Non-echoplanar diffusion-weighted MRI in children and adolescents with cholesteatoma: reliability and pitfalls in comparison to middle ear surgery

Thekla von Kalle,Peter Amrhein,Assen Koitschev

DOI: https://doi.org/10.1007/s00247-015-3287-y

2015-02-13

Pediatric Radiology

Abstract:BackgroundCurrently, there is only limited and contradictory evidence of the role of diffusion-weighted MRI (DW-MRI) in the management of children with cholesteatoma.ObjectiveTo provide surgically controlled data that may allow to replace second-look surgery by non-echoplanar DW-MRI in children.Materials and methodsFifty-five children and adolescents with a median age of 8.6 years (2.2-17.7 years) underwent 61 preoperative half-Fourier acquisition single-shot turbo spin-echo (HASTE) DW-MRI of their petrous bone. Surgical interventions followed within 24 h (79%), within 5 months (20%) or at 18 months (1 case).ResultsSurgery detected a cholesteatoma or retraction pocket in 41 of 61 cases (67%). In 49 cases (80%), the MR result was confirmed by surgical findings. Two MR findings were false-positive and 10 false-negative (including cholesteatomas <4 mm). HASTE DW-MRI alone had a sensitivity of 76% and a specificity of 90%. The positive predictive value was 94%, the negative predictive value 64%. In combination with preoperative otoscopy, sensitivity was 90% and negative predictive value 82%.ConclusionDW-MRI correctly detected the majority of lesions but could not reliably exclude small cholesteatomas and empty retraction pockets. We would therefore not generally recommend MR as a substitute for second-look surgery.

pediatrics,radiology, nuclear medicine & medical imaging