COMMENTARY
Dibenzazepine Agen s in Epilepsy: How Does
Eslica bazepine Ace a e Di e ?
Cha lo e Law hom .Jukka Pel ola .Rob McMu ay .
Emma Dodd .Vicen e Villanue a
Recei ed: Augus 6, 2018 / Published online: Sep embe 24, 2018
ÓThe Au ho (s) 2018
ABSTRACT
Ca bamazepine (CBZ), oxca bazepine (OXC),
and eslica bazepine ace a e (ESL) belong o he
dibenzazepine amily o an iepilep ic d ugs and
a e all hough o p ima ily ac as sodium
channel blocke s (SCBs). Howe e , ESL is s uc-
u ally dis inc om CBZ and OXC, esul ing in
di e ences in me abolism, pha macokine ics,
and pha macodynamics. Despi e a lack o di ec
compa a i e da a, e idence o po en ial di e -
ences in e ec i eness and ole abili y wi hin
he dibenzazepine amily has eme ged om
s udies in which pa ien s being ea ed wi h one
dibenzazepine agen ha e ecei ed adjunc i e
ea men wi h ano he (ha ing achie ed
insu icien seizu e con ol wi h he i s ) o
ha e ansi ioned om one dibenzazepine
agen o ano he because o lack o e ec i eness
o poo ole abili y. Mos o hese s udies ha e
been conduc ed in he eal-wo ld clinical p ac-
ice se ing. ESL has been shown o be e ec i e
as adjunc i e he apy in pa ien s who ha e
p e iously achie ed inadequa e seizu e con ol
wi h CBZ, indica ing ha he use o di e en
dibenzazepine agen s in combina ion can p o-
ide addi i e e ec i eness bene i s, which may
e lec unde lying di e ences in hei mecha-
nisms o ac ion. Simila ly, ESL mono he apy
can be e ec i e in pa ien s who ha e swi ched
om ano he dibenzazepine, such as CBZ o
OXC, because o inadequa e e icacy. The e is
also conside able e idence o demons a e ha
pa ien s ansi ioning om OXC o CBZ o ESL
as a esul o ad e se e en s expe ience
imp o emen s in ole abili y, which may also
be associa ed wi h imp o emen s in quali y o
li e, ale ness, and/o lipid p o iles. Cu en
e idence he e o e demons a es ha ESL di e s
om o he dibenzazepine agen s in e ms o
e ec i eness and ole abili y.
Funding: Eisai L d.
Keywo ds: An iepilep ic d ug; Ca bamazepine;
Dibenzazepine; Epilepsy; Eslica bazepine
ace a e; Oxca bazepine
Enhanced digi al ea u es To iew enhanced digi al
ea u es o his a icle go o h ps://doi.o g/10.6084/
m9. igsha e.7082696.
C. Law hom (&)
Aneu in Be an Uni e si y Heal h Boa d,
Newpo , UK
e-mail: [email p o ec ed]
J. Pel ola
Tampe e Uni e si y Hospi al, Tampe e, Finland
R. McMu ay E. Dodd
Eisai Eu ope L d, Ha ield, He o dshi e, UK
V. Villanue a
Re ac o y Epilepsy Uni , Neu ology Se ice,
Hospi al Uni e si a io y Poli e
´cnico La Fe, Valencia,
Spain
Neu ol The (2018) 7:195–206
h ps://doi.o g/10.1007/s40120-018-0111-2
INTRODUCTION
Sodium channel blocke s (SCBs) ha e played a
cen al ole in he pha macological manage-
men o ocal and gene alized onic–clonic
seizu es since pheny oin was i s used as an
an iepilep ic d ug (AED) in 1936 [1]. In addi ion
o pheny oin, AEDs ha a e p ima ily SCBs in
hei mechanism o ac ion include lamo igine
(LTG), lacosamide, u inamide, and he diben-
zazepine ca boxamide amily o AEDs, com-
p ising ca bamazepine (CBZ), oxca bazepine
(OXC), and eslica bazepine ace a e (ESL) [1].
Despi e sha ing a common mechanis ic a ge —
he ol age-ga ed sodium channel—SCBs di e
in hei speci ic mechanisms o ac ion. Fo
example, pheny oin, CBZ, and OXC a e
hough o block he as inac i a ion o sodium
channels, whe eas lacosamide and ESL a e
hough o ac on he slow inac i a ed s a e o
sodium channels [1–3]. Such mechanis ic a i-
a ions may unde lie he obse ed di e ences in
e icacy o SCBs, and help o explain why SCBs
may show addi i e e ec s when used in com-
bina ion, compa ed wi h when used as single
agen s alone [1]. Fu he mo e, SCBs di e
ma kedly om each o he in hei ole abili y/
sa e y p o iles, pa icula ly in e ms o hei
a iable p opensi y o enzyme induc ion [1,4].
The objec i e o his a icle is o e iew cu -
en e idence o di e ences in e ec i eness
and ole abili y be ween ESL and o he agen s
wi hin he dibenzazepine amily o AEDs, and
o discuss he implica ions o hese di e ences
in clinical p ac ice. This a icle is based on p e-
iously conduc ed s udies. All p ocedu es pe -
o med in s udies conduc ed by he au ho s
in ol ing human pa icipan s we e in acco -
dance wi h he e hical s anda ds o he ins i u-
ional and/o na ional esea ch commi ee and
wi h he 1964 Helsinki decla a ion and i s la e
amendmen s o compa able e hical s anda ds;
in o med consen was ob ained om all indi-
idual pa icipan s included in hese s udies.
The a icle also con ains s udies wi h human
pa icipan s o animals ha we e no pe o med
by any o he au ho s.
OVERVIEW OF THE DIBENZAZEPINE
FAMILY
CBZ was disco e ed by chance in he 1950s
[5,6] and, o e subsequen decades, became
widely accep ed as he i s d ug o choice o
ocal epilepsy [7] and he s anda d compa a o
o Eu opean egula o y s udies in newly diag-
nosed epilepsy [1]. OXC, a ke o analogue o CBZ
[8], was in oduced in he 1980s, wi h he aim
o imp o ing he pha macokine ic p o ile and
ole abili y o CBZ [6]. In he 1990s, es e s o i s
majo me aboli e we e syn hesized in o de o
p o ide u he imp o emen s in pha maco-
kine ics and ole abili y, which esul ed in he
de elopmen o ESL [6,9].
Pha macology and Po en ial Clinical
Co ela es
CBZ, OXC, and ESL all ha e a dibenzazepine
nucleus bea ing he 5-ca boxamide subs i u e,
bu ESL is s uc u ally dis inc om CBZ and
OXC a he 10,11-posi ion (Fig. 1)[10,11],
esul ing in di e ences in me abolism, pha -
macokine ics, and pha macodynamics (Table 1)
[15]. Unlike CBZ, ESL is no me abolized o
ca bamazepine-10,11-epoxide, he me aboli e
which may play a majo ole in causing ad e se
e ec s, such as d owsiness, dizziness, and
diplopia [6,9,16,17]. CBZ’s ole abili y is u -
he a ec ed by i s po en enzyme induc ion
p ope ies, which educe he du a ion and
ac ion o many d ugs [4] and may addi ionally
con ibu e o he de elopmen o como bidi-
ies, including ascula disease, os eopo osis,
and sexual dys unc ion, ia e ec s on enzymes
in ol ed in endogenous me abolic pa hways
[4,18].
ESL’s me abolism also di e s om ha o
OXC. OXC apidly unde goes hyd oxyla ion o
an ac i e 10-monohyd oxy de i a i e, com-
p ising an enan iome ic mix u e o (S)-lica -
bazepine (‘‘eslica bazepine’’) and (R)-
lica bazepine (in a a io o 4:1), wi h a small
amoun o pa en compound emaining in he
plasma [8,11,19]. ESL is s e eoselec i ely
me abolized p ima ily o (S)-lica bazepine,
which accoun s o app oxima ely 94% o
196 Neu ol The (2018) 7:195–206
plasma d ug exposu e ollowing o al adminis-
a ion o ESL, wi h (R)-lica bazepine and
oxca bazepine accoun ing o app oxima ely
5% and less han 1%, espec i ely ( a io o
(S)-lica bazepine o (R)-lica bazepine, app oxi-
ma ely 19:1; Fig. 1)[11,19]. Impo an ly, his
s e eoselec i e me abolism o ESL a oids he
ea ly peaks in (R)-lica bazepine and OXC
Fig. 1 Chemical s uc u e and main me abolic pa hways
o CBZ, OXC, and ESL. CBZ ca bamazepine, ESL
eslica bazepine ace a e, OXC oxca bazepine. Rep in ed by
pe mission om Sp inge Na u e, Neu o he apeu ics
(Eslica bazepine ace a e (BIA 2-093)), Luis Almeida,
Pa ı
´cio Soa es-da-Sil a, 2007
Table 1 Summa y compa ison o me abolism, pha macokine ics, and pha macodynamics o CBZ, OXC, and ESL
[3,11–14]
CBZ OXC ESL
Me abolism
Rou e Hepa ic Hepa ic Hepa ic
P ima y
bio ans o ma ion
pa hway
Epoxide bio ans o ma ion o
CBZ 10,11-epoxide
Cy osolic educ ion o
10-monohyd oxy de i a i e
Hyd olysis o (S)-lica bazepine
Pha macokine ics
Time o maximum
plasma concen a ion
12 h 4.5 h 2–3 h
Time o s eady s a e 1–2 weeks 2–3 days 4–5 days
Hal -li e o ac i e
me aboli e
6 h 9.3 h 20–24 h
Pha macodynamics
Main ac i e me aboli e CBZ 10,11-epoxide 10-Monohyd oxy de i a i e
a
(S)-Lica bazepine
P ima y mechanism o
ac ion
Fas inac i a ion o ol age-
ga ed sodium channels
Fas inac i a ion o ol age-
ga ed sodium channels
Slow inac i a ion o ol age-
ga ed sodium channels
CBZ ca bamazepine, ESL eslica bazepine ace a e, OXC oxca bazepine
a
An enan iome ic mix u e o (S)-lica bazepine and (R)-lica bazepine in a a io o 4:1
Neu ol The (2018) 7:195–206 197
concen a ions obse ed in plasma and ce e-
b ospinal luid ollowing immedia e- elease
OXC adminis a ion, which a e hough o be
associa ed wi h OXC- ela ed side e ec s, such as
dizziness and headache [19]. Consequen ly, ESL
may be associa ed wi h ewe neu ological/psy-
chia ic side e ec s han OXC [13,14,20].
E idence om he apeu ic d ug moni o ing
da abases has demons a ed ha ESL is associ-
a ed wi h ewe in e ac ions wi h o he AEDs
han CBZ and OXC [21]. The pha macokine ic
p o ile o ESL was shown o be una ec ed by
enzyme-inducing AEDs and alp oa e, and ESL
did no a ec he me abolism o LTG; by con-
as , CBZ clea ance was ound o be inc eased
by enzyme-inducing AEDs and alp oa e, and
LTG clea ance was inc eased by bo h OXC and
CBZ [21]. Fu he mo e, once-daily dosing o ESL
may imp o e he likelihood o adhe ence o ESL
ea men in compa ison wi h CBZ and OXC,
since adhe ence o a once-daily AED dosing
egimen has been shown o be signi ican ly
be e han o a wice- o h ee- imes daily
dosing egimen [22].
Pha macodynamic Di e ences
ESL di e s pha macodynamically om CBZ,
since ESL is hough o ac p ima ily by
enhancing he slow inac i a ion o ol age-
ga ed sodium channels, whe eas CBZ is hough
o al e he as inac i a ion o hese channels
[3]. As p e iously men ioned, al hough OXC
and ESL sha e a common ac i e me aboli e (es-
lica bazepine), his accoun s almos exclusi ely
o ESL’s pha macodynamic e ec , whe eas i is
he monohyd oxy de i a i e o OXC (comp is-
ing a subs an ial p opo ion o (R)-lica bazepine
as well as eslica bazepine) ha accoun s o i s
pha macodynamic e ec (Fig. 1)[8,11,19]. The
appa en a ini y o eslica bazepine o ol age-
ga ed sodium channels in he inac i a ed and
es ing s a es is conside ably lowe han ha o
CBZ and (R)-lica bazepine, sugges ing ha esli-
ca bazepine has enhanced selec i i y o
inhibi ing apidly i ing ac i e ‘‘epilep ic’’ neu-
ons [23–26]. This may in pa explain he
obse ed e icacy o ESL in he p esence o CBZ
esis ance in expe imen al models o epilepsy,
which ha e demons a ed signi ican add-on
e ec s when ESL is used in combina ion wi h
CBZ [27,28]. These pha macodynamic di e -
ences a e hough o in luence he clinical
e ec i eness and ole abili y o he membe s o
he dibenzazepine amily, CBZ, OXC, and ESL.
DIFFERENCES IN EFFECTIVENESS:
EVIDENCE FROM CLINICAL
STUDIES
Few clinical s udies ha e di ec ly compa ed he
e icacy and ole abili y o membe s o he
dibenzazepine amily. Resul s om he S an-
da d And New An iepilep ic D ugs (SANAD)
s udy sugges ed a non-signi ican ad an age o
CBZ compa ed wi h OXC [7]. ESL was no
included as i was no app o ed a he ime
SANAD was conduc ed. In a phase III andom-
ized con olled ial, ESL was shown o be non-
in e io o con olled- elease CBZ (CBZ-CR) as
mono he apy o pa ien s wi h newly diagnosed
ocal epilepsy [29]. Seizu e eedom a es du ing
he en i e 6-mon h e alua ion pe iod o he
pe p o ocol popula ion ( he p ima y endpoin
o he ial) we e 71.1% wi h ESL and 75.6%
wi h CBZ-CR [29].
Despi e he lack o di ec compa a i e da a,
e idence o po en ial di e ences in e ec i e-
ness (and ole abili y) wi hin he dibenzazepine
amily has eme ged om s udies in which
pa ien s ha e ecei ed adjunc i e ea men
wi h a dibenzazepine AED. These pa ien s had
achie ed insu icien seizu e con ol wi h
ano he AED, o we e ansi ioned om one
dibenzazepine agen o ano he , because o lack
o e ec i eness o poo ole abili y. The majo -
i y o his e idence—as ou lined below—has
eme ged om he eal-wo ld se ing, high-
ligh ing he impo ance o clinical p ac ice
s udies in u he complemen ing e idence
ob ained om clinical ials.
A pos hoc analysis was conduc ed o 1049
pa ien s en olled in h ee phase III, mul icen e ,
double-blind, andomized, placebo-con olled
ials, in which adul s wi h ocal seizu es,
despi e ea men wi h one o h ee AEDs,
ecei ed adjunc i e ea men wi h ESL [30].
E icacy was compa ed be ween pa ien s who
198 Neu ol The (2018) 7:195–206
did (n= 611) and did no (n= 438) ecei e CBZ
as a concomi an AED [30]. Seizu e equency
o e a 12-week main enance pe iod ( he p i-
ma y endpoin ) was signi ican ly educed,
compa ed wi h placebo, in pa ien s ea ed wi h
ESL 800 and 1200 mg/day, no only when
adminis e ed wi hou CBZ (p 0.01 and
p 0.05, espec i ely) bu , in e es ingly, also
when adminis e ed wi h CBZ (p 0.01 and
p 0.0001, espec i ely) [30]. In pa ien s no
ecei ing CBZ, esponde a es ( esponse
de ined as a leas 50% seizu e equency
educ ion) we e 26% wi h placebo, 39% wi h
ESL 800 mg/day, and 40% wi h 1200 mg/day
(Fig. 2)[30]. In hose ecei ing CBZ, esponde
a es we e 19% wi h placebo, 34% wi h ESL
800 mg/day and 47% wi h ESL 1200 mg/day
(Fig. 2)[30]. These da a sugges ha ESL was
e icacious in bo h subse s o pa ien s, wi h o
wi hou concomi an CBZ. Pa ien s who did no
achie e seizu e con ol wi h CBZ may s ill
bene i om ei he swi ching o o adding in
adjunc i e ESL.
These indings a e suppo ed by he esul s o
eal-wo ld clinical p ac ice s udies. The ESLI-
BASE s udy was a 1-yea , mul icen e , e o-
spec i e, non-in e en ional s udy, which
e alua ed he long- e m e icacy and sa e y o
adjunc i e ESL he apy in 327 pa ien s wi h
ocal epilepsy in a clinical p ac ice se ing in
Spain [31]. Six y pa ien s included in he ESLI-
BASE s udy we e being ea ed wi h CBZ a he
s a o he s udy [31]. O hese, 45 we e an-
si ioned om CBZ o ESL (because o lack o
e icacy [n= 28] and ole abili y p oblems
[n= 17] wi h CBZ; CBZ o ESL a io, 1:1.5) and
15 we e ea ed wi h bo h AEDs [31]. Among
hose who ansi ioned om CBZ o ESL
(n= 45), 13.3% we e seizu e ee du ing he
3 mon hs p io o inclusion [31]. A e
12 mon hs o ESL ea men , 11.4% o pa ien s
we e seizu e ee, 38.7% had esponded o ESL
ea men ( esponse de ined as a leas 50%
seizu e equency educ ion om baseline), and
20.0% had expe ienced wo sening o seizu es
[31]. Simila ly, 50 pa ien s we e ecei ing OXC
be o e he s a o he s udy, o whom 48 we e
ansi ioned o ESL (because o ad e se e en s
[AEs; n= 26] and lack o e icacy [n= 22]; OXC
o ESL a io, 1:1) and wo we e ea ed wi h bo h
AEDs. Among pa ien s who ansi ioned om
OXC o ESL (n= 48), 16.7% we e seizu e ee in
he 3 mon hs p io o inclusion [31]. A e
12 mon hs o ESL ea men , 31.3% we e seizu e
ee, 45.9% we e esponde s, and 16.7% had
expe ienced wo sening o seizu es [31].
A single-cen e , obse a ional, desc ip i e,
c oss-sec ional s udy in es iga ed he e ec i e-
ness and sa e y/ ole abili y o ESL in 61 con-
secu i e pa ien s wi h d ug- esis an epilepsy
o e a mean ollow-up du a ion o 4.7 mon hs
[32]. These included 13 pa ien s who ansi-
ioned o e nigh om CBZ o ESL (CBZ o ESL
a io, 1:1 o 1:1.3; up o a maximum ESL dose o
1600 mg/day) and 12 pa ien s who ansi ioned
o e nigh om OXC o ESL (OXC o ESL a io,
1:1.1) [32]. Eigh o he 13 pa ien s who ansi-
ioned om CBZ we e ollowed up o mo e
han 3 mon hs a e ini ia ing ESL and he
median numbe o seizu es/mon h inc eased by
33.3% in hese pa ien s (p= no signi ican )
[32]. In con as , 11 o he 12 pa ien s who
ansi ioned om OXC we e ollowed up o
mo e han 3 mon hs and he median numbe o
seizu es/mon h dec eased by 66.7% in hese
pa ien s (p= 0.017) [32].
The EARLY-ESLI s udy was a mul icen e ,
e ospec i e, 1-yea , obse a ional s udy con-
duc ed in 16 hospi als in Spain, which included
Fig. 2 Responde a es in pa ien s included in he ESL
phase III adjunc i e he apy ials who we e no ecei ing
CBZ o who we e ecei ing CBZ. Response was defined as
a leas 50% seizu e equency educ ion om baseline
[30]. CBZ ca bamazepine, ESL eslica bazepine ace a e
Neu ol The (2018) 7:195–206 199

253 pa ien s wi h ocal epilepsy, aged 18 yea s
o o e , in whom ESL ea men was ini ia ed
because o ailu e wi h he i s AED
mono he apy [33]. The p ima y easons o ESL
ini ia ion we e lack o e icacy (n= 157) and
ole abili y p oblems (n= 74) wi h he i s AED
mono he apy [33]. A o al o 28 pa ien s con-
e ed om CBZ mono he apy o ESL
mono he apy a some poin du ing ollow-up.
Th ee o ou pa ien s who con e ed o ESL
because o lack o e icacy wi h CBZ we e seizu e
ee a 1 yea [33]. Simila ly, 12 pa ien s con-
e ed om OXC mono he apy o ESL
mono he apy, and one o ou pa ien s who did
so because o lack o e icacy was seizu e ee a
1yea [33].
Di e ences in E ec i eness: Summa y
Despi e limi ed di ec head- o-head da a, ESL
has been shown o be e ec i e as adjunc i e
he apy in pa ien s who ha e p e iously
achie ed inadequa e seizu e con ol wi h CBZ
[30]. This indica es ha he use o di e en
dibenzazepine agen s in combina ion can p o-
ide addi i e e ec i eness bene i s, which may
e lec unde lying di e ences in he mecha-
nisms o ac ion o his amily o AEDs
[1,27,28]. Simila ly, his may in pa explain
why ESL mono he apy can be e ec i e in
pa ien s who ha e swi ched om ano he
dibenzazepine, such as CBZ o OXC, because o
inadequa e e icacy [31–33]. O e all, he e i-
dence ou lined he e demons a es ha pa ien s
who ha e no achie ed adequa e seizu e con ol
wi h CBZ o OXC may expe ience imp o emen
in seizu e con ol wi h ESL, when used ei he as
an adjunc i e ea men o as mono he apy.
TOLERABILITY PROFILES: DO THESE
DIFFER?
In he phase III ial ha compa ed ESL wi h
CBZ-CR as mono he apy o pa ien s wi h
newly diagnosed ocal epilepsy, he sa e y p o-
ile o ESL was gene ally simila o ha o CBZ-
CR [29]. The incidence o ea men -eme gen
AEs judged o be a leas possibly ea men -
ela ed was lowe wi h ESL han wi h CBZ-CR
(42.1% s 51.5%), as was he incidence o
ea men -eme gen AEs leading o discon inu-
a ion (14.0% s 18.4%) [29]. Fewe pa ien s
ea ed wi h ESL e sus CBZ-CR expe ienced
ea men - ela ed inc eases in gamma-glu amyl
ans e ase (3.5% s 13.1%) and alanine
amino ans e ase (0.5% s 2.2%) [29]. In addi-
ion, ewe pa ien s ea ed wi h ESL e sus
CBZ-CR discon inued because o alle gic de -
ma i is (0.5% s 1.7%) [29]. The incidence o
hypona emia judged o be a leas possibly
ea men - ela ed was highe wi h ESL han
wi h CBZ-CR (2.5% s 1.0%), bu only one
pa ien (in he CBZ-CR g oup) discon inued
because o his AE [29]. In he pos hoc analysis
o 1049 pa ien s en olled in h ee ESL phase III
adjunc i e he apy ials, in which ou comes
we e compa ed be ween pa ien s who did
(n= 611) and did no (n= 438) ecei e CBZ as a
concomi an AED, he incidence o ea men -
eme gen AEs was highe in pa ien s ea ed
wi h CBZ (50% wi h placebo; 71% wi h ESL)
han in hose who we e no ea ed wi h CBZ
(40% wi h placebo; 55% wi h ESL) [30].
Se e al eal-wo ld clinical p ac ice s udies
ha e in es iga ed he impac on ole abili y o
ansi ioning pa ien s who we e expe iencing
OXC- o CBZ- ela ed AEs o ESL [31–37]. In a
single-cen e , e ospec i e cha e iew o 23
pa ien s wi h ocal epilepsy who we e ansi-
ioned o e nigh om OXC o ESL because o
ole abili y p oblems, an e alua ion o he
e ec s o he ansi ion was made a e 1 and
3 mon hs o ESL ea men [34]. The mos
in ole able OXC- ela ed AEs we e somnolence,
dizziness, and diplopia [34], consis en wi h he
a o emen ioned pa e n o OXC me abolism
[19]. OXC- ela ed AEs had esol ed in 14/23
(60.9%) and 15/23 (65.2%) pa ien s 1 and
3 mon hs a e ansi ion, espec i ely [34].
Two- hi ds (66.5%) o OXC- ela ed AEs occu -
ed in he mo ning and 93.4% o hese esol ed
a e ansi ioning om OXC o ESL [34]. The
only ESL- ela ed AE epo ed by a leas 10% o
pa ien s was headache, and, impo an ly, all
pa ien s con inued ESL ea men h oughou
he s udy pe iod [34].
A e ospec i e, single-cen e s udy was
conduc ed in which 21 pa ien s wi h d ug-
200 Neu ol The (2018) 7:195–206
esis an ocal epilepsy on a s able dose o
immedia e- elease OXC o a leas 4 weeks we e
ansi ioned o e nigh o ESL because hey
expe ienced pe sis en seizu es wi h OXC bu
we e unable o ole a e inc eased OXC dosing
because o AEs [35]. Tole abili y (assessed using
he Ad e se E en s P o ile [AEP]), quali y o li e
(assessed using he Quali y o Li e in Epilepsy
In en o y-10 [QOLIE-10]), and ale ness (asses-
sed as eac ion ime using a sub es o he Tes
Ba e y o A en ion Pe o mance e sion 2.3)
we e compa ed immedia ely be o e and 5 days
a e ansi ioning om OXC o ESL [35]. Fol-
lowing ansi ion o ESL, he e we e signi ican
imp o emen s in mean sco es o AEP
(p 0.001), QOLIE-10 (p= 0.001), and ale ness
(p 0.05) [35]. O e all, AEP o al sco es, QOLIE-
10 o al sco es, and ale ness sco es imp o ed
o 100.0%, 81.0%, and 76.2% o pa ien s,
espec i ely [35]. Se um sodium le els (mea-
su ed on days 0 and 5) emained s able, indi-
ca ing ha he likelihood o de eloping
hypona emia did no inc ease a e swi ching
ea men o e nigh [35].
In a single-cen e , obse a ional, desc ip i e,
c oss-sec ional s udy o 61 consecu i e pa ien s
wi h d ug- esis an epilepsy who we e ea ed
wi h ESL o e a mean ollow-up du a ion o
4.7 mon hs, eigh o 13 pa ien s who we e
ansi ioned o e nigh om CBZ o ESL expe-
ienced AEs, as did ou o 12 pa ien s who
ansi ioned o e nigh om OXC o ESL, bu all
AEs we e ansien [32]. Two pa ien s who
ansi ioned om OXC o ESL epo ed ha he
AEs hey expe ienced wi h OXC (dizziness and
d owsiness) imp o ed a e ansi ioning o ESL
[32]. O e all, dec eased sodium le els we e
obse ed in 4/61 (6.6%) pa ien s, none o whom
was aking OXC o CBZ [32]. None o hese
sodium alues we e below 125 mmol/L [32]. In
addi ion, hypona emia esol ed ( om 128 o
138 mmol/L) in one pa ien and emained he
same (133 mmol/L) in ano he pa ien a e
ansi ioning om OXC o ESL [32]. In a mul-
icen e , e ospec i e case se ies o 29 elde ly
pa ien s ([65 yea s) wi h ocal seizu es who
we e ea ed wi h ESL, he ole abili y p o ile
imp o ed in ou o i e pa ien s who ansi-
ioned om CBZ o OXC o ESL because o AEs
[36]. O e all, 3/29 (10.3%) pa ien s de eloped
hypona emia, bu only one de eloped symp-
oms (con usion) ha esol ed a e wi hd awal
o ESL [36].
In he EARLY-ESLI s udy, 14 pa ien s con-
e ed om CBZ mono he apy o ESL
mono he apy because o ole abili y p oblems
wi h CBZ, and only wo o hese pa ien s
epo ed AEs a 1 yea [33]. In addi ion, o he
i e pa ien s con e ed om OXC mono he apy
o ESL mono he apy because o ole abili y
p oblems wi h OXC, h ee had no AEs wi h ESL
a 1 yea [33]. A o al o 137 pa ien s wi hd ew
o ESL mono he apy om ea men wi h o he
AEDs du ing EARLY-ESLI, o whom h ee (2.2%)
de eloped hypona emia; hypona emia was
symp oma ic in only one o hese pa ien s
(113 mEq/L) [33]. In he ESLIBASE s udy, he
s udy popula ion included 26 pa ien s who
we e ansi ioned o ESL om OXC because o
OXC- ela ed AEs and 17 pa ien s who ansi-
ioned om CBZ because o CBZ- ela ed AEs
[31]. O e all, 15/26 (57.7%) pa ien s p e iously
ea ed wi h OXC no longe had AEs a e
ansi ioning o ESL, and 8/17 (47.1%) p e i-
ously ea ed wi h CBZ no longe had AEs a e
ansi ioning o ESL [31]. Thus, a e ansi ion
om ei he OXC o CBZ o ESL, app oxima ely
50% o pa ien s no longe expe ienced AEs. In
he o e all ESLIBASE popula ion, hypona emia
( anging om 116 o 128 mEq/L) was epo ed
in 9/327 (2.8%) pa ien s and led o ea men
discon inua ion in ou pa ien s [31]. Hypona-
emia was asymp oma ic in eigh pa ien s; he
emaining pa ien epo ed con usion ha
esol ed apidly a e ESL discon inua ion and
did no equi e hospi al admission [31].
Nonadhe ence o AEDs is a common and
challenging p oblem in epilepsy, wi h nonad-
he ence a es o up o 60% epo ed in some
popula ions [37]. In he EARLY-ESLI s udy,
adhe ence imp o ed in all o he 10 pa ien s
who con e ed om CBZ o ESL mono he apy
because o adhe ence p oblems wi h CBZ [33].
Simila ly, o i e pa ien s who con e ed om
OXC mono he apy o ESL mono he apy
because o ole abili y p oblems wi h OXC,
h ee had no AEs wi h ESL a 1 yea , and o h ee
pa ien s who con e ed because o adhe ence
p oblems wi h OXC, wo had no adhe ence
p oblems wi h ESL a 1 yea [33].
Neu ol The (2018) 7:195–206 201
Di e ences Wi hin he Dibenzazepine
Family in Thei Impac on Lipid Ma ke s
T ea men wi h enzyme-inducing AEDs,
including CBZ, has been associa ed wi h
inc eased se um lipid le els [4]. Al hough (S)-
lica bazepine is a weak induce o cy och ome
P450 3A4 and u idine 50-diphospho-glucu ono-
syl ans e ase [14], i is a less po en enzyme
induce han CBZ, which induces a b oade
ange o oxida i e and conjuga ing enzymes [1].
Lipid p o iles a e pa icula ly ele an in ligh o
epidemiological e idence demons a ing ha
pa ien s wi h epilepsy ha e signi ican ly ele-
a ed a es o ca dio ascula and ce eb o ascu-
la disease [4]. In a s udy in which 12 pa ien s
wi h epilepsy ansi ioned om CBZ o OXC,
cy och ome P450 enzyme sys em unc ion
(assessed by e alua ing an ipy ine pha macoki-
ne ics) was shown o no malize a e 2 mon hs,
esul ing in a signi ican educ ion in se um
o al choles e ol le els and a de ec able bu no
s a is ically signi ican educ ion in low-densi y
lipop o ein choles e ol le els (p= 0.057), wi h
se um concen a ions o high-densi y lipop o-
ein choles e ol and iglyce ides emaining
unchanged [38]. Seizu e con ol was gene ally
simila ollowing he ansi ion om CBZ o
OXC, wi h only one pa ien expe iencing an
inc ease in seizu e equency [38]. Two u he
s udies ha e speci ically in es iga ed he impac
o ESL on pa ien s’ lipid p o iles, including he
e ec o ansi ioning om CBZ o OXC o ESL
[39,40]. A e ospec i e coho s udy o 36 adul
pa ien s assessed lipid le els be o e and a e
ini ia ing ESL as adjunc i e he apy o e an
a e age ollow-up o 11 mon hs [39]. A e a
leas 6 mon hs o ESL ea men , he e we e
signi ican educ ions in he le els o o al
choles e ol ( om 191.3 o 179.7 mg/dL;
p 0.0001) and low-densi y lipop o ein
choles e ol ( om 114.6 o 103.1 mg/dL;
p 0.0001), and a signi ican inc ease in he
le el o high-densi y lipop o ein choles e ol
( om 57.5 o 63.9 mg/dL; p 0.0001) [39]. A
second obse a ional, e ospec i e, single-cen-
e , coho s udy in es iga ed he impac o ESL
ea men on he lipid p o iles o 108 pa ien s
o e a median du a ion o 23.1 mon hs ( ange
3–41 mon hs) [40]. O hese, 51.9% had
swi ched o ESL om p io ea men wi h CBZ
o OXC [40]. In he o e all popula ion, mean
o al choles e ol le els dec eased signi ican ly
om p e ious pa hological ([200 mg/dL) o
no mal ( 200 mg/dL) alues du ing ESL ea -
men (p= 0.015) [40]. In he subg oup o
pa ien s who ansi ioned om CBZ o OXC o
ESL, he e was a signi ican dec ease in median
iglyce ide le els (p 0.001) and mean low-
densi y lipop o ein choles e ol (p= 0.014), and
a ma ked dec ease in mean o al choles e ol
(p= 0.053) [40]. The p opo ion o pa ien s wi h
hype choles e olemia also dec eased signi i-
can ly a e ansi ioning o ESL ( om 57.1% o
37.5%; p= 0.018), as did he p opo ion o
pa ien s wi h hype iglyce idemia ( om 10.8%
o 1.8%; p= 0.011) (Fig. 3)[40]. Bo h s udies
he e o e sugges ha swi ching o ESL is o
alue when conside ing lipid p o iles.
Fig. 3 Pe cen age o pa ien s wi h no mal o ele a ed
le els o o al choles e ol and iglyce ides be o e and a e
swi ching ea men om CBZ o OXC o ESL. CBZ
ca bamazepine, ESL eslica bazepine ace a e, OXC oxca -
bazepine. Adap ed om Epilepsy Resea ch, 115, M. Ley, A.
P incipe, J. Jime
´nez-Conde, R. Rocamo a, Assessing long-
e m e ec s o eslica bazepine ace a e on lipid me abolism
p ofile, sodium alues and li e unc ion es s,
pp. 147–152, Copy igh (2015), wi h pe mission om
Else ie
202 Neu ol The (2018) 7:195–206
Tole abili y Di e ences: Summa y
In a phase III mono he apy ial, ESL was shown
o ha e a gene ally simila ole abili y p o ile o
CBZ-CR, bu wi h lowe incidences o ea -
men - ela ed AEs and AEs leading o discon in-
ua ion [29]. In clinical ials o ESL adjunc i e
he apy, ea men egimens ha included CBZ
we e shown o be less well ole a ed han hose
ha did no , despi e a aining clinical e icacy
[30]. The e is also conside able e idence o
demons a e ha pa ien s ansi ioning om
OXC o CBZ o ESL because o AEs expe ience
imp o emen s in ole abili y [31–36]. In e es -
ingly, imp o emen s in quali y o li e and/o
ale ness we e also epo ed [35]. Pa ien s an-
si ioning om OXC o CBZ o ESL ha e addi-
ionally demons a ed signi ican
imp o emen s in lipid p o iles [40]. T ansi ion-
ing om CBZ o OXC o ESL does no appea o
be associa ed wi h an inc eased likelihood o
de eloping hypona emia; howe e , i is good
p ac ice o moni o o he po en ial de elop-
men o hypona emia wi h all dibenzazepine
agen s, pa icula ly in he elde ly. Blood le el
moni o ing is no ou inely equi ed o diben-
zazepine AEDs [12–14]. Howe e , dec eased
pla ele o whi e blood cell coun s may occu
wi h CBZ ea men and i is he e o e ecom-
mended ha a comple e blood coun should be
ob ained be o e s a ing CBZ ea men and
pe iodically he ea e [12].
CLINICAL IMPLICATIONS
ESL has been shown o be e ec i e in pa ien s
who ha e p e iously no achie ed adequa e
seizu e con ol wi h CBZ o OXC, when used
ins ead o hese agen s. ESL has also been shown
o imp o e e ec i eness when added o CBZ
ea men . Such e idence is consis en wi h
indings om expe imen al models ha ha e
e ealed di e ences in he mechanism o ac ion
be ween dibenzazepine agen s [3,27,28], sup-
po ing he no ion o ‘‘ a ional poly he apy’’,
whe eby agen s wi h di e en mechanisms o
ac ion may syne gize when used in combina-
ion o p o ide enhanced e ec i eness [41].
Ra ional poly he apy mus , howe e , also ake
accoun o he inc eased isk o ad e se e ec s
and pha macokine ic in e ac ions as a pa ien ’s
d ug bu den is inc eased, and hus balance he
po en ial bene i o inc eased e ec i eness wi h
poly he apy agains he po en ial isk o
inc eased ole abili y p oblems [42]. The co-
adminis a ion o mo e han one AED is likely
o limi dosing o he newly added d ug. In
pa ien s who ail i s -line mono he apy, and
ha e s a ed hei second mono he apy, he e is
e idence o suppo swi ching om CBZ o
OXC o ESL. E idence om clinical p ac ice
s udies also demons a es ha he e a e di e -
ences in ole abili y wi hin he dibenzazepine
amily, since pa ien s who ha e expe ienced
CBZ- o OXC- ela ed AEs may expe ience
imp o emen s in ole abili y when ansi ioned
o ESL. Fu he mo e, i may be app op ia e o
ansi ion pa ien s om CBZ o ESL i hey a e
expe iencing, o a isk o de eloping, me abolic
p oblems esul ing om CBZ induc ion o
enzymes in ol ed in endogenous me abolic
pa hways (e.g., hype choles e olemia, os eo-
po osis, sexual dys unc ion) [4,43,44]. Pub-
lished guidance has ou lined he
me hodological conside a ions associa ed wi h
ansi ioning pa ien s om CBZ o OXC o ESL
[43,44]. Taken oge he , and despi e he lack o
di ec compa a i e da a, cu en e idence
demons a es ha ESL di e s om o he
dibenzazepine AEDs, highligh ing he impo -
ance o adap ing and ailo ing ea men o
each pa ien ’s speci ic needs.
ACKNOWLEDGEMENTS
Funding. Eisai L d unded edi o ial suppo
o he p epa a ion o his manusc ip and he
jou nal’s a icle p ocessing cha ges. All au ho s
had ull access o all o he da a in his s udy
and ake comple e esponsibili y o he in eg-
i y o he da a and accu acy o he da a
analysis.
Au ho ship. All named au ho s mee he
In e na ional Commi ee o Medical Jou nal
Edi o s (ICMJE) c i e ia o au ho ship o his
a icle, ake esponsibili y o he in eg i y o
Neu ol The (2018) 7:195–206 203