Diagnosis, mastoidectomy, and ossicular reconstruction for a destructive middle-ear lesion
Cholesteatoma is not a tumor but an abnormal growth of keratinizing squamous epithelium trapped within the middle ear and mastoid. Left undiagnosed, it behaves like a slow, locally destructive mass — expanding, accumulating debris, and eroding the bone around it. Early otoscopic and radiologic recognition is what allows surgery before irreversible damage occurs.
Most acquired cholesteatoma begins with chronic Eustachian tube dysfunction. Negative middle-ear pressure draws the thin, collagen-deficient pars flaccida (Shrapnell's membrane) inward, forming a retraction pocket. If the pocket cannot self-clean, exfoliated keratin accumulates layer upon layer inside it — the pocket becomes a closed sac of dead squamous debris.
Once trapped, keratinocytes continue to proliferate and desquamate, generating a slowly enlarging, onion-skinned mass. A hyperproliferative, cytokine-active matrix (the outer squamous epithelial lining) and perimatrix (subepithelial connective tissue rich in inflammatory cells) actively secrete collagenases and osteoclast-activating cytokines — this is why cholesteatoma erodes bone even without infection, unlike a passive cyst.
Key Insight: cholesteatoma growth is not passive accumulation — it is biologically active. Matrix keratinocytes and perimatrix inflammatory cells release IL-1, TNF-α, and matrix metalloproteinases that recruit osteoclasts, so bone erosion can outpace the physical growth of the debris mass itself.
On otoscopy, the hallmark finding is a retraction pocket or perforation in the pars flaccida (attic) or posterosuperior pars tensa, often with visible white/pearly keratin debris, foul discharge, or granulation tissue at the margin. A congenital cholesteatoma instead presents as a pearly-white mass medial to an intact, normal-appearing tympanic membrane, classically in the anterosuperior quadrant, in a patient with no history of perforation, otorrhea, or prior ear surgery.
Audiometry typically shows a conductive hearing loss from ossicular involvement; a mixed or sensorineural component raises concern for labyrinthine fistula or inner-ear invasion. Any unilateral conductive loss with a retraction pocket should be treated as cholesteatoma until proven otherwise.
Non-contrast high-resolution CT (HRCT) of the temporal bone is the diagnostic workhorse: it defines a non-dependent soft-tissue opacity in the epitympanum or mastoid antrum, with characteristic bony changes — scutal blunting, erosion of the lateral epitympanic wall, widening of the aditus ad antrum ("Prussak's space" expansion), and ossicular erosion or displacement.
CT cannot reliably distinguish cholesteatoma matrix from granulation tissue or cholesterol granuloma on density alone, so diffusion-weighted MRI is increasingly used pre-operatively (and for surveillance) because cholesteatoma keratin restricts water diffusion and lights up distinctively, while inflammatory tissue does not.
Untreated cholesteatoma does not stay contained. It expands centrifugally along the path of least resistance — through the attic into the mastoid antrum and air-cell system — and it erodes every structure it contacts: the scutum, the ossicular chain, the bony labyrinth, the tegmen, and the facial nerve canal.
The ossicular chain erodes in a predictable pattern dictated by blood supply and matrix contact. The long process of the incus is the most frequently and most severely eroded structure — it has the most tenuous blood supply of the three ossicles and lies directly in the path of attic-based cholesteatoma. The stapes suprastructure (crura) is next most vulnerable; the footplate itself is more resistant. The malleus handle and head are eroded less often and later in the disease course, usually only with extensive attic disease.
Erosion occurs through a combination of direct pressure necrosis from the expanding matrix, osteoclast-mediated resorption driven by perimatrix cytokines, and impaired periosteal blood flow beneath the advancing matrix sheet.
From the epitympanum, disease advances through the aditus ad antrum into the mastoid antrum and the honeycomb of mastoid air cells, progressively hollowing out and thinning the surrounding trabecular bone. As pneumatized cells are destroyed, the mastoid can become a single confluent cavity packed with keratin debris.
Critical adjacent structures at risk include the tegmen tympani/mastoideum (roof, separating the middle ear from the middle cranial fossa — erosion risks dural exposure or CSF leak), the bony labyrinth (erosion of the lateral semicircular canal produces a labyrinthine fistula and positional vertigo), the fallopian canal housing the facial nerve, and the sigmoid sinus plate.
Key Insight: because cholesteatoma expands along low-resistance air-cell tracts rather than uniformly, the size of the visible otoscopic pocket routinely underestimates the true extent of disease — a small attic perforation can mask an extensively eroded mastoid on CT.
As erosion progresses, patients develop worsening conductive hearing loss (ossicular discontinuity), chronic foul-smelling otorrhea (secondary bacterial/fungal colonization of the debris), and — once the labyrinth, facial nerve, or intracranial structures are involved — vertigo, facial nerve paresis, and rarely intracranial complications such as meningitis, epidural abscess, or sigmoid sinus thrombosis.
This natural history is the entire rationale for surgery: cholesteatoma is not managed medically. Antibiotics can quiet secondary infection temporarily, but only complete surgical removal of the matrix halts the destructive process.
Once the decision to operate is made, the central strategic choice is whether to preserve the bony posterior external auditory canal wall (canal wall up, "intact canal wall") or remove it to exteriorize the mastoid into the ear canal (canal wall down, "open cavity"). Neither approach is universally superior — each trades recurrence risk against long-term cavity morbidity.
Canal wall up (CWU) mastoidectomy preserves the posterior bony canal wall, maintaining a normal-appearing ear canal and tympanic membrane position, avoiding the need for lifelong cavity cleaning, and allowing patients to swim and wear hearing aids without cavity-related complications.
Its drawback is visualization: hidden recesses (the sinus tympani, facial recess, and anterior epitympanum) are harder to inspect and clear completely behind an intact wall, so residual microscopic matrix is more likely to be left behind — the dominant reason CWU carries a higher recurrence rate and usually mandates a planned second-look procedure.
Canal wall down (CWD) mastoidectomy removes the posterior canal wall, converting the mastoid and middle ear into a single, open, exteriorized cavity continuous with the ear canal (often combined with a meatoplasty to widen the canal opening). This maximizes direct visualization of the epitympanum and mastoid, allowing more complete matrix removal in a single stage and lowering recurrence risk.
The tradeoff is a permanent mastoid "cavity" that requires periodic office debridement, is prone to intermittent discharge or debris accumulation, complicates hearing-aid fitting, and generally restricts water exposure indefinitely.
Surgeons weigh several factors: extent of disease (extensive attic/mastoid disease, an already poorly pneumatized or sclerotic mastoid, or a low-lying tegmen/sigmoid sinus favor CWD), status of the contralateral ear (a CWD cavity is more tolerable if the only-hearing ear), patient factors (ability to attend follow-up cavity care, occupational water exposure, MRI surveillance access), and surgeon experience.
Hybrid and modified techniques (e.g., canal wall reconstruction/obliteration with cartilage or bone pâté) attempt to combine CWD-level clearance with CWU-level cavity comfort, and are increasingly favored where expertise allows.
Key Insight: the approach is a tradeoff, not a hierarchy — CWD trades a lower recurrence rate for a lifelong cavity-care burden, while CWU trades a more "normal" ear for a materially higher chance of needing a second operation to find residual disease.
| Product | Indication | Trial Design | Key Result |
|---|---|---|---|
| Canal Wall Up (CWU) | ~20–40% residual/recurrent disease | None — normal canal, no chronic debridement | Routine (protocol DWI-MRI at 1–3 yrs) or planned second-look surgery |
| Canal Wall Down (CWD) | ~5–15% residual/recurrent disease | Ongoing — periodic office cleaning of open mastoid cavity | Cavity can often be inspected directly in clinic; MRI less essential |
| CWD + Obliteration | Low, similar to standard CWD | Reduced — cavity partly filled (cartilage/bone pâté/fat) | Smaller cavity, faster epithelialization, still needs imaging surveillance |
Whichever approach is chosen, the single most important determinant of surgical success is not the mastoidectomy technique itself but the completeness of matrix removal. Cholesteatoma recurs because matrix is left behind, not because the wrong cavity shape was chosen — every fold, recess, and ossicular surface must be inspected and cleared.
Complete matrix removal requires meticulous, patient dissection under the operating microscope (and increasingly, angled rigid endoscopes for recesses the microscope cannot see around corners into). The matrix is carefully separated as an intact sheet wherever possible from the ossicles, dura, facial nerve sheath, and labyrinthine windows, working from the periphery inward so that no fragment is left stranded and unaccounted for.
High-risk hidden sites include the sinus tympani (a deep recess medial to the facial nerve, poorly seen even with wide exposure), the facial recess, the anterior epitympanic space, and the region around an already-eroded stapes footplate or a labyrinthine fistula, where aggressive matrix stripping risks profound sensorineural hearing loss.
Key Insight: when matrix overlies a labyrinthine fistula or the exposed facial nerve, surgeons often deliberately leave a thin island of matrix in place ("staged" removal at second-look) rather than risk labyrinthine or facial nerve injury — a calculated tradeoff between completeness and safety.
It is essential to distinguish two failure modes. Residual cholesteatoma is matrix inadvertently left behind at the index operation that regrows — a technical/visualization problem. Recurrent cholesteatoma is a new retraction pocket that develops later from ongoing Eustachian tube dysfunction, forming an entirely new lesion even after complete initial clearance — a disease-biology problem.
This distinction matters clinically: residual disease is reduced by better technique, better exposure (CWD, endoscopes), and second-look surgery; recurrent disease is reduced by reconstructing a stable, well-supported tympanic membrane and canal wall that resists re-retraction (e.g., with cartilage grafting of the attic or posterosuperior quadrant).
Because residual matrix is often microscopic and invisible even to a meticulous surgeon, many centers plan a routine second-look procedure 6–18 months after CWU mastoidectomy, re-opening the ear specifically to search for and remove any residual pearl of matrix before it can enlarge — and often to complete ossicular reconstruction at the same time once the cavity is confirmed disease-free.
Diffusion-weighted MRI has reduced, but not eliminated, the need for routine second-look surgery: it reliably detects residual cholesteatoma pearls down to about 2–3 mm, so many practices now use serial MRI surveillance instead of automatic re-exploration, reserving surgery for imaging-confirmed recurrence.
Removing the cholesteatoma cures the disease, but it typically leaves a disrupted or absent ossicular chain and a large conductive hearing loss. Ossiculoplasty — bridging the gap with a prosthesis — is what converts a disease-free but deaf ear back into a hearing one.
A Partial Ossicular Replacement Prosthesis (PORP) is used when the stapes suprastructure (the crura) is intact — the prosthesis bridges from the tympanic membrane (or malleus handle) to the stapes head/capitulum, a short and mechanically efficient path that preserves the natural lever and impedance-matching action of the stapes.
A Total Ossicular Replacement Prosthesis (TORP) is required when the entire stapes suprastructure has been eroded or must be removed, leaving only the mobile footplate. The prosthesis then spans directly from the tympanic membrane to the footplate — a longer path with a wider contact area on the footplate, which is mechanically less efficient and carries a higher risk of footplate fixation, fistula, or sensorineural injury, explaining its comparatively worse average hearing outcome.
Key Insight: hearing outcome after ossiculoplasty depends more on the status of the stapes superstructure and the presence of a mobile, intact footplate than on the specific prosthesis material used — an intact stapes arch is the single best predictor of a good result.
Modern prostheses are most commonly hydroxyapatite, titanium, or polymer/ceramic composites, chosen for biocompatibility, low extrusion risk, and (for titanium) a favorable stiffness-to-mass ratio for sound transmission. A cartilage cap is frequently placed between the prosthesis head and the tympanic membrane graft to reduce the risk of extrusion through thin overlying skin.
Reconstruction timing is debated: single-stage reconstruction (at the time of matrix removal) avoids a second anesthetic and gets the patient hearing sooner, but risks the prosthesis being displaced or extruded if disease recurs and a revision is needed. Staged reconstruction — placing the prosthesis at a planned second-look procedure once the ear is confirmed disease-free — sacrifices several months of hearing but improves long-term prosthesis stability and audiometric outcome, and is generally preferred after canal wall up surgery.
Surgical hearing success is measured by the postoperative air-bone gap (ABG) — the difference between air-conduction and bone-conduction thresholds on audiometry, representing how much conductive loss remains after reconstruction. A postoperative ABG of ≤20 dB is the commonly used benchmark for "successful" ossiculoplasty (American Academy of Otolaryngology–Head and Neck Surgery Committee guidelines).
Factors reducing success include an absent or immobile stapes footplate, mucosal disease/adhesions in the middle ear, eustachian tube dysfunction causing recurrent negative pressure on the reconstructed chain, and prosthesis extrusion. Even when the disease is completely eradicated, roughly a quarter to half of ears (more with TORP) will not fully close the gap to normal — hearing restoration, unlike disease clearance, is never guaranteed by surgery alone.