Learning Objectives:

Aim: In many clinics non-EPI DWI has replaced second look surgery because of its high negative predictive value. In our institution, follow-up DWI is performed at least twice after surgery. Aim of this study was to determine the yield of the second follow-up MR-DWI (MR-DWI2) after in patients with a negative first follow-up MR-DWI (MR-DWI1) and an absence of clinical otoscopic suspicion of recurrence or residual cholesteatoma.

Methods: Between 2006 and 2013 we retrospectively included 45 ears in 44 patients which had undergone cholesteatoma surgery, had a negative MR-DWI1 performed 6–24 months after surgery, an MR-DW2 performed at least 6 months after MR-DWI1 and an absence of clinical suspicion of recurrence or residual cholesteatoma between surgery and MR-DWI2. Two radiologists independently scored MR-DWI1 and MR-DWI2. Descriptive analysis were used for determining the yield of MR-DWI2. Interobserver agreement was calculated using Cohen's kappa statistics.

Results: In 14 of 45 ears (31%) MR-DW2 was equivocal (n = 6, 13%) or positive (n = 8, 18%). Interobserver agreement indicated substantial agreement (κ = 0.75). Patients with a positive MR-DW2 were younger of age compared to those with an equivocal or negative MR-DW2. In the group of 8 patients with positive MR-DW2, 6 were operated on with surgical confirmation of cholesteatoma in 5 of these patients. In 1 patient only fatty tissue was found.

Conclusion: The most important finding of this study is that 31% of MR-DW2 showed equivocal or positive evidence of cholesteatoma despite clinical and MR-DW1 follow-up. Given the known high sensitivity and specificity of non-EPI DWI, good quality of the included DWI examinations and high interobserver agreement in our study, it seems very unlikely this can be explained by a missed cholesteatoma larger than 2–3 mm on MR-DW1. It is also striking that patients with a positive follow-up MR-DW2 are younger of age. This may influence follow-up strategies in the future.