8Imaging Ossicular Erosion, Fixation, and Dehiscence

FWhy CT, and what it can and cannot show

The preoperative imaging study for the middle ear is non-contrast high-resolution CT (HRCT) of the temporal bone, acquired at sub-millimetre slice thickness with a bone reconstruction algorithm and reviewed in multiple planes. It is the only routine test that resolves the bonymiddle ear — the ossicles, the otic capsule, the scutum, and the thin walls of the facial and semicircular canals. MRI, by contrast, does not depict cortical ossicular bone; its role is to characterise soft tissue (chiefly cholesteatoma on diffusion-weighted sequences), not to grade an eroded ossicle. So when the question is “what is the state of the chain and the bony anatomy around it,” CT is the workhorse [2023].

It is just as important to understand what CT cannotreliably tell you. The structures that most often decide a reconstruction — the slender long process of the incus and the stapes superstructure — are the very ones that sit at or below the resolution limit. A CT that does not show a defect in those segments is indeterminate, not negative. The honest preoperative posture is therefore: let CT confirm aeration, define soft-tissue extent, map bony landmarks, and warn of dangerous variants, while accepting that the definitive ossicular grade remains an operative palpation finding.

What a good HRCT reliably contributes before ossiculoplasty:

• Aeration and soft tissue. Is the middle ear and attic ventilated, or filled with mucosa, effusion, granulation, or a soft-tissue mass suggesting cholesteatoma?
• Bony landmarks and danger zones.The position of the tegmen, sigmoid sinus, jugular bulb, and the course of the facial nerve — the anatomy that shapes the surgical approach and its risks.
• Gross ossicular status. Presence, displacement, or major destruction of the malleus and incus, and the relationship of the chain to an overlying mass.
• Erosion of the bony walls.Scutum, lateral semicircular canal, facial canal, and tegmen — the complications that change consent and operative strategy.

FReading ossicular erosion and discontinuity

The long process of the incus is the Achilles heel of the chain: it is the most frequently eroded ossicular segment in chronic otitis media, undermined by a slender pedicle and a tenuous, effectively end-arterial blood supply. On CT, erosion shows as shortening, thinning, or frank absenceof the descending long process, loss of the small lenticular “dot” that normally marks the incudostapedial joint, and a widened gap toward the stapes head. The mirror finding — discontinuity— is inferred when a segment is missing or when the expected articulation is replaced by a gap.

How well does CT actually perform against the operating microscope? In direct CT-versus-operative correlation in chronic otitis media, the bulky malleus and incus are detected with high sensitivity and near-perfect specificity, but the stapes superstructure is the least reliable, with sensitivity falling into the mid-70s [2024]. The pattern is intuitive: the thicker the ossicle, the more dependable the CT call; the thinner and more obliquely oriented the segment, the more it is smeared out by partial-volume averaging. This is exactly why a “negative” CT for long-process or stapes erosion should never be read as an intact chain.

Two practical reading habits follow. First, trace the chain segment by segment in more than one plane: the malleus handle and incus body in the axial “ice-cream cone,” then the long process descending toward the oval window, then the stapes. Second, read the company the erosion keeps. Scutum blunting with a Prussak-space mass and medially displaced ossicles points to acquired cholesteatoma; a clean, aerated middle ear with isolated long-process thinning is more typical of post-inflammatory necrosis. The erosion pattern, not just its presence, foreshadows what the reconstruction will face.

TFixation: otosclerosis and the fissula ante fenestram

Erosion breaks the chain; fixation stiffens it. The commonest cause of a fixed footplate is fenestral otosclerosis, and its CT signature is a focal hypodense (otospongiotic) plaque at the fissula ante fenestram— the islet of connective tissue just anterior to the oval window, between it and the cochleariform process. Active, immature foci are conspicuous as lucencies; with footplate involvement the plate itself may look thickened. The Symons–Fanning grading scheme stratifies disease by site — purely fenestral versus retrofenestral(cochlear) involvement — and has good inter- and intra-observer agreement, which makes it a useful structured way to report the extent[2009].

The crucial caveat is that CT sensitivity for otosclerosis is incomplete. As an otospongiotic focus matures and re-mineralises, its density approaches that of the surrounding otic capsule and it becomes hard to see; sclerotic-phase disease with a smooth bony contour can be radiologically silent. A normal CT therefore does notexclude otosclerosis — the diagnosis in a patient with a maximal conductive loss, an intact mobile drum, a type A tympanogram, and absent reflexes remains substantially clinical and audiometric, with CT confirming and grading rather than ruling out.

Fixation can also arise proximalto the stapes — tympanosclerotic ossification of the anterior malleal ligament or cementing of the malleus head in the epitympanum — producing a conductive loss with a continuous-looking but immobile chain. CT may show calcific deposits in the attic, but here too imaging is a hint, not a verdict: the level of the block (footplate versus epitympanic) is confirmed by intraoperative palpation and determines whether the operation is a stapedotomy or a malleus mobilisation.

TBony dehiscence: facial canal and superior canal

Two bony dehiscences change the operation and the consent even when the chain itself is intact. The first is dehiscence of the tympanic (horizontal) segment of the facial canal, just above the oval window and lateral to the cochleariform process. CT may show absence of the thin bony cover with the nerve apposed to the middle-ear cavity. Detection is imperfect because that cover is near the resolution limit, but preoperative CT of the tympanic facial canal correlates well with operative findings when axial, coronal, and sagittal planes are reviewed together, and forewarns the surgeon of an exposed nerve before any drilling around the oval window [2011]. A reported or suspected dehiscence raises the threshold for facial-nerve monitoring and careful dissection.

The second is superior semicircular canal dehiscence (SCD)— absence of bone over the apex of the superior canal at the arcuate eminence. SCD matters to the ossiculoplasty surgeon chiefly as a mimic: it creates a low-frequency pseudoconductiveair–bone gap by opening a “third window” into the labyrinth, yet the reflexes are intact and the symptoms (sound- or pressure-induced vertigo, autophony) are characteristic. Recognising it on CT prevents a fruitless middle-ear exploration for a gap that no ossiculoplasty can close.

SCD imaging is, however, exquisitely technique-dependent, and it is the cleanest illustration of how slice thickness drives the diagnosis of any subtle bony defect. On thick (1.0 mm) sections, partial-volume averaging overcalls dehiscence: the positive predictive value of an apparent dehiscence is only about 50%, rising to roughly 93% with 0.5 mm collimation [2003]. Reformatting along the canal in the Poschl and Stenvers planes pushes overall accuracy close to 99% [2017]. The radiologic appearance must always be reconciled with symptoms and physiologic testing before SCD is diagnosed.

TTechnique, pitfalls, and mimics

Almost every diagnostic error in middle-ear CT is, at root, a technique or anatomyproblem. The SCD data above generalise: thick sections smear small structures, so subtle erosion is missed and subtle dehiscence is invented. The defences are the same throughout the temporal bone — sub-millimetre acquisition, a bone algorithm, and disciplined multiplanar (and where needed oblique) reformatting before any small structure is called normal or abnormal.

The recurring pitfalls worth naming explicitly:

The last row deserves emphasis. The temporal bone is riddled with normal sutures, fissures, small canals, and vascular variants— the petrotympanic fissure, the singular canal, an aberrant carotid, a high or dehiscent jugular bulb — that can masquerade as pathology to the unwary. Knowing the normal roadmap is the prerequisite for calling anything abnormal [2005]. When in doubt, the question to ask is not “is there a line here?” but “does this line track a known normal structure, and does it persist across planes?”

CFrom the report to the operative plan

Imaging earns its place only when it changes a decision. For the reconstructive surgeon, each CT finding maps to a concrete operative consequence:

• Long-process or lenticular erosion with an intact, mobile stapes→ anticipate a short-gap reconstruction (bone cement re-bridging, incus interposition, or a partial ossicular prosthesis) that exploits a preserved superstructure, and have it on the tray.
• Absent stapes superstructure→ plan for a total ossicular prosthesis onto the footplate; but because CT under-reads the stapes, confirm by palpation and keep both PORP and TORP available[2024].
• Fissula ante fenestram focus / footplate fixation→ this is stapes-surgery territory, not a gap to be bridged; grade by site and counsel accordingly, remembering CT can be normal in mature disease[2009].
• Tympanic facial canal dehiscence→ raise the threshold for nerve monitoring, modify dissection around the oval window, and consent specifically for facial-nerve risk [2011].
• Superior canal dehiscence→ recognise the pseudoconductive gap, avoid a needless middle-ear exploration, and redirect to vestibular work-up [2003].
• Scutum blunting with a Prussak mass→ treat as cholesteatoma until proven otherwise; plan an atticotomy or canal-wall procedure and expect ossicular involvement.

The unifying principle is one of calibrated trust. CT is excellent for aeration, soft-tissue extent, bony landmarks, gross ossicular destruction, and the dangerous dehiscences — provided it is acquired and read with the right technique. It is weakest precisely where the reconstruction is decided: the thin long process and the stapes. The skilled surgeon therefore reads the scan to prepare and to avoid surprises, while reserving the final ossicular grade for the moment the chain is palpated under the microscope.