10The Facial Nerve in the Middle Ear: Course and Surgical Hazards

FWhy the facial nerve is the surgeon’s first thought

Of all the structures crowded into the middle ear, the facial nerve is the one whose injury a patient notices in the mirror. A misplaced curette can deafen an ear, but a deafened ear is private; a facial palsy is worn on the face for life. For this reason the seventh cranial nerve is the structure most feared during ossicular surgery, and a careful otologist orients to it before touching anything else. The nerve does not merely pass near the operative field — for much of its intratemporal journey it forms the very wall against which the surgeon works, separated from the instrument by no more than a wafer of bone, and sometimes by nothing at all [2016].

Within the temporal bone the facial nerve takes the longest bony course of any nerve in the body, running from the internal auditory canal to the stylomastoid foramen through a narrow channel called the fallopian canal. It changes direction twice, and at each turn it disappears behind a landmark, so that the surgeon who has lost track of the nerve has usually lost it at a corner. Mapping its segments is therefore not an academic exercise but the mental scaffold on which every safe middle-ear and mastoid manoeuvre is built.

FThe intratemporal course: five segments

The intratemporal facial nerve is conventionally divided into five parts. The labyrinthine segmentis the first and the shortest — only a few millimetres — running from the fundus of the internal auditory canal forwards to the geniculate ganglion. It is also the narrowest part of the canal, which is why oedema here, in trauma or in idiopathic palsy, throttles the nerve so readily. At the geniculate ganglion the nerve makes its first sharp turn (the first genu) and gives off the greater petrosal nerve; the ganglion may be dehiscent on the floor of the middle cranial fossa, a relation that matters in middle-fossa and lateral skull-base work[2016].

From the ganglion the tympanic (horizontal) segmentruns backwards along the medial wall of the tympanic cavity, immediately above the oval window and below the bulge of the lateral semicircular canal. This is the segment the ossiculoplasty surgeon meets. At the posterior end of the cavity the nerve makes its second turn — the second genu, lying just behind the pyramidal eminence — and descends as the vertical mastoid segment to exit at the stylomastoid foramen. Along the way the mastoid segment gives the nerve to stapedius and, lower down, the chorda tympani, which loops back up into the middle ear between the incus and the malleus.

TThe tympanic segment over the oval window

For ossiculoplasty the tympanic segment is the segment that matters, because it runs directly over the destination of every prosthesis. A partial prosthesis is set on the stapes head; a total prosthesis is set on the footplate within the oval window. The horizontal facial nerve forms the roof of that window, separated from it by little more than a millimetre — micro-CT studies place the nerve a mean of around 3 mm above the posterior edge of the oval window, with dehiscences clustering just above and slightly behind it [2016]. To work on the footplate is therefore to work in the immediate shadow of the nerve.

The clinical consequence is direct. When disease — granulation, cholesteatoma matrix, or tympanosclerosis — sits in the oval window niche, the temptation is to curette it away. But the curette in that exact spot is the commonest path to an iatrogenic palsy, because the nerve so often lies there unprotected. The trained eye learns to read the landmark: a smooth white tube of bone is the intact canal, whereas a soft, pinkish, sausage-shaped structure with fine longitudinal vessels is the bare nerve, and it changes everything about how the niche is cleared [2016].

TThe dehiscent canal: how common, and where

Natural absence of the bony canal — a dehiscence— is not a rarity but an expectation. A systematic review and meta-analysis of nearly six thousand cholesteatoma operations put the pooled prevalence of facial canal dehiscence at roughly a quarter of ears, and found that around 94% of these dehiscences lie in the tympanic segment[2023]. In other words, in cholesteatoma surgery the question is less whether the nerve might be exposed than whether it is exposed here, today. Anatomical and micro-CT series in undiseased bones report dehiscence of the tympanic segment in well over half of specimens, usually as small elliptical gaps in the inferior wall above the oval window, and frequently bilateral[2016].

Two practical lessons follow. First, dehiscence is commoner in adults than children, in revision than primary surgery, and in cholesteatoma than in mucosal disease — and it rises with the duration of disease, so the long-standing, previously operated cholesteatoma is exactly the ear in which the nerve is most likely to be bare and most likely to be scarred into the matrix [2020, 2009]. Second, imaging cannot be relied upon to warn you. High-resolution CT of the temporal bone is useful but imperfect for the facial canal: reported false-negative rates of nearly a fifth and high false-positive rates mean that a “normal” canal on CT does not license a careless curette [2012]. The surgeon must assume dehiscence until the operative view proves otherwise.

TThe mastoid segment and the facial recess

The vertical mastoid segment is the surgeon’s concern whenever access to the posterior mesotympanum is gained from behind. The posterior tympanotomy— the window drilled through the facial recess in canal-wall-up surgery and cochlear implantation — is opened in the triangle bounded medially by this descending nerve and laterally by the chorda tympani. Here the bone is thinned deliberately down toward the nerve, and the margin for error narrows as the recess deepens. In canal-wall-down cavities the same nerve is encountered when the facial ridge is lowered to exteriorise the mastoid; lowering the ridge improves the cavity but carries the nerve closer to the burr at every pass [2016].

It is no surprise, then, that when iatrogenic facial palsy does occur, the segments implicated mirror this anatomy. In a 20-year referral series of iatrogenic injuries, the tympanic segment was the most frequently injured part, involved in around three-quarters of cases, with the second genu next and the mastoid segment third; in nearly half of patients more than one segment was damaged, reflecting injuries that begin at a corner and run on[2023]. The chart below shows this distribution.

These data also explain why the second genu earns special respect. It is the point at which the nerve turns out of the surgeon’s line of sight, just behind the pyramidal eminence, so that a burr advancing confidently along the horizontal nerve can run onto the genu before the surgeon registers the change of direction [2023].

CAvoiding iatrogenic injury at ossiculoplasty

The principles that keep the nerve safe are few, old, and non-negotiable. The first is orientation before dissection: identify the lateral semicircular canal, the cochleariform process, the oval window and the second genu, and hold the line of the nerve in mind before any instrument approaches the medial wall. The second is technique appropriate to the bone you find. Over an intact canal a curette may be used with care; over a dehiscent or prolapsed nerve it must not. Sharp, tangential dissection along the axis of the nerve, lifting disease off the nerve rather than pressing toward it, replaces the curette; bipolar diathermy and any sustained pressure on the nerve are simply forbidden [2016].

The third principle is the willingness to compromise the reconstruction for the nerve. A dehiscent, slightly prolapsed tympanic segment can overhang the oval window and obstruct the ideal vertical seating of a total prosthesis. Faced with that conflict, nerve integrity always wins: accept a less-than-perfect prosthesis angle, choose a partial reconstruction onto a mobile stapes if the footplate cannot be safely reached, or stage the ossiculoplasty for a later, cleaner ear. A good hearing result built over a damaged nerve is no result at all [2023, 2020].

Two adjuncts support, but never replace, this discipline. Intraoperative facial nerve monitoring gives an audible warning when the nerve is mechanically or electrically stimulated and is widely used, particularly in revision and cholesteatoma surgery; but a stable trace must never be read as permission to push, because monitoring detects stimulation, not safety, and a sharp transection can occur with little warning[2023]. Preoperative CTflags the high-risk ear — the previously operated, the cholesteatomatous, the ear with a fistula — but its unreliability for the canal itself means the operative finding always overrides the scan[2012]. The safest surgeons are not those with the best images but those who assume the nerve is bare, work as though it is, and are content to leave the ear with imperfect hearing and a moving face.