Ra ionale and design o a mul icen e, andomized,
placebo-con olled ial o mi abeg on, a Be a3-
ad ene gic ecep o agonis on le en icula mass
and dias olic unc ion in pa ien s wi h s uc u al hea
disease Be a3-le en icula hype ophy (Be a3-LVH)
Anne-Ca he ine Pouleu
1
, S e an Anke
5,6
, Dulce B i o
7
, Oana B os eanu
8
, Di k Hasencle e
23
, Ba ba a
Casadei
9
, F ank Edelmann
10,11,12
, Ge asimos Filippa os
13
, Damien G uson
2
, Igna ios Ikonomidis
13
, Renaud
Lhommel
3
, Masliza Mahmod
14
, S e an Neubaue
14
, Alexand e Pe su
1
, Be nha d L. Ge be
1
, S e an Piechnik
14
,
Bu ke Pieske
10,11,12,15
, Elisabe h Pieske-K aighe
9
, Faus o Pin o
6
, Pio Ponikowski
16,17
, Michele Senni
18
,
Jean-Noël T ochu
19,20
, Nancy Van O e s ae en
1
, Rol Wach e
21,22
and Jean-Luc Balligand
4
*
1
Ca dio ascula Depa men , Cliniques Uni e si ai es Sain -Luc, Uni e si é ca holique de Lou ain, B ussels, Belgium;
2
Clinical Biology Depa men , Cliniques Uni e si ai es
Sain -Luc, Uni e si é ca holique de Lou ain, B ussels, Belgium;
3
Nuclea Medicine Depa men , Cliniques Uni e si ai es Sain -Luc, Uni e si é ca holique de Lou ain, B ussels,
Belgium;
4
Depa men o Medicine, Pole o Pha macology and The apeu ics (FATH), Ins i u de Reche che Expé imen ale e Clinique (IREC), Cliniques Uni e si ai es Sain -Luc,
Uni e si é ca holique de Lou ain, B1.53.09, 52 a enue Mounie , 1200, B ussels, Belgium;
5
Inno a i e Clinical T ials, Depa men o Ca diology and Pneumology, Uni e si y
Medical Cen e Gö ingen (UMG), Gö ingen, Ge many;
6
Di ision o Ca diology and Me abolism—Hea Failu e, Cachexia and Sa copenia, Depa men o Ca diology, Be lin
B andenbu g Cen e o Regene a i e The apies, Cha i é Uni e si y o Medicine, Be lin, Ge many;
7
Depa men o Ca diology, CHLN, CCUL (Ca dio ascula Cen e), AIDFM,
Hospi al de San a Ma ia, Uni e sidade de Lisboa, Lisbon, Po ugal;
8
Clinical T ial Cen e Leipzig—ZKS, Facul y o Medicine, Leipzig Uni e si y, Leipzig, Ge many;
9
Di ision o
Ca dio ascula Medicine, Radcli e Depa men o Medicine, Uni e si y o Ox o d, John Radcli e Hospi al, Ox o d, UK;
10
Depa men o In e nal Medicine and Ca diology,
Cha i é—Uni e si ä smedizin Be lin—Campus Vi chow Klinikum, Be lin, Ge many;
11
Ge man Cen e o Ca dio ascula Resea ch (DZHK), Pa ne Si e, Be lin, Ge many;
12
Be lin Ins i u e o Heal h (BIH), Be lin, Ge many;
13
Na ional and Kapodis ian Uni e si y o A hens, School o Medicine and Depa men o Ca diology, Hea Failu e Uni ,
A hens Uni e si y Hospi al A ikon, A hens, G eece;
14
Ca dio ascula Imaging Co e Labo a o y, Ox o d Cen e o Clinical Magne ic Resonance Resea ch (OCMR), 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;
15
Depa men o In e nal Medicine and Ca diology, Ge man Hea Ins i u e,
Be lin, Ge many;
16
Depa men o Hea Diseases, W ocław Medical Uni e si y, W ocław, Poland;
17
Ca diology Depa men , Mili a y Hospi al, W ocław, Poland;
18
Depa men Ca dio ascula Medicine, Ca diology Di ision, Papa Gio anni XXIII Hospi al, Be gamo, I aly;
19
Ins i u du ho ax, Cen e Hospi alie Uni e si ai e de Nan es,
Nan es, F ance;
20
Medical School, Uni e si y o Nan es, Nan es, F ance;
21
Clinic o Ca diology and Pneumology, Uni e si y o Gö ingen Medical Cen e, Gö ingen,
Ge many;
22
DZHK (Ge man Cen e o Ca dio ascula Resea ch), Pa ne Si e Gö ingen, Gö ingen, Ge many;
23
Ins i u e o Medical In o ma ics, S a is ics & Epidemiology—
IMISE, Facul y o Medicine, Leipzig Uni e si y, Leipzig, Ge many
Abs ac
Aims P og essi e le en icula (LV) emodelling wi h ca diac myocy e hype ophy, myoca dial fib osis, and endo helial
dys unc ion plays a key ole in he onse and p og ession o hea ailu e wi h p ese ed ejec ion ac ion. The Be a3-LVH
ial will es he hypo hesis ha he β
3
ad ene gic ecep o agonis mi abeg on will imp o e LV hype ophy and dias olic
unc ion in pa ien s wi h hype ensi e s uc u al hea disease a high isk o de eloping hea ailu e wi h p ese ed ejec-
ion ac ion.
Me hods and esul s Be a3-LVH is a andomized, placebo-con olled, double-blind, wo-a med, mul icen e, Eu opean, pa -
allel g oup s udy. A o al o 296 pa ien s will be andomly assigned o ecei e ei he mi abeg on 50 mg daily o placebo o e
12 mon hs. The main inclusion c i e ion is he p esence o LV hype ophy, ha is, inc eased LV mass index (LVMi) o inc eased
wall hickening by echoca diog aphy. The co-p ima y endpoin s a e a change in LVMi by ca diac magne ic esonance imaging
and a change in LV dias olic unc ion (assessed by he E/e0 a io). Seconda y endpoin s include mi abeg on’s e ec s on ca diac
fib osis, le a ial olume index, maximal exe cise capaci y, and labo a o y ma ke s. Two subs udies will e alua e mi abeg on’s
e ec on endo helial unc ion by pulse ampli ude onome y and b own a ac i i y by posi on emission omog aphy using
17F-fluo odeoxyglucose. Mo bidi y and mo ali y as well as sa e y aspec s will also be assessed.
ORIGINAL RESEARCH ARTICLE
© 2018 The Au ho s. ESC Hea Failu e published by John Wiley & Sons L d on behal o he Eu opean Socie y o Ca diology.
ESC HEART FAILURE
ESC Hea Failu e 2018; 5: 831–842
Published online 22 June 2018 in Wiley Online Lib a y (wileyonlinelib a y.com) DOI: 10.1002/eh 2.12306
This is an open access a icle unde he e ms o he C ea i e Commons A ibu ion-NonComme cial-NoDe i s License, which pe mi s use and dis ibu ion in any medium,
p o ided he o iginal wo k is p ope ly ci ed, he use is non-comme cial and no modifica ions o adap a ions a e made.
Conclusions Be a3-LVH is he fi s la ge-scale clinical ial o e alua e he e ec s o mi abeg on on LVMi and dias olic unc-
ion in pa ien s wi h LVH. Be a3-LVH will p o ide impo an in o ma ion abou he clinical cou se o his condi ion and may
ha e significan impac on ea men s a egies and u u e ials in hese pa ien s.
Keywo ds β
3
ad ene gic ecep o ; Mi abeg on; Hype ensi e s uc u al hea disease; Hea ailu e wi h p ese ed ejec ion ac ion
Recei ed: 28 Ma ch 2018; Accep ed: 22 Ap il 2018
*Co espondence o: Jean-Luc Balligand, Depa men o Medicine, Pole o Pha macology and The apeu ics, Ins i u de Reche che Expé imen ale e Clinique, Cliniques
Uni e si ai es Sain -Luc, Uni e si é ca holique de Lou ain, B1.53.09, 52 a enue Mounie , 1200 B ussels, Belgium. Tel: +32- 2- 764 5260; Fax: +32- 2- 764 5369.
Email: jl.balligand@uclou ain.be
Lis o con ibu o s in Appendix.
In oduc ion
Hea ailu e (HF) ep esen s a majo and g owing public
heal h bu den, a ec ing 2–3% o adul s in de eloped coun-
ies.
1
I a ec s p edominan ly he elde ly, wi h o e 80% o
HF hospi aliza ions occu ing in pe sons o e 65 yea s o
age.
2
Up o hal o HF cases occu in he se ing o p ese ed
le en icula (LV) ejec ion ac ion (HFpEF), a p opo ion
ha will con inue o ise a an ala ming a e o a ound 1%
pe yea in pa because o he p og essi e ageing o he pop-
ula ion
.2–5
Besides heal hca e expendi u e, HFpEF pu s a
hea y bu den on he quali y o li e o (mos ly elde ly) pa-
ien s, wi h a loss o au onomy and he discom o o e-
pea ed hospi aliza ions. The e o e, HFpEF is a ch onic,
cos ly, debili a ing disease.
Fu he mo e, symp oma ic HF is only he su ace o he
eme ging HF epidemic. The ageing popula ion along wi h in-
c easing a es o hype ension, diabe es, and obesi y c ea es
a g owing pool o indi iduals a pa icula ly high isk o HF
de elopmen . The Ame ican College o Ca diology/Ame ican
Hea Associa ion HF s aging model emphasizes iden i ica ion
o hese asymp oma ic a isk pa ien s wi hou (S age A) o
wi h (S age B) e idence o ca diac emodelling o acili a e
p e en a i e ac ion p io o p og ession o symp oma ic HF
(S age C).
6
Thus, he de ec ion and coun e ac ion o asymp-
oma ic a isk pa ien s may be impo an o educe he inci-
dence o clinical HFpEF.
Despi e he g owing incidence o HFpEF o e he las
15 yea s, he e a e cu en ly no p o en e ec i e he apies.
Indeed, despi e encou aging esul s om some ALDOs e one
ecep o blockade in Dias olic Hea Failu e (ALDO-DHF)
7
bu
no all Ni a e’s E ec on Ac i i y Tole ance in Hea Failu e
wi h P ese ed Ejec ion F ac ion (NEAT-HFpEF)
8
Phase II i-
als, all ou come Phase III ials ha e been neu al so a
[Pe indop il in Elde ly People wi h Ch onic Hea Failu e
(PEP-CHF), Candesa an in Hea ailu e: Assessmen o Re-
duc ion in Mo ali y and mo bidi y (CHARM)-P ese ed s udy,
I besa an in Hea Failu e wi h P ese ed Ejec ion F ac ion (I-
PRESERVE) and T ea men o P ese ed Ca diac Func ion
Hea Failu e wi h an Aldos e one An agonis ial
(TOPCAT)].
9–12
The p Ese eD le en icula ejec Ion ac-
ion ch onic hea Failu e wi h i ab adine s udY (EDIFY) ial
included 179 pa ien s in New Yo k Hea Associa ion (NYHA)
Classes II and III, in sinus hy hm, wi h hea a e o ≥70 b.
p.m., and i ab adine (o placebo) was i a ed o 7.5 mg b.i.
d. No e idence o imp o emen was ound in any o he h ee
co-p ima y endpoin s (E/e0 a io, 6 min walking es and
N- e minal p o b ain na iu e ic pep ide (NT-p oBNP).
13
Mo e
ecen ly, in he P ospec i e compa ison o ARNI wi h ARB on
Managemen O hea ailU e wi h p ese ed ejec ioN ac-
Tion (PARAMOUNT) ial, he new d ug LCZ696 (combining
alsa an and sacubi il, a nep ilysin inhibi o ) was es ed
agains alsa an alone in pa ien s wi h HF and LV ejec ion
ac ion (LVEF) ≥45% (80% o whom we e in NYHA Class II);
he p ima y endpoin was a change in NT-p oBNP, a ma ke
o LV wall s ess, om baseline o 12 weeks.
14
The esul s
showed a signi ican lowe ing o NT-p oBNP a 12 weeks in
he LCZ696 g oup s. alsa an alone, bu i was no sus ained
a 36 weeks. Ano he ial, SOluble guanyla e Cyclase s imula-
oR in heA T ailu E pa ien S wi h PRESERVED EF (SOCRATES-
PRESERVED) s udy, es ed he soluble guanylyl cyclase, s imu-
la o , e icigua , agains placebo in symp oma ic (NYHA Clas-
ses II–IV) pa ien s wi h LVEF ≥45% and le a ial (LA)
enla gemen , who expe ienced a ecen acu e decompensa-
ion e en . Despi e lack o an e ec on NT-p oBNP and LA
olume, he pa ien - epo ed symp oms and unc ional limi-
a ions, assessed by he Kansas Ci y Ca diomyopa hy Ques-
ionnai e sco e, we e imp o ed wi h he highe wo doses
o e icigua compa ed wi h placebo.
15,16
No ably, bo h PARAMOUNT and SOCRATES-PRESERVED
es ed d ugs known o inc ease in acellula (including ca -
diac myocy e) cyclic guanosine monophospha e (cGMP).
Despi e mi iga ed esul s in hese wo ials, hei pa ial e i-
cacy aises he in e es o he apeu ic s a egies ac ing on
he same pa hway, albei h ough di e en pha macody-
namic mechanisms. Acco dingly, he Be a3-LVH ial will
p o ide a p oo o concep in humans o he clinical e icacy
o β
3
ad ene gic ecep o (β
3
AR) ac i a ion o a enua e/
p e en ca diac emodelling. β
3
AR is exp essed in se e al hu-
man issues, including bladde muscle, and also ca diac and
ascula issues.
17–19
β
3
AR couples o he ni ic oxide (NO)/
cGMP pa hway, esul ing in co ona y asodila a ion, and
aises cGMP in human myoca dium, wi h a esul ing e ec
on ca diac myocy es ha is an ipa he ic o classical β
1–2
AR
posi i e ino opic e ec s.
19,20
In p eclinical s udies, ac i a ion
o β
3
AR dec eases myoca dial hype ophy and ib osis in
832 A.-C. Pouleu e al.
ESC Hea Failu e 2018; 5: 831–842
DOI: 10.1002/eh 2.12306
esponse o neu oho monal o haemodynamic s esses,
wi hou comp omising LV unc ion.
21,22
As ample e idence
now poin s o he ad e se e ec s o sus ained ac i a ion o
β
1–2
AR, leading o ecep o desensi iza ion/in e naliza ion,
loss o con ac ile/ equency ese e, ad e se emodelling,
calcium o e load, and myocy e loss, we easoned ha ac i a-
ion o he unc ionally an ipa he ic β
3
AR would p o ec
agains such dele e ious e ec s o ch onic ad ene gic s imu-
la ion. The ial will es he e ec o mi abeg on, a β
3
AR-se-
lec i e agonis ha was de eloped and ma ke ed o clinical
use in o e ac i e bladde disease, on LV mass and dias olic
unc ion.
23
The e o e, his ial will examine he ‘d ug
epu posing’o mi abeg on o HFpEF, a highly p e alen dis-
ease a ec ing mos ly elde ly pa ien s.
Me hods
S udy objec i es
The p ima y objec i e o his ial is o de e mine whe he he
β
3
AR-speci ic agonis mi abeg on is supe io o placebo in de-
c easing LV mass and/o imp o ing dias olic unc ion in pa-
ien s wi h LV s uc u al emodelling wi h o wi hou
symp oms o HF (NYHA Class ≤II).
Besides he p ima y e ec o mi abeg on on LV hype o-
phy, i s e ec on o he indica o s o HFpEF, ha is, ca diac
ib osis, LA olume index, maximal exe cise capaci y, and lab-
o a o y ma ke s, will be analysed.
S udy design
Be a3-LVH is a wo-a med, p ospec i e, andomized, placebo-
con olled, double-blind, mul icen ic Eu opean Phase IIb
clinical ial. Pa icipa ing ial cen es will sc een all consecu-
i e ou pa ien s and inpa ien s o en y inclusion and exclu-
sion c i e ia. Pa ien s who ul il s udy en y c i e ia will be
andomized o ecei e ei he mi abeg on o placebo ( an-
domiza ion a io 1:1).
O e a 36 mon h pe iod, we plan o ec ui a o al o 296
pa ien s om 10 clinical ial si es in eigh Eu opean coun-
ies. A s udy low cha is shown in Figu e 1.
Inclusion and exclusion c i e ia
The s udy inclusion and exclusion c i e ia a e lis ed in Table 1.
The main inclusion c i e ia a e as ollows: p esence o LVH by
echoca diog aphy, ha is, inc eased LV mass index by echo-
ca diog aphy (≥95 g/m
2
o emale; ≥115 g/m
2
o male) o
end-dias olic wall hickness ≥13 mm in a leas one wall seg-
men , in he absence o gene ic hype ophic ca diomyopa-
hy and signi ican al ula disease.
2,27
I is an icipa ed ha mos o hese pa ien s will ha e
sys emic hype ension; i so, hey will be equi ed o be
on s able he apy acco ding o cu en guideline algo-
i hms (including s able medica ion o a leas 4 weeks
be o e inclusion) and a well-con olled hype ension. This
pa ien popula ion is known o be mos likely o de elop
p og essi e hype ensi e ca diac emodelling and/o
HFpEF.
28
Figu e 1 Synopsis o he Be a3-LVH ial.
A mul icen e, andomized, placebo-con olled ial o mi abeg on 833
ESC Hea Failu e 2018; 5: 831–842
DOI: 10.1002/eh 2.12306
Table 1 Inclusion and exclusion c i e ia
Inclusion c i e ia
Age be ween 18 and 90 yea s
Mo phological signs o s uc u al ca diac emodelling by echoca diog aphy, ha is, inc eased LV mass index (95 g/m
2
o highe o
emale; 115 g/m
2
o highe o male subjec s o end-dias olic wall hickness ≥13 mm in a leas one wall segmen
1
W i en in o med consen : o subjec s unable o ead and/o w i e, o al in o med consen obse ed by an independen wi ness is
accep able i he subjec has ully unde s ood o al in o ma ion gi en by he in es iga o . The wi ness should sign he consen o m on
behal o he subjec .
No e: pa ien s a e allowed o ake a β
1–2
-blocke , o he han he d ugs lis ed in he exclusion c i e ia
Exclusion c i e ia
Uncon olled hype ension wi h sys olic BP ≥160 mmHg and/o dias olic BP ≥100 mmHg (confi med a h ee consecu i e o fice
measu emen s in si ing posi ion); i so, he pa ien may be e-sc eened a e op imiza ion o an i-hype ensi e ea men .
Hype ensi e pa ien s no unde s able he apy acco ding o cu en guideline algo i hm (including s able medica ion o a leas
4 weeks be o e inclusion)
24
Documen ed ischaemic ca diac disease is as ollows:
•cu en angina pec o is,
•ischaemia on s ess es ,
•un ea ed co ona y s enosis >50%,
•his o y o AMI,
•CABG (<3 mon hs p io o sc eening), o
•PTCA less han 3 mon hs p io o sc eening.
Pa ien s wi h uncon olled ecu en pe sis en and pe manen AF acco ding o AHA/ACC/ESC guidelines
25
(wi h a HR >100 pe minu e,
RACE II
26
). I AF wi h HR >100 pe minu e, he pa ien may be e-sc eened a e ea men o a e con ol.
His o y o hospi aliza ion o o e hea ailu e wi hin las 12 mon hs
Pa ien s a e hea ansplan a ion
His o y o high-deg ee impulse conduc ion blocks (g ea e han second-deg ee AV block Type 2)
Hype ophic o dila ed ca diomyopa hy
EF <50%, ega dless o symp oms
Significan al ulopa hy (less han 1 cm
2
ao ic al e a ea o significan mi al al e insu ficiency a Dopple echoca diog aphy) and/o
p e ious al ula su ge y
Congeni al al ulopa hies
Pa ien s wi h a known his o y o QT p olonga ion (QT >450 ms) o pa ien s wi h documen ed QT p olonga ion (QT >450 ms) while
aking medicinal p oduc s known o p olong he QT in e al
NYHA Class >II
BMI ≥40 kg/m
2
Hype hy oidism/hypo hy oidism
Known o he cause (i.e. COPD) o espi a o y dys unc ion. Pa ien s unde posi i e p essu e (CPAP) ea men o sleep apnoea synd ome
may be included, p o ided hey ha e been e ficien ly con olled unde egula ea men o a leas 1 yea be o e inclusion in he s udy
Mode a e enal impai men defined as eGFR <30 mL/min
Abno mal li e unc ion es s (AST o ALT >2× uppe no mal limi o pa ien s wi h known hepa ic impai men defined as Child–Pugh
Class B o highe )
Type I diabe es, complica ed Type II diabe es (i.e. wi h documen ed co ona y mac oangiopa hy, con e (c ) exclusion c i e ion 1, o
documen ed o he ascula complica ion)
Pa ien s wi h anaemia (male: Hb <13.0 g/L; emale: Hb <12.0 g/L)
Pa ien s wi h bladde ou le obs uc ion
Pa ien s using an imusca inic choline gic d ugs o ea men o OBD
Cu en use o digi alis, bup anolol, p op anolol, and nebi olol (known o in e e e wi h β
3
AR signalling)
Pa ien s con inuously ea ed wi h sildenafil o o he PDE5 inhibi o s
Cu en use o an i ungal azole de i a i es (fluconazole, i aconazole, miconazole, posaconazole, and o iconazole) (known inhibi o s
o CYP3A4, he main me abolize o mi abeg on)
Cu en ea men wi h mi abeg on o indica ion o u u e ea men wi h mi abeg on due o o he indica ions
Con aindica ion o MRI (e.g. defib illa o , e omagne ic de ices, o se e e claus ophobia)
P egnan o nu sing women
Pa icipa ion in any o he in e en ional ial: pa ien s unable o gi e in o med consen (people unde legal gua dianship)
Women o child-bea ing po en ial wi hou highly e ec i e con acep i e measu es
Con aindica ion o mi abeg on (e.g. hype sensi i i y)
β
3
AR, β
3
ad ene gic ecep o ; AF, a ial fib illa ion; AHA/ACC/ESC, Ame ican College o Ca diology/Ame ican Hea Associa ion/Eu opean
Socie y o Ca diology; ALT, alanine ansaminase; AMI, acu e myoca dial in a c ion; AST, aspa a e ansaminase; AV, a io en icula ;
BP, blood p essu e; CABG, co ona y a e y bypass g a ; COPD, ch onic obs uc i e pulmona y disease; CPAP, con inuous posi i e ai way
p essu e; EF, ejec ion ac ion; eGFR, es ima ed glome ula fil a ion a e; Hb, haemoglobin; HR, hea a e; LV, le en icula ; MRI, mag-
ne ic esonance imaging; NYHA, New Yo k Hea Associa ion; OBD, o e ac i e bladde disease; PTCA, pe cu aneous ansluminal co ona y
angioplas y; RACE II, Ra e Con ol E ficacy in Pe manen A ial Fib illa ion.
In case o disc epancy be ween ambula o y and in o fice blood p essu e, he o fice assessmen p e ails. GFR (mL/min/
1.73 m
2
) = 175 × (S
c
)
1.154
× (age)
0.203
× (0.742 i emale) × (1.212 i A ican Ame ican), om h p://nkdep.nih.go /lab-e alua ion/
g /es ima ing.sh ml#md d-s udy-equa ion. In case o cu en ea men wi h one o he excluded d ugs, pa ien s can be e-sc eened a -
e a washou pe iod o h ee hal -li es.
834 A.-C. Pouleu e al.
ESC Hea Failu e 2018; 5: 831–842
DOI: 10.1002/eh 2.12306
Du a ion o he ial
Each pa ien will be ea ed wi h placebo/mi abeg on o e a
pe iod o 12 mon hs. Thus, ial pa icipa ion encompasses
52 weeks pe pa ien . The e will be no long- ime ollow-up.
A sa e y phone call will be pe o med 4 weeks a e a pa ien
has s opped aking s udy medica ion.
S udy medica ion
Mi abeg on will be es ed agains placebo in pa ien s ecei -
ing con en ional backg ound he apy (e.g. o hype ension),
which will be p esc ibed a he disc e ion o he ea ing phy-
sician. Bup anolol, p op anolol, o nebi olol (known o in e -
e e wi h β
3
AR signalling) a e no allowed and lis ed as
exclusion c i e ion. O he medica ions no allowed in he ial
a e lis ed in Table 1. Adminis a ion o he i s dose o he
s udy d ug is pa o he andomiza ion p ocedu e and will
be supe ised by he local in es iga o . The e is no
up- i a ion planned. The s udy d ug will be labelled as
equi ed by he ICH-GCP Guideline E6 (Eu opean Commission
2/3/2010).
29
S udy endpoin s
P ima y endpoin s
We de ine wo equally anked, p ima y endpoin s, in o de o
assess bo h s uc u al and unc ional aspec s o LV emodel-
ling (Table 2):
•Change in LV mass index (in g/m
2
,defined as LV mass
di ided by body su ace) measu ed a baseline and 6
and 12 mon hs a e andomiza ion. Ca diac magne ic
esonance (CMR) is pe o med locally acco ding o a
s anda dized p o ocol, and LV mass index will be
measu ed in he cen al CMR co e lab. Reg ession o LV
hype ophy, which is eflec ed by educ ion in he LV
mass index, is known o be associa ed wi h a ou able
clinical ou comes.
30
•Change in LV dias olic unc ion, assessed as he a io o
peak ea ly ansmi al en icula filling eloci y o ea ly
dias olic issue Dopple eloci y (E/e0) measu ed a
baseline and 6 and 12 mon hs a e andomiza ion. This
pa ame e will be assessed by echoca diog aphy, pe -
o med locally acco ding o a s anda dized p o ocol, and
will be measu ed in he cen al echo co e lab. E/e0is an
es ablished indica o o dias olic unc ion and has been
shown o eliably de ec changes in unc ional pe o -
mance.
31–35
In addi ion, change in E/e0has been shown
o be associa ed wi h he change in sel - epo ed physical
unc ioning.
36
Fu he mo e, Be a3-LVH will in es iga e se e al seconda y
endpoin s, as well as he sa e y o he s udy medica ion
(Table 2).
Key measu emen s
Ca diac magne ic esonance
All pa icipan s will unde go a CMR scan pe o med on he
same 1.5T o 3T magne ic esonance sys em in each cen e.
Gi en he a ia ion o magne ic esonance sys ems be ween
cen es, acquisi ions a e pe o med wi h sequences p e e ed
by each cen e, subjec o he quali y checks by he cen al
co e lab o assu e ha he images a e i o pu pose.
Typically, a e s anda dized planning, CMR will be ac-
qui ed wi h an elec oca diog am-ga ed, b ea h-hold, wo-
dimensional, s eady-s a e ee p ecession cine sequence as
p e iously desc ibed.
37
Long-axis iews and sho -axis cine
s ack will be used o he calcula ion o he LV mass and
unc ion.
Table 2 P ima y, seconda y, and sa e y endpoin s
P ima y endpoin s
•Change in LV mass index measu ed a baseline and 6 and 12 mon hs a e andomiza ion
•Change in dias olic unc ion, assessed as he a io o peak ea ly ansmi al en icula filling eloci y o ea ly dias olic issue Dopple eloci y
(E/e0) measu ed a baseline and 6 and 12 mon hs a e andomiza ion
Seconda y endpoin s
•CMR endpoin s (all measu ed in he cen al CMR co e lab)
○Ca diac fib osis a baseline and a 12 mon hs
○LAVI a baseline and a 12 mon hs
•Labo a o y pa ame e s a baseline and a 3, 6, and 12 mon hs
○se um bioma ke s (Galec in3, GDF15, NT-p oBNP, and hsTnT)
○me abolic pa ame e s ( as ing glucose, modified HOMA es , HbA1c, and se um lipids)
•Maximal exe cise capaci y (peak VO
2
) a baseline and 12 mon hs
Sa e y endpoin s
•Incidence, se e i y, and equency o ad e se and se ious ad e se e en s
•Mo ali y
CMR, ca diac magne ic esonance; HbA1c, glyca ed haemoglobin; HOMA, homeos a ic model assessmen ; LAVI, le a ial olume index;
LV, le en icula ; NT-p oBNP, N- e minal p o b ain na iu e ic pep ide.
A mul icen e, andomized, placebo-con olled ial o mi abeg on 835
ESC Hea Failu e 2018; 5: 831–842
DOI: 10.1002/eh 2.12306

P e-con as T1 mapping images will be acqui ed a he
basal, mid-ca i y, and apical le els. A bolus o gadolinium-
based con as agen (0.15 mmol/kg) and a 10 mL saline lush
will be adminis e ed ia a cannula in he pa ien ’s a m. Pos -
con as T1 measu emen s, a he exac same h ee sho -axis
cu s as he p e-con as T1 maps, will ollow app oxima ely 5,
15, and 30 min a e injec ion o con as , using an app op i-
a e pos -con as T1 mapping sequence.
38
La e gadolinium enhancemen images a e acqui ed using
he clinical sequences o choice by each cen e, subjec o
sui abili y checks by he cen al co e lab. Fo quan i ica ion
o ex acellula olume,
39
blood sampling o haema oc i will
be ob ained on he same day as he CMR scan.
All he exams will be analysed (in a blinded ashion) by
he CMR co e lab. Rep oducibili y will be in es iga ed and
epo ed.
Echoca diog aphy
Two-dimensional and M-mode images will be acqui ed in ac-
co dance wi h cu en Ame ican Socie y o Echoca diog aphy
guidelines.
40
Pa icipa ing si e sonog aphe s will pe o m ade-
qua e echo examina ions on he basis o he echo manual,
ollowed by submission o a ce i ica ion echo. Echo Co e Lab
will assess es echoca diog ams o app op ia eness and com-
ple eness o imaging quali y acco ding o he s udy echo
manual, including op imal image quali y. Echoca diog ams a
andomiza ion, 6 and 12 mon hs, will be ead cen ally by
he blinded academic echoca diog aphy co e lab o analysis
o he p ima y, seconda y, and explo a o y echo pa ame e s
acco ding o a p e-speci ied analysis plan. Con en ional analy-
ses including wo-dimensional Dopple and issue Dopple will
be pe o med by co e lab sonog aphe s blinded o clinical in-
o ma ion and ea men assignmen using an o line
endo -independen pla o m (Tom ec, Munich, Ge many).
Ca diopulmona y exe cise es ing
A e e ence labo a o y will ac as he blinded co e lab o all
aspec s ela ed o ca diopulmona y exe cise es ing. A inal
s anda d ope a ing p ocedu e o ca diopulmona y exe cise
es ing is issued by he e e ence labo a o y. Ae obic capaci y
(peak VO
2
) es ing will be pe o med on bicycle acco ding o
a amp es p o ocol (10 W/min) a e an ini ial wo k a e a
20 W. All es s will be symp om limi ed, wi h s ong encou -
agemen o achie e a espi a o y exchange a io ha is
>1.10. C i e ia o discon inua ion o he exe cise es a e
de ined as ecommended by he Eu opean Socie y o Ca diol-
ogy.
41
A s anda d 12-lead elec oca diog am will be moni-
o ed con inuously o hea a e, ST-segmen changes, and
a hy hmias. Blood p essu e will be eco ded a es and hen
e e y 2 min. Ven ila o y exchange (VE), oxygen up ake (VO
2
),
and o he ca diopulmona y a iables such as VE/VCO
2
slope
will be acqui ed by a e aging b ea h-by-b ea h measu e-
men s o e 10 s in e als. Peak hea a e and wo kload will
be eco ded immedia ely upon he end o exe cise. Peak
VO
2
is de ined as he maximum alue o he las h ee 10 s
a e ages du ing exe cise, and anae obic h eshold will be de-
ec ed using he V-slope me hod.
42
Ch ono opic ese e will
be also eco ded. VO
2
assessmen s will be pe o med a in-
clusion and Mon h 12.
Biome ic aspec s (see he Suppo ing
in o ma ion o ex ended desc ip ion)
Randomiza ion
Randomiza ion o pa ien s be ween ac i e d ug and placebo
is pe o med cen ally ia a secu e web-based ool using a
modi ied minimiza ion p ocedu e wi h s ochas ic componen
acco ding o Pocock in a 1:1 p opo ion.
43
S a is ical desc ip ion o he ial hypo hesis
This ial aims o demons a e ha mi abeg on as add-on o
s anda d ea men compa ed wi h s anda d ea men alone
imp o es a leas one o he wo p ima y endpoin s o e
12 mon hs.
The Hochbe g me hod will be used o adjus o endpoin
mul iplici y.
44
I bo h P- alues a e below 0.05, we will claim
e icacy in bo h p ima y endpoin s; i o he wise he smalles
P- alue is below 0.025, we will claim e icacy in he espec i e
p ima y endpoin . This p ocedu e con ols he amily-wise
e o a e in he s ong sense a a wo-sided signi icance
le el o 5%.
Planned me hods o analysis
The ull analysis se (also called modi ied in en ion- o- ea
popula ion) will include all andomized pa ien s wi h alid in-
o med consen and a leas one alid measu emen o he
p ima y endpoin s (baseline and 6 o 12 mon hs). A pe -
p o ocol se will also analyse all pa ien s belonging o he in-
en ion o ea wi hou majo iola ions o he s udy
p o ocol.
Fo p ima y and seconda y endpoin s, mean changes om
baseline mean will be analysed using a epea ed measu emen
linea mixed model wi hou in e cep con aining he ixed, ca -
ego ical e ec s o isi (baseline and 6 and 12 mon hs), ea -
men (ac i e d ug/placebo), ea men by isi in e ac ion,
a ial ib illa ion (yes/no), diabe es melli us (yes/no), and a
pa ien -speci ic, isi andom e ec ( h ee-dimensional no -
mal wi h a gene al uns uc u ed a iance–co a iance ma ix).
Sensi i i y analyses (speci ied in he s a is ical analysis
plan) will include (i) he ea lie model es ic ed o he pe
p o ocol popula ion and (ii) analysis o co a iance wi h base-
line alues as co a ia es and andomiza ion g oup as ac o in
all andomized pa ien s wi h baseline and 12 mon hs mea-
su emen s and wi h impu a ion o missing alues by las in-
o ma ion ca ied o wa d.
Addi ional baseline sou ces o a iabili y will be explo ed
du ing he blinded e iew o he da a, o example, age,
836 A.-C. Pouleu e al.
ESC Hea Failu e 2018; 5: 831–842
DOI: 10.1002/eh 2.12306
gende , and NYHA class, and included in explo a i e mul i a -
ia e analyses as app op ia e. Explo a o y subg oup analyses
will include use o a be a-blocke in he s anda d ea men
(yes/no); his is o es he hypo hesis ha di e en ial egula-
ion o he exp ession and coupling o he β
3
ecep o s may
occu unde β
1
AR blockade.
45
All CMR and echo seconda y endpoin s as well as peak VO
2
will be analysed along he same lines as he p ima y end-
poin s. All measu emen s will be pe o med cen ally in cen-
al co e labs (ca diac magne ic esonance imaging and
echoca diog aphy), and ep oducibili y will be in es iga ed
and epo ed. Ad e se and se ious ad e se e en s will be
compa ed by χ
2
es s. Odds a ios wi h 95% con idence in e -
als will be p o ided. All analyses will be p e-speci ied in a de-
ailed s a is ical analysis plan, which will be inalized be o e
unblinding he da a.
Sample size
We in es iga e wo equally anked, p ima y endpoin s. We
conse a i ely plan sample sizes o a signi icance le el o
2.5%. We base ou sample size calcula ion on he pa ame e
assessing dias olic unc ion, E/e0, because eliable and consis-
en planning da a o his pa ame e a e a ailable in he li -
e a u e.
7,36,46
Typically, E/e0dec eases du ing ollow-up in
ea ed pa ien s, while i inc eases in con ol pa ien s, leading
o mean di e ences o he baseline- o- ollow-up changes o
up o 2 be ween con ol and ea men g oup, wi h a ypical
baseline mean o abou 12.
In ou ial, we aim o de ec a di e ence o 1.2
be ween ac i e d ug and placebo g oup. This di e ence
oughly co esponds o 5 poin s on he SF-36 physical unc-
ion scale,
36
hus indica ing a mode a e bu pa ien -
ele an di e ence. Based on he aw da a o he ALDO-
DHF
7
and Exe cise aining in Dias olic Hea Failu e (Ex-
DHF)-Pilo
36
ials men ioned ea lie , which we e a ailable
o addi ional analysis, we assume a s anda d de ia ion o
3. This is in line wi h he sample size assump ions o he
Dene a Ion o he enAl Sympa he ic ne es in hea T ail-
u e wi h nO mal L Ejec ion ac ion (DIASTOLE) ial.
47
Wi h hese assump ions, a o al o 272 pa ien s ha e o
be analysed o achie e a powe o 85% a a signi icance
le el o 2.5% using a wo-sided - es (NQue y Ad iso ®
7.0).
Because he e a e no da a on LVMI in ou speci ic a ge
popula ion, we canno ully speci y a planning scena io.
Howe e , wi h 272 pa ien s, an e ec size in he magni ude
o 0.4 is de ec able wi h a powe o a leas 85% a a
signi icance le el o 2.5%. In p e ious ials such as ALDO-
DHF
7
and ex-DFH pilo ,
36
he d op-ou a e was low
(ALDO-DHF 5% in 12 mon hs, ex-DHF-Pilo 3% in 6 mon hs,
Kosmala e al. 1% in 6 mon hs
46
). Thus, we expec a d op-
ou a e no exceeding 8%. Taking his in o accoun , 296 pa-
ien s will be andomized.
The Be a3-LVH is a Phase IIb ial and in es iga es end-
poin s ela ed o ca diac emodelling. The ial does no ad-
d ess ha d clinical endpoin s and is no designed no
powe ed o de ec di e ences in long- e m clinical ou come.
Howe e , clinical e en s a e collec ed (such as dea h, ca dio-
ascula dea h, HF hospi aliza ions, and new-onse hea
ailu e).
Subs udies
Two subs udies will also assess he e ec o mi abeg on on
endo helial unc ion by pulse ampli ude onome y,
coupled o measu emen s o ni osyla ed haemoglobin and
b own a ac i i y by posi on emission omog aphy–
compu ed omog aphy using 17F- luo odeoxyglucose. Fo
each p ojec , a sepa a e p o ocol is p o ided (Suppo ing
In o ma ion).
E hics
The in es iga ion con o ms wi h he p inciples ou lined in he
Decla a ion o Helsinki (B Med J 1964; ii: 177). Eu opean
(Volun a y Ha moniza ion P ocedu e), na ional, and locally
appoin ed e hics commi ees ha e app o ed he esea ch
p o ocol, and in o med consen will ha e been ob ained om
all he s udy subjec s.
S udy o ganiza ion
The p incipal in es iga o , he s udy coo dina o , and he Clin-
ical T ial Cen e Leipzig a e esponsible o all aspec s o he
s udy p o ocol and amendmen s. The S ee ing Commi ee
gua an ees scien i ic o e sigh and consul ing in all s udy-
ela ed aspec s. A Da a Sa e y and Moni o ing Boa d
ope a es independen ly o he o he s udy commi ees and
o he sponso . The Da a Sa e y and Moni o ing Boa d will e-
iew he p og ess o he ial and, unde blinded condi ions,
con ol he sa e y o he pa ien s en olled in Be a3-LVH.
Names and a ilia ions o all pa icipan s in ol ed in Be a3-
LVH a e lis ed in he Appendix.
Discussion
A majo con ibu o o HFpEF is myoca dial emodelling, o
example, hype ophy and ib osis, as well as cellula
unc ional/s uc u al modi ica ions leading o impai men in
unc ional p ope ies (including elaxa ion) and LV dis ensibil-
i y. Un o una ely, despi e he g owing incidence o HFpEF
o e he las 15 yea s, he e a e cu en ly no e idence-based
ea men s a egies ha will change i s e olu ion. This pu s
A mul icen e, andomized, placebo-con olled ial o mi abeg on 837
ESC Hea Failu e 2018; 5: 831–842
DOI: 10.1002/eh 2.12306
mo e emphasis on new s a egies and a ge s ha may p e-
en he p og ession o emodelling owa ds he de elopmen
o LV dys unc ion and symp oma ic HFpEF. This ial is de-
signed o assess he clinical e icacy o a no el he apeu ic
concep : β
3
AR ac i a ion o a enua e/p e en ca diac
emodelling.
Ra ionale o a ge ing β
3
ad ene gic ecep o
The unde lying mechanis ic concep is buil on he p eclini-
cal demons a ion o he coupling o β
3
AR o he NO
syn hase/cGMP pa hway, wi h an expec ed p o ec ion om
myoca dial emodelling
19,20
(see also Figu e 2). The hypo h-
esis was es ed in a ansgenic mouse model wi h ca diac
myocy e-speci ic exp ession o he human β
3
AR; hese mice
(and hei li e ma e con ols) we e submi ed o a numbe
o in e en ions all leading o myoca dial s ess (i.e. mini-
pump o i.p. in usions o isop o e enol o angio ensin II,
ansao ic cons ic ion). The esul s uni o mly showed p o-
ec ion o he ansgenic mice om he de elopmen o
pa hological emodelling con a y o he wild- ype con-
ols.
21,22
Impo an ly, his was no a he expense o LV
unc ion, which emained no mal. The β
3
AR may hen be
an a ac i e a ge o p e en ad e se emodelling in he
ace o ch onic ad ene gic s imula ion, all he mo e because
i is dis inc i ely esis an o homologous desensi iza ion
( oden and human β
3
ARs lack consensus sequences o
phospho yla ion by βARK o PKA) and e ains coupling o
downs eam signalling in he pa hological hea , as demon-
s a ed in human diseased myoca dium ex i o.
17
Mo eo e ,
con a y o β
1–2
AR, i s exp ession inc eases in he diseased
myoca dium. On he basis o obse a ions in ansgenic
mice, one can assume ha his β
3
AR up egula ion is a p o-
ec i e mechanism in he ace o myoca dial s ess. How-
e e , as he β
3
AR is ypically ac i a ed by highe
ca echolamine concen a ions ( han β
1/2
AR), i is possible
ha his p o ec i e pa hway is no maximally ec ui ed e en
in ci cums ances o pa hophysiological ad ene gic ac i a ion.
This would lea e a he apeu ic ma gin o addi ional ac i a-
ion by a po en and speci ic β
3
AR agonis , such as
mi abeg on.
Figu e 2 Ta ge s o he he apeu ic e ec o mi abeg on. As β
3
ad ene gic ecep o agonis , mi abeg on is expec ed o ac i a e β
3
ad ene gic ecep-
o s in adipocy es (le ), esul ing in inc eased adipocy e ‘b owning’, ene gy expendi u e, and pe iphe al insulin sensi i i y; in endo helial cells o he
ascula u e (cen e; including co ona y esis ance a e ies), he eby inc easing endo helium-dependen asodila a ion, myoca dial pe usion, and pa a-
c ine ni ic oxide-media ed signalling; and in ca diac myocy es ( igh ), esul ing in an ioxidan and cyclic guanosine monophospha e-media ed p o ec-
i e e ec s agains emodelling and imp o ed elaxa ion. Al oge he , hese e ec s a e expec ed o p e en myoca dial ischaemia and imp o e dias olic
unc ion.
838 A.-C. Pouleu e al.
ESC Hea Failu e 2018; 5: 831–842
DOI: 10.1002/eh 2.12306
Based on olde e idence,
18
he e we e conce ns ha he
he apeu ic po en ial o β
3
AR agonis s would be limi ed by
ca diodep essi e e ec s, ollowing he demons a ion ha ,
ex i o, ca diac β
3
AR s imula ion was associa ed wi h nega-
i e ino opy in human en icula samples. Howe e , his
was only obse ed a high concen a ions o agonis s ha
exceed clinically ele an plasma le els. Tha adminis a ion
o BRL37344, a β
3
AR agonis , in la ge animal models o HF
does no ansla e o dec eased ca diac pe o mance
in i o,
48
mi iga es hese conce ns. In addi ion, adminis a-
ion o mi abeg on in a small g oup o pa ien s wi h se e e
HF EF (ejec ion ac ion <40%) esul ed in imp o emen o
LV unc ion in he BE a 3 Agonis s T ea men in HF (BEAT-
HF) ial (see he succeeding ex s).
49
Impo an ly, β
3
AR also media e an ioxidan e ec s
22
ha ,
unlike p e ious he apeu ic app oaches wi h guanylyl cyclase
s imula o s ( e icigua ) o PDE5 inhibi o s (sildena il), would
p o ec he NO/cGMP signalling om oxida i e deg ada ion
and p ese e i s e icacy in emodelling myoca dium wi h
p e ailing oxidan s ess. This an ioxidan e ec also con ib-
u es o dec ease pa ac ine p o- ib o ic signalling.
22
This an i-
ib o ic e ec is likely o p e en u he deg ada ion o LV
compliance leading o HFpEF, as ib osis is a pa hogenic com-
ponen o dias olic dys unc ion.
50
In addi ion, β
3
AR was also
shown o a enua e he oxida i e inac i a ion o he Na-K-
ATPase pump in animal models, he eby educing Na o e -
load in he ailing hea ; his may also con ibu e o co ec
dias olic dys unc ion.
51
These e ec s add up wi h hose on
co ona y NO elease and asodila a ion media ed by β
3
AR ac-
i a ion on co ona y mic o ascula endo helial cells, which
would con ibu e pa ac ine e ec s on LV elaxa ion and in-
c ease co ona y pe usion, he eby a oiding u he ischae-
mic damage.
No ably, ac i a ing β
3
AR may p o ide a mo e egula ed
and a ge ed ac i a ion o cGMP downs eam pa hways
han s imula o s o ac i a o s o he soluble guanylyl cy-
clase, which need ca e ul dose inding o a oid de imen al
sys emic hypo ension. Mi abeg on was also well ole a ed
in combina ion wi h β
1–2
AR-blocke s in he BEAT-HF ial,
an associa ion also allowed in he p esen Be a3-LVH ial;
such combina ion would no only a oid o - a ge agonism
a β
1–2
AR bu may e en ein o ce he e ec s on down-
s eam NO syn hase/cGMP signalling.
45
In he Be a3-LVH
ial, only unspeci ic β
1–2–3
AR-blocke s, such as p op anolol
o bup anolol, a e no allowed, as well as nebi olol which,
aside om speci ic β
1
AR blockade, is also endowed wi h
β
3
AR agonis ac i i y. No e ha commonly used β
1–2
AR-
blocke s p e iously es ed in RCT o HF (e.g. bisop olol,
me op olol, and ca edilol) exhibi a leas 100- old less a -
ini y o β
3
AR and so would lea e i una ec ed.
52
Finally,
one can expec indi ec ca dio ascula bene i s om ac i a-
ion o β
3
AR in ex aca diac a ge s, such as beige/b own
a , wi h ensuing imp o emen s in pe iphe al me abolism
and insulin sensi i i y, jus i ying ou subs udies on
endo helial unc ion, me abolic pa ame e s, and b own
a ac i a ion by luo odeoxyglucose–posi on emission
omog aphy.
Sa e y and ole abili y o mi abeg on in he a ge
popula ion
Clinical da a om he mi abeg on clinical s udies in o e ac-
i e bladde disease did no aise majo conce ns in e ms
o sa e y and ole abili y (see he Suppo ing In o ma ion
o mo e de ails). In heal hy olun ee s, mi abeg on causes
a dose-dependen inc ease in hea a e (3–6 h pos -dosing)
and in sys olic blood p essu e (24 h a e age), which has been
a ibu ed o ba o e lex ac i a ion seconda y o sho - e m
hypo ensi e e ec s. In he clinical ial popula ions, in agg e-
ga e, his ansla ed o an inc ease o app oxima ely 1 b.p.m.
in hea a e and an inc ease in sys olic blood p essu e
o <1 mmHg, which was no associa ed wi h inc eased ca -
dio ascula complica ions (Eu opean Medicines Agency
mi abeg on EPAR epo EMA/706651/2012), a leas up o
8–12 weeks in he ini ial Phase II ials. Al hough sa e y and
ole abili y s udies wi h 1 yea ea men du a ion con i med
his, pos -ma ke ing su ey iden i ied an inc eased isk o
ca dio ascula complica ions in pa ien s wi h uncon olled
hype ension (an exclusion c i e ion in he p esen ial).
23
The- i s -in-man andomized ial o a β
3
AR agonis in
ch onic HF (BEAT-HF) was ecen ly conduc ed in 70 pa ien s
wi h NYHA Classes II and III HF and LVEF <40% a sc eening
echoca diog aphy.
49
Pa ien s ecei ed mi abeg on o placebo
o 6 mon hs as add-on o op imized s anda d he apy. The
p ima y endpoin o an inc ease in LVEF a e 6 mon hs as
measu ed by compu ed omog aphy was no eached. Explo -
a o y analysis indica ed ha β
3
AR s imula ion by mi abeg on
inc eased LVEF in pa ien s wi h se e e HF. In ha s udy, ea -
men wi h mi abeg on appea ed sa e and did no cause p o-
longa ion o he QT in e al in he en i e coho o in he
subg oup wi h LVEF <40%. The explo a o y da a indica e ha
mi abeg on inc eases con ac ili y in he s a e o mo e di-
la ed le en icles, a he han induc ion o emodelling
wi h educed dias olic dimensions. Howe e , LV mass was
no epo ed.
Conclusions
Hea ailu e wi h p ese ed ejec ion ac ion is a common,
disabling, and cos ly disease. Howe e , no es ablished
he apeu ic s a egies exis . Be a3-LVH is he i s clinical ial
o assess he e ec o mi abeg on, a β
3
ad ene gic ecep o
agonis on LV mass and dias olic unc ion in pa ien s wi h
s uc u al hea disease in o de o p e en p og ession o
A mul icen e, andomized, placebo-con olled ial o mi abeg on 839
ESC Hea Failu e 2018; 5: 831–842
DOI: 10.1002/eh 2.12306