Recently, Stephenson and Saliba [1] reported their data on preoperative CT-scanning in cholesteatoma patients with labyrinthine fistula in the article ‘‘Prognostic indicators of hearing after complete resection of cholesteatoma causing a labyrinthine fistula’’. The authors state that high-resolution preoperative CT-scan is very precise in diagnosing labyrinthine fistulas and that its radiologic size helps to predict the type of the fistula. This study is generally well designed and well reported, but we have several comments. The authors mention: ‘‘a further study is needed to identify if a diffusion-weighted sequence for magnetic resonance imaging of labyrinthine cholesteatomatous fistula can predict the effect of the depth on the postoperative hearing outcome after a complete removal of the cholesteatoma’’. Although we consider non-echo-planar diffusion-weighted MRI as an essential tool in the preoperative work-up (and postoperative follow-up) of the cholesteatoma patient, its usefulness for the work-up of labyrinthine fistula in the cholesteatoma patient is limited [2]. Indeed, this sequence does not allow for discrimination of both cholesteatoma and other tissues, such as bone. However, other MRI sequences can be useful. Earlier studies on T1and T2-weighted imaging, gadolinium-enhanced spin-echo, three-dimensional Fourier transformation constructive interference in steady state (3DFT-CISS), and fluid-attenuated inversion recovery (FLAIR) were reported, respectively by Casselman et al. [3], Smadja et al. [4], and Sone et al [5]. In our department, we routinely perform preoperative temporal bone cone-beam CT-scans combined with a full MRI protocol (including non-echo-planar diffusion-weighted MRI) in the cholesteatoma patient [2]. In case of a lateral canal fistula, as diagnosed on CT, we use T2-weighted imaging to evaluate the presence of the labyrinthine fluid. If the labyrinthine fluid in the part of the lateral semicircular canal adjacent to the cholesteatoma is absent (no signal on T2 MRI), this will probably be due to reactive fibrosis. In such cases, the risk for perioperative sensorineural hearing loss is lower because the matrix and perimatrix of the cholesteatoma are not directly in contact with the membranous labyrinth. In case the fluid is still present (high signal on T2 MRI), the risk for a perioperative sensorineural hearing loss is higher because the matrix and perimatrix of the cholesteatoma can be in direct contact with the fluid contents of the membranous labyrinth, depending on the extent of the fistula. Our preoperative counseling of the patient is based on this policy [6]. A further study on the use of preoperative MRI in lateral canal fistula is definitely feasible and should, in our opinion, focus on T2-weighted imaging instead of diffusionweighted imaging. Additionally, the role of the cone-beam CT-scan should be further studied because of the increased spatial resolution and the significant decrease in radiation dosage. V. Van Rompaey E. Offeciers (&) University Department of Otorhinolaryngology and Head and Neck Surgery, Sint Augustinus Hospital, Oosterveldlaan 24, Wilrijk, 2610 Antwerp, Belgium e-mail: erwin.offeciers@gza.be