Learning Objectives:

Introduction: While all otologists operate on cholesteatoma, little analysis exists of the patterns of disease spread which logically should dictate surgical approaches. Nor has comparison been made between different parts of the world.

Method: A detailed data-base was prospectively collected on every cholesteatoma surgery in Tygerberg Hospital between 2003 and 2016 (n = 516). This included inter alia the cholesteatoma's origin from the tympanic membrane, and its presence/not in all parts of the mesotympanum, epitympanum and mastoid. This data is analysed for the various sites and subsites and the site of origin of the cholesteatoma.

Results: Surprisingly, the tympanic membrane origin was: pars flaccida 27%, pars tensa posterior-superior marginal 29%; pars tensa central 30 %; and indeterminate 13% (some had multiple origins).

Cholesteatoma involved the epitympanum in 78%, mesotympanum in 77%, and mastoid in only 52%. Subsite involvement in the epitympanum was: posterior 86%; middle, superficial to ossicles 69%; middle deep to ossicles 54%; anterior 65%, and anterior epitympanic recess 9%.

In the mesotympanum it was central over promontory in 76%; posterior-superior in 72%; in the facial recess and sinus tympani in 48%, anterior in 22% and in the hypotympanum in 12%.

In the mastoid it was restricted to the antrum in 52% and more widely in 26%.

Differences are described in the pattern of spread dependent on the site of origin from tympanic membrane.

Conclusion: This data shows cholesteatoma is primarily a condition of the meso- and epitympanum. One thus needs to be able to remove it from the complex subsites of the middle ear, including retrotympanum, and rehabilitate the mesotympanum. The frequent involvement of the ossicles in the epitympanum demands techniques to mitigate recidivism here. Our South African data is significantly different from what the literature describes - the question is whether this disease is different in different regions.