Even though a ‘second look’ remains a gold standard for detection of residual cholesteatoma after intact canal wall techniques, non-echo-planar diffusion weighted MRI is considered a reasonable alternative to avoid further surgery. However, to establish or exclude a cholesteatoma de novo by this method remains controversial. This study assessed the diagnostic performance of diffusion weighted MRI to pick up cholesteatomas in unclear clinical situations, such as aural polyposis, retracted tympanic membrane, atypical tympanic membrane with possible implantation or congenital cholesteatoma, stenosed external ear canal and no evident disease clinically. This was compared with accuracy of diffusion weighted MRI in cases of clinically evident cholesteatoma.
The results in these two situations were compared and showed a statistical significance. The diffusion weighted MRI was 80% accurate in cases with persistent inflammatory tissue. The accuracy fell to 45% in cases with tympanic membrane retraction. Cholesteatoma was noted in eight out of 26 patients who underwent surgery despite absence of it in this imaging modality.
The authors therefore contend that diffusion weighted MRI has limitations. There are false positive and false negative results which can be misleading in making decisions for surgery. There is no reason why this modality should not be used and may well be complementary. However, the results of the MRI should be used only in context with the clinical findings. Judgements on clinical grounds should not be overruled by use of this modality.