Objectives: To define predictors of residuals and retraction pockets (RP) in children operated on for congenital cholesteatoma (CC).
Design and setting: Retrospective review (1996-2005), academic center.
Patients: One hundred seventeen patients treated for CC corresponding to modified Derlacki's criteria were included (median age, 6.5 yr). No case of RP at time of diagnosis, with a mean follow-up of 2.5 years after last surgery.
Main outcome measures: Clinical and surgical data influencing outcome. Multivariate analysis.
Results: Two groups were defined after CC removal: group I (12 cases), no second look required and no case of subsequent re-intervention; group II (105 cases), planned second look always performed (mean delay, 12.1 mo), no difference of sex ratio (M/F = 2). Group I patients were younger than in group II (3.3 vs. 5.9 yr, P < .001). All of them had a normal contralateral eardrum and a disclosure of CC by routine examination (vs. 19% in group II, P < .001). In group I, the mass occupied one or two anterior quadrants (41.6% and 58.4%, respectively) versus more than two quadrants in 46.6% in group II. Residuals and RP rates were 41% and 15%, respectively (only in group II). Predictors for residuals were atticotomy (odds ratio [OR] 2.9, 95% confidence interval [CI] 1.3-6.7) and destruction of stapes (OR 4.3, 95% CI 1.7-10.5). Predictors for RP were eustachian tube extension (OR 6.8, 95% CI 1.7-26.8) and nonreconstructed atticotomy (OR 5.9, 95% CI 1.1-30.9).
Conclusions: Young children with small CC had no recurrences. Residuals were more frequent in case of atticotomy and stapes destruction. RP occurred especially in cases of eustachian tube extension and if cartilage tympanoplasty was not performed. Tympanic and canal wall reinforcement should be considered in extensive CC.