15Temporal Bone Surgical Anatomy for the Otologist

FA bone you operate inside

Most surgical fields are surfaces you work upon; the temporal bone is a solid you work within. The otologist does not so much expose the middle ear as tunnel toward it, removing bone in a planned sequence while the structures that matter most — the facial nerve, the labyrinth, the dura, the sigmoid sinus — remain hidden inside the very bone being drilled. There is no retraction that brings them into view; there is only a mental three-dimensional map, continuously updated from landmark to landmark, that tells the surgeon where each lies relative to the bur tip [2016].

For the reconstructive surgeon this map is not an academic indulgence. An ossiculoplasty is only as durable as the disease clearance beneath it, and the recesses that shelter residual cholesteatoma — above all the sinus tympani — lie precisely where the anatomy is hardest to read. The whole purpose of this module is to assemble the mastoid, the attic, the facial recess, and the sinus tympani into a single coherent picture, oriented as you meet them at operation rather than as they appear in a flat textbook diagram.

FFrom the surface to the antrum

Every mastoid operation begins on the lateral surface of the bone, where a shallow triangular depression behind and above the ear canal — MacEwen’s triangle, the suprameatal or mastoid fossa — marks the way in. It is bounded by the supramastoid crest above, the suprameatal (Henle’s) spine in front, and a line dropped from the posterior canal wall behind. Drilling here, in the adult, reaches the mastoid antrumat a depth of roughly 1.5–2 cm; in the infant, in whom the mastoid process is undeveloped, the antrum lies far more superficially and the facial nerve emerges more laterally, which is why the same landmarks carry different margins of safety at different ages[2016].

The antrum is the central air-cell chamber of the mastoid, the hub from which the pneumatised cell tracts radiate. As the surgeon enlarges the cavity, two structures are sought as fixed points of reference: the lateral (horizontal) semicircular canal, a smooth pale prominence on the medial wall of the antrum, and the fossa incudis, the small notch that cradles the short process of the incus. These two landmarks are the surgeon’s anchor, because the facial nerve’s second genu lies in a constant relationship just inferolateral to the lateral canal [2016].

FThe epitympanum and the aditus

From the antrum, a short passage called the aditus ad antrum leads forward into the epitympanum, or attic — the part of the tympanic cavity that sits above the level of the tympanic membrane. The threshold of the aditus is guarded by the body and short process of the incus sitting in the fossa incudis; the head of the malleus lies just beyond. The bulk of the ossicular mass, in other words, resides in the attic rather than at membrane level, which is why attic disease so readily erodes the ossicular chain and why the attic must be opened to deal with it [2016].

The attic is not one open loft but a compartmentalised space. A bony ridge called the cog descends from the tegmen, and the tensor fold stretches across in front of it; together with the malleal and incudal folds they form an epitympanic diaphragm that divides the attic into anterior and posterior compartments and channels its ventilation through a narrow gap, the tympanic isthmus. When the isthmus is blocked and the diaphragm complete, the attic ventilates selectively poorly even when the eustachian tube is working normally — the substrate of selective attic retraction and cholesteatoma that Marchioni termed the epitympanic dysventilation syndrome[2010, 1964]. Reading the cog and tensor fold therefore tells the surgeon not only where the anterior epitympanic recess lies but why disease has collected there.

TThe facial nerve as the master landmark

No structure organises temporal-bone surgery as completely as the facial nerve. Its intratemporal course is conventionally read in segments, and the surgeon follows them as a route rather than memorising them as a list. The labyrinthine segment is the shortest and narrowest, the bottleneck where the nerve is most vulnerable to compressive oedema. It turns at the geniculate ganglion, whose floor is often dehiscent, then runs backward as the tympanic (horizontal) segmentabove the oval window and below the lateral semicircular canal. Here the bony canal is naturally dehiscent in a substantial minority of ears, leaving the nerve covered by mucosa alone — and this is the single commonest site of iatrogenic injury[2016].

At the second genu, inferolateral to the lateral semicircular canal and just below the fossa incudis, the nerve turns downward into the mastoid (vertical) segment and descends to the stylomastoid foramen, signposted along its lower course by the digastric ridge. Reported rates of iatrogenic facial palsy are of the order of 0.6–3.7% in primary mastoidectomy and several-fold higher in revision surgery, with most injuries clustering in the tympanic segment and at the second genu — precisely where landmarks are most easily lost [2016]. The disciplined habit is to identify the nerve from the constant landmarks — lateral semicircular canal, fossa incudis, digastric ridge — rather than from the nerve itself, and to keep its expected line continuously in mind.

TThe retrotympanum: facial recess and sinus tympani

Behind the tympanic membrane, the descending facial nerve divides the posterior compartment into two recesses whose contrast is one of the most clinically loaded facts in otology. The facial recess lies lateralto the nerve, a triangle bounded medially by the facial canal, laterally by the chorda tympani, and superiorly by the fossa incudis. Because it opens toward the surgeon, it can be drilled from the mastoid — the posterior tympanotomy— to deliver a view into the mesotympanum, the round window, and the stapes without removing the posterior canal wall. The window is narrow, and its width tracks the angle at which the chorda leaves the facial nerve (a mean of roughly 24°), so an unusually acute angle leaves little room between the two nerves[2010].

The sinus tympani lies medial to the facial nerve, scooped into the labyrinthine wall between the ponticulus above and the subiculumbelow, with the pyramidal eminence as its lateral lip. This is the great blind spot of the middle ear: it points away from every line of sight the operating microscope can offer, and its depth varies enormously between ears. The depth is graded by Marchioni’s classification — a shallow type A, an intermediate type B that reaches the facial canal, and a deep type C that extends medial and posterior to the nerve. A type C sinus can shelter cholesteatoma matrix that no straight instrument can see or reach, and it is the recess most associated with residual disease after canal-wall-up surgery[2011, 2022].

Because a deep sinus tympani sits beyond the microscope’s reach, two strategies have reshaped its management. The angled endoscope, passed transcanal, looks around the pyramidal eminence directly into the sinus; and for the deepest type C sinuses a retrofacial approach— drilling behind the descending facial nerve from the mastoid — opens the recess from the other side. Knowing in advance, from the preoperative scan, that a sinus is type C tells the surgeon that microscopic transmastoid clearance alone will not be enough[2022].

CHolding the map during reconstruction

For the clinician, all of this converges on a single principle: hearing reconstruction is built on a cleared and correctly read cavity, and the geography of the temporal bone dictates both the clearance and the reconstruction. The choice between a canal-wall-up and a canal-wall-downoperation is in large part a judgement about whether the hidden recesses can be reliably cleared with the wall preserved; the facial recess is the access route that makes canal-wall-up surgery and posterior ossicular work possible, and the sinus tympani is the recess that most often forces a wider exposure, an endoscope, or a retrofacial detour before the cavity can be declared clean[2010, 2022].

Only then does the reconstruction itself begin, and it too is governed by the map. The oval window on the medial wall, with the tympanic facial nerve immediately above it, is the destination of every prosthesis: the footplate for a total reconstruction, the stapes head for a partial one. A prosthesis set toward a misidentified window, or over a footplate beneath retained matrix in the sinus tympani, is a prosthesis destined to fail. The same lateral semicircular canal and facial canal that guided the dissection now define the corridor in which the prosthesis must sit without touching the nerve [2016].

Mastery of temporal-bone anatomy, then, is less the recall of named structures than the ability to hold them in three dimensions and rotate the picture as the operation rotates the bone: surface to antrum, antrum to attic, attic to the facial nerve, and the nerve dividing the recesses where disease hides and reconstruction either succeeds or quietly fails [2016, 2016].