In 2005 we started using non-EPI DWI sequences in cholesteatoma imaging. At a certain point in time, we noticed that the ENT surgeons in our hospital were almost completely relying on the imaging findings when screening for residual and/or recurrent cholesteatoma in patients having previously undergone surgery for cholesteatoma. The non-EPI DWI sequence, in combination with conventional MRI sequences, had replaced the need for second-look surgery in this patient population. Therefore, we decided to review our diagnostic accuracy on cholesteatoma imaging to prove and substantiate that the imaging findings were highly reliable and reproducible. Furthermore, we wanted to evaluate our imaging protocol, as there were recent published data suggesting the sufficiency of only very limited MRI sequences for postoperative cholesteatoma imaging.

The results of this study confirmed the hypothesis that MRI has a very high accuracy rate in postoperative patients with cholesteatoma. These findings are in agreement with published literature. However, interestingly, a few false-positives were recorded during the research—in contrast with other literature. The false-positive findings concerned, almost exclusively, patients who had been operated on using fat transplant during primary cholesteatoma surgery. This fat transplant may result in increased diffusion-weighted signal intensity on the non-EPI DWI sequence. If the non-EPI DWI sequence would be the only sequence used in the screening for residual and/or recurrent cholesteatoma, as suggested in several recent articles, the use of a fat transplant would produce false-positive results. In centers using this surgical technique, the imaging protocol for screening for residual and/or recurrent cholesteatoma should be completed with conventional T1- and T2-weighted sequences, and should not be restricted to the currently recommended imaging protocol of a non-EPI DWI sequence only.

To maintain avoidance of unnecessary surgery in postoperative patients who undergo screening for residual and/or recurrent cholesteatoma, our aim is to establish a false-positive rate as low as possible and a sensitivity as high as possible based on the imaging findings. In our clinical practice, fat transplant is sometimes being used as a surgical technique in cholesteatoma surgery (fat from the abdominal wall is placed in the mastoidectomy cavity). Therefore, in our hospital we do not use the limited MRI sequence protocol recommended worldwide in postoperative patients with cholesteatoma. Instead, we use conventional T1- and T2-weighted sequences in addition to the non-EPI DWI to decrease the risk of misdiagnosis.

Furthermore, during our literature study, we came across recent literature from De Foer et al, which stated that delayed postgadolinium T1-weighted imaging did not add diagnostic accuracy to the non-EPI DWI for evaluating residual and/or recurrent cholesteatoma. Based on this literature, we adjusted the imaging protocol and excluded routine administration of gadolinium for the screening of residual and/or recurrent cholesteatoma. As a consequence, the imaging time and the cost of the imaging protocol have been reduced.

I presented the results of the study at the European Congress of Radiology (2011) before publication. After publication, I presented the results of the study at the Annual Meeting of the American Society of Neuroradiology (2012). Both oral presentations brought up an interactive discussion afterwards. Furthermore, after publication I have had email contact with neuroradiologists all over the world with correspondence about the imaging protocol and so on.

We are expanding the non-EPI DWI sequence in cholesteatoma imaging into a pediatric population and I am devoting myself to a fellowship in pediatric radiology.

 

Read this article at AJNR.org . . .