Elec oca diog am pheno ypes in hype ophic
ca diomyopa hy caused by dis inc
mechanisms: apico-basal epola iza ion
g adien s s. Pu kinje-myoca dial coupling
abno mali ies
Au o e Lyon
1
, Al onso Bueno-O o io
1
, E nes o Zacu
2
, Rina A iga
3
, Vicen e G au
2
,
S e an Neubaue
3
, Hugh Wa kins
3
, Blanca Rod iguez
1
*
†
, and Ana Minchole´
1†
1
Depa men o Compu e Science, Uni e si y o Ox o d, Wol son Building, Pa ks Rd, Ox o d OX1 3Q, UK;
2
Ins i u e o Biomedical Enginee ing (IBME), Depa men o
Enginee ing Science, Uni e si y o Ox o d, Ox o d, UK; and
3
Di ision o Ca dio ascula Medicine, Radcli e Depa men o Medicine, Uni e si y o Ox o d, Ox o d, UK
Recei ed 9 Sep embe 2018; edi o ial decision 17 Sep embe 2018; accep ed 27 Sep embe 2018
Aims To iden i y key s uc u al and elec ophysiological ea u es explaining dis inc elec oca diog am (ECG) pheno ypes
in hype ophic ca diomyopa hy (HCM).
............................ ............. ............. .................. ............. ............. ............................... ............. ............. .................. ............. .........
Me hods
and esul s
Human hea – o so ana omical models we e cons uc ed om ca diac magne ic esonance (CMR) images o HCM
pa ien s, ep esen a i e o ECG pheno ypes iden i ied p e iously. High pe o mance compu ing simula ions using bido-
main models we e conduc ed o dissec key ea u es explaining he ECG pheno ypes wi h inc eased HCM Risk-SCD
sco es, namely G oup 1A, cha ac e ized by no mal QRS bu in e ed T wa es la e ally and coexis ence o apical and
sep al hype ophy; and G oup 3 wi h ma ked QRS abno mali ies (deep and wide S wa es la e ally) and sep al hype -
ophy. Hype ophic ca diomyopa hy abno mali ies cha ac e ized om CMR, such as hype ophy, issue mic os uc-
u e al e a ions, abno mal conduc ion sys em, and ionic emodelling, we e selec i ely included o assess hei in luence
on ECG mo phology. Elec oca diog am abno mali ies could no be explained by inc eased wall hickness no by local
conduc ion abno mali ies associa ed wi h ib e disa ay o ib osis. In e ed T wa e wi h no mal QRS (G oup 1A)
was ob ained wi h inc eased apico-basal epola iza ion g adien caused by ionic emodelling in sep um and apex.
La e al QRS abno mali ies (G oup 3) we e only eco e ed wi h abno mal Pu kinje-myoca dium coupling.
............................ ............. ............. .................. ............. ............. ............................... ............. ............. .................. ............. .........
Conclusion Two ECG-based HCM pheno ypes a e explained by dis inc mechanisms: ionic emodelling and ac ion po en ial p o-
longa ion in hype ophied apical and sep al a eas lead o T wa e in e sion wi h no mal QRS complexes, whe eas ab-
no mal Pu kinje-myoca dial coupling causes abno mal QRS mo phology in V4–V6. These indings ha e po en ial impli-
ca ions o pa ien s’ managemen as hey poin owa ds di e en a hy hmia mechanisms in di e en pheno ypes.
䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏䊏
Keywo ds Hype ophic ca diomyopa hy •Compu e modelling •Pe sonalized simula ions •Ca diac magne ic eso-
nance imaging •Elec oca diog aphy
* Co esponding au ho . Tel: þ44 1865 610806; ax: 00441865273839. E-mail add ess: blanca. od ig[email p o ec ed]
†
The las wo au ho s se ed as join senio au ho s.
V
CThe Au ho (s) 2018. Published by Ox o d Uni e si y P ess on behal o he Eu opean Socie y o Ca diology.
This is an Open Access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License (h p://c ea i ecommons.o g/licenses/by/4.0/), which pe mi s
un es ic ed euse, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed.
Eu opace (2018) 20, iii102–iii112 ORIGINAL RESEARCH
doi:10.1093/eu opace/euy226
Downloaded om h ps://academic.oup.com/eu opace/a icle/20/suppl_3/iii102/5202143 by Uni e sidad de Za agoza. Biblio eca Uni e si a ia use on 26 Feb ua y 2021
In oduc ion
Hype ophic ca diomyopa hy (HCM) has emained a challenge due
o he ex eme he e ogenei y in i s clinical cou se. Al hough mos
pa ien s a e asymp oma ic, HCM is a majo cause o sudden ca diac
dea h (SCD) in young people due o en icula a hy hmias. These
li e- h ea ening a hy hmias a e e ec i ely abo ed by implan able
ca dio e e -de ib illa o s (ICDs) and he ocus o isk s a i ica ion is
o a ge high- isk pa ien s wi h an ICD. The elec oca diog am
(ECG) is abno mal in he majo i y o pa ien s om abno mal Q
wa es, ST segmen s, T-wa es o wide QRS.
1,2
Howe e , he elec o-
ca diog aphic signa u e o HCM has been inconclusi e o eliable
isk s a i ica ion.
3
P og ess owa ds isk s a i ica ion was achie ed in ecen wo k,
4
whe e ou HCM pheno ypes we e iden i ied using compu a ional
analysis o he ECG. The g ea es de ia ion om a no mal ECG oc-
cu ed in wo o he ou ECG pheno ypes: G oup 1A and G oup 3.
5
Pa ien s in G oup 1A had no mal QRS mo phology, in e ed T wa es
la e ally, inc eased HCM Risk-SCD sco es, and coexis ing sep al and
apical hype ophy. Pa ien s in G oup 3 had QRS abno mali ies wi h
deep wide S wa es la e ally and sep al hype ophy. The s uc u al
and elec ophysiological sou ces o hese di e en ECG pheno ypes
in HCM emain unclea and complex, as hey may include hype o-
phy, ib e disa ay, ib osis, ionic emodelling, as well as Pu kinje ab-
no mali ies.
6–8
A be e unde s anding o he elec ophysiological
and s uc u al basis o he HCM pheno ypes may help o imp o e
pe sonalized pa ien managemen , as he mechanisms unde lying a -
hy hmic isk may di e .
The goal o his s udy is o iden i y he key HCM s uc u al and
elec ophysiological ea u es unde lying he pa ien s’ pheno ypes
wi h mo e abno mal ECGs. We hypo hesise ha abno mali ies in he
mic os uc u e such as ib e disa ay may explain abno mal elec ical
ac i a ion and ma ked QRS abno mali ies in G oup 3, and ha apical
HCM ionic emodelling may lead o p ima y epola iza ion p olonga-
ion and in e ed T wa es in G oup 1A. To achie e hese aims, we
exploi he powe o high pe o mance compu ing simula ions o he
ECG using human biophysically-de ailed o so- en icula models
pe sonalized o he pa ien s’ ca diac magne ic esonance (CMR)
images.
9
Me hods
Hype ophic ca diomyopa hy
pheno ypes
This s udy ocused on he in e p e a ion o he ECG mo phologies o
pa ien s om G oup 1A and 3 iden i ied in Re .,4which exhibi ed he
g ea es de ia ion om a no mal ECG. As a e e ence, simula ions we e
also conduc ed o G oup 1B, which had no mal QRS and T wa e mo -
phologies. One ep esen a i e pa ien was chosen om each o he HCM
pheno ypic g oups based on he p esence o he cha ac e is ic ECG ea-
u es (G oup 1B: no mal QRS, up igh T wa es in V4–V6, G oup 1A: no -
mal QRS, in e ed T wa es in V4–V6, G oup 3: sho R du a ion and
ampli ude, long S du a ion and ampli ude in V4–V6) (Figu e 1A), and he
p esence o he cha ac e is ic hype ophic a eas on he CMR images
(G oup 1A: mixed pa e n o sep al and apical hype ophy, G oups 1B
and 3: sep al hype ophy) (Figu e 1B).
Ca diac magne ic esonance-based
hea – o so models
Bi en icula hea , o so, lungs, and ibs meshes we e de i ed om
CMR scans o each pa ien by image p ocessing, and i ual ECG elec o-
des we e posi ioned on he o so su ace in s anda d placemen o simu-
la e he 12-lead ECG (see Supplemen a y ma e ial online,1.1).
Human en icula memb ane kine ics we e ep esen ed by modi ied
e sions o he human en icula elec ophysiological O’Ha a–Rudy
model o heal hy
10
and HCM ca diomyocy es.
7
The models a e
biophysically-de ailed and include ep esen a ion o sodium, calcium, and
po assium ionic cu en s as well as calcium dynamics. Hype ophic ca di-
omyopa hy- ela ed ionic emodelling was ep esen ed by al e ing so-
dium, calcium, and po assium cu en s as desc ibed in Re .
7
Myoca dial
and o so conduc i i ies and en icula ib e a chi ec u e we e se as de-
sc ibed in Supplemen a y ma e ial online, 1.2. Elec ophysiological he e -
ogenei ies we e in oduced o ep esen he dispe sion o ac ion
po en ial du a ion (APD) in en icula epola iza ion (see
Supplemen a y ma e ial online, 1.3). To simula e sinus hy hm, endoca -
dial ac i a ion was ini ia ed in oo nodes (o si es o ea lies ac i a ion)
connec ed wi h a as ac i a ing endoca dial laye as in Re .,9 ep esen ing
a phenomenological model o Pu kinje- en icula coupling connec ed o
he es o he wo king myoca dium (see Supplemen a y ma e ial online,
1.2). The loca ion o ea ly ac i a ion nodes in all ou hea geome ies
we e selec ed o ollow co esponden ana omical loca ions by mapping
homologous ana omical egions, such as inse ion poin s, apex, and la e al
walls, among all geome ies. The elec ophysiological model was he e-
o e kep he same in all cases in o de o isola e he e ec o each HCM-
ela ed ea u e on he ECG, a he han aiming a ma ching he clinical
ECG in a pa ien -speci ic app oach. Va iabili y in body conduc i i ies,
which has been shown o s ongly a ec QRS ampli ude
9
may be espon-
sible o di e ences in wa e ampli ude be ween simula ed and clinical
ECGs. Analysis and compa isons o he ECG signals we e he e o e
made in e ms o o e all mo phology and pola i y.
Hype ophic ca diomyopa hy
abno mali ies
Se e al hypo heses we e es ed o explain he ECG pheno ypes in HCM,
based on CMR-based ea u es and HCM elec ophysiological and
Wha ’s new?
•Simula ions using mul iscale whole-o gan compu e models
based on ca diac magne ic esonance images dissec he in lu-
ence o s uc u al and elec ophysiological abno mali ies on
he elec oca diog am (ECG) in hype ophic ca diomyopa hy
(HCM).
•Inc eased wall hickness, ib e disa ay, bundle block, o myo-
ca dial abno mali ies do no accoun o he di e en ECG
pheno ypes iden i ied in HCM.
•An abno mal Pu kinje-myoca dium coupling is he only mecha-
nism we ound ha could explain he QRS abno mali ies wi h
deep and wide S wa es in la e al leads obse ed in one pa ien
g oup.
•Apico-basal epola iza ion he e ogenei ies due o ionic emod-
elling in coexis ing egions o sep al and apical hype ophy ex-
plain he no mal QRS and in e ed T wa es obse ed in he
pheno ype associa ed wi h inc eased sudden ca diac dea h isk
sco e.
ECG pheno ypes in HCM caused by dis inc mechanisms iii103
Downloaded om h ps://academic.oup.com/eu opace/a icle/20/suppl_3/iii102/5202143 by Uni e sidad de Za agoza. Biblio eca Uni e si a ia use on 26 Feb ua y 2021
s uc u al unde s anding (Figu e 1C). The ana omy (inc eased wall hick-
ness) was in insically included in he hea – o so models as hey we e
pe sonalized om CMR images. Abno mali ies in issue mic os uc u e
we e also in es iga ed. Fib e disa ay was modelled as an iso opic con-
duc ion in he egions o low ac ional aniso opy ob ained om di u-
sion enso imaging (DTI) da a. Dec eased conduc ion eloci y (such as
due o ib osis, known o a ec HCM hea s
6
) was modelled in hose
a eas as a ou - ime educ ion in issue conduc i i ies, esul ing in educed
conduc ion eloci y o 20 cm/s. Con e sely, hype ophied cells ha e
been shown o lead o inc eased conduc ion eloci y in ea lie s ages o
le en icula hype ophy.
11
Such an e ec was he e o e addi ionally
in es iga ed by a ou - ime inc ease in conduc i i ies, leading o an in-
c eased conduc ion eloci y in he egions o hype ophy.
Bundle b anch blocks ha e also been epo ed in HCM,
12,13
and he e -
ec o a ious blocks in he conduc ion sys em on he ECG we e he e-
o e es ed (le bundle b anch block, LBBB; le en icula pos e io and
an e io hemiblocks; igh bundle b anch block, RBBB) by dele ing associ-
a ed oo nodes in he model.
9
In addi ion, slu ed S wa e mo phologies
Figu e 1 In es iga ion o he e ec o HCM abno mali ies on he ECG. (A) ECG mo phologies o he HCM pheno ypes s udied: G oup 1B wi h
no mal ECG mo phology in V4–V6, G oup 3 wi h QRS abno mali ies wi h deep and wide S wa es in V4–V6, and G oup 1A wi h no mal QRS mo -
phology bu in e ed T wa es in V4–V6. (B) Pe sonalized 3D olume ic meshes o he hea and o so a e compu ed o each ep esen a i e pa ien .
(C) The e ec o HCM abno mali ies (ana omy, issue mic os uc u e, conduc ion sys em, and ionic emodelling) on he ECG a e in es iga ed using
compu e simula ions. CMR, ca diac magne ic esonance; ECG, elec oca diog am; HCM, hype ophic ca diomyopa hy.
iii104 A. Lyon e al.
Downloaded om h ps://academic.oup.com/eu opace/a icle/20/suppl_3/iii102/5202143 by Uni e sidad de Za agoza. Biblio eca Uni e si a ia use on 26 Feb ua y 2021
in V5–V6 may be ypical o RBBB, sugges ing i s po en ial implica ion in
he QRS abno mali ies o G oup 3 pa ien s. Abno mal Pu kinje-
myoca dium coupling, as epo ed in HCM pa ien s,
8
was induced by a
misma ch be ween he as endoca dial laye and he myoca dium in o-
duced by doubling conduc i i ies. Fas p opaga ion egions ha e been
shown capable o c ea ing unidi ec ional p opaga ion block because o
sou ce-sink misma ch.
14
Finally, HCM ionic emodelling was in oduced
in he egions o hype ophy, as desc ibed in Re .
7
This consis ed in an al-
e a ion o he cellula ionic p ope ies, mainly including an inc ease o
he la e sodium and he L- ype calcium cu en , educ ion o he po as-
sium cu en s, and emodelling o he calcium handling subsys em.
Resul s
E ec o ca diac ana omy on he
elec oca diog am pheno ypes in
hype ophic ca diomyopa hy
The ana omical e ec s o ca diac shape, hype ophy dis ibu ion,
and ex en on he ECG we e in es iga ed by pa ien -speci ic geome-
ies de i ed om CMR images (Figu e 2). Despi e cha ac e is ic dis-
ibu ions o hype ophy in each g oup, he simula ions yielded
simila QRS mo phology and up igh T wa es in all g oups, which
we e consis en wi h he clinical ECG om G oup 1B. As a sensi i i y
analysis, we also analysed he in luence o a ying bo h he
in acellula and ex acellula conduc i i ies on he shape o he QRS.
Conduc i i y changes did no lead o mo phological abno mali ies
such as na ow all R wa es and deep wide S wa es as seen in G oup
3(seeSupplemen a y ma e ials online, 2.1). The e o e, inc eased
wall hickness alone did no explain abno mali ies in QRS mo phol-
ogy in G oup 3, o he T wa e in e sion in G oup 1A.
E ec o abno mali ies in issue
mic os uc u e on he QRS complex
E ec o local iso opic elec ical p opaga ion due o ib e
disa ay
Hype ophic ca diomyopa hy hea s exhibi s uc u al abno mali ies
such as ib e disa ay, which co esponds o a disa angemen o he
ca diac ib es and a loss o he no mal elec ical aniso opy p ope y
o he ca diomyocy es. This may a ec en icula ac i a ion, and in
u n he QRS complex, as obse ed in G oup 3 pa ien s. Howe e ,
he in oduc ion o iso opic elec ical conduc ion in a eas wi h low
ac ional aniso opy om DTI made li le change o he simula ed
ECGinG oup3(seeSupplemen a y ma e ial online,2.2).
E ec o al e ed conduc ion eloci y by ib osis and
hype ophy
We hen e alua ed he po en ial impac o slowed elec ical conduc-
ion in ou di e en egions o he myoca dium on he ECG
Figu e 2 E ec o he ana omy on he ECG pheno ypes obse ed in HCM. Simula ed ECG o ep esen a i e pa ien s in G oup 1B (A), G oup 3
(B), and G oup 1A (C) o leads V1–V6 along wi h hei pe sonalized hea ana omy. A ows indica e a eas o hype ophy. ECG, elec oca diog am;
HCM, hype ophic ca diomyopa hy.
ECG pheno ypes in HCM caused by dis inc mechanisms iii105
Downloaded om h ps://academic.oup.com/eu opace/a icle/20/suppl_3/iii102/5202143 by Uni e sidad de Za agoza. Biblio eca Uni e si a ia use on 26 Feb ua y 2021
Figu e 3 E ec o changes in conduc ion speed (due o ib osis, hype ophy) on he QRS. (Top) Clinical ECG o G oup 3 pa ien (black). (A)
Simula ed QRS wi h slow conduc ion in a ious loca ions o he myoca dium (mid-sep um, le en icula basal ee wall, apex and sep um om
base o apex), highligh ed in ed. (B) Simula ed QRS wi h as conduc ion in a ious loca ions o he myoca dium (mid-sep um, le en icula basal
ee wall, apex and sep um om base o apex), highligh ed in ed. Simula ed QRS in con ol condi ions is ep esen ed in blue. Black a ows ep esen
lead o in e es discussed in he ex . ECG, elec oca diog am; LV, le en icula .
iii106 A. Lyon e al.
Downloaded om h ps://academic.oup.com/eu opace/a icle/20/suppl_3/iii102/5202143 by Uni e sidad de Za agoza. Biblio eca Uni e si a ia use on 26 Feb ua y 2021

(Figu e 3A). Slowing conduc ion in a mid-sep al egion yielded abno -
mal QRS mo phologies in sep al and an e io leads V1–V4. Slowing
conduc ion in he le en icula basal ee wall did no a ec he
QRS in he sep al leads V1–V3, despi e a smoo he up igh S slope o
he baseline. QRS in leads V4–V6 showed a simila mo phology bu
inc eased wid h. Slowing conduc ion in he apex a ec ed mos ly
leads V4–V6 wi h se e ely abno mal pa e ns o wide QRS. Finally,
slowing down conduc ion in a egion co e ing he en i e sep um
om base o apex led o se e e QRS abno mali ies in sep al leads
V2–V4, wi h educed QRS ampli ude and inc eased wid h, bu simila
mo phologies in leads V5–V6. The abno mali ies in leads V1–V4 wi h
he en i e sep um a ec ed we e milde han in he case o he mid
sep al egion being a ec ed. This was explained by he la ge ac i a-
ion delays in he an e io o he en icles caused by he egion o
slow conduc ion in he mid sep um, which ex ended o he an e io .
These we e no p esen in he en i e sep um case (see ac i a ion
maps in Supplemen a y ma e ial online, 2.3, Figu e S4). Impo an ly
none o he changes in oduced esul ed in he deep S wa es in leads
V4–V6 obse ed in G oup 3.
We hen conside ed ha he inc eased su ace- o- olume a io o
hype ophic myocy es could also lead o an inc ease in elec ical
p opaga ion speed.
11
Fas e conduc ion in bo h apical and sep al
egions inc eased he QRS ampli ude and educed he QRS wid h in
leads V3–V4, bu did no lead o a deep S wa e in leads V5 and V6
(Figu e 3B). Al oge he , hese esul s showed ha posi i e o nega i e
changes in he conduc ion speed in local egions o he myoca dium
did no p oduce a deep S wa e in leads V4–V6, and he e o e could
no explain he abno mali ies obse ed in G oup 3 pa ien s.
E ec o abno mali ies in he conduc ion
sys em on he QRS complex
E ec o al e ed ea ly ac i a ion si es
P esence o b anch blocks, such as LBBB o RBBB, has been
epo ed in HCM.
12,13
We, he e o e, in es iga ed he e ec on
he ECG o hype ophic en icles o a ious combina ions o ac-
i a ion blocks (Figu e 4). Simula ion esul s showed ha a ious
combina ions o ac i a ion blocks in he le en icle a ec ed he
du a ion o he ac i a ion, and he e o e, he wid h o QRS com-
plexes, bu did no change he o e all mo phology o he QRS in
la e al leads. Blocking he ac i a ion in he le en icle ee wall
(LBBB) led o wide QRS complexes in V1–V6 due o slow ac i a-
ion o he le en icle. Knocking ou he pos e io b anch o he
le en icle did no a ec he mo phology o he QRS in any leads,
sligh ly inc easing he ampli ude o he QRS in V2–V4. Knocking
ou he an e io b anch o he le en icle had e y li le e ec on
he QRS, sligh ly inc easing he QRS wid h, and did no a ec wa e
pola i ies. Finally, we in es iga ed he e ec o blocking he ea ly
ac i a ion in he igh en icle (RBBB). This led o a sligh inc ease
in ampli ude and wid h o he S wa es in la e al leads, bu main-
ained he p ominence o he R wa e in V4–V6. The p eco dial
leads V1–V3 we e s ongly a ec ed and showed abno mali ies no
obse ed on he ECG o G oup 3 pa ien s.
We concluded ha al e ing he ac i a ion sequence o he le and
igh en icles did no lead o he mo phological abno mali ies ob-
se ed in he la e al leads V4–V6 combined wi h no mal QRS mo -
phology in leads V1–V3.
E ec o abno mal Pu kinje-myoca dium coupling
The abno mal Pu kinje-myoca dium coupling, modelled as desc ibed
in Me hods sec ion, led o a pa chy ac i a ion and a di use ac i a ion
sequence compa ed wi h he con ol simula ion (Figu e 5Aand B).
This led o he p esence o la e-ac i a ed a eas in he myoca dium
compa ed wi h he con ol simula ion, especially in he igh en icle
base and endoca dium, and he apex (Figu e 5C). This ansla ed in o
deep and wide S wa es in la e al leads V4–V6 o he ECG, wi hou a -
ec ing he mo phology o he sep al leads (Figu e 5D). The e o e, an
abno mal coupling be ween b anches o he Pu kinje sys em and he
en icula endoca dium was he only way o ep oduce he QRS
pola i y abno mali ies in G oup 3, wi h abno mal deep and wide S
wa e in he la e al leads. A ec ing he myoca dial p opaga ion o he
ea ly ac i a ion sequence did no explain hem.
E ec o hype ophic ca diomyopa hy
ionic emodelling on Twa e pola i ies
In o de o in es iga e he pheno ype o G oup 1A pa ien s, we hy-
po hesized ha HCM ionic emodelling, esul ing in longe APD in
he egion o hype ophy (combined sep um and apex) may ans-
la e in an in e sion o he T wa e in leads V4–V6. The HCM ionic
emodelling desc ibed in Re .
7
was modelled in he sep um and apex,
in ag eemen wi h he mixed pa e n o hype ophy o hese
pa ien s. This led o an inc eased APD in he egion o hype ophy
(455 ms wi h HCM ionic emodelling s. 324 ms wi hou ). The simu-
la ed T wa es showed in e sions in he la e al leads V4–V6, while he
T wa es in o he leads emained una ec ed (Figu e 6A). This epli-
ca ed he mo phology o he clinical ECG o G oup 1A pa ien s in all
leads. The QRS complex, ep esen ing he ac i a ion phase o he
ca diac cycle, was no a ec ed by HCM ionic emodelling. The ionic
emodelling modelled in he apex only also led o in e ed T wa es
in he la e al leads (Figu e 6B). In e es ingly, ionic emodelling mod-
elled only in he sep um (Figu e 6C) did no yield in e ed T wa es in
leads V4–V6, con i ming he c ucial ole o he apical hype ophy in
G oup 1A pheno ype. Indeed, G oup 1A was he only one exhibi ing
mixed apico-basal hype ophy and in e ed T wa es, while he o he
g oups only exhibi ed sep al hype ophy. This con i med ha apico-
basal he e ogenei ies due o ionic emodelling in he sep al and apical
hype ophied egion may be esponsible o he p ima y epola iza-
ion abno mali ies obse ed in G oup 1A, exp essed as in e ed T
wa es and no mal QRS mo phology.
Discussion
High pe o mance compu ing simula ions using CMR-in o med hu-
man o so- en icula models ha e p o ided s ong e idence ha
he wo pheno ypes displaying mo e abno mal ECGs in HCM may
be explained by wo dis inc mechanisms, wi h po en ial ele ance o
pa ien s’ managemen . Hype ophic ca diomyopa hy pa ien s wi h
in e ed T wa e and no mal QRS (G oup 1A) we e a ec ed by
apico-basal epola iza ion g adien s caused by ionic emodelling in
sep al and apical hype ophy egions. Hype ophic ca diomyopa hy
pa ien s wi h QRS abno mali ies in he la e al leads (G oup 3) we e
a ec ed by abno mal coupling be ween Pu kinje b anches and myo-
ca dium. These simula ions ha e also shown ha ana omical and
ECG pheno ypes in HCM caused by dis inc mechanisms iii107
Downloaded om h ps://academic.oup.com/eu opace/a icle/20/suppl_3/iii102/5202143 by Uni e sidad de Za agoza. Biblio eca Uni e si a ia use on 26 Feb ua y 2021
Figu e 4 E ec o ea ly ac i a ion si es knock-ou s on he QRS. (Top) Clinical ECG o G oup 3 pa ien (black). Changes in he ac i a ion sequence
wi h di e en blocking combina ions (le bundle b anch block, pos e io block, an e io block, and igh bundle b anch block). Ac i a ion maps in
bo h en icles a e shown along wi h he speci ic simula ed ECG ( ed), no mal simula ion (blue) in leads V1–V6. Black a ows ep esen lead o in e -
es discussed in he ex . ECG, elec oca diog am.
iii108 A. Lyon e al.
Downloaded om h ps://academic.oup.com/eu opace/a icle/20/suppl_3/iii102/5202143 by Uni e sidad de Za agoza. Biblio eca Uni e si a ia use on 26 Feb ua y 2021
mic os uc u al abno mali ies alone could no explain he dis inc
ECG abno mali ies iden i ied in each HCM pheno ypes.
Hype ophic ca diomyopa hy s uc u al
abno mali ies: hype ophy, disa ay and
ib osis
Ven icula ana omical abno mali ies alone did no yield any o he
ECG abno mali ies iden i ied in he HCM pheno ypes, ei he QRS
abno mali ies o T wa e in e sions. The e o e, his sugges s ha le
en icula hype ophy may no be a majo d i e in he iden i ica-
ion o he dis inc ECG pheno ypes in HCM and may no , on i s
own, s ongly con ibu e owa ds highe a hy hmic isk in HCM
pa ien s, as highligh ed in p e ious s udies.
15
The e ec s o o he
HCM abno mali ies, such as ib e disa ay (modelled as iso opic
elec ical p opaga ion) o changes in conduc ion speed due o ib o-
sis o hype ophy, we e simula ed. Howe e , none o hese changes
in myoca dial conduc ion ep oduced he QRS abno mali ies ob-
se ed in G oup 3. This ag ees wi h s udies o HCM amilies exhibi -
ing myo ib e disa ay in he absence o en icula hype ophy,
whe e no QRS abno mali ies we e epo ed.
16
Al e ing he ac i a-
ion sequence by modelling a ious ac i a ion knock-ou s did no
yield hese abno mal mo phologies ei he .
Abno mal Pu kinje-myoca dium
coupling
A ec ing he Pu kinje-myoca dium coupling was he only way o e-
p oduce he abno mali ies in QRS complex cha ac e is ic o G oup
3. We modelled a educed coupling in he Pu kinje b anches coupled
o he endoca dium, which led o a pa chy ac i a ion wi h a eas o
la e ac i a ion, and ansla ed o deep and wide S wa es in la e al
leads V4–V6 on he ECG. This is in ag eemen wi h a s udy epo ing
la ge Pu kinje-muscle junc ion delays in HCM pa ien s.
8
These ind-
ings may illus a e an abno mal Pu kinje conduc ion in hese pa ien s,
also in ag eemen wi h s udies epo ing 23% o HCM pa ien s wi h
abno mal His-Pu kinje conduc ion sys em.
17
Cases o HCM pa ien s
wi h abno mal His-Pu kinje sys em we e also epo ed in Re .,
18
wi h he men ion o a pa ien exhibi ing a deep S wa e mo phology in
V6. In asi e s udies such as endoca dial mapping may help in es iga e
Figu e 5 E ec o an abno mal Pu kinje-myoca dium coupling on he QRS. (A) Pa chy ac i a ion in he simula ion wi h abno mal Pu kinje-myoca -
dium coupling compa ed wi h he no mal conduc ion. (B) Ac i a ion maps o he simula ion wi h abno mal Pu kinje-myoca dium coupling show di -
use ac i a ion. (C) The simula ion wi h abno mal Pu kinje-myoca dium coupling shows egions o la e ac i a ion (blue) compa ed wi h he no mal
simula ion. (D) Modelling o al e ed as p opaga ion sys em ( ed) wi h no mal simula ion (blue) in leads V1–V6. Clinical ECG is shown in black. ECG,
elec oca diog am.
ECG pheno ypes in HCM caused by dis inc mechanisms iii109
Downloaded om h ps://academic.oup.com/eu opace/a icle/20/suppl_3/iii102/5202143 by Uni e sidad de Za agoza. Biblio eca Uni e si a ia use on 26 Feb ua y 2021
Figu e 6 E ec o ionic emodelling in he egion o hype ophy on he T wa e. (Top) Clinical ECG o G oup 1A pa ien (black). (A) Simula ed
ECG wi h ionic emodelling in he sep um and apex ( ed a ea) o leads V1–V6. (B) Simula ed ECG wi h ionic emodelling in he apex ( ed a ea) o
leads V1–V6. (C) Simula ed ECG wi h ionic emodelling in he sep um ( ed a ea) o leads V1–V6. ECG, elec oca diog am; HCM, hype ophic
ca diomyopa hy.
iii110 A. Lyon e al.
Downloaded om h ps://academic.oup.com/eu opace/a icle/20/suppl_3/iii102/5202143 by Uni e sidad de Za agoza. Biblio eca Uni e si a ia use on 26 Feb ua y 2021