Cholesteatoma is a mass formed by the keratinizing, squamous epithelium in the tympanic cavity and/or mastoid and subepithelial connective tissue, and progressive accumulation of keratin debris.1 Although cholesteatomas are not true tumors, they can proliferate, and erode and destroy the structure of the middle ear, resulting in hearing loss and serious intracranial and extracranial complications. The annual incidence of middle ear cholesteatoma is approximately 12.6 per 100,000 in adults, and 3 per 100,000 in children.2 Among them, the incidence of binaural disease is about 4.4 %–17.1 %.3,4

In children, Bilateral Middle Ear Cholesteatoma (BMEC) can cause bilateral hearing loss and requires long-term medical treatment. It can have a serious effect on children’s learning and social adaptation and may affect the speech development of young children. The treatment strategy of BMEC needs more consideration. Published literature on BMEC is limited to a few small series and case reports published during previous decades.5–8 Currently no standard protocols are available for the management of BMEC in children. Retaining or improving hearing and reducing the recurrence rate make management challenging. In this paper, we review previous cases of BMEC treated at our institution and propose a treatment algorithm based on our experience to improve the understanding of the disease and provide guidance for clinical diagnosis and treatment.

Between 2016 and 2023, a total of 320 pediatric with cholesteatoma were treated in our department, including 12 cases of BMEC with 4 girls and 8 boys. Their characteristics are shown in Table 1. The mean age at diagnosis was 6.5 ± 2.6-years (rang from 2.5 to 9.8), with only one child aged younger than 3-years. The mean duration of follow-up was 4.28 ± 2.58-years.

Four children had CC in one ear and AC in the other ear and 8 children had AC in both ears. The course of disease in the bilateral ears of the same child is different, but there is no significant difference statistically (p = 0.264, t = 1.178). Three children (Cases 6, 10, and 11) had a different course of disease in bilateral ears and two children (Cases 5 and 8) had unilateral middle ear cholesteatoma at the time of their first visit and subsequently developed a cholesteatoma in the contralateral ear. The extent of the lesions differed in bilateral ears in 6 children (Cases 4, 7, 8, 9, 10 and 11) were different. Nine children had maxillary developmental deformities (75%) (cleft palate in 3 cases and occult submucosal cleft palate in 6 cases), all of nine children were accompanied by diploetic or sclerotic type mastoid. Fig. 2 shows the details of a BMEC in a child (Case 12) with occult submucosal cleft palate. There was only 1 case of pneumatolytic type in bilateral mastoid.

Fig. 2.

Case 12. (A) Otoscope and CT scan on temporal bone. (B) An inverted V-shaped depression at the junction of the soft palate and the hard palate. (C) The three-dimensional reconstruction of the maxillary CT scan confirmed the absence of posterior nasal spine. (D) The hand drawing shows a complete posterior nasal spine structure.

The surgical treatment varied according to the severity of the lesion. The specific surgical methods and ossiculoplasty are shown in Table 1. Only one child (Case 7) underwent simultaneous binaural operation, and the other 11 children underwent staged surgery. The median interval between the first and second surgery was 3-months (range: 1–12 months). Six children (7-ears) (Cases 1, 2, 4, 5, 6, and 9) experienced a recurrence. In the revision operation due to recurrence, 6-ears underwent CWD and one ear underwent CWU surgery. The ossiculoplasty may be performed in the first or second-look stage. The final procedure method used for ossiculoplasty is shown in Table 1. The mean number of operations per patient was 2.67 ± 0.65. The mean preoperative PTA was 41.23 ± 16.50 dB, and the mean long-term postoperative PTA was 37.79 ± 14.42 dB. There was no significant difference in hearing results between preoperative and postoperative patients (p = 0.291, t = 1.082). No postoperative complications such as sensorineural deafness or facial paralysis were observed in any of the children.

There was no significant difference in postoperative hearing between CC and AC (p = 0.355, F = 4.268). There were significant differences in postoperation hearing among different surgical methods (p = 0.006, H = 12.630). In pairwise comparison, there were significant differences in postoperative hearing between the CWU group and the group undergoing recurrent surgery (p = 0.003). There was no significant difference between the other groups. There were significant differences in postoperation hearing among different ossiculoplasty methods (p = 0.001, H = 17.590). In a pairwise comparisons, there was significant difference in hearing outcomes between the without ossiculoplasty group and PORP group (p = 0.006). There was significant difference between the without ossiculoplasty group and no need ossiculoplasty group (p = 0.013). And among the other groups there were no significant differences (Table 2).

This paper reports 12 cases of BMEC in children. To our knowledge, this is the largest case series on children to date with complete long-term follow-up data. This study revealed that the age at diagnosis was relatively young, with a median age of 6.5-years (considering the delay in diagnosis, the age of onset is likely to have been younger). In addition, bilateral ears had different course, different degrees of lesion, and lesions at different locations. The BMEC in this study included both congenital and acquired, and the pathogenic mechanisms are likely to have been multifactorial. In our cases, palatal malformations, including cleft palate and occult submucosal cleft palate, were important risk factors. According to previous studies, children with cleft palate are prone to middle ear cholesteatoma and otitis media because of abnormalities in Eustachian tube anatomy and function, viscosity of middle ear effusion, and frequency of middle ear infection.9 The incidence of cholesteatoma in children with cleft palate is between 1.5% and 5.9%, which is 200 times higher than that in the general population.10–12 It is worth paying attention to that children with occult submucosal cleft palate are also susceptible to BMEC. In this study, 3 children had cleft palate, and 6 children had occult submucosal cleft palate. Occult submucosal cleft palate is caused by the defect of a part of the posterior nasal spine of the hard palate.13 Physical examination shows an inverted V-shaped depression at the junction of the soft palate and the hard palate when making an “ah” sound (Fig. 2B), with or without speech impediments. The diagnosis of occult submucosal cleft palate in children is often overlooked because the defect is not easily detected. The middle ear problems associated with occult submucosal cleft palate have not been reported. The palate should be carefully monitored during physical examination. If signs of occult submucosal cleft palate are present, attention should be paid to middle ear problems in the early stage. In addition, habitual sniffing extraction caused by patulous eustachian tube may be related to the pathogenesis of BMEC14,15 owing to the presence of total closing failure, with an intermittent free transfer of respiratory pressure to the middle ear, the sniffing procedure maintains a negative intratympanic pressure, gradually forming an attic retraction pocket.16 However, in our children, no patulous eustachian tube was found. Additional etiologies of BMEC requires analysis of a larger samples.

In the case of binaural lesions, repeated surgeries and bilateral hearing loss can have a significant impact on children’s development, and complete removal of the lesion and protection of hearing are particularly important. In terms of the surgical treatment strategy, the first choice is between simultaneous or staged surgery. In the traditional experience, bilateral chronic middle ear disease is typically performed by staging surgery for the risk of iatrogenic bilateral sensorineural hearing loss, bilateral facial nerve injury, and temporary bilateral hearing loss during the recovery period in simultaneous binaural surgery. However, there are also arguments for the advantages of simultaneous binaural surgery, such as reduces the cost of surgery and improves convenience for patients and surgeons. We tend to believe that staging surgery is preferred. But if you are quite confident in the surgery for the first ear won’t be injured, simultaneous binaural surgery can be considered. Of the 12 children in our group, 11 underwent binaural staged surgeries and only one underwent simultaneous surgery. The simultaneous surgery was performed because one of his first ear is CC with a very limited small lesion and without any structural damage, the surgery of the ear went very smoothly and quick under otoscopy. In all staged surgeries, the more severe side is selected first. For the Klemens et al.17 suggested another situation that surgeons should consider before surgery of bilateral: Are the anatomical and pathological conditions of the first ear sufficiently unusual or difficult that it is safer, for the patient, to address the second ear while all the relationships are fresh in the surgeon’s mind?. We haven't encountered this unusual situation, but it's worth considering. In summary, simultaneous binaural surgery in children requires more caution and certainty.

In this study, there was no statistical difference in hearing changes before and after surgery, but the PTA of long-term hearing after surgery was better than that before surgery. And there was no significant difference in hearing outcomes between CC and AC. The hearing outcomes were worse in the group that did not undergo ossiculoplasty and worse in the group that underwent recurrent surgery. And there was no significant difference in hearing prognosis between CWU, CWD and TEES groups. There was No statistical difference between PORP, TORP, Cartilage and did not need ossiculoplasty groups. However, the limitation is that the sample size of each group is small. The specific surgical method and the type of ossiculoplasty need to be personalized according to the patient's condition and intraoperative states. Other factors in the actual situation may also lead to unsatisfactory postoperative hearing recovery, such as extensive lesion destruction, congenital eustachian tube dysfunction, poor middle ear ventilation or postoperative ossicle displacement. Early use of hearing aids should be considered in children with poor hearing in both ears (moderate or higher hearing loss) after stabilization of the ear condition after surgery.

The high recurrence rate of the surgeries (7/24, 29.2%) shown in this study may be related to the presence of certain risk factors in these children. In this study, the recurrence time was 0.8‒3-years after surgery and the longest follow-up was 9-years. About the postoperative follow-up, International Pediatric Oncology Group (IPOG) guidelines18 recommend a minimum of 5-years. Most of the children had no interruption in follow-up. We recommend that children with BMEC be followed-up for more than 5-years or even for life. Based on our experience, we summarized a treatment flow chart for BMECs (Fig. 3).

Fig. 3.

Flow chart of clinical treatment for BMEC in children. BMEC: bilateral middle ear cholesteatoma.

In our cohort, BMEC presented at a significantly younger age and exhibited a high prevalence of concomitant palatal cleft anomalies or occult submucous cleft palates, necessitating particular clinical vigilance in this pediatric population. A staged bilateral surgical intervention was prioritized in most cases, with the more symptomatic ear addressed first. Furthermore, given the elevated recurrence risk in these patients, long-term follow-up surveillance is imperative.

Xiaoxu Wang: 0000-0002-2329-8730

Lining Guo: 0000-0002-3660-4864

Enxia Tian: 0009-0002-7199-5394

Min Chen: 0000-0001-8935-5606

Wei Liu: 0000-0002-1927-5499

Xin Ni: 0000-0003-0857-8512

Jie Zhang: 0000-0002-1621-7779

The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.