Abstract
Objective
To assess outcomes of transcanal endoscopic ear surgery (TEES) for congenital cholesteatoma.
Study Design
Case series with chart review of children who underwent TEES for congenital cholesteatoma over a 10-year period.
Setting
Three tertiary referral centers.
Methods
Cholesteatoma extent was classified according to Potsic stage; cases with mastoid extension (Potsic IV) were excluded. Disease characteristics, surgical approach, and outcomes were compared among stages. Outcomes measures included residual or recurrent cholesteatoma and audiometric data.
Results
Sixty-five cases of congenital cholesteatoma were included. The mean age was 6.5 years (range, 1.2-16), and the mean follow-up was 3.9 years (range, 0.75-9.1). There were 19 cases (29%) of Potsic stage I disease, 10 (15%) stage II, and 36 (55%) stage III. Overall, 24 (37%) patients underwent a second-stage procedure, including 1 with Potsic stage II disease (10%) and 21 (58%) with Potsic stage III disease. Eight cases (12%) of residual cholesteatoma occurred. One patient (2%) developed retraction-type (“recurrent”) cholesteatoma. Recidivism occurred only among Potsic stage III cases. Postoperative air conduction hearing thresholds were normal (<25 dB HL) in 93% of Potsic stage I, 88% of stage II, and 36% of stage III cases.
Conclusion
TEES is feasible and effective for removal of congenital cholesteatoma not extending into the mastoid. Recidivism rates were lower with the TEES approach in this large series than in previously reported studies. Advanced-stage disease was the primary risk factor for recidivism and worse hearing result. As minimally invasive TEES is possible in the youngest cases, children benefit from early identification and intervention.
Congenital cholesteatoma has a variety of presentations, from a small pearl in the middle ear to an extensive open cyst with associated destructive inflammatory changes. Congenital cholesteatoma occurs behind an intact tympanic membrane in a patient with no history of perforation or otorrhea. 1 It is theorized to result from postnatal persistence of epidermoid rests in the middle ear and often arises in the anterosuperior quadrant of the middle ear near the cochleariform process.2-4
The most widely used staging system for congenital cholesteatoma was proposed by Potsic et al in 2002. 5 Stages I and II describe single or multiquadrant disease without ossicular involvement or mastoid extension. Stage III indicates ossicular involvement, and stage IV describes disease extension into the mastoid. Stage I-III disease may be treated by transcanal endoscopic ear surgery (TEES), while stage IV disease often requires a combined approach tympanomastoidectomy.
As a disease originating in the middle ear, congenital cholesteatoma is particularly amenable to the use of endoscopic techniques for resection. The endoscopes provide minimally invasive transcanal access to the middle ear, with wide dynamic visualization, which can include angled endoscopes for looking into recesses, such as the protympanum, epitympanum, and retrotympanum.6,7 These advantages have been shown to enhance completeness of cholesteatoma resection when endoscopic ear surgery is compared with microscopic techniques. 8
Several series have shown that pediatric cholesteatoma can be successfully managed with TEES.9-11 However, the literature on the use of endoscopic techniques for resection of congenital cholesteatoma has been limited to small series.12,13 This study assesses the outcomes of TEES for congenital cholesteatoma over 10 years of experience from 3 tertiary pediatric medical centers. Recidivism rates and hearing outcomes are described.
Methods
Approval for this study was obtained from the Institutional Review Board at Ann & Robert H. Lurie Children’s Hospital of Chicago, The Hospital for Sick Children in Toronto, and the University of Modena. Pediatric patients (age <18 years) who underwent TEES as primary surgery for cholesteatoma at the participating institutions between January 1, 2011, and September 1, 2019, were retrospectively reviewed. Congenital cholesteatoma was defined by the presence of cholesteatoma behind an intact tympanic membrane in a patient with no history of perforation or otorrhea. 1 Patients who underwent a transcanal endoscopic-only technique were included (class 3 in Cohen et al classification). 14 Patients with mastoid extension (Potsic stage IV) and/or those requiring mastoidectomy (class 2b) were excluded. Patients were identified from prospectively maintained databases (Toronto, Chicago) or by searching an operative database (Modena). All cases were reviewed for inclusion and exclusion criteria.
The following data were collected: age, sex, disease characteristics (extent, type of cholesteatoma [pearl, closed, open], and ossicular involvement), details of surgery (duration, need for canal widening and/or atticotomy, type and extent of tympanic membrane reconstruction, and use of ossicular chain reconstruction), incidence of residual or recurrent disease, and hearing results of pre- and postoperative audiograms. Cholesteatoma discovered behind an intact tympanic membrane on second look or during clinical follow-up was classified as residual disease. Cholesteatoma arising from a retraction pocket that developed postoperatively was classified as recurrence. Hearing results were recorded as air conduction pure tone average (PTA), bone conduction PTA, and air-bone gap (ABG) per the average of 500, 1000, 2000, and 4000 Hz.
Consistency in data reporting was achieved across participating institutions with the use of well-defined classification systems for reporting disease stage and surgical approach. Extent of disease was classified via Potsic stages. 5 Extent and approach to surgery were classified with the International Otology Outcome Group’s SAMEO-ATO classification of tympanomastoid surgery. 15 Potsic and SAMEO-ATO classifications were recorded in a subset of operative notes and determined retrospectively based on operative findings and descriptions in all other cases.
Data were analyzed with Stata 11 (StataCorp LLC). Fisher’s exact and 1-way analysis of variance tests were used to examine differences in characteristics, treatment, and outcomes of Potsic stage I-III disease. Paired t tests were used to compare pre- to postoperative hearing results. P values <.05 were considered significant.
Results
Patient Characteristics
The study population comprised 65 ears in 64 patients. One patient had bilateral congenital cholesteatomas, which were excised under a single anesthetic. Patient characteristics are summarized in Table 1 . The mean age at first surgery was 6.5 years (range, 1.2-16). The majority of patients had preoperative computer tomography scans (79%). The mean length of follow-up was 3.9 years (range, 0.75-9.1).
Patient Demographics.
Abbreviation: CT, computed tomography.
Disease Characteristics
There were 19 cases (29%) of Potsic stage I, 10 (15%) of stage II, and 36 (55%) of stage III disease. Disease characteristics varied by stage ( Table 2 ). Children presenting with Potsic stage III disease were significantly older than those with stage I and II disease. The majority (79%) of stage I cases were classified as pearl-type cholesteatoma, while the proportion of infiltrative disease increased by stage (5% of Potsic stage I vs 64% of stage III). By definition, there were no cases of ossicular involvement in stage I and II disease, while the majority of stage III cases (89%) had ossicular erosion present. Four Potsic stage III cases did not have ossicular erosion but required removal of ≥1 involved ossicles for disease resection.
Disease Characteristics for Potsic Stage I-III Disease. a
Abbreviation: CT, computed tomography.
Values are presented as mean (95% CI) or No. (%). Bold indicates significance (P < .05).
Surgical Treatment
Details of surgery are summarized in Table 3 , including relevant aspects of the SAMEO-ATO classification of tympanomastoid surgery. 15 Length of surgery was significantly longer for Potsic stage III disease as compared with stage I and II cases. All cases were primary surgery (SAMEO-ATO S1) and endoscopic transcanal (A1). The scutum was removed in 35% of cases for access to the epitympanum (M2a) and typically repaired with tragal cartilage. No mastoid obliteration was performed (Ox). Adequate access for endoscopic dissection of cholesteatoma through the ear canal was present in all cases regardless of age, and canalplasty was not required in any patient. Widening of the posterior portion of the tympanic sulcus was required for visualization of the ossicular chain in 1 patient with Potsic stage I (5%) and the majority of stage III cases (56%). The majority of patients with stage I (84%) and II (60%) disease had the tympanic membrane preserved, while the majority (67%) of patients with stage III disease required grafting. Graft materials included soft tissue (perichondrium or fascia), cartilage, or Biodesign porcine collagen graft (Cook Medical). The decision to proceed with ossiculoplasty was based on surgeon’s preference, including perception of risk of residual disease. Among 33 patients with Potsic stage III disease and without an intact ossicular chain, 16 (48%) were reconstructed at the initial procedure, and 17 (52%) were left unreconstructed.
Surgical Treatment for Potsic Stage I-III Disease. a
Abbreviation: TM, tympanic membrane.
Values are presented as mean (95% CI) or No. (%). Bold indicates significance (P < .05).
Indicates International Otology Outcome Group’s SAMEO-ATO classification.
Biodesign porcine collagen graft (Cook Medical).
Types of ossicular chain reconstruction: Osi incus to stapes (1), Osm malleus to stapes head (3), Osd tympanic membrane directly to stapes head (1), Ofm malleus to footplate (8), Oft tympanic membrane to footplate (3).
A planned second-stage operation was performed in patients with higher risk of residual disease, including most patients with Potsic stage III disease, open cholesteatoma, or spillage at initial surgery. Patients at low risk for residual cholesteatoma were followed clinically with a second stage only if there was otoscopic evidence of disease or a decision to have ossiculoplasty. Overall, 24 (37%) patients underwent a second-stage procedure at a median interval of 13 months (range, 3.9-59): 1 patient with Potsic stage II disease (10%) and 23 (64%) with Potsic stage III disease. Two second-stage procedures were prompted by clinical findings of recidivism and performed at a delayed interval (27 and 59 months). The remaining 20 second-stage procedures were performed within 22 months. An additional 6 patients underwent postoperative diffusion-weighted magnetic resonance imaging or computer tomography scans for surveillance. The remaining 35 ears (54%) were surveilled clinically.
Recidivism
No ears with Potsic stage I or II disease had residual or recurrent cholesteatoma ( Table 4 ). Overall, there were 9 cases of recidivism, all of which occurred among ears with Potsic stage III disease. Eight ears (12%) had residual cholesteatoma discovered during a second-look procedure. One ear developed retraction-type cholesteatoma 12 months postoperatively, which was classified as recurrent disease (5-year recurrence rate, 2% [95% CI, 0%-5%]; Kaplan-Meier survival analysis). A retraction pocket grew into the supratubal recess underneath the cartilage graft used for scutum reconstruction. Potsic stage III cases were examined for differences in disease characteristics and surgical details between cases without recidivism and cases with residual and recurrent cholesteatoma. No differences were found in type of cholesteatoma, frequency of ossicular or stapes involvement, need for atticotomy, extent of repair, graft type, or ossicular chain reconstruction status.
Outcomes of Potsic Stage I-III Disease. a
Abbreviations: ABG, air-bone gap; PTA, pure tone average.
Values are presented as mean (95% CI) or No. (%). Bold indicates significance (P < .05).
Normal hearing after surgery: air conduction <25 dB HL.
Audiometric Outcomes
Audiometric data were available for the majority of patients but not all, owing to compliance. Pre- and postoperative air conduction PTAs were not significantly different (31 vs 27 dB HL, n = 51, P = .3). Pre- and postoperative ABGs were also similar (24 vs 21 dB, n = 43, P = .8). Pre- and postoperative air conduction PTA and ABG were different among Potsic I-III groups, with worse pre- and postoperative hearing results measured among Potsic stage III cases ( Table 4 ). The majority of patients with Potsic stage I and II (93% and 88%, respectively) had normal hearing after surgery (defined as air conduction PTA <25 dB HL), as opposed to only 36% with Potsic stage III. Among 16 patients who underwent ossicular chain reconstruction, there was a mean difference in pre- to postoperative improvement in air conduction PTA of 10 dB (95% CI, 0-20) and a mean postoperative ABG of 20 dB (95% CI, 15-25).
Discussion
TEES has numerous benefits when compared with traditional open approaches: parents and patients prefer the minimally invasive approach; postoperative pain and morbidity are decreased; and increased overall health-related quality of life has been demonstrated.16-19 In addition to these benefits, numerous studies have cited enhanced detection of cholesteatoma and lower rates of recidivism.10,20-23 In this study, favorable rates of recidivism were observed with the TEES technique, with a residual rate of 12% and a recurrence rate of 2% among 65 cases of stage I-III congenital cholesteatoma. Recidivism has been reported in roughly 35% of congenital cholesteatoma cases (Potsic I-IV) managed by microscopic surgery.5,24
Consistent with prior publications, we used the Potsic staging system. One limitation of this system is the potential for variable interpretation of stage III: whether “ossicular involvement” implies contact of cholesteatoma to an ossicle, ossicular erosion, or surgeon’s judgment that ossicle removal is required. Nevertheless, an advantage of the Potsic system is that it delineates 4 stages of cholesteatoma that correlate with outcome. We also analyzed cases using the classification system of the European Academy of Otology and Neurotology/Japanese Otological Society (data not shown), which has been approved with international consensus and is intended for acquired as well as congenital cholesteatoma. 25 This system was of limited predictive utility, as only 2 stages of disease were present in this series.
Kobayashi et al published the first series of children with congenital cholesteatoma (N = 12) managed by TEES. 12 The majority (92%) of cases were Potsic I and II disease, and there was 1 case of recidivism that occurred in the sole patient with stage III disease. Park et al reported on 25 children with congenital cholesteatoma managed with TEES. 13 The majority of this series was based on Potsic stage I and II disease (80%), with 5 cases of stage III disease. The authors reported 1 case of recidivism (4%) in a patient with stage II disease, and there were no complications. Both series acknowledged the limitations of small case numbers and short follow-up (mean, 23 and 24 months); however, the authors concluded that TEES has real advantages and can be used safely for congenital cholesteatoma with comparable surgical outcomes to microscopic techniques.12,13 Multiple other series established the feasibility and favorable outcomes of TEES for pediatric cholesteatoma, though they did not distinguish congenital cases from acquired ones or included few congenital cases.10,21-23,26
The present study represents the largest series to date of congenital cholesteatoma managed by TEES and provides additional evidence for the effectiveness of this technique for the management of congenital cholesteatoma without mastoid extension. The majority of cases were Potsic stage III (55%), few of which were included in prior series. Outcomes varied by Potsic stage, which is well known. All but 1 patient with Potsic stage I and II disease were followed clinically and spared postoperative imaging and second-look procedures, and none had residual or recurrent disease. Recidivism was low even among advanced disease (Potsic stage III), with a residual rate of 22% and overall recidivism rate of 25%. These rates compare favorably to those of residual cholesteatoma in Potsic stage III disease, up to 50%.5,27
No differences were found for disease characteristics or surgical approach between Potsic stage III cases that had residual or recurrent disease and those that did not. Risk factors that have been suggested for residual disease include open-type cholesteatoma, atticotomy, and stapes involvement.24,28 Development of retraction-type cholesteatoma occurs in a minority of patients after surgery for congenital cholesteatoma. Eustachian tube dysfunction, disease extension into the eustachian tube, and nonreconstructed atticotomy have been reported as risk factors for retraction-type cholesteatoma.17,24,29
Recent consensus recommendations from the International Pediatric Otolaryngology Group stated that a microscope and endoscope are required for congenital cholesteatoma surgery; however, the present study shows that the microscope is typically unnecessary. 30 Optimal management of congenital cholesteatoma via TEES is accomplished with modifications to standard approaches. Meatal skin incisions may be extended anterosuperiorly and the tympanomeatal flap based inferiorly to expose the anterosuperior quadrant of the middle ear, which is most commonly involved in congenital cholesteatoma. 31 As the tympanomeatal flap is raised, gentle inferior traction of the tympanic membrane with a cupped forceps can be used to “deglove” the lateral process and handle of the malleus ( Figure 1 ). This technique provides optimal access to the anterosuperior quadrant and protympanum. On occasion, anteroinferior extension of disease may necessitate detachment of the fibrous attachments of the tympanic membrane from the umbo via microscissors. With a long experience of this technique in our departments, we have not seen postoperative lateralization of the tympanic membrane.32,33 Careful dissection and inspection of the anterior surface of the tensor tympani tendon and cochleariform process are required to address the typical origin of congenital cholesteatoma. TEES is particularly advantageous in allowing for visualization of this area. 34 According to the curvature of the ear canal and medialization of the umbo, it may be advantageous to elevate the annulus in the anteroinferior quadrant to adequately inspect this area with a 30° or 45° endoscope.
Two images demonstrate “degloving” of the malleus in (A) Potsic stage I and (B) Potsic stage III disease.
Small congenital cholesteatoma (Potsic stage I) is ideally suited to the 1-handed endoscopic approach, as dissection is typically easy and there is usually no requirement to place a graft for reconstruction. When cholesteatoma extends beyond the lateral semicircular canal in the mastoid (Potsic stage IV), it typically cannot be reached with TEES, though it is sometimes possible to extract well-encapsulated disease by delivering it into the attic by using angled instruments and endoscopes. For cholesteatoma of intermediate extent, the choice of approach depends on many factors. TEES may be used for cholesteatoma that extends no farther than the mastoid by extending scutum removal into an atticoantrostomy via a curette, ultrasonic bone removal, or a drill with a protected shaft. It is necessary to reconstruct the bone defect to prevent the risk of recurrent disease from new retraction. Cholesteatoma trapped in the medial attic under an intact ossicular chain cannot be readily removed with TEES through the mesotympanic route. Options to preserve the ossicular chain include the creation of a large-enough atticoantrostomy to work around the intact ossicular chain or the use of a combined approach tympanomastoidectomy to access the medial attic transmastoid with an endoscope or microscope. 35 Alternatively, the ossicular chain can be disarticulated and ossiculoplasty performed. When choosing from these alternatives, it is important to consider the small gain in hearing achieved by maintaining the intact chain (in the order of 10 dB) as compared with the greater postoperative morbidity associated with a more invasive approach. 36
Several potential challenges and disadvantages of TEES are worth noting. A longer learning curve has been suggested, though this may be less relevant to current and future generations of surgeons who receive early training in endoscopic techniques. 37 One recent series even demonstrated a shorter operative time than a microscopic technique for the management of cholesteatoma. 17 There is a loss of binocular vision that is afforded by the microscope; however, this is overcome by the dynamic visualization provided by the endoscope. Finally, TEES requires a 1-handed technique, which can be limiting particularly in the setting of bleeding. Techniques to control bleeding during TEES have been described. 38
We acknowledge several limitations of this study. Though the length of follow-up was adequate to detect recidivism in most cases (mean, 3.9 years), 3% of patients had <1 year of follow-up. It is possible that patients may have recidivism that has not yet presented or is otherwise unknown by their surgical teams due to loss of follow-up. This series of congenital cholesteatomas managed by TEES includes cases from multiple surgeons across institutions and represents a significant advancement from smaller, single-surgeon series. Still, the skill and decision making of the surgeons in this series cannot be controlled for and thus may limit the generalizability of these results to other surgeons.
Conclusion
We present the largest series to date of congenital cholesteatoma treated by TEES. This series provides evidence for the efficacy of TEES in treating congenital cholesteatoma that does not extend into the mastoid (Potsic stages I-III). Lower disease stages were associated with a shorter operating time, a decreased need for scutum or bone removal, less need and extent of tympanic membrane reconstruction, and improved hearing outcomes. There were no cases of residual or recurrent disease among Potsic stage I and II cases. Recidivism was low even for higher-stage disease (Potsic stage III).
Footnotes
This article was presented at the World Congress of Endoscopic Ear Surgery; June 13, 2019; Boston, Massachusetts.
