Background: Recurrent cholesteatoma after surgical excision occurs frequently in children. Until recently, a surgical second look was mandatory and considered as standard reference. MRI including a delayed T1 sequence after gadolinium injection and diffusion-weighted imaging (DWI) has proved its efficiency but has been evaluated mainly in adults.
Objective: Our purpose was to evaluate the accuracy of DWI to diagnose recurrence of cholesteatoma in children.
Materials and methods: We evaluated prospectively with MRI 20 ears in 18 children who had had surgery for cholesteatoma. We compared DWI and delayed T1-weighted images following gadolinium administration with intraoperative or follow-up findings. We calculated the sensitivity and specificity of each sequence for the diagnosis of recurrent cholesteatoma.
Results: Sensitivity to diagnose recurrent cholesteatoma was 87% for both DWI and delayed post-gadolinium sequences, specificity was 71% and 83%, respectively. Adding both sequences, the sensitivity was 87%, the specificity 100%. There was one false negative probably due to small size recurrence.
Conclusion: In our series, DWI was reliable to diagnose recurrent cholesteatoma in children and allows avoiding surgery when negative. However, because small recurrences less than 5 mm may be missed, follow-up must be prolonged (5 years).