Otolaryngol Clin N Am

39 (2006) 1081–1094

Congenital Cholesteatoma: Theories,

Facts, and 53 Patients
Marc Bennett, MD*, Frank Warren, MD,
Gary C. Jackson, MD, David Kaylie, MD
The Otology Group, Otolaryngology Head & Neck Surgery,
Vanderbilt University, 300 20th Avenue North,
Suite 502, Nashville, TN 37203, USA

Although the first written report of congenital cholesteatoma (CC) was

over 100 years ago in 1885, its true incidence, etiology, and pathogenesis still
remains a controversial. CC of the temporal bone can be found intradural,
most commonly at the cerebropontine angle, or extradural in the middle ear
or mastoid. Congenital cholesteatoma of the middle ear was first described
by Howard House [1] in 1953. Later, Derlacki and Clemis [2] described six
cases of CC and established the clinical criteria for the diagnosis. These in-
clude a pearly white mass medial to an intact tympanic membrane, a normal
pars tensa and flaccida, and no history of otorrhea, perforation, or previous
otologic procedure. Levenson revised the criteria by adding that previous
bouts of otitis media or effusion should not be exclusion criteria.
Once thought to be relatively rare, CC of the middle ear is now thought
to be on the rise, and accounts for 2% to 5% of all cholesteatomas [3]. There
are multiple theories to the pathophysiology of congenital cholesteatomas of
the middle ear. Proposed mechanisms are inclusion, migration, or invasion
of squamous epithelium, epithelial rests from faulty embryogenesis, or meta-
plasia of normal epithelium. Patients with CC may present with a variety of
complaints including conductive hearing loss, but the most common presen-
tation is an asymptomatic white mass medial to an intact tympanic
membrane. Early detection of CC is critical, limiting the size of the retro-
tympanic mass and reducing the risks and complications from surgery.
Treatment remains surgical removal. This study will discuss the diagnosis,
classification, treatment, and theories of pathogenesis, as well as report
the 53 cases of CC treated at our institution.

* Corresponding author.
E-mail address: marc.bennett@vanderbilt.edu (M. Bennett).

0030-6665/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.otc.2006.08.001 oto.theclinics.com
1082 BENNETT et al

Incidence
The true incidence of CC is difficult to determine. Initially thought to be
rare, the incidence seems to be on the rise [4]. The incidence of CC of the mid-
dle ear is estimated to be between 1% to 5% of all cholesteatomas in most
published series [4–6]. Earlier treatment of otitis media and allergies is reduc-
ing the number of acquired cholesteatomas, and consequently increasing the
percentage of CC. There are also many reasons that the number of reported
cases of CC has increased over the last 30 years. Heightened awareness of the
condition by pediatricians and otolaryngologists has lead to earlier diagnosis
and intervention, avoiding the tympanic membrane perforation, which
would eventually occur as the natural progression of the disease [7]. This
would preclude the diagnosis of CC. Routine audiologic screening has also
identified children with conductive hearing losses at early ages. Workup of
these children with better otoscopic equipment may identify masses medial
to the tympanic membrane that previously would have been unseen. In ad-
dition, incidental masses are occasionally seen on CT scans of the temporal
bone ordered in children with a conductive hearing loss.

Presentation and growth

The clinical presentation of any ear mass depends on its size, location,
and histology. Earlier diagnosis decreases the overall size of the CC and re-
duces the likelihood of ossicular erosion. The most common presentation of
a CC is a white retrotympanic middle ear mass [5], although they may be
discovered incidentally during the time of routine otologic evaluation or
during a myringotomy. Improved American health care and preventive
medicine for children has lead to a substantially earlier diagnosis of CC at
an average age of 4 to 5 years old [5,7].
The most common location of CCs is the anterior superior quadrant of
the tympanic membrane, followed by the posterior–superior quadrant
[5,7]. Lesions discovered at later ages are more likely to be located in the
posterior mesotympanum, as continued growth of the cholesteatoma is gen-
erally in a posterior direction of spread. The growth follows a natural course
as described by Koltai [8]. Although growth can proceed inferiorly toward
the hypotympanum, it more commonly extends posteriorly along the medial
surface of the ossicles. Once in the posterior superior mesotympanum, the
expansion can involve the incudostapedial joint or stapes superstructure,
but usually spares the footplate. Continued expansion proceeds toward
the facial recess, sinus tympani, and eventually the mastoid air cells. It is
rare for the cholesteatoma to invade the otic capsule bone or labyrinth.
As growth continues, symptoms become more common. Both large ante-
rior and posterior lesions present with conductive hearing loss for different
lesions. Posterior located lesions affect the ossicles much more frequently,
causing a conductive hearing loss from ossicular movement impairment or
 CONGENITAL CHOLESTEATOMA 1083

discontinuity. Anterior lesions may compromise the function of the Eusta-

chian tube and cause a conductive hearing loss through middle ear effusion.
Otalgia and otorrhea are rare, but nearly 50% of patients describe episodes
of previous otitis media [9]. If cholesteatoma extends to invade the laby-
rinth, patients may suffer from vertigo or sensorineural hearing loss. Despite
early reports describing a high percentage of facial nerve palsy at presenta-
tion, facial nerve dysfunction is relatively uncommon from middle ear CC
[10]. In fact, facial paresis mandates exclusion of malignancies, metastases,
facial nerve neuromas, and other diagnoses.

Histology
CC or epidermoid cyst is a stratified squamous epithelial lined cyst filled
with keratin debris. Like acquired cholesteatoma, the cyst forms as the result
of progressive desquamation of the epithelium. The congenital form of choles-
teatoma is indistinguishable by histology from the acquired form; therefore, it
is the clinical picture that is important in distinguishing the two entities.

Imaging
Pediatric patients with conductive hearing loss and a normal otomicro-
scopic examination require radiologic evaluation to evaluate for the presence
of middle ear anomalies like CC. Because plain radiographs are nonspecific,
high-resolution CT and MRI are the most commonly used imaging modalities
[11]. CT is generally used as the first imaging modality because of its superior
bony definition. CT cannot only confirm the location of a middle ear mass, but
can accurately determine the size of the lesion. CC is usually seen as a hypo-
dense expansile lesion, round to oval in shape, with well-defined margins
that do not enhance with contrast. The lack of enhancement helps distinguish
the cholesteatoma from other lesions like neuromas, glomus tumors, sarco-
mas, or meningiomas. Unlike patients with chronic ear disease, the mastoid
air cells are usually well aerated and nonsclerotic [12].
MRI further enhances the evaluation by distinguishing the cholesteatoma
from other middle ear masses such as neuromas, adenomas, schwannomas,
or metastasis. On T1 weighted images, CC appears as a homogenous lesion
that is hypointense to brain, but can also appear isointense [12]. On T2
weighted images, signal intensity is usually high like cerebral spinal fluid
(CSF). There is usually no enhancement with gadolinium. Diffusion
weighted imaging can be used to help distinguish cholesteatoma from other
cystic masses as well [11].

Staging
CC of the middle ear is often staged by its location and relationship with
surrounding structures. Derlacki and Clemis [13] are credited with the first
1084 BENNETT et al

classification system for CC. They classified the lesions into petrous pyra-
mid, mastoid, or tympanic. Recently, two staging systems for CC of the
middle ear have been suggested. The first, by Potsic [14], suggested the
following stages:
Stage 1: Single quadrant with no ossicular or mastoid involvement
Stage 2: Multiple quadrants with no ossicular or mastoid involvement
Stage 3: Ossicular involvement but no mastoid involvement
Stage 4: Mastoid extension
In their experience, 40% of the cases were stage I, 14% were stage II,
23% were stage III, and 23% were stage IV. There was a statistically signif-
icant association between the stage, hearing outcomes, and residual disease.
Nelson suggested classification of CC into three categories [15]:
Type 1: Mesotympanum with no incus or stapes erosion
Type 2: Mesotympanum or attic with ossicular erosion but no mastoid
extension
Type 3: Mesotympanum with mastoid extension
In their experience, 15% were type I, 59% were type II, and 26% were
type III. Again, recurrence rates were correlated with the clinical stage.
Although no type I patients recurred, 34% and 55% of patients with type
II and III lesions respectively recurred.
The purpose of staging systems should be to (1) aid the clinician in
preoperative planning of treatment, (2) I ndicate prognosis, (3) facilitate
exchange of information between different clinicians, and (4) evaluate the
results of treatment.
Both Potnic’s and Nelson’s staging systems accomplish these goals. The
major difference between systems is that Potnic’s system separates Nelson’s
type I lesions into two categories, depending on whether it occupies more
than one quadrant of the tympanic membrane. Both systems have excellent
correlation between stage and recurrence.

Surgery
Surgical management of the CC requires complete removal of the matrix
or exteriorization to prevent recurrence. The goal of surgery is complete
removal of disease with optimal hearing outcomes. For complete removal,
intraoperative dissection must be complete around the matrix, often
mandating a mastoidectomy for better visualization.
As expected, the majority of patients with isolated anterior mesotympa-
num lesions can be adequately approached through a standard tympano-
plasty. Posterior lesions involving the ossicles often require a mastoidectomy
for circumferential visualization of the cholesteatoma. Most surgeons advo-
cate an intact canal wall mastoidectomy because it allows for better hearing
 CONGENITAL CHOLESTEATOMA 1085

reconstruction and eliminates both water restrictions and the periodic

debridements that can be necessary after a canal wall down procedure [16].
However, a canal wall down mastoidectomy is performed if the patient
had [17]:
(1) unreconstructible external auditory canal defects
(2) labyrinthine fistulas
(3) poor health
(4) poor compliance
In CC patients, it is rarely indicated as the patients are usually in good
health, the cholesteatoma by definition cannot erode into the external audi-
tory canal, and the lesions rarely affect the otic capsule bone. Middle ear
reconstruction is used if ossicles are eroded or removed to allow complete
excision of the lesion. Cholesteatoma matrix is frequently found close to
the facial nerve, but rarely is mobilization of the nerve needed for complete
excision of matrix. Middle cranial fossa approaches may be required for
perigeniculate lesions [18].
There is some controversy over whether there should be a planned second
look in CC. Limited anterior CC not involving the ossicles has an extremely
low rate of residual disease and, therefore, a single stage is frequently suffi-
cient. However, if there is any concern about residual disease, a ‘‘second
look’’ is certainly warranted. For more extensive lesions the indications
for a second-look procedure are less clear and must be individualized for
each patient. Recurrence in a canal wall down mastoidectomy is fairly con-
spicuous, and therefore serial exams are all that is necessary. Long-term fol-
low-up is important for all patients. A routine second look is no longer the
general policy at our institution. Patients are reexplored for worsening hear-
ing or obvious recurrence.

Pathophysiology
The accepted cause of CC remains controversial. The competing theories
of pathogenesis fall into four categories: implantation, invagination or inva-
sion, metaplasia, and epithelial rest formation.

Epithelial rest
The most commonly accepted and quoted theory on the origin of CC is
the epithelial rest theory. This theory is based on Teed’s initial observation
of an epidermal structure found in a 5-month human fetus in ‘‘the dorsal
lateral pole of the tympanum, just medial to the neck of the malleus’’ [19].
Paparella [20] explained these rests as ectodermal implants in the fusion
plates between the first and second branchial arches that appear around
10 weeks at the junction of the first branchial cleft and pouch systems in
the anterior mesotympanum near the geniculate ganglion. The structure is
1086 BENNETT et al

distinct from surrounding tissue, and is located at the transition from simple
cuboidal epithelium of the tympanic cavity to the pseudostratified epithe-
lium of the anterior mesotympanum and Eustachian tube [21]. The exact
function of the rests is unknown, but Levenson [21] postulated that it aided
in middle ear and tympanic membrane development. Initially dormant, it
undergoes rapid proliferation before resorption around 33 weeks’ gestation.
CC is thought to form if resorption is incomplete. Levenson [21] proposed
that the epidermal rests fail to undergo involution because of chronic
irritation.
Michaels [22] histologically confirmed these rests of epithelial cells in
54% of the fetal temporal bones examined and claimed their persistence
led to CCs.
McGill [23] further confirmed their presence, but suggested that interme-
diate forms of the epithelial rests between 33 weeks and early childhood
must be found. To date, there is no documentation of epithelial rests beyond
33 weeks of gestation up to birth.
Although the middle ear epidermoid provides a satisfactory source for
the CC, it does not explain the existence of CCs found outside the anterior
superior quadrant of the tympanic membrane where these rests are tradi-
tionally found. However, recent studies have found that the locations of ep-
ithelial rests vary. Although most are found in the anterosuperior annular
region, they can also be in seen posterosuperior, posteroinferior, and ante-
roinferior regions of the lateral wall of the tympanic cavity.

Invaginaton
Aimi was the first to suggest migration of normal squamous epithelium
from the external canal through the tympanic ring and into the middle
ear as the source of CC. His theory holds that small inflammatory injury
of the tympanic membrane near the neck of the malleus causes invagination
of the epithelium that progresses to form a CC. This event may occur in
utero or during childhood development. The retracted tympanic membrane
is adherent to the malleus or incus. The eardrum is loosed from the ossicles
and torn, leaving a small remnant of keratinized epithelium adherent to the
bone along with or without a small perforation. The perforation heals but
the retained epithelium forms a cholesteatoma over time. Alternatively,
Reudi [24] postulated that external auditory canal ectoderm may penetrate
the tympanic membrane in utero and migrate into the middle ear. Small in-
flammatory injuries to the tympanic membrane produce small perforations
in the epithelium through which squamous epithelium invades into the mid-
dle ear.
Although there is no histopathologic evidence of this theory, it provides
a plausible hypothesis for CC. It would help explain lesions not located in
the anterior–superior quadrant of the tympanic membrane as the invagina-
tion could occur anywhere in the tympanic membrane.
 CONGENITAL CHOLESTEATOMA 1087

Implantation
Friedberg and Sheehy were the first to suggest that CC was the result of
implantation of squamous epithelium to the middle ear either from trauma
or an unrecognized and subsequently healed tympanic membrane perfora-
tion. This may account for some cases of cholesteatoma that appear to
meet the criteria to be classified as CC. This theory also helps explain
why lesions can be isolated in many different sites. However, because this
theory requires an insult to the tympanic membrane, it by definition
excludes the classification of CC.

Metaplasia
Squamous metaplasia of the inflamed middle ear epithelium is not un-
common, and can occasionally be seen even in nonpathologic ears [25]. Re-
cently, Sade [25] has found squamous metaplasia in cases of chronic otitis
media. In addition, other studies have shown that retinoic acid depletion
can induce squamous differentiation of middle ear epithelium in cultures
[26]. If metaplastic squamous epithelium becomes keratinizing, a cholestea-
toma may form from the accumulation of keratin. Serial sections of middle
ear mucosa have revealed keratin production in cells not connected to sur-
face epithelium. However, Friedberg noted the location of metaplasia
would be random and not explain the high frequency of lesions in the
anterior superior tympanic cavity; however, this area near the eustacian
tube may be more prone to inflammation and therefore have a higher
rate of CC.

Current series
Material and methods
Chart review
Between March 1971 and December 2003, over 3000 chronic ear surgeries
were performed at our institution. A computerized otologic database was
used to identify 53 patients who had a history of CC. Cholesteatoma was
considered to be congenital if there was no history of otologic surgery,
otorrhea, or perforation, and no tympanic membrane abnormality on
examination.
Charts of these patients were reviewed, and patients were classified ac-
cording to the following data: age of patient, location and extent of choles-
teatoma, type of surgery performed, audiologic outcomes, development of
recurrent perforation or cholesteatoma, and intraoperative complications.
Disease was defined by size as occupying the middle ear, Eustachian tube, ep-
itympanum, ossicles, or mastoid. Recurrence was defined visible disease or
cholesteatoma visualization at subsequent surgical procedures or clinic visits.
1088 BENNETT et al

Demographics
Ninety patients were identified in the database with the diagnosis of CC.
Unfortunately, 37 had previous procedures performed elsewhere before be-
ing evaluated at our institution. They were excluded because it could not be
determined conclusively whether they had a CC. Fig. 1 shows the number of
patients in each pediatric age group. The average age of presentation was
4.7  3.1 years old. There were 22 female patients and 31 males. Thirty-two
were left ears and 21 right ears. No patient had bilateral CC. In our series,
CC was 2.48% of all primary cholesteatomas, 1.80% of all cholesteatoma
surgeries, and 8.31% of cholesteatomas in patients less than 18 years old.

Disease location
Table 1 lists the general location of our patients. Twenty-nine patients
had disease limited to the middle ear or epitympanum without mastoid in-
volvement. Twenty-four patients had cholesteatoma extension into the mas-
toid, while 12 patients had disease extension into the Eustachian tube.
Twelve patients had ossicular erosion requiring reconstruction. Only two
of these patients had disease extension into the mastoid. We classified our
patients according to Nelson’s staging and found 17 patients with type I dis-
ease, 12 patients with type II, and 24 patients with type III.

Surgeries performed
Surgeries were directed at removing cholesteatoma and optimizing hear-
ing for all patients. Table 2 lists the procedures performed and the relative
percentages. Twenty-four patients with extension into the mastoid under-
went a complete mastoidectomy. In addition, 23 patients with posterior
mesotympanic disease with inadequate transcanal exposure underwent

14
Series1

12
Number of Patients

10

0
1 2 3 4 5 6 7 8 9 10 11 >12
Age

Fig. 1. Age at presentation.

 CONGENITAL CHOLESTEATOMA 1089

Table 1
Location of cholesteatoma
Nelson Number of
Site of cholesteatoma stage patients Percentage
Middle ear and eustacian tube I 0 0.0
Eptiympanum only I 0 0.0
Tympanic membrane I 1 1.9
Mastoid only I 2 3.8
Middle ear, epitympanum, eustacian tube I 2 3.8
Middle ear and epitympanum I 4 7.5
Middle ear only I 8 17.0
Middle ear and epitympanum with ossicular invt II 4 7.5
Middle ear with ossicular involvement II 8 17.0
Mastoid, middle ear, epitympanum, eust. tube III 14 18.9
Mastoid, middle ear, epitympanum III 10 22.6

mastoidectomy procedures for better visualization. Twelve patients had dis-

ease involving the middle ear ossicles and requiring reconstruction. Seven
patients had erosion of the incus that required partial ossicular chain recon-
struction prosthesis (PORP) with cartilage overlay. Two of these patients
each underwent partial ossicular chain prosthesis reconstruction with
a Black PORP, hydroxyapatite PORP, or Kurz prosthesis. One patient
had an incus interposition. Five patients had eroded stapes superstructures
requiring total ossicular chain reconstruction prosthesis (TORP) with carti-
lage overlay. Two patients underwent successful revision surgery for tym-
panic membrane perforation.

Audiologic data
Audiologic data was analyzed for the patients using preoperative and
postoperative, and long-term (O1 year) pure-tone average air–bone gaps
(PTA-ABG) obtained from four frequencies (500, 1000, 2000, and 3000
Hz) according to AAO-HNS guidelines and listed in Table 3. Hearing re-
sults were also described by the stage of the cholesteatoma. For all patients,
regardless of the stage or surgery type, the average preoperative, postoper-
ative, and last PTA-ABG were compared. The results showed the difference

Table 2
Procedures performed
Procedures performed Number of patients Percentage
Tympanoplasty no OCR 1 1.9
Tympanoplasty with OCR 1 1.9
Tympanoplasty with mastoidectomy and OCR 11 20.8
Tympanoplasty with mastoidectomy no OCR 36 67.9
Tympanoplasty with modified radical mastoidectomy 4 7.5
Abbreviation: OCR, ossicular chain reconstruction.
1090 BENNETT et al

Table 3
Hearing outcomes
Preoperative Postoperative Most recent
All patients 24.8  13.4 30.3  15.5 31.6  16.3
No OCR 24.0  14.0 30.5  16.1 31.0  17.4
Stage I 16.6  12.9 25.0  12.7 27.2  16.0
Stage II 30.6  15.6 27.1  16.0 35.9  12.3
Stage III 32.8  12.8 32.6  15.3 35.1  17.0
Abbreviation: OCR, ossicular chain reconstruction.

between preoperative and both postoperative and last PTA-ABG was statis-
tically worse using the standard paired t-test (P ¼ 0.001). The hearing results
of patients undergoing an ossicular chain reconstruction were also analyzed.
There was no significant difference between preoperative, postoperative, and
last PTA-ABG using the standard paired t-test (P ¼ 1.1). For patients with
intact ossicular chains, the postoperative and last PTA-ABG were statisti-
cally better than preoperative PTA-ABG using the standard paired t-test
(P ¼ 0.001). Twenty-one patients did not have long-term follow-up or serial
audiograms. The hearing results by stages is shown in Table 3. Patients with
stage 3 disease had worse hearing than stages 1 or 2.

Surgical findings
Thirteen patients had intraoperative findings listed in Table 4. Although
eight patients had evidence of facial nerve exposure, no patients had injuries
to the facial nerve. In addition, facial nerve exposure was not related to dis-
ease extent. Three patients with mastoid extension of cholesteatoma had du-
ral exposure from the cholesteatoma, but no CSF leaks were encountered.
There was one case of oval window fistula, but no other labyrinthine fistu-
las. There was one dehiscent carotid artery seen intraoperatively without
complication.

Table 4
Intraoperative complications
Intraoperative complications Number of patients Percentage
Facial nerve injury 0 0.0
Dural exposure 3 5.7
Oval window fistula 1 1.9
Facial nerve exposure 8 13.2
Semicircular canal fistula 0 0.0
Dehiscent carotid artery 1 1.9
Death 0 0.0
CSF Leak 0 0.0
Fixed footplate 0 0.0
Abbreviation: CSF, cerebral spinal fluid.
 CONGENITAL CHOLESTEATOMA 1091

Long-term complications
Only 6 of the 53 patients had long-term complications from the CC sur-
gery as listed in Table 5. Four patients had tympanic membrane perforation
in the operated ear. Two elected to undergo revision surgery and were suc-
cessfully closed. One patient had delayed facial paresis (House-Brackmann
VI), which resolved with steroids. One patient had a prosthesis extrusion
with healed tympanic membrane. No patients had recurrence of their
cholesteatoma.

Discussion
Although considerable controversy exists over the existence of CC, there
exists a set of pediatric patients with normal eustachian tube function and
normal tympanic membranes with a retrotympanic cholesteatoma. Congen-
ital and acquired cholesteatomas share many similarities, but key differences
in the pathophysiology and location of these lesions mandate separate treat-
ment protocols.
Preoperative workup includes a thorough examination and audiogram.
Because these lesions develop behind a normal appearing tympanic mem-
brane, the physical examination may be completely normal or reveal a retro-
tympanic white mass. Audiologic evaluation usually reveals a conductive
hearing loss. Because CT shows the size and location of the CC, it is often
helpful in determining the extent of disease and surgical intervention that
will be required; however, it is not absolutely necessary in the evaluation
of patients with a clearly visible retrotympanic mass. MRI adds little to
the evaluation of the middle ear CC unless the diagnosis is unclear.
Unlike acquired cholesteatomas, which develop from posterior tympanic
membrane retraction pockets, most CCs start in the anterior mesotympa-
num. This anterior location often allows for cholesteatoma dissection

Table 5
Postoperative complications
Postoperative complications Number of patients Percentage
Recurrent cholesteatoma 0 0.0
Recurrent perforation 4 7.5
Prosthesis extrusion 1 1.9
Otorrhea 1 1.9
Delayed facial paresis 1 1.9
Mastocutaneous fistula 0 0.0
Granuloma 0 0.0
Meningitis 0 0.0
Abscess 0 0.0
Profound hearing loss 0 0.0
Perichondritis 0 0.0
Hematoma 0 0.0
Wound infection 0 0.0
1092 BENNETT et al

without disarticulation of the incudostapedial joint, therefore, improving

chances of postoperative hearing results. We start each case via a postauric-
ular tympanoplasty approach. Lesions found to extend into the ossicular
mass can usually be approached using this approach. Increased exposure
can be created with gentle curetting of the anterior medial scutum. If poste-
rior exposure is still inadequate, an intact canal wall mastoidectomy can be
performed. The little additional morbidity and improved exposure of an
intact canal wall mastoidectomy is favorable to a transcanal atticotomy.
McGrew [27] noted no increase in postoperative complications of facial pa-
resis, CSF leak, or meningitis in patients undergoing a tympanoplasty with
mastoidectomy as opposed to those undergoing a tympanoplasty alone. In
addition, our experience is that the need for canal wall down procedures is
relatively rare and should only be performed if there is a labyrinthine fistula
or children with either poor health or likely to have poor follow-up. When
performed, the basic principles of a wide meatoplasty, low facial ridge, and
complete eradication of mastoid air cells apply [28]. Often, these cavities
become self-cleaning and do not require life-long care.
Like other chronic ear disease, surgery for CC is relatively safe. The ideal
timing of surgery remains unclear. Early detection and surgical intervention
reduces the risk of ossicular involvement and improves hearing outcomes.
Although there is no lower age limit to chronic ear surgery, generally the
children should be at least 10- to 12-months-old for both anesthetic and sur-
gical consideration. At this age, the external auditory canal and middle ear
cleft are large enough to accommodate surgery and the risk of anesthesia is
minimal. The two most likely unexpected intraoperative findings are dural
exposure and facial nerve exposure. Unplanned facial nerve injury is a dev-
astating complication in these young patients. Care must be exercised when
dissecting around the facial nerve to ensure there is a healthy bony covering.
Although Hough observed nearly 30% of patients with CC had a dehiscence
in the tympanic segment of the facial nerve, our rates were much lower.
The pathophysiology of CCs does not involve Eustacian tube dysfunc-
tion; therefore, recurrence rates should be lower than acquired forms of
the disease. In fact, although typical rates of recurrence for cholesteatomas
are around 10% to 40% for children, we did not experience a recurrence in
the 32 patients with long-term follow-up. Because of low recurrence rates,
we do not routinely stage these patients and only reexplore these patients
for obvious recurrence, worsening or poor hearing.
There are several explanations for why the hearing results do not appear
more favorable. Over half of the patients had air–bone gaps less than 20 dB
PTA preoperatively. If preoperative hearing is not that bad, it is hard to
make significant improvement postoperatively. Patients with stage II or
III disease had slight improvements in their hearing after surgery. This
also underscores the fact that it is more likely to improve hearing in a patient
that has poorer hearing preoperatively. The patients with stage 1 disease had
significantly better hearing than those with stage II or III disease. There are
 CONGENITAL CHOLESTEATOMA 1093

inherent difficulties with serial audiograms in children, especially those un-

der 4 or 5 years old. They are prone to inconsistencies in their audiograms,
which may account for some inaccuracies. A significant difference could also
be accounted for by lack of follow-up. As a tertiary otology referral center,
nearly half of 53 patients sought follow-up care with referring physicians. It
is likely that these patients had better outcomes or revision surgery or fol-
low-up would have been arranged. In addition, our local patients without
complaints and good hearing often skip follow-up appointments, and there-
fore would not have postoperative audiograms.

Further readings
Doyle K, Luxford W. Congenital aural cholesteatoma: results of surgery in 60 cases. Laryngo-
scope 1995;105:263–7.
Eavey RD. Abnormalities of the neonatal ear: otoscopic observations, histologic observations,
and a model for contamination of the middle ear by cellular contents of amniotic fluid.
Laryngoscope 1993;103(1 Pt 2 Suppl 58):1–31.
Grundfast KM, et al. The inferiorly based superior tympanomeatal flap for removal of congenital
cholesteatoma. Laryngoscope 1990;100:1341–3.
Karmody C, et al. The origin of congenital cholesteatoma. Am J Otolaryngol 1998;19:292–7.
Nelson M, et al. Congenital cholesteatoma. Arch Otolaryngol Head Neck Surg 2002;128:810–4.
Piza J, et al. Meconium contamination fo the neonatal ear. J Pediatr 1989;115:910–4.
Robert Y. Congenital cholesteatoma of the temporal bone: MR findings and comparison with
CT. Am J Neurorad 1995;19:755–61.
Selesnick SH, Lynn-Macrae AG. The incidence of facial nerve dehisence at surgery for cholestea-
toma. Otolaryngol Neurotol 2001;22(2):129–32.
Thakkar K, et al. Congenital cholesteatoma Isolated to the mastoid. Otolaryngol Neurotol 2006;
27:282–3.
Tos M. A new pathogenesis of mesotympanic cholesteatoma. Laryngoscope 2000;110:1890–7.

References
[1] House HP. An apparent primary cholesteatoma. A case report. Laryngoscope 1953;63(3):
712–3.
[2] Derlacki EL, Clemis JD. Congenital cholesteatoma of the middle ear and mastoid. Ann Otol
Rhinol Laryngol 1965;74(3):706–27.
[3] Paparella M, Rybak L. Congenital cholesteatoma. Otolaryngol Clin North Am 1978;11:
113–20.
[4] Kazahaya K, Potsic WP. Congenital cholesteatoma. Curr Opin Otol Head Neck Surg 2004;
12:398–403.
[5] Darrouzet V, et al. Congenital middle ear cholesteatomas in children: our experience in 34
cases. Otolaryngol Head Neck Surg 2002;126:34–9.
[6] House JW, Sheehy JL. Cholesteatoma with intact tympanic membrane: a report of 41 cases.
Laryngoscope 1980;90:70–6.
[7] Potsic W, et al. CongenitalcCholesteatoma: 20 years experience at the Children’s Hospital of
Philadelphia. Otolaryngol Head Neck Surg 2002;126:409–13.
[8] Koltai PJ, et al. The natural history of congenital cholesteatoma. Arch Otolarygol Head
Neck Surg 2002;128:804–9.
[9] Friedberg J. Congenital cholesteatoma. Laryngoscope 1994;104:1–23.
1094 BENNETT et al

[10] Cannoni M. Congenital cholesteatoma of the petrous bone. Rev Laryngol Otol Rhinol
(Bord) 1989;110(1):33–42.
[11] Jackler RK, Parker DA. Radiographic differential diagnosis of petrous apex lesions. Am J
Otolaryngol 1991;13:561–73.
[12] Peron DL, Schuknecht HF. Congenital cholesteatoma with other anomalies. Arch Otolar-
yngol 1975;101:498.
[13] Derlacki EL, Harrison WH, Clemis JD. Congenital cholesteatoma of the middle ear and
mastoid. A 2nd report presenting 7 additional cases. Laryngoscope 1968;78:1050–78.
[14] Potsic WP, et al. A staging system for congenital cholesteatoma. Arch Otolaryngol Head
Neck Surg 2002;128:1009–12.
[15] Nelson M, et al. Congenital cholesteatoma: classification, management, and outcomes. Arch
Otolaryngol Head Neck Surg 2002;128:810–4.
[16] Dodson EE, Hashisaki GT, et al. Intact canal wall mastoidectomy with tympanoplasty for
cholesteatoma in children. Laryngoscope 1998;108(7):977–83.
[17] Jackson CG, Glasscock ME, Nissen AJ. Open mastoid procedures: contemporary indica-
tions and surgical technique. Laryngoscope 1985;95:1037–43.
[18] House WF, Hitselberger WE, Horn KL. The middle fossa transpetrous approach to the
anterior–superior cerebellopontine angle. Am J Otol 1986;7:1–4.
[19] Teed RW. Cholesteatoma verum tympani: its relationship to the first epibrachial placode.
Arch Otolaryngol 1936;24:455–74.
[20] Paparella MM. Congenital cholesteatoma. Otolaryngol Clin North Am 1978;11:113–20.
[21] Levenson MJ, et al. Congenital cholesteatoma in children. Laryngoscope 1988;98:949–55.
[22] Michaels L. Origin of congenital cholesteatoma from a normally occurring epidermoid rest
in the developing middle ear. Int J Pediatr Otorhinolaryngol 1989;15:51–65.
[23] Liang J, et al. Immunohistochemical characterization of the epidermoid formation in the
middle ear. Laryngoscope 2003;113:1007–14.
[24] Ruedi L. Cholesteatoma formation in the middle ear in animal experiments. Acta Otolaryngol
1959;50(3–4):233–40.
[25] Sade J, BaBiacki A, Pinkus G. The metaplastic and congenital origin of cholesteatoma. Acta
Otolaryngol 1983;96:119–29.
[26] Choi JY, et al. Retinoic acid depletion induces keratinizing squamous differentiation in
human middle ear epithelial cell cultures. Acta Otolaryngol 2003;123(4):466–70.
[27] McGrew BM, Jackson CG, Glasscock ME. Impact of mastoidectomy on simple tympanic
membrane perforation repair. Laryngoscope 2004;114(3):506–11.
[28] Jackson CG, et al. A surgical solution for the difficult chronic ear. Am J Otolaryngol 1996;
17(1):7–14.