OtolaryngologyHead and Neck Surgery144(5) 662 675 American Academy of OtolaryngologyHead and Neck Surgery Foundation 2011Reprints and permission: sagepub.com/journalsPermissions.navDOI: 10.1177/0194599811399241http://otojournal.org

No sponsorships or competing interests have been disclosed for this article.

Abstract

Background. Hearing loss associated with congenital cyto-megalovirus (CMV) infection occurs in 0.2 to 0.6 per 1000 neonates.

Objective. The primary goal of this systemic review was to test the following null hypotheses: (1) antiviral therapy has no impact on congenital CMV-related sensorineural hearing loss and (2) surgical therapy has no impact on congenital CMV-related sensorineural hearing loss.

Data Sources. Computerized searches of MEDLINE and EMBASE databases through September 2010 were performed, supplemented with manual searches and inquiries to topic experts.

Review Methods. Studies were included based on review of 387 studies according to criteria developed a priori. Data extrac-tion was performed by independent reviewers and focused on relevant audiologic measurements, study designs, and potential confounders.

Results. Criterion-meeting studies (n = 19) included a total of 446 participants. The largest randomized controlled trial (RCT) suggested a significant protective effect of intravenous ganciclovir against deterioration of hearing in neonates with central nervous system manifestations of CMV infection. It also, however, suggested a 3-fold increase in neutropenia. The second RCT suggested that there may be no significant ben-efit of intravenous ganciclovir for normal-hearing infants with asymptomatic congenital CMV. Additional prospective and retrospective data evaluated the impact of oral therapy and cochlear implantation in affected patients.

Conclusion. Although results are mixed, the highest level of evidence suggests that antiviral therapy confers a protective benefit on neonates with hearing loss and symptomatic CMV. Cochlear implantation can result in advancement of speech and language skills, but there are mixed results compared with nonCMV-infected patients.

Keywords

hearing loss, cytomegalovirus, antiviral, systematic review

Received November 16, 2010; revised January 4, 2011; accepted January 13, 2011.

Congenital hearing loss is estimated to occur in 1 to 4 per 1000 live births in the United States.1,2 Congenital hearing loss may result in developmental delay, lan-guage impairment, academic deficiency, and lower perceived health status.3 Medical care for children with hearing impair-ment has been estimated to cost 3 times that of normal-hearing children.3

Cytomegalovirus (CMV) is the most common congenital infection in the United States, with the economic impact exceeding $2 billion in this country alone.4-6 Data from cohorts followed in the United States, Europe, and Canada suggest that congenital CMV occurs in 3 to 12 per 1000 newborns.5 Hearing loss is the most common manifestation of congenital CMV infection, occurring in 15% to 65% of affected patients.5 The hearing loss may manifest in infants with or without cen-tral nervous system manifestation of CMV. It may also occur at birth and be detectable during newborn hearing screening, or it may occur years later. CMV-associated hearing loss is estimated to occur in 0.2 to 0.6 per 1000 and may account for up to 10% to 60% of congenital hearing loss.5

399241OTOXXX10.1177/0194599811399241Shin et alOtolaryngologyHead and Neck Surgery The Author(s) 2010

Reprints and permission:sagepub.com/journalsPermissions.nav

1Department of Otology and Laryngology, Harvard Medical School, Massachusetts Eye and Ear Infirmary, Boston, Massachusetts, USA2Division of Head and Neck Surgery, Los Angeles Medical Center, Southern California Permanente Medical Group, Los Angeles, California, USA3Division of Pediatric Infectious Disease, Los Angeles Medical Center, Southern California Permanente Medical Group, Los Angeles, California, USA

Corresponding Author:Jennifer J. Shin, MD, SM, Division of Head and Neck Surgery, Los Angeles Medical Center, Southern California Permanente Medical Group, Los Angeles, California, USA, and Department of Otology and Laryngology, Harvard Medical School, Massachusetts Eye and Ear Infirmary, 243 Charles Street, Boston, MA 02114 Email: jennifer_shin@meei.harvard.edu

Medical and Surgical Interventions for Hearing Loss Associated with Congenital Cytomegalovirus: A Systematic Review

Jennifer J. Shin, MD, SM1,2, Donald G. Keamy Jr, MD, MPH1 and Evan A. Steinberg, MD3

Literature Review

at IMSS on May 26, 2015oto.sagepub.comDownloaded from

Shin et al 663

A variety of studies have been performed to assess the potential impact of proposed therapies for hearing loss associ-ated with congenital CMV, with mixed reports of efficacy. Proposed therapies have the potential for serious adverse effects. Medical therapy may result in neutropenia, thrombo-cytopenia, and liver dysfunction. Surgical intervention intro-duces risks of facial nerve dysfunction, cerebrospinal fluid leak, and wound problems. Thus, it is critical to understand whether data suggest that intervention confers a true benefit to affected patients. This systematic review was therefore under-taken with the primary goal of testing the following null hypotheses: (1) antiviral therapy (intravenous, oral, or a com-bination) has no impact on congenital CMV-related sensori-neural hearing loss and (2) surgical therapy has no impact on congenital CMV-related sensorineural hearing loss. In addi-tion, secondary inquiries were made into potential optimal therapies and ages of intervention.

MethodsSearch StrategyA combination of computerized and manual searches was performed to identify all relevant data. A computerized PubMed search of MEDLINE ranging from 1966 to September 2010 was performed. The medical subject headings Cytomegalovirus and Cytomegalovirus Infections were exploded and the articles collected into a first group. Next, articles mapping to the subject headings Hearing Loss, Hearing Loss, Sensorineural, Hearing Loss, High-Frequency, Hearing Loss, Unilateral, Hearing Loss, Bilateral, Hearing Loss, Sudden, and Deafness were

exploded and collected into a second group. The 2 groups were cross-referenced, yielding 180 articles. In addition, a PubMed search from 1966 to September 2010 via the Endnote interface was performed using keywords Cytomegalovirus and Hearing in all fields. This search yielded 372 articles. A similar search was performed in EMBASE through September 2010, cross-referencing the exploded Emtree terms Cytomegalovirus or Hearing loss, which yielded 73 refer-ences. Reference lists for relevant narrative reviews and crite-ria-meeting publications were searched manually for additional studies. In addition, topic experts were contacted via phone, email, or letter to determine if any additional stud-ies or unpublished data could be identified. Studies were considered for inclusion in the systematic review by 2 inde-pendent physicians. The titles of all of the studies from these combined search methods (n = 387) were evaluated according to the a priori inclusion/exclusion criteria described below. This title evaluation then yielded 134 potential abstracts, which were reviewed in more detail according to the same criteria. Ultimately, 77 full articles were evaluated against a priori criteria in detail (Figure 1).

Inclusion/Exclusion CriteriaThe articles identified by the computerized and manual search strategy described above were evaluated to identify those that met the following inclusion criteria: (1) patients with congeni-tal CMV infection, (2) intervention with medical (intravenous or oral) or surgical therapy, (3) outcome measured in terms of hearing thresholds and (4) hearing outcomes reported in com-parison to a control group or in comparison to before treatment.

PubMed search of MEDLINE (n = 180)All fields search of MEDLINE (n = 372)EMBASE search (n = 73)Manual search (n = 13)Field experts (n = 6)

Citations initially identified*

Excluded based on a priori criteria (n = 253)

Included based on a priori criteria (n = 134)

Title search

Included based on a priori criteria (n = 77)

Excluded based on a priori criteria (n = 57)Abstract search

Inclusion in final analysis Included based on a priori criteria (n = 19)

Excluded based on a priori criteria (n = 58)Complete article search

Figure 1. Flow diagram showing the stages of identification of studies for the systematic review.

at IMSS on May 26, 2015oto.sagepub.comDownloaded from

664 OtolaryngologyHead and Neck Surgery 144(5)

Congenital CMV infections were confirmed via urine, saliva, or blood laboratory analysis of the affected infant. Hearing results were quantified and described via auditory brainstem response (ABR) or age-appropriate booth audiometry. Articles were excluded if (a) no hearing results were reported, (b) hear-ing results were reported only before or only after intervention but not both, (c) hearing results were unclear relative to the timing of intervention, (d) only maternal CMV status was analyzed, or (e) only diagnostic measures for CMV were evaluated. Letters, abstracts, and brief reports were included if sufficient information was present to determine that the inclu-sion/exclusion criteria were clearly met. This process yielded 19 criterion-meeting studies.

Data ExtractionData extraction was focused on items relevant to the study results, potential sources of heterogeneity among those results, and study identification (author, year of publication, full reference citation). Extracted data included (1) the num-ber/percentage with maintenance of hearing thresholds, improvement in hearing thresholds, deterioration in hearing thresholds; (2) the number of subjects in each group; (3) the P value, confidence interval, standard error of the mean, proportions, or descriptive statistics reported; and (4) the follow-up time. Data collection also included multiple poten-tial sources of heterogeneity among studies: (1) age at inter-vention, (2) means of CMV diagnosis, (3) details of the medical (antiviral type, dose, route of delivery, duration) or surgical intervention regimen, (4) details of the control regi-men if applicable, (5) audiologic criteria used for stratifica-tion of data, (6) additional manifestations of CMV, (7) primary study endpoints, (8) morbidity/complications of therapy, and (9) study design. Two reviewers experienced in systematic reviews and clinical treatment of CMV-affected infants evaluated the data independently using standardized tables.

Quantitative Data AnalysisAn a priori plan was made to perform a meta-analysis if the data were appropriate, meaning that study designs, outcome measures, and follow-up periods were similar enough. There proved, however, to be wide variation in study design, audio-metric criteria, and reporting parameters, so a quantitative meta-analysis was not performed.

ResultsStudy CharacteristicsThe 19 criterion-meeting studies included a total of 446 par-ticipants, but some of these were lost to follow-up, leaving 365 patients whose complete audiologic results could be ana-lyzed. Studies evaluating medical therapy included interven-tion with intravenous ganciclovir, oral ganciclovir, oral valganciclovir, or a combination thereof. Studies evaluating surgical intervention focused on cochlear implantation for severe to profound sensorineural hearing loss associated with congenital CMV. Study data and characteristics are described in Tables 1 to 6.

Impact of Antiviral Therapy on HearingAmong studies assessing the impact of antiviral therapy, there were 2 randomized controlled trials (RCTs), 2 prospective cohort studies, 3 retrospective case series, and 4 case reports (Tables 1-5).

Highest level of evidence. The larger randomized controlled trial6 was conducted by the National Institute of Allergy and Infectious Diseases Collaborative Antiviral Study Group. This RCT evaluated neonates with symptomatic congenital CMV disease involving the central nervous system. Hearing was assessed with ABR at baseline, 6 months, and 12 months. Randomization effectively balanced age, sex, race, prematu-rity, and multiple other potential confounders between antivi-ral and control groups. Infants were treated before 1 month of age with intravenous ganciclovir for 6 weeks. A placebo was not administered due to ethical concerns associated with maintaining an intravenous catheter for 6 weeks in a neonate, but the audiologist reviewing ABR results was masked to the status of the treatment arm. Forty-two infants completed the study; 100 were initially enrolled (58% attrition rate). A priori power calculations were not described in detail, although an initial plan to recruit 130 patients was described. The study was terminated early by the data safety and monitoring board due to favorable results in the interim analysis, as well as chal-lenges in patient accrual and follow-up. The results suggested that 6 weeks of intravenous ganciclovir resulted in a signifi-cantly improved prevention of deterioration of hearing in both the raw and adjusted analyses (P < .01). While 41% of neo-nates in the control group had worsened hearing at the 6-month follow-up, none of the neonates who underwent ganciclovir therapy had hearing deterioration. This significant decrease in risk persisted at the 12-month follow-up and regardless of whether a best ear or total ear analysis was performed. There was, however, also a higher rate of neutropenia in the ganci-clovir group (63% vs 21%, P < .01).

The second, smaller RCT7 evaluated the impact of intrave-nous ganciclovir on neonates 10 days of age with asymptom-atic congenital CMV. Randomization effectively balanced confounders between groups. All patients had normal hearing at the outset, and 18 of 23 completed the study. Hearing dete-rioration did not occur in any of the infants treated with ganci-clovir, while 25% of those in the control group had progressive hearing loss. The difference was not statistically significant, although an a priori power analysis was not reported. Neutropenia occurred in 2 (11.1%) of the infants in the ganci-clovir group.

Treatment of symptomatic infants with intravenous ganciclovir only. Multiple studies addressed the impact of treatment of infants affected by symptomatic CMV disease with intrave-nous ganciclovir only. The larger RCT6 discussed in detail above (see the Highest Level of Evidence section) used intravenous ganciclovir 6 mg/kg/dose every 12 hours for 6 weeks and showed a significant decrease in risk of progres-sion of hearing loss as compared with no treatment. A pro-spective phase 2 cohort study8 utilized 6 weeks of 8 or 12 mg/kg/d of intravenous ganciclovir for 6 weeks in infants

665

Tabl

e 1.

Med

ical

Tre

atm

ent:

Ran

dom

ized

Con

trol

led

Tria

ls

Ref

eren

ce:

Kim

berl

in e

t al

(20

03)6

Lack

ner

et a

l (20

09)7

Leve

l (de

sign

):1

(ran

dom

ized

con

trol

led

tria

l)1

(ran

dom

ized

con

trol

led

tria

l)Sa

mpl

e si

zea :

42 p

atie

nts

(100

), 85

ear

s (2

00)

18 (

23)

patie

nts

Out

com

es

Best

Ear

A

sses

smen

tH

eari

ng

Impr

ovem

ent

Mai

nten

ance

of

Nor

mal

Hea

ring

Mai

nten

ance

of

Abn

orm

al H

eari

ngH

eari

ng

Det

erio

ratio

nH

eari

ng

Impr

ovem

ent

Mai

nten

ance

of

Nor

mal

Hea

ring

Mai

nten

ance

of

Abn

orm

al H

eari

ngH

eari

ng

Det

erio

ratio

n

Ant

ivir

al24

% (

n =

6/2

5)60

% (

n =

15/

25)

16%

(n

= 4

/25)

0% (

n =

0/2

5)0%

(n

= 0

/10)

100%

(n

= 1

0)0%

(n

= 0

/10)

0% (

n =

0/1

0)C

ontr

ol29

% (

n =

5/1

7)29

% (

n =

5/1

7)0%

(n

= 0

/17)

41%

(n

= 7

/17)

0% (

n =

0/8

)75

% (

n =

6/8

)0%

(n

= 0

/8)

25%

(n

= 2

/8)

P.0

5C

oncl

usio

nA

ntiv

iral

reg

imen

has

sig

nific

ant

effe

ct o

n pr

even

ting

dete

rior

atio

n of

hea

ring

in

adju

sted

and

una

djus

ted

anal

ysis

Tren

d to

war

d le

ss h

eari

ng d

eter

iora

tion

with

ant

ivir

al t

hera

py b

ut n

o si

gnifi

cant

di

ffere

nce

Follo

w-u

p tim

e6

mo,

1 y

Mea

n of

7.1

y (

SD, 2

.2; r

ange

, 3.2

-10.

3)

Stud

y D

esig

n

Incl

usio

n

crite

ria

Neo

nate

s (

1 m

o of

age

) w

ith s

ympt

omat

ic c

onge

nita

l CM

V d

isea

se in

volv

ing

the

CN

S, w

ith c

onfir

med

isol

atio

n of

CM

V fr

om a

uri

ne s

peci

men

. All

had

C

NS

dise

ase

such

as

(1)

mic

roce

phal

y, (2

) in

trac

rani

al c

alci

ficat

ions

, (3

) ab

norm

al C

SF fo

r ag

e, (

4) c

hori

oret

initi

s, an

d/or

(5)

hea

ring

def

icits

. Inf

ants

>

32 w

k ge

stat

ion,

>12

00 g

at

birt

h

Neo

nate

s

1.5

mg/

dL, H

IV, h

ydra

nenc

epha

lyM

icro

ceph

aly,

hydr

ocep

haly,

ven

tric

ulom

egal

y, ch

orio

retin

itis

or o

ther

oph

thal

mol

ogic

al

sym

ptom

s, he

pato

sple

nom

egal

y, th

rom

bocy

tope

nia,

neut

rope

nia,

anem

ia, j

aund

ice,

or

hear

ing

diso

rder

at

the

outs

etR

ando

miz

atio

n ef

fect

iven

ess

No

sign

ifica

nt d

iffer

ence

in a

ge, s

ex, r

ace,

pre

mat

urity

, ges

tatio

nal a

ge, w

eigh

t,

head

cir

cum

fere

nce,

abn

orm

al C

T o

r C

SF, A

LT, p

late

let

coun

t, bi

lirub

in,

sple

nom

egal

y, he

mal

omeg

aly,

base

line

ABR

, AN

C g

rade

No

sign

ifica

nt d

iffer

ence

in a

ge, s

ex, h

eari

ng a

t th

e ou

tset

Age

at

inte

rven

tion

1 m

o1

0 d

of li

fe

Mas

king

A s

ingl

e au

diol

ogis

t w

ho w

as m

aske

d to

ran

dom

izat

ion

revi

ewed

all

ABR

sN

ot r

epor

ted

Ant

ivir

al

regi

men

de

tails

Gan

cicl

ovir

6 m

g/kg

/dos

e IV

eve

ry 1

2 h

for

6 w

k; if

AN

C fe

ll be

low

50

0 ce

lls/m

m3 ,

ganc

iclo

vir

was

hel

d un

til a

ret

urn

to >

750

cells

/mm

3 , w

hen

it w

as r

esum

ed a

t th

e fu

ll do

se. I

f neu

trop

enia

rec

urre

d, t

he d

ose

was

de

crea

sed

by 5

0% u

ntil

the

AN

C r

ose

abov

e 50

0 ce

lls/m

m3 .

If ne

utro

peni

a pe

rsis

ted

at t

he 5

0% d

osag

e, g

anci

clov

ir w

as d

isco

ntin

ued

Gan

cicl

ovir

10

mg/

kg IV

for

21 d

ays.

If ne

utro

peni

a or

oth

er a

dver

se e

ffect

s oc

curr

ed,

the

dose

was

low

ered

to

5 m

g/kg

; the

rapy

was

sto

pped

if s

ide

effe

cts

did

not

reso

lve

Con

trol

reg

imen

de

tails

No

trea

tmen

t. A

pla

cebo

arm

was

not

use

d be

caus

e of

eth

ical

con

cern

s ov

er

mai

ntai

ning

intr

aven

ous

acce

ss fo

r 6

wk

No

trea

tmen

t

(con

tinue

d)

at IMSS on May 26, 2015oto.sagepub.comDownloaded from

666 Ref

eren

ce:

Kim

berl

in e

t al

(20

03)6

Lack

ner

et a

l (20

09)7

Leve

l (de

sign

):1

(ran

dom

ized

con

trol

led

tria

l)1

(ran

dom

ized

con

trol

led

tria

l)Sa

mpl

e si

zea :

42 p

atie

nts

(100

), 85

ear

s (2

00)

18 (

23)

patie

nts

Endp

oint

sT

he p

rim

ary

stud

y en

dpoi

nt w

as im

prov

emen

t by

1 g

rada

tion

in A

BR a

t th

e 6-

mo

follo

w-u

p (e

g, m

oder

ate

impa

irm

ent

impr

oved

to

mild

impa

irm

ent)

or

norm

al A

BR a

t bo

th t

ime

poin

ts. N

onpr

imar

y en

dpoi

nts:

thro

mbo

cyto

peni

a, he

patit

is, o

rgan

omeg

aly,

chor

iore

tiniti

s, ra

te o

f gro

wth

, dea

th.A

fter

the

thi

rd

inte

rim

ana

lysi

s, th

e D

SMB

reco

mm

ende

d ea

rly

term

inat

ion

base

d on

favo

rabl

e pr

elim

inar

y st

udy

resu

lts a

nd in

rec

ogni

tion

of c

halle

nges

in p

atie

nt a

ccru

al a

nd

follo

w-u

p

Hea

ring

loss

, alth

ough

chi

ldre

n w

ere

orig

inal

ly r

ando

miz

ed fo

r a

prev

ious

stu

dy. H

eari

ng

thre

shol

d w

as a

sses

sed

usin

g ea

r m

icro

scop

y, m

iddl

e-ea

r im

peda

nce

test

ing,

and

beha

vior

al o

bser

vatio

n au

diom

etry

unt

il 4

y of

age

; pur

e-to

ne a

udio

met

ry w

as u

sed

for

olde

r ch

ildre

n

Aud

iolo

gic

crite

ria

used

ABR

thr

esho

lds:

the

low

est

inte

nsity

leve

l at

whi

ch w

ave

V c

ould

be

dete

cted

and

re

plic

ated

. The

follo

win

g de

finiti

ons

wer

e us

ed: n

orm

al h

eari

ng, 0

to

20 d

B; m

ild

hear

ing

loss

, 21

to 4

5 dB

; mod

erat

e he

arin

g lo

ss, 4

6 to

70

dB; s

ever

e he

arin

g

loss

, >71

dB

Sens

orin

eura

l hea

ring

loss

was

def

ined

as

a m

edia

n se

nsor

ineu

ral d

ecre

ase

in h

eari

ng

of >

10 d

B he

arin

g lo

ss a

t lo

w (

125-

1000

Hz)

, mid

dle

(100

0-40

00 H

z), o

r hi

gh (

4000

-16

,000

Hz)

freq

uenc

ies

and

was

gra

ded

as m

ild (

25 t

o 40

dB

hear

ing

loss

), m

oder

ate

(41

to 6

5 dB

hea

ring

loss

), se

vere

(66

to

96 d

B he

arin

g lo

ss),

or p

rofo

und

(>96

db

hear

ing

loss

)Po

wer

Sam

ple

size

cal

cula

tions

pro

ject

ed 1

30 e

nrol

led

patie

nts

to p

rovi

de 1

00 e

valu

able

pa

tient

sN

ot r

epor

ted

Mor

bidi

ty/

com

plic

atio

ns63

% o

f gan

cicl

ovir

rec

ipie

nts

deve

lope

d gr

ade

3 or

4 n

eutr

open

ia, c

ompa

red

with

21

% o

f pat

ient

s in

the

con

trol

gro

up (

P

2.0

mg

Not

oth

erw

ise

spec

ified

Age

at

time

of

trea

tmen

t