propranolol_versus_corticosteroids_in_the.27

PEDIATRIC/CRANIOFACIAL

Propranolol versus Corticosteroids in theTreatment of Infantile Hemangioma:A Systematic Review and Meta-Analysis

Ali Izadpanah, M.D., C.M.,M.Sc.

Arash Izadpanah, M.D.,C.M., B.Sc.

Jonathan Kanevsky, B.Sc.Eric Belzile, M.Sc.

Karl Schwarz, M.D., M.Sc.

Winnipeg, Manitoba, and Montreal,Quebec, Canada

Background: Infantile hemangiomas are benign vascular neoplasms that cancause numerous functional or cosmetic problems. The authors reviewed thepathogenesis of hemangioma and compared the efficacy and complicationsrelated to therapy with propranolol versus corticosteroids.Methods: A comprehensive review of the literature was conducted from 1965to March of 2012 using MEDLINE, PubMed, Ovid, Cochrane Review database,and Google Scholar. All articles were reviewed for reports of clinical cases,reported side effects, doses, duration of treatment, number of patients, andresponse rate to treatment.Results: A total of 1162 studies were identified. Of those, only 56 articles metinclusion criteria after review by two independent reviewers (A.I. and J.K.). Forthe meta-analysis, 16 studies comprising 2629 patients and 25 studies comprising795 patients were included. Less than 90 percent of patients treated with cor-ticosteroids responded to therapy, compared with 99 percent of patients treatedwith propranolol after 12 months of follow-up. Meta-analysis demonstrated thecorticosteroid studies to have a pooled response rate of 69 percent versus thepropranolol response rate of 97 percent (p � 0.001).Conclusions: Propranolol is a relatively recent therapy of hemangiomas withfewer side effects, a different mechanism of action, and greater efficacy thancurrent first-line corticosteroid therapy. Many of these studies do not have thesame patient population or duration/regimen of treatment for hemangiomas;however, based on available data in the literature, it appears that propranololcould be an emerging and effective treatment for infantile hemangiomas. Fur-ther randomized controlled trials are recommended. (Plast. Reconstr. Surg.131: 601, 2013.)CLINICAL QUESTION/LEVEL OF EVIDENCE: Therapeutic, III.

Hemangiomas are the most common truebenign vascular neoplasm of infancy, withan incidence of 1.0 to 2.6 percent in Cau-

casian infants.1 They are frequently encounteredbenign vascular neoplasms that can cause numer-ous functional or cosmetic deformities.

Early intervention is indicated when the lesioncauses visual field disruption, respiratory obstruc-

tion, congestive heart failure, severe hemorrhage,or serious disfigurement.1 Corticosteroid therapyhas been considered as the first-line treatment;however, because of recent successful reports ofpropranolol, a beta-blocker, the current standardfirst-line treatment of these lesions is on the vergeof change.2 Thus, the aim of this article is to sys-tematically review the existing published data re-garding the treatment of infantile hemangiomascomparing propranolol and corticosteroids and tosummarize and perform meta-analysis on the cur-rent existing literature.

From the Division of Plastic and Reconstructive Surgery,University of Manitoba; the Division of Plastic and Recon-structive Surgery, McGill University Health Center; and theDepartment of Clinical Epidemiology and Community Stud-ies, Saint Mary’s Hospital Center.Received for publication May 9, 2012; accepted September21, 2012.Dr. Arash Izadpanah and Mr. Kanevsky contributed equallyto this work.Copyright ©2013 by the American Society of Plastic Surgeons

DOI: 10.1097/PRS.0b013e31827c6fab

Disclosure: The authors have no financial interestto declare in relation to the content of this article. Noexternal funding was received.

www.PRSJournal.com 601

METHODSA systematic literature review from 1946 to March

of 2012 using the PubMed database of the NationalCenter for Biotechnology Information, National Li-brary of Medicine (Bethesda, Md.), MEDLINE,Google Scholar, and Cochrane database was com-pleted for all relevant published studies describingthe use of propranolol or corticosteroids in the treat-ment of infantile hemangiomas. Key words usedwere “hemangioma,” “corticosteroid,” “proprano-lol,” “beta-blockers,” “vascular malformations,” “cav-ernous hemangioma,” “capillary hemangioma,” and“cherry hemangioma.” All articles were reviewed forreports of clinical cases, side effects, used doses, du-ration of treatment, number of patients, and re-sponse rate to treatment. Inclusion criteria includedstudies with a clear indication of effect of propran-olol or corticosteroids in the treatment of heman-giomas in pediatric population. Studies without aclear report of outcomes or those reporting the useof steroids and propranolol for treatment of non-cutaneous hemangiomas (visceral tissues, i.e., liver)were excluded.

Data analysis was performed using R software2.13.0, package META (function metaprop andforest). Meta-analysis was carried out first by usingthe Q statistic to test between-study homogeneity:homogeneity was rejected when the Q statistic pvalue was less than 0.10. Depending on whetherhomogeneity was accepted or rejected, we usedthe fixed or the random effect model to computethe combined proportion of “success” and the 95percent confidence interval. The I2 measure was

used to show the proportion of inconsistency be-tween the studies that could not be explained bychance alone. A value of p � 0.05 was consideredto be statistically significant.

Forest plot was graphed and contains individ-ual studies representing the horizontal solid linewith their confidence intervals, and the surface ofthe gray square proportional to the weight of thestudy in the pooled estimate. Finally, meta-regres-sion was performed to compare the two groups.

RESULTSOne thousand one hundred sixty-two studies

were identified (Fig. 1). Of those, only 56 articlesmet our inclusion criteria after review by two in-dependent reviewers (A.I. and J.K.) (Tables 1 and2).3–62 Only 40 studies met final inclusion criteriafor meta-analysis (Fig. 1).

Efficacy and Complication Profile ofCorticosteroids

Nineteen studies comprising 2697 patients re-ported use of oral or locally administered corti-costeroids (Table 1). Of these, 2451 patients (90.8percent) were treated for lesions located in thehead and neck, including airway hemangiomas.The remaining 246 patients (9.2 percent) weretreated for lesions located on the trunk orextremities.3 The most common dosage of corti-costeroid was 2 to 3 mg/kg/day.3–5 Eight studiesinvestigated the effect of intralesional injectionof corticosteroid with a response rate of 66.4percent.6–10 Duration of therapy was variable de-

Fig. 1. Flow chart of selection of articles for inclusion in systematic review.

Plastic and Reconstructive Surgery • March 2013

602

Tab

le1

.Li

sto

f19

Stu

die

sU

sin

gO

ralS

tero

ids

for

Trea

tmen

to

fHem

ang

iom

as

Stud

yN

o.of

Cas

esH

eman

giom

aL

ocat

ion

Pre

viou

sT

hera

pyD

ose

(mg/

kg/d

ay)

Dur

atio

nof

The

rapy

Res

pons

eR

ate

(%)

Mea

sure

ofR

egre

ssio

nL

evel

ofE

vide

nce

Com

plic

atio

ns,N

o.of

Patie

nts

(%)

Ble

ian

dC

hia

nes

e,19

9923

3022

faci

al;

5su

bglo

ttic

;3

extr

emit

ies

Non

e2–

4N

R25

Vis

ual

appe

aran

ceof

lesi

on(s

ign

ific

ant

decr

ease

inpi

gmen

tati

onan

dsi

ze)

III

Evi

den

cede

laye

dm

ilest

ones

,3

(10)

;en

docr

ine

chan

ges,

4(1

3);

moo

nfa

cies

,7

(23)

Sada

nan

dW

olac

h,

1996

560

20or

bita

l;6

subg

lott

ic;

34fa

cial

Non

e3–

52–

3m

o93

Vis

ual

appe

aran

ceof

lesi

on;

exce

llen

t(i

nvo

luti

onw

ith

in1–

2w

kw

ith

no

regr

owth

);go

od(s

low

erin

volu

tion

wit

h,

mul

tipl

eco

urse

sof

trea

tmen

t);

faile

d(n

ore

spon

seto

ther

apy)

III

Moo

nfa

cies

,32

(53)

;gr

owth

reta

rdat

ion

,1

(1.6

);os

teop

oros

is,

1(1

.6)

*Boo

net

al.,

1999

462

50ce

rvic

ofac

ial,

5su

bglo

ttic

;3

thor

acic

;1

hep

atic

;1

peri

nea

l;2

uppe

rex

trem

ity

Non

e2–

32

wk

NR

Que

stio

nn

aire

III

Moo

nfa

cies

,44

(70)

;pe

rson

alit

ych

ange

s,18

(29)

;fu

nga

lin

fect

ion

,4

(6);

grow

thre

tard

atio

n,

22(3

5);

low

wei

ght

gain

,26

(41)

Pan

dey

etal

.,20

093

1127

1058

hea

dan

dn

eck;

69tr

unk

Non

e1–

2N

R89

Exc

elle

nt

(�75

%re

gres

sion

wit

hou

tan

ysi

gnif

ican

tsc

arri

ng)

;go

od(5

0–75

%re

gres

sion

wit

h/

wit

hou

tsc

arri

ng)

;po

or(2

5–50

%re

gres

sion

wit

h/w

ith

out

scar

rin

g);

no

resp

onse

(�25

%or

no

regr

essi

on)

III

Hyp

erte

nsi

on,

50(4

);gr

owth

reta

rdat

ion

,58

(5);

moo

nfa

cies

,58

(5)

*Arg

enta

etal

.,19

826

1L

arge

glut

eal

hem

angi

oma

Non

e1

3w

k10

0V

isua

lap

pear

ance

ofle

sion

(sig

nif

ican

tde

crea

sein

pigm

enta

tion

and

size

)

VN

R

(Con

tinue

d)

Volume 131, Number 3 • Treatment of Infantile Hemangioma

603

Tab

le1

.(C

onti

nued

)

Stud

yN

o.of

Cas

esH

eman

giom

aLo

catio

nPr

evio

usT

hera

pyD

ose

(mg/

kg/d

ay)

Dur

atio

nof

The

rapy

Res

pons

eR

ate

(%)

Mea

sure

ofR

egre

ssio

nLe

velo

fEv

iden

ceC

ompl

icat

ions

,No.

ofPa

tient

s(%

)

Ch

enet

al.,

2000

2415

5H

ead

and

nec

kN

one

0.1–

3m

l(1

0m

g/m

l�4

inje

ctio

ns)

5m

o60

Visu

alap

pear

ance

ofle

sion

(dec

reas

eof

atle

ast5

0%of

volu

me)

III

Cus

hin

goid

appe

aran

ce,

2(1

.2),

cuta

neo

usat

roph

y,5

(3.2

)C

how

dri

etal

.,19

9425

7448

hea

dan

dn

eck,

11tr

unk,

15lim

bs

Non

e0.

5–6,

l(1

0m

g/m

l�1–

7in

ject

ion

s)

3–4

mo

43E

xcel

len

t(�

75%

regr

essi

on);

good

(50–

75%

regr

essi

on);

fair

(25–

50%

);po

or(�

25%

)

III

Cus

hin

goid

appe

aran

ce,

2(2

.7)

*Iw

anak

aet

al.,

1994

265

Eye

lidan

dor

bit

Non

e2–

3m

g/kg

/day

(ora

l),

40m

gtr

iam

cino

lone

�8

mg

Ora

lst

eroi

ds,

1w

k;in

ject

ion

s,2–

3m

o

40V

isua

lap

pear

ance

ofle

sion

(dec

reas

eof

atle

ast

75%

ofvo

lum

e)

IVN

R

Kus

hn

er,

1979

2725

Peri

orbi

tal

Non

e40

mg

tria

mci

nolo

ne�

6m

gbe

tam

etha

sone

�2

inje

ctio

ns

3–4

wk

84M

arke

d,(�

80%

regr

essi

on);

mod

erat

e(2

0–80

%);

min

imal

(�20

%re

gres

sion

)

III

NR

Pras

etyo

no

and

Djo

enae

di,

2011

874

9H

ead

and

nec

kU

nsu

cces

sful

syst

emic

ster

oid

orla

ser

trea

tmen

t

1–2

ml

tria

mci

nol

one

5%�

3in

ject

ion

son

aver

age

10–1

5w

k71

Exc

elle

nt

(�75

%re

gres

sion

);go

od(5

0–75

%re

gres

sion

);fa

ir(2

5–50

%);

poor

(�25

%)

IIFa

ilure

toth

rive

,1.

7%;

inje

ctio

n-

site

ulce

rati

on,

1.4%

Kel

lyet

al.,

2010

2816

14h

ead

and

nec

k,2

uppe

rex

trem

ity

Non

e2.

53

mo

50V

isua

lap

pear

ance

ofle

sion

(dec

reas

eof

atle

ast

30%

ofvo

lum

e)

IIL

ymph

ocyt

ede

crea

se,

16(1

00),

grow

thre

tard

atio

nZ

hou

etal

.,20

1029

23N

AN

A3–

53

mo

87E

xcel

len

t(�

75%

regr

essi

on);

good

(50–

75%

regr

essi

on);

fair

(25–

50%

);po

or(�

25%

)

III

Cus

hin

goid

appe

aran

ce,

8(3

4),

poor

appe

tite

,5

(21)

Ch

anth

arat

anap

iboo

n,

2008

916

019

uppe

rex

trem

ity,

7lo

wer

extr

emit

y,13

4h

ead

and

nec

k

Non

e1–

2m

g/kg

tria

mci

nolo

ne�

5in

ject

ions

4–12

wk

90N

AII

IN

R

Ros

sler

etal

.,20

0830

3828

head

and

neck

,3

uppe

rex

trem

ity,4

trun

k,1

peri

neal

,2

visc

eral

Prev

ious

lase

rth

erap

y

24

mo

86Ex

celle

nt(�

75%

regr

essio

n);g

ood

(50–

75%

regr

essio

n);

fair

(25–

50%

);po

or(�

25%

)

III

Beh

avio

ralc

hang

e,6

(25)

,gro

wth

reta

rdat

ion,

3(8

);hy

pert

ensio

n,2

(5)

(Con

tinue

d)


604

Tab

le1

.(C

onti

nued

)

Stud

yN

o.of

Cas

esH

eman

giom

aLo

catio

nPr

evio

usT

hera

pyD

ose

(mg/

kg/d

ay)

Dur

atio

nof

The

rapy

Res

pons

eR

ate

(%)

Mea

sure

ofR

egre

ssio

nLe

velo

fEv

iden

ceC

ompl

icat

ions

,No.

ofPa

tient

s(%

)

Pope

etal

.,20

0731

10H

ead

and

nec

kN

one

2(m

g/kg

/day

)IV

puls

eof

30m

g/kg

/m

o

3m

o70

Vis

ual

anal

ogue

scal

e1

Hyp

erte

nsi

ons,

4(2

0);

abn

orm

alco

rtis

ol,

16(8

0)

10H

ead

and

nec

kN

one

IVpu

lse

of30

mg/

kg/m

o3

mo

12V

isua

lan

alog

uesc

ale

Hyp

erte

nsi

on,

4(2

0);

abn

orm

alco

rtis

ol,

16(8

0)Ja

lilet

al.,

2006

1025

NR

Non

e2

2m

o32

Vis

ual

appe

aran

ceof

lesi

on(d

ecre

ase

ofat

leas

t50

%of

volu

me)

IO

vera

ll,20

%

25N

RN

one

1–5

mg/

kgtr

iam

cino

lone

�6

mo

6m

o44

Ove

rall,

24%

Gre

ene

and

Cou

to,

2011

3267

Paro

tid

Non

e2–

39

mo

83V

isua

lap

pear

ance

ofle

sion

;re

gres

sion

,st

abili

zati

onor

no

resp

onse

III

NR

Al-S

ebei

han

dM

anou

kian

,20

0033

14Su

bglo

ttic

Non

e0.

5–2

9m

o90

Vis

ual

appe

aran

ceof

lesi

onby

bron

chos

cope

(dec

reas

eof

atle

ast

50%

ofvo

lum

e)

III

Cus

hin

goid

face

,1

(7)

Nar

cyet

al.,

1985

3421

Subg

lott

icN

one

222

days

33V

isua

lap

pear

ance

ofle

sion

bybr

onch

osco

pe(d

ecre

ase

ofat

leas

t50

%of

volu

me)

III

NR

NR

,n

otre

port

ed;

NA

,n

otap

plic

able

.


605

Tab

le2

.C

om

par

iso

no

f35

Stu

die

sU

sin

gP

rop

ran

olo

lfo

rTr

eatm

ent

ofH

eman

gio

mas

Ref

eren

ceN

o.of

Cas

esH

eman

giom

aL

ocat

ion

Pre

viou

sT

hera

pyA

vera

geD

ose

(mg/

kg/d

ay)

Dur

atio

nof

The

rapy

Res

pons

eR

ate

(%)

Mea

sure

ofR

egre

ssio

nL

evel

ofE

vide

nce

Com

plic

atio

ns,N

o.of

Patie

nts

(%)

Pric

eet

al.,

2011

3559

49h

ead

and

nec

kN

R2

7.9

mo

100

Vis

ual

appe

aran

ceof

lesi

on(s

ign

ific

ant

decr

ease

inpi

gmen

tati

onan

dsi

ze)

III

Hyp

ogly

cem

ia,

1(1

.4)

Sch

upp

etal

.,20

1136

5542

hea

dan

dn

eck;

13tr

unk

and

limbs

5co

rtis

one;

28la

ser

25.

898

Vis

ual

appe

aran

ceof

lesi

on(s

ign

ific

ant

decr

ease

inpi

gmen

tati

onan

dsi

ze)

IIC

old

extr

emit

ies,

6(1

0.9)

;fa

tigu

e,4

(7.2

);as

thm

aag

grav

atio

n,

2(3

.6)

Bag

azgo

itia

etal

.,20

1137

7160

hea

dan

dn

eck;

11tr

unk

and

limbs

4pr

edn

ison

e;4

surg

ical

debu

lkin

g

23

mo

100

Vis

ual

appe

aran

ceof

lesi

on(s

ign

ific

ant

decr

ease

inpi

gmen

tati

onan

dsi

ze)

III

Slee

pdi

stur

ban

ce,

10(1

4)

Jin

0078

NR

NR

27.

6m

o98

.7V

isua

lap

pear

ance

ofle

sion

(sig

nif

ican

tde

crea

sein

pigm

enta

tion

and

size

)

IISl

eep

dist

urba

nce

s,12

(15.

4)

Lv

etal

.,20

1238

3737

hea

dan

dn

eck

4pu

lse

dye

lase

r;1

inte

rfer

on;

1cr

yoth

erap

y;1

oral

pred

nis

one

23

mo

100

Vis

ual

appe

aran

ceof

lesi

on(s

ign

ific

ant

decr

ease

inpi

gmen

tati

onan

dsi

ze)

III

Dia

rrh

ea,

9(2

4)

Fuch

sman

etal

.,20

1139

3939

hea

dan

dn

eck

9pr

edn

ison

e2

8.5

94V

isua

lap

pear

ance

ofle

sion

(sig

nif

ican

tde

crea

sein

pigm

enta

tion

and

size

III

Slee

pdi

stur

ban

ces,

5(1

2.8)

Zah

eret

al.,

2011

4030

30h

ead

and

nec

kN

R2

2–14

mo

97.7

Vis

ual

appe

aran

ceof

lesi

on(s

ign

ific

ant

decr

ease

inpi

gmen

tati

onan

dsi

ze)

IIN

R

deG

raaf

etal

.,20

1141

2828

hea

dan

dn

eck

5pr

edn

ison

e;1

vin

cris

tin

e;1

puls

edy

ela

ser;

1su

rgic

alde

bulk

ing

2.2

4.5–

17m

o10

0V

isua

lap

pear

ance

ofle

sion

(sig

nif

ican

tde

crea

sein

pigm

enta

tion

and

size

)

IIH

ypog

lyce

mia

,2

(7);

bron

chia

lh

yper

reac

tivi

ty,

3(1

0);

sym

ptom

atic

hyp

oten

sion

,1

(3);

seiz

ure,

1(3

)Sc

hie

stl

etal

.,20

1142

2525

hea

dan

dn

eck

2pu

lse

dye

lase

r2

10.5

100

Vis

ual

anal

ogue

scal

eII

IA

sym

ptom

atic

hyp

oten

sion

,6

(24) (C

ontin

ued)


606

Tab

le2

.(C

onti

nued

)

Ref

eren

ceN

o.of

Cas

esH

eman

giom

aLo

catio

nPr

evio

usT

hera

pyA

vera

geD

ose

(mg/

kg/d

ay)

Dur

atio

nof

The

rapy

Res

pons

eR

ate

(%)

Mea

sure

ofR

egre

ssio

nLe

velo

fEv

iden

ceC

ompl

icat

ions

,No.

ofPa

tient

s(%

)

Mis

soi

etal

.,20

1143

1717

peri

ocul

arN

R2

6.8

mo

100

Vis

ual

appe

aran

ceof

lesi

on(s

ign

ific

ant

decr

ease

inpi

gmen

tati

onan

dsi

ze)

III

Asy

mpt

omat

icbr

adyc

ardi

a,1

(6)

Al

Dh

aybi

etal

.,20

1144

1818

peri

ocul

arN

R2

2m

o94

Vis

ual

anal

ogue

scal

eII

IH

ypot

ensi

on,

1(5

);di

arrh

ea,

3(1

5);

inso

mn

ia,

6,(3

0)H

ogel

ing

etal

.,20

1145

207

faci

al;

3pe

rior

bita

l;3

nas

alti

p;4

lip;

1to

rso;

1lim

b

4re

ceiv

edor

alco

rtic

oste

roid

26

mo

78B

linde

dvo

lum

em

easu

rem

ents

and

blin

ded

scor

ing

ofph

otog

raph

s

IB

ron

chio

litis

,4

(21)

;co

ldex

trem

itie

s,1

(5);

ulce

rati

on,

1(5

)H

olm

eset

al.,

2010

4631

NR

NR

33

mo

100

Blin

ded

scor

ing

ofph

otog

raph

sII

NR

Zvu

lun

ovet

al.,

2011

4742

NR

NR

2.1

3.6

mo

100

Vis

ual

anal

ogue

scal

eII

IN

R

Ch

aiet

al.,

(in

pres

s)48

2722

hea

dan

dn

eck;

3tr

unk

NR

22.

75–5

.75

mo

100

Vis

ual

appe

aran

ceof

lesi

on(s

ign

ific

ant

decr

ease

inpi

gmen

tati

onan

dsi

ze)

IIN

R

Ber

tan

det

al.,

2011

49

123

eyel

id,

2ch

eek,

2n

ose,

2fo

reh

ead,

1pa

roti

d,1

lip

NR

2.7

10.6

100

Vis

ual,

Dop

pler

ultr

asou

nd,

MR

III

ISl

eep

dist

urba

nce

,6

(50)

Tan

etal

.,20

1150

159

hea

dan

dn

eck,

6tr

unk

and

limbs

NR

1.5

3.5

100

Vis

ual

appe

aran

ceof

lesi

on(s

ign

ific

ant

decr

ease

inpi

gmen

tati

onan

dsi

ze)

IIN

R

Lea

ute-

Lab

reze

etal

.,00

1110

faci

al,

1fo

rear

m4

oral

cort

icos

tero

ids

2N

R10

0V

isua

lap

pear

ance

ofle

sion

(sig

nif

ican

tde

crea

sein

pigm

enta

tion

and

size

)

IVN

R

Th

elet

san

eet

al.,

2009

51*

1Fa

cial

,or

oph

aryn

x,la

ryn

x,PH

AC

ES

syn

drom

e

Hig

h-d

ose

ster

oid

resp

onse

2N

R10

0V

isua

lap

pear

ance

ofle

sion

(sig

nif

ican

tde

crea

sein

pigm

enta

tion

and

size

)

VN

R

Itan

ian

dFa

kih

,20

0952

*

1U

pper

eyel

idN

R2

NR

100

Vis

ual

appe

aran

ceof

lesi

on(s

ign

ific

ant

decr

ease

inpi

gmen

tati

onan

dsi

ze;

decr

ease

doc

ular

occl

usio

n)

VN

R

(Con

tinue

d)


607

Tab

le2

.(C

onti

nued

)

Ref

eren

ceN

o.of

Cas

esH

eman

giom

aLo

catio

nPr

evio

usT

hera

pyA

vera

geD

ose

(mg/

kg/d

ay)

Dur

atio

nof

The

rapy

Res

pons

eR

ate

(%)

Mea

sure

ofR

egre

ssio

nLe

velo

fEv

iden

ceC

ompl

icat

ions

,No.

ofPa

tient

s(%

)

Big

orre

etal

.,20

0914

*4

Faci

al(p

arot

idea

l,h

emif

acia

l)1

peri

nea

l

Ster

oids

,n

ore

spon

se10

(ace

buto

lol)

2m

o10

0V

isua

lap

pear

ance

ofle

sion

(sig

nif

ican

tde

crea

sein

pigm

enta

tion

and

size

;de

crea

sed

ocul

aroc

clus

ion

)

IVN

R

Den

oyel

leet

al.,

2009

15*

2Su

bglo

ttic

Vin

cris

tin

ean

dco

rtic

oste

roid

,n

ore

spon

se

NR

NR

100

En

dosc

opy

(dec

reas

edsu

bglo

ttic

sten

osis

)IV

NR

Pere

zet

al.,

2010

22*

1Fa

cial

NR

29

mo

100

Vis

ual

appe

aran

ceof

lesi

on(s

ign

ific

ant

decr

ease

inpi

gmen

tati

onan

dsi

ze)

VN

R

Buc

kmill

eret

al.,

2010

5332

22fa

cial

,10

trun

k/ex

trem

ity

Faile

dre

spon

seto

intr

ales

ion

alst

eroi

d,de

bulk

ing

wit

hca

rbon

diox

ide

lase

r,an

dvi

ncr

isti

ne

2N

R50

Phot

ogra

phic

anal

ysis

and

pare

nt

ques

tion

nai

re(1

6ex

celle

nt

resp

onde

rs,

no

follo

w-u

pn

eede

d;15

requ

ired

follo

w-u

p;1

had

no

resp

onse

)

III

Som

nol

ence

,6,

(27.

3);

gast

roes

oph

agea

lre

flux

,2

(9.1

);al

lerg

icra

sh,

1(4

.5);

resp

irat

ory

syn

cyti

alvi

rus

exac

erba

tion

,1

(4.5

)B

onif

azi

etal

.,20

1054

115

nos

e,lip

,fo

reh

ead,

oral

cavi

ty,

chee

k;6

paro

tide

al,

nip

ple

ankl

e,n

ose,

paro

tide

al,

han

dan

dfo

rear

m,

met

amer

ic

NR

2(1

%to

pica

l)N

R10

0V

isua

lap

pear

ance

ofle

sion

(sig

nif

ican

tde

crea

sein

pigm

enta

tion

and

size

)

IVN

R

Zim

mer

man

etal

.,20

1011

*

1Pa

roti

deal

Ster

oid,

no

resp

onse

24

mo

100

Vis

ual

appe

aran

ceof

lesi

on(n

earb

yco

mpl

ete

regr

essi

on)

VN

R

San

set

al.,

2009

5532

face

,fo

rear

m,

thor

ax,

nec

k,pa

roti

deal

13of

32pa

tien

tsre

ceiv

edan

dfa

iled

tore

spon

dto

ster

oids

2N

R10

0V

isua

lap

pear

ance

ofle

sion

(sig

nif

ican

tde

crea

sein

pigm

enta

tion

,si

ze,

and

har

dnes

s),

ultr

asou

nd

(th

ickn

ess)

,R

esis

tivi

tyIn

dex

III

NR

(Con

tinue

d)


608

Tab

le2

.(C

onti

nued

)

Ref

eren

ceN

o.of

Cas

esH

eman

giom

aLo

catio

nPr

evio

usT

hera

pyA

vera

geD

ose

(mg/

kg/d

ay)

Dur

atio

nof

The

rapy

Res

pons

eR

ate

(%)

Mea

sure

ofR

egre

ssio

nLe

velo

fEv

iden

ceC

ompl

icat

ions

,No.

ofPa

tient

s(%

)

Law

ley

etal

.,20

0956

*2

Eye

lid,

hem

angi

omat

osis

Ster

oids

and

prop

ran

olol

2N

R10

0V

isua

lap

pear

ance

ofle

sion

(sig

nif

ican

tde

crea

sein

pigm

enta

tion

and

size

)

IVL

eth

argy

,h

ypog

lyce

mia

,1

(50)

Mah

adev

anet

al.,

2011

57

10Su

bglo

ttic

Ster

oid,

no

resp

onse

29–

12m

o10

0E

ndo

scop

ic(d

ecre

ased

subg

lott

icst

enos

is)

III

NR

Leb

oula

nge

ret

al.,

2010

19

14Su

bglo

ttic

Ster

oid,

no

resp

onse

2–3

6m

o10

0E

ndo

scop

ic(d

ecre

ased

subg

lott

icst

enos

is)

III

Ast

hm

a,1

(7)

Mis

try

and

Tzi

fa,

2010

58*

1Su

prag

lott

icC

oncu

rren

tst

eroi

dth

erap

y

113

wk

100

Vis

ual

appe

aran

ceof

lesi

on(s

ign

ific

ant

decr

ease

inpi

gmen

tati

onan

dsi

ze)

VN

R

Cus

hin

get

al.,

2011

5949

Hea

dan

dn

eck

No

prev

ious

ster

oid

ther

apy

23–

4m

o10

0V

isua

lap

pear

ance

ofle

sion

(sig

nif

ican

tde

crea

sein

pigm

enta

tion

and

size

)

III

NR

Bre

ur*

1Pe

rior

bita

lC

oncu

rren

tst

eroi

dth

erap

y

26

mo

100

Vis

ual

appe

aran

ceof

lesi

on(s

ign

ific

ant

decr

ease

inpi

gmen

tati

onan

dsi

ze)

VH

ypog

lyce

mia

Pavl

akov

icet

al.,

2010

211

Ch

est

wal

lN

opr

evio

usst

eroi

dth

erap

y

25

mo

100

Vis

ual

appe

aran

ceof

lesi

on(s

ign

ific

ant

decr

ease

inpi

gmen

tati

onan

dsi

ze)

VH

yper

kale

mia

Hol

lan

det

al.,

2010

203

1ey

elid

,2

nas

alti

pN

opr

evio

usst

eroi

dth

erap

y

1–2

3w

k–4

mo

NR

NR

IVH

ypog

lyce

mia

NR

,n

otre

port

ed.

*Stu

dies

that

wer

ere

view

edbu

tn

otin

clud

edin

met

a-an

alys

isbe

caus

eth

eydi

dn

otm

atch

incl

usio

ncr

iter

ia.


609

pending on the time period to improvement, mostcommonly 4 to 8 weeks (Table 1).4 Sixteen studieswere included in the final meta-analysis, with anoverall efficacy of 69.1 percent (I2 � 95 percent,p � 0.0001) (Fig. 2). When intralesional studieswere omitted, the overall efficacy rate was seen tobe 71.0 percent compared with 97 percent forpropranolol (p � 0.0001) (Fig. 2).

The most common side effects reported werealtered growth (6 percent) and moon facies (5 per-cent). Other reported side effects included osteo-porosis, fungal infection, and hypertension (Table2). Side effects were encountered in 475 of 2697patients (17.6 percent), with a hemangioma resolu-tion rate of 84.5 and 66.4 percent for systemic andlocal administration, respectively (p � 0.0001).

Fig. 2. Meta-analysis comparing corticosteroids and propranolol in treatment of cutaneous infantile hemangiomas.


610

Efficacy and Complications Profile ofPropranolol

Thirty-five studies comprising 799 patientswith an average reported age of 12.4 � 6.2 monthswere identified and included initially for the effectof propranolol, of which 600 patients had lesionsaffecting the head and neck and the remaining199 had lesions on the trunk or extremities (Table2). The most commonly used dosage was 2 mg/kg/d in divided doses of three or four times daily.11

Duration of therapy varied between 4 weeks and12 months secondary to variable response rate.

All studies other than one reported a responserate of greater than 90 percent. A single studyreported a 50 percent response rate following theinitial course of propranolol and a final 95 percentresponse rate after additional courses.12 Overall,98.9 percent of patients with infantile hemangio-mas showed evidence of resolution of their lesionsfollowing treatment with beta-blockers (Table 2).Twenty-four studies finally met our inclusion cri-teria for meta-analysis. Studies with fewer than 10subjects were omitted. Meta-analysis demon-strated an overall response rate of 97.3 percent(I2 � 68 percent, p � 0.0001) (Fig. 1).

Overall, 96 of 699 patients (13.7 percent) ex-perienced side effects of propranolol therapy (Ta-ble 2). There was no statistically significant differ-ence observed between studies with fewer orgreater than 30 subjects. A comparison betweenpropranolol and corticosteroids in response ratedemonstrated statistical significance (71.0 percentversus 97.3 percent; p � 0.0001).

DISCUSSIONPropranolol is a nonselective beta-blocker

commonly used to treat hypertension, congenitalcardiac abnormalities, and arrhythmias in the pe-diatric population.13 It is an emerging therapy forhemangioma associated with greater efficacy ascompared with steroids after either systemic orlocal administration. In our systematic review, wehave demonstrated that propranolol was evenused in some refractory cases after initial cortico-steroid therapy (Tables 1 and 2), with an overallaccumulative success rate of 91.0 percent for pa-tients treated with only propranolol and an overallrate of 98.9 percent.14,15 Performing meta-analysisdemonstrated a resolution rate of 97 percent com-pared with 71 percent for propranolol versus sys-temic corticosteroids.

Propranolol is thought to alter the course ofhemangioma growth by inducing vasoconstrictionand down regulation of angiogenic factors such as

vascular endothelial growth factor and basic fibro-blast growth factor.2 Furthermore, it up-regulatesapoptosis of capillary endothelial cells. In addi-tion, propranolol has been reported to be a se-lective inhibitor of matrix metalloproteinase 9,which further supports the antiangiogenic prop-erties of propranolol.16 Liggett et al.17 describe acertain protective genetic polymorphism in G pro-tein–coupled receptor kinases, GRK5-Leu41, anendogenous form of beta-blockade present in 40percent of African Americans as a potential ex-planation for the decreased incidence of heman-giomas in the African American population.17,18

These observations and the success of propranololin treating hemangioma could suggest that beta-adrenergic pathways have a critical role in angio-genesis and progression of hemangioma. How-ever, occasional regrowth of the lesion has beenreported on cessation of therapy, which can beattributed to the effect of propranolol to maintainhemangiomas in a regressed state rather than in-ducing involution.19 After examining reportedcomplications, our review demonstrated that thesteroid therapy complication rate is more thandouble when compared with propranolol (23 per-cent versus 9.6 percent).

Boon et al.4 reported that the majority of chil-dren who experienced growth retardation second-ary to corticosteroid therapy returned to their pre-treatment growth curve for height by 24 monthsafter cessation of a 2-week steroid course. How-ever, numerous courses of steroid therapy can lastmore than 2 weeks, leading to serious long-termside effects.5

In contrast, propranolol side effects should bemonitored closely and dosage should be adjustedin the event of developing side effects such as lowmean blood pressure of less than 50 mmHg, heartrate less than 90 beats per minute, or glucose levelsless than 4 mM (72 mg/dl).20 Thus, many studiesrecommend that, before initiation of therapy, pa-tients should undergo a thorough cardiac workupincluding an electrocardiogram and monitor forany signs of bradycardia or hypoglycemia duringtreatment initiation.7,20,21

Some studies suggest that most of the improve-ment with propranolol treatment occurs in thedeep components of the lesion, whereas steroidtherapy usually results in improvement of super-ficial components.12 Thus, propranolol could bean effective adjunct to corticosteroid therapy. Un-like local corticosteroids, use of local propranololwas not shown to be as effective as systemicallyadministered propranolol.1,22


611

This study is not short of limitations. This re-view includes only published data that could havebeen affected by positive effect publication bias. Inaddition, the effects of treatment on cutaneoushemangiomas are subjective, which could also ad-versely affect the outcomes. Another major draw-back of this review is the absence of any random-ized controlled trials comparing propranolol tocorticosteroid therapy. A randomized controlledtrial with a set dosage and duration of therapy withappropriate follow-up is necessary to further val-idate the observed therapeutic benefits of pro-pranolol compared with corticosteroids.

CONCLUSIONSPropranolol is an emerging, popular therapy

for treatment of hemangiomas, with promisingresults and fewer side effects as compared withcorticosteroids. The discovery of propranolol’sutility in treating hemangiomas has led to a betterunderstanding and treatment of the disease. Thepresent systematic review suggests that proprano-lol therapy could be potentially superior to thecurrent standard first-line therapy with corticoste-roids. However, further randomized controlledstudies are needed to evaluate and compare thelong-term effects of beta-blocker therapy for in-fantile hemangiomas.

Ali Izadpanah, M.D., C.M., M.Sc.Plastic and Reconstructive Surgery

McGill University Health Centre760 Upper Lansdowne Avenue

Westmount, Quebec H3Y 1J8, [email protected]

REFERENCES1. Burns AJ, Navarro JA, Cooner RD. Classification of vascular

anomalies and the comprehensive treatment of hemangio-mas. Plast Reconstr Surg. 2009;124 (1 Suppl):69e–81e.

2. Leaute-Labreze C, Taıeb A. Efficacy of beta-blockers in in-fantile capillary haemangiomas: The physiopathological sig-nificance and therapeutic consequences (in French). AnnDermatol Venereol. 2008;135:860–862.

3. Pandey A, Gangopadhyay AN, Gopol SC, et al. Twenty years’experience of steroids in infantile hemangioma: A develop-ing country’s perspective. J Pediatr Surg. 2009;44:688–694.

4. Boon LM, MacDonald DM, Mulliken JB. Complications ofsystemic corticosteroid therapy for problematic hemangi-oma. Plast Reconstr Surg. 1999;104:1616–1623.

5. Sadan N, Wolach B. Treatment of hemangiomas of infantswith high doses of prednisone. J Pediatr. 1996;128:141–146.

6. Argenta LC, Bishop E, Cho KJ, Andrews AF, Coran AG.Complete resolution of life-threatening hemangioma by em-bolization and corticosteroids. Plast Reconstr Surg. 1982;70:739–744.

7. Kushner BJ. The treatment of periorbital infantile heman-gioma with intralesional corticosteroid. Plast Reconstr Surg.1985;76:517–526.

8. Prasetyono TO, Djoenaedi I. Efficacy of intralesional steroidinjection in head and neck hemangioma: A systematic review.Ann Plast Surg. 2011;66:98–106.

9. Chantharatanapiboon W. Intralesional corticosteroid ther-apy in hemangiomas: Clinical outcome in 160 cases. J MedAssoc Thai. 2008;91(Suppl 3):S90–S96.

10. Jalil S, Akhtar J, Ahmed S. Corticosteroids therapy in themanagement of infantile cutaneous hemangiomas. J Coll Phy-sicians Surg Pak. 2006;16:662–665.

11. Zimmermann AP, Wiegand S, Werner JA, Eivazi B. Propran-olol therapy for infantile haemangiomas: Review of the lit-erature. Int J Pediatr Otorhinolaryngol. 2010;74:338–342.

12. Buckmiller L, Dyamenahalli U, Richter GT. Propranolol forairway hemangiomas: Case report of novel treatment. Laryn-goscope 2009;119:2051–2054.

13. Nguyen J, Fay A. Pharmacologic therapy for periocular in-fantile hemangiomas: A review of the literature. Semin Oph-thalmol. 2009;24:178–184.

14. Bigorre M, Van Kien AK, Valette H. Beta-blocking agent fortreatment of infantile hemangioma. Plast Reconstr Surg. 2009;123:195e–196e.

15. Denoyelle F, Leboulanger N, Enjolras O, Harris R, Roger G,Garabedian EN. Role of Propranolol in the therapeutic strat-egy of infantile laryngotracheal hemangioma. Int J PediatrOtorhinolaryngol. 2009;73:1168–1172.

16. Annabi B, Lachambre MP, Plouffe K, Moumdjian R, BeliveauR. Propranolol adrenergic blockade inhibits human brainendothelial cells tubulogenesis and matrix metalloprotei-nase-9 secretion. Pharmacol Res. 2009;60:438–445.

17. Liggett SB, Cresci S, Kelly RJ, et al. A GRK5 polymorphismthat inhibits beta-adrenergic receptor signaling is protectivein heart failure. Nat Med. 2008;14:510–517.

18. Frieden IJ, Drolet BA. Propranolol for infantile hemangio-mas: Promise, peril, pathogenesis. Pediatr Dermatol. 2009;26:642–644.

19. Leboulanger N, Fayoux P, Teissier N, et al. Propranolol inthe therapeutic strategy of infantile laryngotracheal heman-gioma: A preliminary retrospective study of French experi-ence. Int J Pediatr Otorhinolaryngol. 2010;74:1254–1257.

20. Holland KE, Frieden IJ, Frommelt PC, Mancini AJ, Wyatt D,Drolet BA. Hypoglycemia in children taking propranolol forthe treatment of infantile hemangioma. Arch Dermatol. 2010;146:775–778.

21. Pavlakovic H, Kietz S, Lauerer P, Zutt M, Lakomek M. Hy-perkalemia complicating propranolol treatment of an infan-tile hemangioma. Pediatrics 2010;126:e1589–e1593.

22. Perez RS, Mora PC, Rodrıguez JD, Sanchez FR, de Torres JdeL. Treatment of infantile hemangioma with propranolol (inSpanish). Ann Pediatr (Barc.) 2010;72:152–154.

23. Blei F, Chianese J. Corticosteroid toxicity in infants treatedfor endangering hemangiomas: Experience and guidelinesfor monitoring. Int Pediatr. 1999;14:146–153.

24. Chen MT, Yeong EK, Horng SY. Intralesional corticosteroidtherapy in proliferating head and neck hemangiomas: Areview of 155 cases. J Pediatr Surg. 2000;35:420–423.

25. Chowdri NA, Darzi MA, Fazili Z, Iqbal S. Intralesional cor-ticosteroid therapy for childhood cutaneous hemangiomas.Ann Plast Surg. 1994;33:46–51.

26. Iwanaka T, Tsuchida Y, Hashizume K, Kawarasaki H, UtsukiT, Komuro H. Intralesional corticosteroid injection withshort-term oral prednisolone for infantile hemangiomas ofthe eyelid and orbit. J Pediatr Surg. 1994;29:482–486.

27. Kushner BJ. Local steroid therapy in adnexal hemangioma.Ann Ophthalmol. 1979;11:1005–1009.

28. Kelly ME, Juern AM, Grossman WJ, Schauer DW, Drolet BA.Immunosuppressive effects in infants treated with cortico-


612

steroids for infantile hemangiomas. Arch Dermatol. 2010;146:767–774.

29. Zhou Q, Yang XJ, Zheng JW, Wang YA, Zhang ZY. Short-termhigh-dose oral prednisone on alternate days is safe and ef-fective for treatment of infantile hemangiomas. Oral SurgOral Med Oral Pathol Oral Radiol Endod. 2010;109:166–167.

30. Rossler J, Wehl G, Niemeyer CM. Evaluating systemic pred-nisone therapy for proliferating haemangioma in infancy.Eur J Pediatr. 2008;167:813–815.

31. Pope E, Krafchik BR, Macarthur C, et al. Oral versus high-dose pulse corticosteroids for problematic infantile heman-giomas: A randomized, controlled trial. Pediatrics 2007;119:e1239–e1247.

32. Greene AK, Couto RA. Oral prednisolone for infantile hem-angioma: Efficacy and safety using a standardized treatmentprotocol. Plast Reconstr Surg. 2011;128:743–752.

33. Al-Sebeih K, Manoukian J. Systemic steroids for the man-agement of obstructive subglottic hemangioma. J Otolaryngol.2000;29:361–366.

34. Narcy P, Contencin P, Bobin S, Manac’h Y. Treatment ofinfantile subglottic hemangioma: A report of 49 cases. IntJ Pediatr Otorhinolaryngol. 1985;9:157–164.

35. Price CJ, Lattouf C, Baum B, et al. Propranolol vs cortico-steroids for infantile hemangiomas: A multicenter retrospec-tive analysis. Arch Dermatol. 2011;147:1371–1376.

36. Schupp CJ, Kleber JB, Gunther P, Holland-Cunz S. Propran-olol therapy in 55 infants with infantile hemangioma: Dos-age, duration, adverse effects, and outcome. Pediatr Dermatol.2011;28:640–644.

37. Bagazgoitia L, Hernandez-Martın A, Torrelo A. Recurrenceof infantile hemangiomas treated with propranolol. PediatrDermatol. 2011;28:658–662.

38. Lv MM, Fan XD, Su LX. Propranolol for problematic headand neck hemangiomas: An analysis of 37 consecutive pa-tients. Int J Pediatr Otorhinolaryngol. 2012;76:574–578.

39. Fuchsmann C, Quintal MC, Giguere C, et al. Propranolol asfirst-line treatment of head and neck hemangiomas. ArchOtolaryngol Head Neck Surg. 2011;137:471–478.

40. Zaher H, Rasheed H, Hegazy RA, Hegazy RA, AbdelhalimDM, Gawdat HI. Oral propranolol: An effective, safe treat-ment for infantile hemangiomas. Eur J Dermatol. 2011;21:558–563.

41. de Graaf M, Breur JM, Raphael MF, Vos M, Breugem CC,Pasmans SG. Adverse effects of propranolol when used in thetreatment of hemangiomas: A case series of 28 infants. J AmAcad Dermatol. 2011;65:320–327.

42. Schiestl C, Neuhaus K, Zoller S, et al. Efficacy and safety ofpropranolol as first-line treatment for infantile hemangio-mas. Eur J Pediatr. 2011;170:493–501.

43. Missoi TG, Lueder GT, Gilbertson K, Bayliss SJ. Oral pro-pranolol for treatment of periocular infantile hemangiomas.Arch Ophthalmol. 2011;129:899–903.

44. Al Dhaybi R, Superstein R, Milet A, et al. Treatment ofperiocular infantile hemangiomas with propranolol: Caseseries of 18 children. Ophthalmology 2011;118:1184–1188.

45. Hogeling M, Adams S, Wargon O. A randomized controlledtrial of propranolol for infantile hemangiomas. Pediatrics2011;128:e259–e266.

46. Holmes WJ, Mishra A, Gorst C, Liew SH. Propranolol asfirst-line treatment for infantile hemangiomas. Plast ReconstrSurg. 2010;125:420–421.

47. Zvulunov A, McCuaig C, Frieden IJ, et al. Oral propranololtherapy for infantile hemangiomas beyond the proliferationphase: A multicenter retrospective study. Pediatr Dermatol.2011;28:94–98.

48. Chai Q, Chen WL, Huang ZQ, Zhang DM, Fan S, Wang L.Preliminary experiences in treating infantile hemangiomawith propranolol. Ann Plast Surg. Epublished ahead of printMay 27, 2011.

49. Bertrand J, McCuaig D, Dubois J, Hatami A, Ondrejchak S,Powell J. Propranolol versus prednisone in the treatment ofinfantile hemangiomas: A retrospective comparative study.Pediatr Dermatol. 2011;28:649–654.

50. Tan ST, Itinteang T, Leadbitter P. Low-dose propranolol forinfantile haemangioma. J Plast Reconstr Aesthet Surg. 2011;64:292–299.

51. Theletsane T, Redfern A, Raynham O, Harris T, Prose NS,Khumalo NP. Life-threatening infantile haemangioma: Adramatic response to propranolol. J Eur Acad Dermatol Vene-reol. 2009;23:1465–1466.

52. Itani MH, Fakih H. Response of facial haemangioma to oralpropranolol. BMJ Case Rep. 2009;2009:bcr0120091476.

53. Buckmiller LM, Munson PD, Dyamenahalli U, Dai Y, RichterGT. Propranolol for infantile hemangiomas: Early experi-ence at a tertiary vascular anomalies center. Laryngoscope2010;120:676–681.

54. Bonifazi E, Acquafredda A, Milano A, Montagna O, LaforgiaN. Severe hypoglycemia during successful treatment of dif-fuse hemangiomatosis with propranolol. Pediatr Dermatol.2010;27:195–196.

55. Sans V, de la Rogue ED, Berge J, et al. Propranolol for severeinfantile hemangiomas: Follow-up report. Pediatrics 2009;124:e423–e431.

56. Lawley LP, Siegfried E, Todd JL. Propranolol treatment forhemangioma of infancy: Risks and recommendations. PediatrDermatol. 2009;26:610–614.

57. Mahadevan M, Cheng A, Barber C. Treatment of subglottichemangiomas with propranolol: Initial experience in 10 in-fants. ANZ J Surg. 2011;81:456–461.

58. Mistry N, Tzifa K. Use of propranolol to treat multicentricairway haemangioma. J Laryngol Otol. 2010;124:1329–1332.

59. Cushing SL, Boucek RJ, Manning SC, Sidbury R, Perkins JA.Initial experience with a multidisciplinary strategy for initi-ation of propranolol therapy for infantile hemangiomas.Otolaryngol Head Neck Surg. 2011;144:78–84.

60. Jin YB, Lin XX, Ye XX, et al. A prospective study of pro-pranolol as first-line treatment for problematic infantilehemangioma in China. Zhonghua Zheng Xing Wai Ke Za Zhi.2011;27:170–173.

61. Léauté-Labréze C, Sans-Martin V. Infantile hemangioma.Presse Med. 2010;39:499–510.

62. Breur JM, de Graaf M, Breugem CC, et al. Hypoglycemiaas a result of propranolol during treatment of infantilehemangioma: A case report. Pediatr Dermatol. 2011;28:169 –171.


613

propranolol_versus_corticosteroids_in_the.27

Documents