International Journal of Pediatric Otorhinolaryngology 79 (2015) 698706

Contents lists available at ScienceDirect

International Journal of Pediatric Otorhinolaryngology

journal homepage: www.elsevier.com/locate/ijporl

Retrospective analysis of tympanoplasty in children with cleft palate:

A 24-year experience. II. Cholesteatomatous cases
Gabor Kopcsanyi a,*, Olga Vincze a, Viktor Bagdan b, Jozsef Pytel c
a

Department of Pediatrics, University of Pecs Medical School, Pecs, Hungary

Pollack Mihaly Faculty of Engineering and Information Technology, University of Pecs, Hungary
c
Department of Otorhinolaryngology, Head and Neck Surgery, University of Pecs Medical School, Pecs, Hungary
b

A R T I C L E I N F O

A B S T R A C T

Article history:
Received 24 November 2014
Received in revised form 16 February 2015
Accepted 17 February 2015
Available online 26 February 2015

Background: Contradictory experience has been published on the outcomes of ear surgery in patients
with cleft palate.
Objectives: The authors of this study investigated whether there were differences in the short- and longterm outcomes of tympanoplasty performed due to cholesteatoma in children with or without cleft
palate.
Setting: Tertiary care medical centre.
Methods: The authors retrospectively analyzed the rst authors 24-year experience of paediatric
tympanoplasty using the software programme developed by the fourth author. The outcomes of 268
tympanoplasties on 172 ears with cholesteatoma in 151 NoCleft patients were compared to the
outcomes of 35 tympanoplasties on 20 ears of 19 Cleft patients. The average age of the patients was
10.7  3.6 years and 9.5  2.7 years respectively. The average follow-up time was 4 and 4.1 years.
Results: Preoperative PTA-ABGs (31.22/34.88 dB; p = 0.058), best postoperative PTA-ABGs (17.04/
16.4 dB; p = 0.499), last postoperative PTA-ABGs (19.93/20.98 dB; p = 0.298), the nal hearing
improvement (11.29/13.9 dB; p = 0.193) and postoperative PTA-ABG deterioration with time (2.89/
4.58 dB; p = 0.117) were statistically compared between the NoCleft and Cleft groups. The same
parameters were analyzed separately in the case of tympanoplasty performed with intact ossicular chain
and the different type of columella ossiculoplasty. No signicant differences were found between the
two groups in any of these parameters. However, signicant difference was found in the necessity for
grommet insertion (8-fold difference, p  0), and conversion to open techniques (p  0).
Conclusions: The authors conclude that the achievable audiological outcomes of tympanoplasty in
children with cleft palate and cholesteatoma do not differ signicantly from those of the general child
population. However, this more frequently requires ventilation tube insertion and more frequent followup visits. The latter is ensured by patient care within the frameworks of the Cleft Palate Team. We have
to accept that in some cases Eustachian tube dysfunction caused by the underlying disease (cleft palate)
takes over and we have to resort to open techniques.
2015 Elsevier Ireland Ltd. All rights reserved.

Keywords:
Cholesteatoma
Cleft palate
Paediatric tympanoplasty
Short and long term follow-up

1. Introduction
According to our present knowledge, the only denitive
treatment for middle ear cholesteatoma is surgery. The aim of
the procedure is to eliminate the dry or infected expansive
cholesteatoma potentially leading to complications and to

* Corresponding author at: Department of Pediatrics, University of Pecs Medical

School, 7 Jozsef Attila Street, 7623 Pecs, Hungary. Tel.: +36 72 535 900;
mobile: +36 303960225.
E-mail address: kopcsanyi.gabor@pte.hu (G. Kopcsanyi).
http://dx.doi.org/10.1016/j.ijporl.2015.02.020
0165-5876/ 2015 Elsevier Ireland Ltd. All rights reserved.

establish an intact novomembrane and air-containing middle

ear cleft while improving or at least preserving hearing and
avoiding recurrence. There are two common types of surgery. The
closed technique (canal wall-up tympanomastoidectomy, CWU
TM) aims to preserve the posterior bony canal wall thus creating
anatomical conditions that are similar to normal (swim-proof
ear). The open technique (canal wall-down tympanomastoidectomy, CWD TM) involves the removal of the posterior bony canal
wall, as a result of which the mastoid area remains open towards
the external auditory canal. This allows for the easier inspection of
the mastoid cavity lined with keratinizing squamous epithelium
on ear examination. We try to create a novomembrane capable of

G. Kopcsanyi et al. / International Journal of Pediatric Otorhinolaryngology 79 (2015) 698706

vibration and reconstruct the ossicular chain, if necessary [1,2]. For

good postoperative hearing results it is crucial to maintain the air
content of the residual or complete middle ear cleft by preserving
automatic Eustachian tube function, or if it is not possible, by
ventilation tube (grommet) insertion providing adequate middle
ear pressure [1,3].
In children, maintaining the constant atmospheric pressure of
the middle ear is of outstanding importance due to their frequent
upper respiratory tract catarrhs. Certain craniofacial anomalies
directly affecting the opening mechanism of the Eustachian tube
particularly complicate the situation. The most expressed examples of this problem are the different forms of cleft palate and the
conditions associated with velopharyngeal insufciency (VPI). It
should be kept in mind that Eustachian tube dysfunction is the
most signicant underlying cause of cholesteatoma resulting
from invagination [4]. It is no wonder that cholesteatoma is much
more common in children with cleft palate than in the healthy
population (1.89.2% vs. 0.0030.006%) [5]. Therefore all
available conservative and surgical treatments should be
attempted to restore Eustachian tube function before resorting
to tympanoplasty.
Contradictory data have been published on the outcomes of ear
surgery in cleft patients. Earlier studies mention poor results of
tympanoplasty in these patients [6,7], while a relatively low
number of more recent studies reect a more optimistic view [8,9].
Only one recently published study from Parma focuses exclusively
on cholesteatoma in children with cleft palate [5].
The aim of our present study was to compare the short- and
long-term outcomes of tympanoplasty performed due to cholesteatoma in children without cleft palate (NoCleft patients) and
with cleft palate (Cleft patients).
We have recently published a retrospective, comparative
analysis of our surgical data on paediatric patients with
mesotympanic chronic otitis media based on similar considerations [10]. In mesotympanic cases, no signicant difference was
found between the two groups in anatomical or functional
outcomes. However, grommet insertion was more likely to be
necessary in children with cleft palate.
2. Materials and methods
The authors of this study analyzed the rst authors (G.K.) data
on paediatric tympanoplasty (17 July 199010 March 2014;
Division of Otorhinolaryngology, Department of Pediatrics University of Pecs Medical School). Since 1996 he has performed the
surgeries of Cleft patients in the frameworks of the Cleft Lip and
Palate Team founded in the same year at the Department of
Pediatrics, University of Pecs Medical School. The team is currently
responsible for the care of more than 600 cleft patients. Several
disciplines are involved in the team to help solve the diverse
problems of patients with cleft palate.
Altogether 508 tympanoplasties were performed on 367 ears in
314 patients in the past 24 years. The average age of the patients
was 10.7  3.7 years (2.421.3 years). Our data on mesotympanic
cases were published recently [10].
During the period mentioned above, 303 tympanoplasties were
performed on 192 ears with cholesteatoma in 170 children. The
average age of the patients was 10.6  3.6 years (2.419.5). In our
present study we analyzed the surgical and hearing test data of all
Cleft and NoCleft patients with cholesteatoma. No patient selection
was applied. The rate of assessable audiological results only depended
on whether the patients showed up at an appropriate number of
follow-up visits or not.
The fourth author (J.P.) developed a computer programme for
the detailed analysis of tympanoplasty outcomes and statistical
data processing (Pytel SoftWare 2003) [1012]. According to

699

current guidelines, pure-tone averages and average ABGs were

calculated by averaging the thresholds for pure tones at 0.5, 1, 2
and 3 kHz [13]. The audiometric criteria used were those of the
American Academy of OtolaryngologyHead and Neck Surgery
(AAOHNS). Auditory results were considered good for a
postoperative, pure-tone, average air-bone gap (PTA-ABG) of
20 dB or less.
All surgical data as well as all pre- and postoperative
audiograms were registered in the database of Pytel SoftWare.
Pytel SoftWare makes selections and statistics according to the
desired aspects [10]. Follow-up times were calculated from the last
surgical intervention.
The surgeon has followed the doctrines of the school of ear
surgery marked by the name of Professor Bauer [1,10]. He
underlines the importance of the potentially preoperative restoration of Eustachian tube function. This involves the basically
conservative treatment of patients with chronic rhinosinusitis, the
medical treatment of nasal allergic symptoms, in rare cases the
surgical treatment of suspected septum deviation and in most
patients without cleft palate adenotomy several months before the
ear surgery.
He aims to apply uniform surgical techniques, which basically
means closed techniques. In cholesteatomatous cases he applies
combined transcanal-transmastoid approach. Following the matrix from the ostium of the cholesteatoma he prepares between the
layers of the two mucous membranes while removing the bone
bounding the matrix where necessary (with diamond drill or with
chisel in close proximity of the matrix). This way in most cases he
performs more or less extended lateral atticotomy. He prepares the
cholesteatoma part retrograde to the facial canal from mastoid
exploration [1].
The rst author does not perform reconstruction on the lateral
wall of the atticus with bone or cartilage since he does not believe
that these could resist the persistent decrease in the pressure of the
tympanic cavity. Furthermore, these bones and cartilages make it
impossible to judge the air-tightness of the tympanic cavity behind
them and interfere with the visualization of possible residual
cholesteatoma during postoperative examinations [14,15]. Residual cholesteatoma refers to the squamous epithelium left inside
the middle ear cleft (inadvertently or intentionally).
The reconstruction of the tympanic membrane and the
removed lateral bony atticus wall is carried out with temporal
fascia ap or in rare cases periosteum but always with underlay
technique and with the application of pull-back technique, if
necessary (Bailey, 1976) [10,16].
Ossicular chain reconstruction is carried out with autogenous
cortical bone columellae (ACBC) according to the methods
described by Bauer, as described in our previous publication [10].
The surgeon always schedules a two-stage surgery in patients
with cholesteatoma after a discussion with the patients and their
parents. The rst stage is sanation and tympanic membrane
reconstruction. Normally the second stage involves revision
(residual cholesteatoma removal if necessary) and ossicular chain
reconstruction.
Unfortunately, some patients do not show up at the second
stage in spite of our persuasive attempts probably due to the
achieved dry ears and the cessation of parental concern.
Ventilation tube (grommet) insertion was carried out depending on the results of microscopic and audiological examinations,
irrespective of potential previous operations on the respective ear
and the presence of cleft palate.
Intraoperative grommet insertion was carried out when we
could not achieve Valsalva-positivity preoperatively and there was
no hope that the patient would be able to ingest air into the middle
ear after surgery. Postoperative grommet insertion was carried out
when signicant and progressive novomembrane retraction was

700

G. Kopcsanyi et al. / International Journal of Pediatric Otorhinolaryngology 79 (2015) 698706

observed after wound healing with signs of seromucous discharge

in the tympanic cavity resulting in type C tympanogram and these
processes did not show any improvement within weeks after the
surgery [1,3,10].
According to the view of our team, cleft palate closure should be
carried out at the age of 1.5 years.
In most cases, the statistical comparison of the identical data of
the two groups was carried out with Students t-test (Microsoft
Excel software). Where the number of data was too low, Mann
WhitneyWilcoxon test was applied (SPSS software). Ratio
analyses were carried out with large-sample z and small-sample
chi-square (x2) tests.
3. Ethical considerations
The parameters mentioned in the study were obtained from the
clinical history in a way that all kinds of relationship between these
parameters and personal identication data were irreversibly
extinguished (data anonymization). Regional Research Ethics
Committee of the Medical Center, Pecs licence number: 5193.
4. Results
During the previously mentioned 24 years 303 tympanoplasties
were performed on 192 ears with cholesteatoma in 170 patients. In
the NoCleft group, 268 surgeries on 172 ears of 151 patients and in
the Cleft group 35 surgeries on 20 ears of 19 patients were
performed. The average age of the patients was 10.7  3.6 years (2.4
19.5 years) and 9.5  2.7 years (5.513.7 years) respectively (Fig. 1).
The graft take rate of the fascia (or in rare cases, periosteum)
used for reconstruction was 100%. Small postoperative perforation
only developed in 1 NoCleft patient.
In patients with cholesteatoma, grommet insertion was
necessary in 9 ears in the NoCleft group (3 intra- and 6
postoperative), which means 5% and in 8 ears in the Cleft group
(6 intra- and 2 postoperative) accounting for 40%. This means a
signicant, 8-fold difference in the ratios between the two groups
(p = 0.00000068; p  0).
No signicant differences were found between the two groups
considering the starting point of the cholesteatoma. Apart from 7
congenital cholesteatoma cases in the NoCleft group, in the
majority of cases (113 ears) the orice of the cholesteatoma was
found in the posterior-superior quadrant of the pars tensa
(posterior subligamentary starting point, according to Tos [17]),
while in 52 ears the starting point of the cholesteatoma was

epitympanic, originating from Shrapnells membrane. In the Cleft

group the cholesteatoma developed from the posterior-superior
quadrant in 15 ears and from the epitympanum in 5 ears (smallscale chi-square test, p = 0.230). This corresponds to the previously
published data: pars tensa is a more common starting point for
cholesteatoma in children, while in adults, epitympanic development from the pars accida is more frequent [15,17,18].
As for the extension of the cholesteatoma found during the rst
surgery, advanced progression with expansion to the atticus and
antrum was more common in Cleft patients. In the NoCleft
group cholesteatoma was limited to the windows and their
surroundings, the retrotympanum and its recesses in 70 ears (socalled niche`- cholesteatoma according to the Hungarian nomenclature by Professor Z Szabo). Isolated atticus cholesteatoma was
found in 9 ears; cholesteatoma compromising the atticus and the
tympanic cavity was found in 24 ears; cholesteatoma affecting the
antrum, the tympanic cavity and the atticus was revealed in 23
ears and cholesteatoma affecting the mastoid area apart from the
areas mentioned above was found in 46 cases. The numbers of
cases in the Cleft group were 4, 1, 3, 7 and 5 in the same order of
extension.
Although parents gave their approval to the two-stage surgery,
only 83 of 172 NoCleft ears underwent the second surgery (48%).
Reoperation was performed in 13 cases, mainly columella
ossiculoplasties due to unsatisfactory hearing improvement
(displacement or atrophy of the previous columella) or to gain a
safety inspection (residual cholesteatoma in stage II).
Residual cholesteatoma was found in 32 cases during the new
exploration (stage II, reoperation), small pearl-like lesions were
seen in 30 cases and large residual cholesteatoma in 2 cases. This
means a residual cholesteatoma ratio of 33% regarding repeated
middle ear explorations and 18.6% regarding the total number of
ears. Recurrent cholesteatoma was found and treated in 5% of
patients undergoing a new exploration (5 cases) and in 3% of all
ears. Recurrent cholesteatoma refers to the newly developing
retraction pocket in the postoperative period.
Stage II surgeries were performed on 12 of 20 Cleft ears (60%)
and 3 ears underwent reoperation due to unsatisfactory hearing
improvement or recurrent cholesteatoma. The rate of residual
cholesteatoma was 27% in Cleft patients (pearl-like lesions were
found in 4 cases during the new exploration). This means a rate of
20% regarding all Cleft ears with cholesteatoma. In the Cleft
group, recurrent cholesteatoma was found in 13% (2 cases), which
necessitated the conversion to open technique (recurrent cholesteatoma developed in 10% of all ears).

Fig. 1. Age distribution of patients with cholesteatoma at the time of the rst surgery.

G. Kopcsanyi et al. / International Journal of Pediatric Otorhinolaryngology 79 (2015) 698706

The results clearly show the explicit difference in the rate of

recurrent cholesteatoma between the two groups: the incidence is
almost threefold in the Cleft group (5% vs. 13%), however, the
difference is statistically not signicant (p = 0.219).
As mentioned earlier, we are basically dedicated to closed
techniques. However, we had to apply open techniques on 2 ears in
the NoCleft group due to cholesteatoma affecting the area of the
semicircular canals and spreading below the facial canal close to
the apex of the mastoid process. There was no hope for the
complete removal of the matrix in these areas.
In the Cleft group, we had to apply or convert the surgery to
open techniques in 3 cases. This was due to extensive cholesteatoma associated with signicant sensorineural hearing loss in one
case, while in two cases the reason for resorting to open technique
was recurrent cholesteatoma developing due to uncontrollable
Eustachian tube dysfunction.
The necessity for open technique showed signicant differences
between the two patient groups (2/172 and 3/20); (chi-square test:
p = 0.000069; p  0).
According to the nomenclature of the Pecs school of ear surgery,
type I tympanoplasty is a surgery ending with intact ossicular
chain; type II is a surgery involving short-type columella
ossiculoplasty (stapes head novomembrane columella); type
III refers to the application of long-type columella (mobile stapes
footplate novomembrane) and type IV leaves an interrupted
ossicular chain at the end of the surgery [1].
In the NoCleft group, 172 stage I surgeries were performed on
the ears with cholesteatoma: 61 type I, 18 type II, 2 type III and 91
type IV tympanoplasties.
The aims of stage II surgeries were revision (is there residual
cholesteatoma?) and columella ossiculoplasty. Two type I, 42 type
II, 22 type III and 7 type IV surgeries were performed and in 10
cases the second surgery was meant to provide a safety inspection.
The purpose of the 13 reoperations was further hearing
improvement or a safety inspection due to residual cholesteatoma found during stage II surgery. Two type II, 7 type III, 1 type
IV surgeries were performed and only a safety inspection in
3 cases.
In the Cleft group 20 stage I interventions were performed on
the ears with cholesteatoma. Five type I, 3 type II and 12 type IV

701

surgeries were carried out. In the second stage, 8 short-type and 2

long-type columella ossiculoplasties and in two cases a repeated
revision of the tympanic cavity were performed.
A third surgery was necessary in 3 cases: 2 short-type columella
ossiculoplasties and one revision.
A fourth surgery was performed in 1-1 cases in both groups. In
the NoCleft group it was a safety revision due to a tiny residual
cholesteatoma observed earlier in the tympanic sinus. During the
fourth surgery intact conditions were found.
In the Cleft group a fourth surgery and conversion to open
techniques were necessary because of recurrent cholesteatoma
due to uncontrollable Eustachian tube dysfunction. This condition
developed within a short time after years of air-tight middle ear
and good hearing (Fig. 2).
4.1. Audiological analysis
The most important indicator of hearing outcomes is considered to be the air-bone gap (ABG) [1,10,12]. ABG reects the
success or failure of our interventions. Pytel Software used for
data analysis only evaluates ears with the appropriate number of
audiograms.
4.1.1. Total cholesteatoma material
Maximum postoperative hearing improvement was 14.75 dB in
our total cholesteatoma material. The average preoperative ABG of
31.7  12.21 decreased to 16.95  11.73 dB. According to the general
experience, the audiological outcomes of tympanoplasty deteriorate
by some dBs over the years. In our material the average ABG was
20.07  13.27 on the last measurements. This means an average
improvement of 11.63 dB compared to the preoperative values so far.
Average ABG deterioration with time was 3.12 dB. Considering the
best postoperative results, an air-bone gap of less than 20 dB was
achieved in 75% of the cases. The average follow-up time was 2.2
years. On the last measurements, this value was 66% calculating the
deterioration of average ABG values below 20 dB. The average followup time of the last measurements was 4 years (the rate of best/last
ABG values below 20 dB and the respective follow-up times:
ABG < 20 dB = 75/66%; average follow-up time: 2.2/4 years)
(Table 1).

Fig. 2. Follow-up curve of patient number 228 (B = underlay technique; S = silicon sheet; pb = pull-back technique; col.breve = II = short-type columella; Op.radic. = radical
mastoidectomy). See the text for further explanation.

G. Kopcsanyi et al. / International Journal of Pediatric Otorhinolaryngology 79 (2015) 698706

702

Table 1
The horizontal rows refer to the examined groups, the columns contain the examined parameters (NoCleft: without cleft plate; Cleft: with cleft palate; type I: intact
ossicular chain; type II: short-type, stapes head tympanic membrane columella; type III: long-type, stapes footplate tympanic membrane columella; Columell.: any
columella type). Signicant difference cannot be revealed in the identical audiological values of the two groups.

Cholest. altogether
NoCleft altogether
Cleft altogether
NoCleft type I
Cleft type I
NoCleft type II
Cleft type II
NoCleft type III
Cleft type III
NoCleft Columell.
Cleft Columell.

Number of
operated
ears

Number of
evaluated
ears (%)

Preop.
ABG (dB)

Best
postop.
AGB (dB)

Maximum
improvement
(dB)

Last
postop.
ABG (dB)

Final
improvement
(dB)

Postop.
best/last
deterioration
(dB)

ABG
<20 dB
best/
last (%)

Average
follow-up
period
best/last
(years)

192
172
20
61
5
59
11
25
2
85
13

145
126
19
43
5
50
11
22
2
72
13

31.7  12.21
31.22  12.12
34.88  12.34
23.88  10.68
29.85  7.16
32.88  11.54
35.57  14.34
37.56  8.57
36.12  4
34.61  11.15
35.66  13.28

16.95  11.73
17.04  11.91
16.4  10.4
11.83  6.9
13.62  4.92
14.69  7.97
12.77  5.18
23.65  14.14
27.12  8.5
17.76  11.32
14.98  7.89

14.75
14.18
18.48
12.05
16.23
18.19
22.8
13.91
9
16.85
20.68

20.07  13.27
19.93  13.27
20.98  13.21
13.21  6.33
15.97  5.47
17.98  9.8
17.51  11.22
30.27  15.87
38.75  7
21.82  13.39
20.77  13.17

11.63
11.29
13.9
10.67
13.88
14.9
18.06
7.29
2.63
12.79
14.89

3.12
2.89
4.58
1.38
2.35
3.29
4.74
6.62
11.63
4.06
5.79

75/66
74/66
83/67
92/90
100/80
80/70
90/81
54/31
50/0
70/57
83/68

2.2/4
2.4/4
1.3/4.1
1.4/2.3
0.6/1.9
2.6/4.6
1.8/3.8
2.3/4.2
0.6/13.3
2.5/4.5
1.6/5.2

(76)
(73)
(95)
(70)
(100)
(85)
(100)
(88)
(100)
(85)
(100)

4.1.2. Comparison
The same parameters are discussed below separately in the
same order in the NoCleft and Cleft groups for better comparison.
In the NoCleft group, the average preoperative ABG of
31.22  12.12 dB decreased to 17.04  11.91 dB considering the best
postoperative results, which means an average best improvement of
14.18 dB considering data evaluated by the programme. On the last
measurements, the average postoperative ABG was 19.93  13.27 dB,
which means an average improvement of 11.29 dB compared to the
preoperative values. Postoperative deterioration with time was
2.89 dB (ABG < 20 dB = 74/66%; average follow-up time: 2.4/4 years)
(Fig. 3).
In the Cleft group, the average preoperative ABG of
34.88  12.34 dB decreased to 16.4  10.4 dB considering the best
postoperative values, which means an average best improvement of
18.48 dB. On the last measurements, the average ABG was
20.98  13.21 dB.
Deterioration
with
time
was
4.58 dB
(ABG < 20 dB = 83/67%; average follow-up time: 1.3/4.1 years) (Fig. 4).
The statistical comparison of the data of the two groups
revealed the following results. There was no signicant difference
in preoperative ABGs (p = 0.058) or in the best postoperative ABGs
(p = 0.499). Similarly, no signicant difference was revealed
between the two groups in the last postoperative ABGs measured
(p = 0.298) or in the nal hearing improvement (p = 0.193). As
mentioned earlier, hearing outcomes deteriorate by some dBs with

time even after successful tympanoplasty. Theoretically, we would

assume that this deterioration is more expressed in patients with
cleft palate due to their vulnerable Eustachian tube function.
However, the statistical analysis of our data did not reveal any
signicant difference between the two groups in the extent of
postoperative ABG deterioration (p = 0.117).
4.1.3. Separate analyses
We performed separate analyses of tympanoplasties with intact
ossicular chain and the surgeries including short- or long-type
columella ossiculoplasty.
4.1.3.1. Intact ossicular chain. Altogether 61 NoCleft ears were
operated on with intact ossicular chain, out of which 43 ears had
enough follow-up audiograms for software evaluation. The
preoperative ABG of 23.88  10.68 dB improved to 11.83  6.9 dB
(best postoperative value), which means an average best improvement of 12.05 dB. Average ABG values deteriorated by 1.38 dB with
time therefore the nal improvement achieved to date is 10.67 dB
(ABG <20 dB = 92/90%; average follow-up time: 1.4/2.3 years).
Only 5 Cleft ears were operated on with intact ossicular chain
at the end of the surgery. The average preoperative ABG was
29.85  7.16 dB the best postoperative improvement was 16.23 dB
and the nal improvement achieved so far is 13.88 dB
(ABG < 20 dB = 100/80%; follow-up time: 0.6/1.9 years).

Fig. 3. Distribution of preoperative and best postoperative ABG values; cholesteatoma.

G. Kopcsanyi et al. / International Journal of Pediatric Otorhinolaryngology 79 (2015) 698706

703

Fig. 4. Distribution of best/last postoperative ABG results in cholesteatoma.

The statistical comparison of the identical data did not reveal

signicant differences in any of the parameters: for preoperative
ABG p = 0.097, for postoperative ABG p = 0.259, for the last ABG
measured p = 0.176, for postoperative ABG deterioration p = 0.311,
for the nal improvement p = 0.247.
4.1.3.2. Columella ossiculoplasties. The following section compares
the results of surgeries with columella ossiculoplasties regardless
of the type of the ossiculoplasty (short- or long-type).
Ossicular chain reconstruction was carried out in 85 NoCleft
and 13 Cleft ears.
In the NoCleft group, the average preoperative ABG was
34.61  11.15 dB, which decreased to 17.76  11.32 dB after
columella ossiculoplasty (best postoperative results), which means
a best postoperative improvement of 16.85 dB. On the last
measurements the average ABG was 21.82  13.39 dB, the average
improvement achieved to date is 12.79 dB. ABG deterioration with
time was 4.06 dB (ABG < 20 dB = 70/57%; respective follow-up
times: 2.5/4.5 years).
In the Cleft group columella ossiculoplasty was performed on
13 ears. The average preoperative ABG was 35.66  13.28 dB, which

improved to 14.98  7.89 dB considering the best postoperative

results. The best average improvement was 20.68 dB. ABG deterioration with time was 5.79 dB, therefore the average nal improvement
compared to the preoperative values is 14.89 dB so far
(ABG < 20 dB = 83/68%; follow-up time: 1.6/5.2 years).
The statistical comparison of the identical data of ears with
columella ossiculoplasty did not reveal any signicant difference
in any of the parameters. For preoperative ABG p = 0.278, for the
best postoperative ABG p = 0.271, for the last ABG p = 0.458, for
postoperative deterioration p = 0.231 and for the nal hearing
improvement p = 0.292 (Fig. 5).
4.1.3.3. Short- and long-type ossiculoplasties. We examined the
results of ears with short- and long-type ossiculoplasty comparing
the outcomes in the NoCleft and Cleft groups.
4.1.3.3.1. Short-type columella ossiculoplasties. In the NoCleft
group 59, in the Cleft group 11 short-type columella ossiculoplasties were performed (see Table 1).
No signicant difference was revealed in any of the parameters
with statistical methods. For preoperative values p = 0.184, for the
best postoperative values p = 0.396, for the last ABG measured

Fig. 5. Comparison of average preoperative and last postoperative ABG values of ears with columella ossiculoplasty; cholesteatoma.

704

G. Kopcsanyi et al. / International Journal of Pediatric Otorhinolaryngology 79 (2015) 698706

Fig. 6. Average preoperative and last postoperative ABG in cholesteatoma patients.

p = 0.256, for postoperative deterioration p = 0.235 and for the nal

improvement p = 0.376.
4.1.3.3.2. Long-type columella ossiculoplasties. In general, the outcomes of long-type ossiculoplasty are far below those of short-type
ossiculoplasty. The stapes footplate novomembrane columella
cannot physiologically replace the original chain consisting of
joints. The chronic suppurative process (with or without cholesteatoma), which destroys the superstructure of the stapes,
probably impairs the mobility of the stapes footplate as well. This
is supported by the fact that long-type columella ossiculoplasty
with stapedectomy leads to better outcomes.
Long-type columella ossiculoplasty was performed in 25
NoCleft and only 2 Cleft ears (see Table 1).
The statistical comparison of the identical parameters did not
reveal any signicant difference in this respect, either. For
preoperative ABG p = 0.463, for the best postoperative ABG
p = 0.266, for the last postoperative ABG p = 0.130, for postoperative deterioration p = 0.181 and for the nal hearing improvement
p = 0.419.

5. Discussion
In our material the incidence of cholesteatoma in children with
cleft palate is 3.2%, which corresponds to the data published by
Vincenti et al. [5].
Nearly all publications on cholesteatoma mention the aggressive behaviour of childhood cholesteatoma and the fact that in
children the incidence of residual and recurrent cholesteatoma is
higher than in adults. Quaranta et al. (1986) gave a histopathological explanation for this aggressive behaviour, namely that the
number of perimatrix mononuclear inammatory elements is
higher in childhood cholesteatomas [19]. Bujia et al. found a higher
proliferation index of childhood cholesteatoma compared to that
of adults with the analysis of a monoclonal antibody marker
responsible for cell proliferation (MIB1). Dornelles et al. revealed
the hyperexpression of matrix metalloproteinase (MMP2), a
proteolytic enzyme, in childhood cholesteatoma [20]. These
processes promote the aggressive bone invasion of the cholesteatoma. A useful review of the corresponding literature is provided
by the publication of Nevoux et al. (Armand Trousseau Children
Hospital) [15].

Apart from the possible causes mentioned above, the starting

point of the cholesteatoma may also explain the higher incidence
of residual cholesteatoma observed in children after the surgery
[21]. In children, the most common starting point is the posterior
superior quadrant of the pars tensa. This creates favourable
conditions for the expansion towards the windows, to the
neighbouring retrotympanic area and to its recesses, which are
difcult to reach surgically [17,18].
We completely agree with the authors supporting the application
of pre-planned two-stage surgery in the case of closed techniques
[1,22]. Stage II or the second look (usually one year after stage I)
allows for the early detection and removal of residual cholesteatoma
(normally a very small pearl-like lesion) and creates better
conditions for the necessary columella ossiculoplasty (as for us,
we use the expression second look to denote safety inspection).
Adequate Eustachian tube function is one of the key determinants of the success of tympanoplasty. Its absence results in
invagination cholesteatoma and it is the most important factor in
the development of postoperative recurrent cholesteatoma. Before
surgery we must try every possible option for the restoration of
Eustachian tube function.
When it cannot be ensured preoperatively, Eustachian tube
dysfunction can be prevented by post- or intraoperative ventilation tube insertion [1,3]. Different authors mention different ages
between 4 and 14 years when the frequent upper respiratory tract
catarrhs of children normally cease and Eustachian tube function
becomes reliable [6,7,14].
In children with cleft palate, Eustachian tube dysfunction is
considered to be the result of the abnormal insertion of the tensor
veli palatine muscle [23,24]. In infants with cleft palate, the
incidence of insufcient Eustachian tube function is nearly 100%
[2427].
The different cleft palate repair techniques (tensor veli palatini
preservation, transection and transection with tensor tenopexy) do
not really inuence the necessity for grommet insertion [2729].
Nevertheless, in the majority of cases Eustachian tube function
recovers some years after the cleft palate surgery. Different studies
mention different ages between 5 and 14 years for the completion
of this process [6,26,27,30]. Our experience suggests that this age is
between 7 and 8 years.
The tympanoplasty of children without cleft palate also raises
the question of the age when middle ear ventilation is stable

G. Kopcsanyi et al. / International Journal of Pediatric Otorhinolaryngology 79 (2015) 698706

enough for the surgery to be successful. In this respect, different

studies recommend very different ages [6,7,31].
We consider the Valsalva manoeuvre the simplest method for
the routine evaluation of Eustachian tube function. Our experience
suggests that children are capable of understanding and performing the Valsalva manoeuvre since the age of 5. Of course, we do not
nd it necessary to wait until the age of 5 in cholesteatomatous
cases.
Contradictory opinions have been published about how much
we should rely on the Eustachian tube function of the contralateral
ear. Most authors attribute a good predictive value to intact
contralateral Eustachian tube function [14,23,31]. In our opinion, it
is not a coincidence that cholesteatoma, in the development of
which Eustachian tube dysfunction has a signicant role, develops
in the affected ear. Therefore, we do not think that intact
contralateral Eustachian tube function is a signicant predictive
factor in tympanoplasty outcomes. Tos suggests that Eustachian
tube function may improve by approximately 30% after tympanoplasty and 68% of Valsalva negative patients become Valsalva
positive [3].
Earlier studies report poor outcomes of the tympanoplasties of
cleft patients [6,7]. The low number of more recent studies
published on this topic reects a more optimistic view.
Vartiainen (1992) compared the outcomes of tympanoplasties
performed on 19 ears of 16 Cleft patients with the surgeries of 671
NoCleft patients and did not nd any signicant difference in
hearing results between the two groups. The average age of Cleft
patients was 16.5 years (241) at the time of the surgery, while
that of NoCleft patients was 37.8 years. The average follow-up
period was 6.5 years. In Cleft patients, he recommends the same
indications for surgery as in NoCleft patients but in the former
group he recommends strict lifelong follow-up [8].
Gardner and Dornhoffer (2002) carried out an age- and
procedure-matched comparison of the outcomes of tympanoplasties performed on 26 ears of 20 Cleft patients and the outcomes of
the same surgeries in 52 NoCleft patients. No signicant
difference was found between the two groups in hearing outcomes,
graft take rate or grommet insertion. The average age of the
patients at the time of the surgery was 24 years (255), the average
follow-up period was 25 months [9].
Both publications mentioned above discuss the tympanoplasties performed due to cholesteatomatous and non cholesteatomatous chronic otitis media in one common category.
Apart from our previously mentioned publication, two papers
have been published recently on the outcomes of type I
tympanoplasty in children with cleft palate and mesotympanic
perforation [32,33]. Both publications reported good anatomical
results, however, one of them reported signicantly poorer
functional outcomes in children with cleft palate [33].
There is only one recent publication from Parma that
exclusively discusses the tympanoplasty outcomes of children
with cleft palate and cholesteatoma. Based on this material (18
ears), the authors concluded that in most patients with cleft palate
closed techniques could be applied and the hearing results are
similar to those observed in the general population [5].
The graft take rate was 100% in our material, small perforation
only occurred in only one NoCleft ear.
Our audiological results were analyzed in detail for comparison.
Although the preoperative ABG was slightly worse in the Cleft
group, the difference was not signicant. Considering audiological
outcomes, no signicant difference was found between the two
groups in the other examined parameters including the best
postoperative ABG, the last postoperative ABG, postoperative
deterioration and nal hearing improvement.
The average follow-up time of our last ABG values was 4 and 4.1
years. No signicant difference was revealed in the nal hearing

705

improvement achieved so far, the values were 11.29 dB and

13.9 dB in favour of the Cleft group (Fig. 6).
Examining the extension of cholesteatoma that we operated on
we can conclude that in Cleft patients the involvement of the
atticus and antrum is more expressed than in the NoCleft group,
which means that more extensive cholesteatoma can be expected
in the former group of patients.
Residual cholesteatoma is partly a surgical error, on the other
hand it can also be attributed to the aggressive behaviour of
childhood cholesteatoma invading hardly reachable areas. Several
modern CT and MRI techniques are available for the early detection
and follow-up of residual cholesteatoma (diffusion-weighted
sequence, non-echo-planar diffusion-weighted imaging), a good
summary of which can be found in the publication by Nevoux et al.
[15]. However, some authors prefer clinical and audiological
examinations to regular postoperative imaging techniques
(Haynes) [22].
We found residual cholesteatoma in 33% of NoCleft and in 27%
of Cleft patients during second look surgeries and revisions. For
the complete patient material these percentages were 18.6% and
20%.
On the other hand, explicit difference was found in the rates of
recurrent cholesteatoma between the two groups. During revision,
recurrent processes were found in 5% of NoCleft and 13% of Cleft
patients, however, the difference is statistically not signicant
(p = 0.219).
Signicant difference was revealed in the necessity for
grommet insertion between the two groups. This means an 8fold difference in the rate of grommet insertions.
Similarly, the need for the application of open techniques
showed signicant differences between the two groups.
We should consider the fact that after the tympanoplasty of
patients with cleft palate Eustachian tube obstruction may develop
rapidly even after years of adequate Eustachian tube function and
good anatomical and functional results, which immediately
destroys the achieved outcomes and we have to resort to revision
with open techniques.
The main advantage of patient management in the frameworks
of the Cleft Palate Team is that this way Cleft patients are more
visible due to the common supervision of different disciplines than
patients without cleft palate. This allows for the earlier and more
frequent detection of new Eustachian tube dysfunctions and their
adequate prevention as well.
6. Conclusions
Considering the original question of our publication, we can
conclude that the achievable short- and long-term audiological
outcomes of tympanoplasty in children with cleft palate and
cholesteatoma do not differ signicantly from those observed in
the general child population. However, the price of these results is
the signicantly more frequent grommet insertion with much
more frequent follow-up visits.
The latter is provided by patient care in the Cleft Palate Team
established in our department in 1996.
However, we have to accept that in some cases Eustachian tube
dysfunction caused by the underlying disease (cleft palate) takes
over and we have to resort to open techniques. The facts that in the
majority of patients with cleft palate and cholesteatoma Eustachian tube function can be restored and closed techniques applied
in the general population usually lead to denitive recovery with
advantageous swim-proof ears, give reason for optimism.
Conict of interest
None declared.

706

G. Kopcsanyi et al. / International Journal of Pediatric Otorhinolaryngology 79 (2015) 698706

Acknowledgement
I would like to express my gratitude to Assistant Professor Dr.
Sandor Herman for his valuable help in the eld of statistical
analysis.
References
[1] M. Bauer, Tympanoplastica/Tympanoplasty, Budapest, Medicina, 2003.
[2] H.A.T. Bailey, Symposium: contraindications to tympanoplasty. Part I. Absolute
and relative contraindications, Laryngoscope 86 (1976) 6769.
[3] M. Tos, Tubal function and tympanoplasty, J. Laryngol. Otol. 88 (1974) 113124.
[4] E. Olszewska, M. Wagner, M. Bernal-Sprekelsen, J. Ebmeyer, S. Dazert, H. Hildmann, et al., Etiopathogenesis of cholesteatoma, Eur. Arch. Otorhinolaryngol. 261
(1) (2004) 624.
[5] V. Vincenti, F. Marra, B. Bertoldi, D. Tonni, M.S. Saccardi, S. Bacciu, et al., Acquired
middle ear cholesteatoma in children with cleft palate: experience from 18
surgical cases, Int. J. Pediatr. Otorhinolaryngol. 78 (6) (2014) 918922.
[6] D. Plester, When not to do middle ear surgery, J. Laryngol. Otol. 96 (1982) 585590.
[7] R. Bellucci, Dual classication of tympanoplasty, Laryngoscope 83 (1973) 17541758.
[8] E. Vartiainen, Results of surgery for chronic otitis media in patients with a cleft
palate, Clin. Otolaryngol. 17 (1992) 284286.
[9] E. Gardner, J.L. Dornhoffer, Tympanoplasty results in patients with cleft palate: an
age- and procedure-matched comparison of preliminary results with patients
without cleft palate, Otolaryngol. Head Neck Surg. 126 (2002) 518523.
[10] G. Kopcsanyi, O. Vincze, J. Pytel, Retrospective analysis of tympanoplasty in
children with cleft palate: a 22-year experience: I. Mesotympanic (non-cholesteatomatous) cases, Int. J. Pediatr. Otorhinolaryngol. 78 (4) (2014) 645651.
[11] I. Gerlinger, G. Rath, I. Szanyi, J. Pytel, Myringoplasty for anterior and subtotal
perforations using KTP-532 laser, Eur. Arch. Otorhinolaryngol. 263 (9) (2006)
816819.
[12] M. Bauer, J. Pytel, I. Vona, I. Gerlinger, Combination of ionomer cement and bone
graft for ossicular reconstruction, Eur. Arch. Otorhinolaryngol. 264 (11) (2007)
12671273.
[13] Committee on Hearing and Equilibrium guidelines for the evaluation of results of
treatment of conductive hearing loss, Otolaryngol. Head Neck Surg. 113 (3)
(1995) 186187.
[14] J. Dornhoffer, Cartilage tympanoplasty: indications, techniques, and outcomes in
a 1000-patient series, Laryngoscope 113 (2003) 18441856.
[15] J. Nevoux, M. Lenoir, G. Roger, F. Denoyelle, H. Ducou Le Pointe, E.N. Garabedian,
Childhood cholesteatoma, Eur. Ann. Otorhinolaryngol. Head Neck Dis. 127 (4)
(2010) 143150.
[16] H.A. Bailey Jr., Symposium: methods of reconstruction in tympanoplasty. II.
Maintenance of the anterior sulcus-tympanic membrane relationships in tympanoplastic surgery, Laryngoscope 86 (2) (1976) 179184.

[17] M. Tos, Recurrence and the Condition of the Cavity After Surgery for Cholesteatoma Using Various Techniques, Kugler and Ghedini, 1989.
[18] V. Darrouzet, J.Y. Duclos, D. Portmann, J.P. Bebear, Preference for the closed
technique in the management of cholesteatoma of the middle ear in children:
a retrospective study of 215 consecutive patients treated over 10 years, Am. J.
Otol. 21 (4) (2000) 474481.
[19] A. Quaranta, L. Resta, A. Santangelo, Otomastoid cholesteatoma in children:
histopathological ndings, Int. J. Pediatr. Otorhinolaryngol. 12 (2) (1986)
121126.
[20] C. Dornelles Cde, S.S. da Costa, L. Meurer, L.P. Rosito, A.R. da Silva, S.L. Alves,
Comparison of acquired cholesteatoma between pediatric and adult patients, Eur.
Arch. Otorhinolaryngol. 266 (10) (2009) 15531561.
[21] S.E. Stangerup, D. Drozdziewicz, M. Tos, A. Hougaard-Jensen, Recurrence of attic
cholesteatoma: different methods of estimating recurrence rates, Otolaryngol.
Head Neck Surg. 123 (3) (2000) 283287.
[22] T.R. McRackan, W.M. Abdellatif, G.B. Wanna, A. Rivas, N. Gupta, M.S. Dietrich,
et al., Evaluation of second look procedures for pediatric cholesteatomas, Otolaryngol. Head Neck Surg. 145 (1) (2011) 154160.
[23] L.D. Hartzell, J.L. Dornhoffer, Timing of tympanoplasty in children with chronic
otitis media with effusion, Curr. Opin. Otolaryngol. Head Neck Surg. 18 (6) (2010)
550553.
[24] W.L. Myerhoff, D.A. Shea, C.A. Foster, Otitis media, cleft palate, and middle ear
ventilation, Otolaryngol. Head Neck Surg. 89 (1981) 288293.
[25] W.J. Doyle, J.S. Reilly, L. Jardini, S. Rovnak, Effect of palatoplasty on the function of
the Eustachian tube in children with cleft palate, Cleft Palate Craniofac. J. 23
(1986) 6368.
[26] S. Goudy, D. Lott, J. Candy, R.J. Smith, Conductive hearing loss and otopathology in
cleft palate patients, Otolaryngol. Head Neck Surg. 134 (2006) 946948.
[27] R.K. Sharma, V. Nanda, Problems of middle ear and hearing in cleft children,
Indian J. Plast. Surg. 42 (Suppl.) (2009) S144S148.
[28] P.K. Chaudhuri, E. Bowen-Jones, An otorhinological study of children with cleft
palates, J. Laryngol. Otol. 92 (1978) 2940.
[29] R.L. Flores, B.L. Jones, J. Bernstein, M. Karnell, J. Canady, C.B. Cutting, Tensor veli
palatini preservation, transection, and transection with tensor tenopexy during
cleft palate repair and its effects on Eustachian tube function, Plast. Reconstr.
Surg. 125 (2010) 282289.
[30] T.L. Smith, D. Diruggiero, K. Jones, Recovery of Eustachian tube function and
hearing outcome in patients with cleft palate, Otolaryngol. Head Neck Surg. 111
(1994) 423429.
[31] A.L. James, B.C. Papsin, Ten top considerations in pediatric tympanoplasty,
Otolaryngol. Head Neck Surg. 147 (6) (2012) 992998.
[32] A.M. Metrailer, M.D. Cox, J. Sunde, L.D. Hartzell, P.C. Moore, J.L. Dornhoffer,
Cartilage tympanoplasty in children with cleft palate repair, Otol. Neurotol. 35
(8) (2014) 14711473.
[33] E. Harterink, N. Leboulanger, S. Kotti, E.N. Garabedian, F. Denoyelle, Results of
myringoplasty in children with cleft palate: a patient-matched study, Otol.
Neurotol. 35 (5) (2014) 838843.