER for dival-
proex-DR will likely result in decreased plasma concentra-
tions of VPA and could result in breakthrough symptoms in
certain patients. Alternatively, substitution of divalproex-DR
624 n The Annals of Pharmacotherapy n 2006 April, Volume 40
for divalproex-ER will result in rapidly rising plasma VPA
concentrations, potentially leading to VPA-induced adverse
effects. Therefore, divalproex-DR and divalproex-ER should
not be casually substituted for one another.
Conclusions
VPA absorption from enteric-coated divalproex-DR is
rapid following a lag-time of approximately 2 hours and is
complete within 6–7 hours of dosing. In contrast, VPA ab-
sorption from divalproex-ER starts immediately after ad-
ministration, but occurs at a slow, approximately constant
rate over more than 20 hours. Such different VPA absorp-
tion characteristics for divalproex-DR versus divalproex-
ER can have a meaningful impact on the rate at which
clinically relevant VPA concentrations are either achieved
or sustained. Divalproex-DR should not be mistaken or in-
discriminately substituted for divalproex-ER.
Sandeep Dutta PhD, Associate Director and Associate Research
Fellow, Clinical Pharmacokinetics, Abbott Laboratories, Abbott Park,
IL
Ronald C Reed PharmD, Senior Clinical Research Scientist, Neu-
roscience Global Pharmaceutical Research & Development, Abbott
Laboratories
Robert F O’Dea PhD MD, Director of Clinical Research, Medical
Affairs, Abbott Laboratories
Reprints: Dr. Dutta, Abbott Laboratories, Dept. R4PK, Bldg. AP13A,
100 Abbott Park Rd., Abbott Park, IL 60064-6104, fax 847/938-5193,
Sandeep.Dutta@abbott.com
This study was presented at the American Academy of Neurology
57th Annual Meeting, Miami Beach, FL, April 2005.
Abbott Laboratories, Abbott Park, IL, funded this study.
References
1. Wilder BJ, Karas BJ, Penry JK, Asconape J. Gastrointestinal tolerance of
divalproex sodium. Neurology 1983;33:808-11.
2. Dutta S, Reed RC, Cavanaugh JH. Absolute bioavailability and absorp-
tion characteristics of divalproex sodium extended-release tablets in
healthy volunteers. J Clin Pharmacol 2004;44:737- 42.
3. Dutta S, Reed RC, Cavanaugh JH. A multiphasic absorption model for
characterizing single dose pharmacokinetics of divalproex-ER in healthy
subjects (abstract). Clin Pharmacol Ther 2005;77:P33.
4. Package insert. Depakote (divalproex sodium delayed-release tablets).
North Chicago, IL: Abbott Laboratories, September 2004.
5. Dutta S, Cloyd JC, Granneman GR, Collins SD. Oral/intravenous main-
tenance dosing of valproate following intravenous loading: a simulation.
Epilepsy Res 2003;53:29-38.
6. Reed RC, Dutta S. What is the best strategy for converting from twice-
daily divalproex to a once-daily divalproex ER regimen?: examinations
and answers via computer simulations. Clin Drug Invest 2004;24:509-21.
7. Reed RC, Dutta S. Predicted serum valproic acid concentrations in pa-
tients missing and replacing a dose of extended-release divalproex sodi-
um. Am J Health Syst Pharm 2004;61:2284-9.
8. Pellock JM, Smith MC, Cloyd JC, Uthman B, Wilder BJ. Extended-re-
lease formulations: simplifying strategies in the management of anti-
epileptic drug therapy. Epilepsy Behav 2004;5:301-7.
9. Doughty J, Baker GA, Jacoby A, Lavaud V. Compliance and satisfaction
with switching from an immediate-release to sustained-release formula-
tion of valproate in people with epilepsy. Epilepsy Behav 2003;4:710-6.
10. Qiu Y, Cheskin HS, Engh KR, Poska RP. Once-a-day controlled-release
dosage form of divalproex sodium I: formulation design and in vitro/in
vivo investigations. J Pharm Sci 2003;92:1166-73.
11. Dutta S, Zhang Y, Selness DS, et al. Comparison of the bioavailability of
unequal doses of divalproex sodium extended-release formulation rela-
www.theannals.com

tive to the delayed-release formulation in healthy volunteers. Epilepsy
Res 2002;49:1-10.
12. Sommerville KW, Dutta S, Biton V, et al. Bioavailability of a divalproex
sodium extended-release formulation versus divalproex sodium delayed-
release tablet formulation in adult patients on enzyme-inducing
antiepileptic drugs. Clin Drug Invest 2003;23:661-70.
13. Dutta S, Reed RC. Divalproex to divalproex extended release conver-
sion. Clin Drug Invest 2004;24:495-508.
14. Dutta S, Zhang Y. Bioavailability of divalproex extended-release formu-
lation relative to the divalproex delayed-release formulation. Biopharm
Drug Dispos 2004;25:345-52.
15. Davis SS, Hardy JG, Fara JW. Transit of pharmaceutical dosage forms
through the small intestine. Gut 1986;27:886-92.
16. Madsen JL. Effects of gender, age, and body mass index on gastrointesti-
nal transit times. Dig Dis Sci 1992;37:1548-53.
17. Argenyi EE, Soffer EE, Madsen MT, Berbaum KS, Walkner WO. Scinti-
graphic evaluation of small bowel transit in healthy subjects: inter- and
intrasubject variability. Am J Gastroenterol 1995;90:938- 42.
18. Degen LP, Phillips SF. Variability of gastrointestinal transit in healthy
women and men. Gut 1996;39:299-305.
19. Fischer JH, Barr AN, Paloucek FP, Dorociak JV, Spunt AL. Effect of
food on the serum concentration profile of enteric-coated valproic acid.
Neurology 1988;38:1319-22.
20. Zhou H. Pharmacokinetic strategies in deciphering atypical drug absorp-
tion profiles. J Clin Pharmacol 2003;43:211-27.
21. Smith MC, Centorrino F, Welge JA, Collins MA. Clinical comparison of
extended-release divalproex versus delayed-release divalproex: pooled
data analyses from nine trials. Epilepsy Behav 2004;5:746-51.
22. Wassef AA, Winkler DE, Roache AL, et al. Lower effectiveness of di-
valproex versus valproic acid in a prospective, quasi-experimental clini-
cal trial involving 9,260 psychiatric admissions. Am J Psychiatry
2005;162:330-9.
23. Freitag FG, Collins SD, Carlson HA, et al. A randomized trial of dival-
proex sodium extended-release tablets in migraine prophylaxis. Neurolo-
gy 2002;58:1652-9.
24. Thibault M, Blume WT, Saint-Hilaire JM, et al. Divalproex extended-re-
lease versus the original divalproex tablet: results of a randomized, cross-
over study of well-controlled epileptic patients with primary generalized
seizures. Epilepsy Res 2002;50:243-9.
EXTRACTO
ANTECEDENTES: Queda aún por establecer las características distintivas
de absorción de las formulaciones de divalproex sódico
grastrorresistente convencional, liberación retardada (LR) y la nueva
formulación de liberación sostenida (LS).
OBJETIVO: Diferenciar cualitativa y cuantitativamente las características
de absorción de las formulaciones divalproex-LR y divalproex-LS.
MÉTODOS: Voluntarios sanos (N = 28) recibieron dosis únicas de
comprimidos de 1000 mg de divalproex-LR y de divalproex-LS en
grupos cruzados. Se usaron análisis compartimentales y no
compartimentales para estimar la farmacocinética del ácido valproico
(AVP) a partir de perfiles plasmáticos de concentración-tiempo
determinados mediante una toma intensiva de muestras de sangre
durante 48 horas.
RESULTADOS: El AVP del divalproex-LR no se absorbió en las primeras 2
horas (retraso de absorción) tras administrar la dosis. Tras la liberación
de APV posterior en el intestino, se absorbió alrededor del 63% de la
dosis en menos de 1 hora, es decir, el tiempo medio de absorción fue en
total de 2.9. Se alcanzó la Cmax 4 horas después de haber administrado la
www.theannals.com The Annals of Pharmacotherapy
View publication stats
Absorption Profiles of Divalproex-DR vs -ER
dosis. Se completó la absorción de AVP (~93% de la dosis) en un
periodo de tres vidas medias de absorción (~4.5 h) de retraso tras la
absorción, es decir, de 6 a 7 horas desde la administración de la dosis.
En contra, la absorción de AVP del divalproex-LS comienza
inmediatamente tras la administración, inicialmente a una velocidad
modesta seguida de una absorción lenta y prolongada a una velocidad
constante durante más de 20 h; las concentraciones de AVP a la primera
y segunda hora fueron del 28 y 40% de la Cmax. Aproximadamente, el
53% de la dosis se absorbió en un periodo de 12 horas (tiempo medio de
absorción); la absorción completa sucedió después de más de 20 horas
sin ninguna pérdida de dosis.
CONCLUSIONES: La absorción de AVP de un comprimido de divalproex-
DR gastrorresistente es rápida tras un retraso de 2 horas y se completa
en 6–7 horas tras la administración de la dosis. En contra, la absorción de
AVP de divalproex-ER comienza inmediatamente tras la administración
pero continúa a una velocidad lenta y casi constante durante más de 20
horas.
Violeta Lopez Sanchez
RÉSUMÉ
GENERALITÉS: Les caractéristiques distinctes d’absorption des
formulations conventionnelles à délitage entérique du divalproex de
sodium à libération retardée et des nouvelles formes à libération
prolongée ne sont pas bien reconnues.
OBJECTIF: Différencier de manière quantitative et qualitative les
caractéristiques d’absorption des formulations du divalproex à libération
retardée et prolongée.
MÉTHODES: Des sujets volontaires en bonne santé (N = 28) avec
crossover ont reçu des doses uniques de divalproex à libération retardée
et prolongée. Les analyses non compartimentales et compartimentales
ont été utilisées pour estimer la pharmacocinétique de l’acide valproïque
(VPA) reposant sur les profils de concentrations plasmatiques-temps
déterminés par d’intensifs prélèvements sanguins pendant 48 heures.
RÉSULTATS: Dans le cas du divalproex à libération retardée, VPA ne fut
pas absorbé pendant les 1ères 2 heures (absorption lag time) après
l’administration de doses. Environ 63% de la dose VPA post libérée
dans l’intestin fut absorbée, c’est-à-dire 2.9 heures (temps moyen
d’absorption) à partir de l’administration de doses. Cmax fut atteint
environ 4 heures après l’administration de doses. L’absorption de VPA
fut complète (environ 93%) dans les limites de 3 demi-vies d’absorption
(environ 4.5 h) post-absorption lag time, c’est-à-dire, 6–7 heures à partir
de l’administration de doses. Par contre, en ce qui concerne le
divalproex à libération prolongée, l’absorption de VPA débute
immédiatement après l’administration, initialement à un rythme modéré,
suivi par une absorption lente et prolongée à un rythme constant pendant
>20 heures. Les concentrations de VPA à 1 heure et 2 heures furent
respectivement 28% et 40% de Cmax. Environ 53% de la dose fut
absorbée dans les 12 heures (temps moyen d’absorption). L’absorption
complète fut observée pendant >20 heures sans doses perdues.
CONCLUSIONS: L’absorption de VPA du divalproex à libération retardée
est rapide selon un lag time d’environ 2 heures et est complète dans des
limites de temps d’environ 6–7 heures à partir de l’administration de
doses. Par contre, en ce qui concerne le divalproex à libération
prolongée, l’absorption débute immédiatement après l’administration,
cependant à un rythme lent, approximativement constant pendant >20
heures.
Thierry Youmbi
n 2006 April, Volume 40 n 625