3Cortical Bone Autograft Columellae

FWhy cortical bone? The case for a bony strut

An ossiculoplasty needs a columella— a strut that bridges the mobile stapes (or its footplate) to the drum or malleus and carries vibration across the gap a diseased chain has left behind. The material of that strut can be the patient’s own ossicle, the patient’s own bone or cartilage, donor tissue, or a manufactured prosthesis. Cortical bone autograft— usually a chip of dense bone taken from the mastoid cortexor canal wall already exposed during the operation — sits firmly in the biological camp. It is attractive for three plain reasons: it is rigid, so a sculpted strut holds its length and transmits force without buckling; it is the patient’s own living tissue, so it provokes no immune rejection and almost no foreign-body reaction; and it is already in the surgical field, needing no separate donor incision and no implant cost.

Biological grafts of this kind have long been valued for one statistic above all: a very low extrusion rate. A rigid manufactured head plate resting against the drum tends, over months, to erode through and be expelled; an autograft, being self-tissue, is far better tolerated and is reported to extrude at a fraction of the rate seen with early alloplastic materials. That tolerance is the historical reason natural materials — sculpted incus, cortical bone and cartilage — dominated reconstruction before the modern titanium-and-cartilage era, and it is why they remain a sound choice when prostheses are unavailable, unaffordable or simply not preferred [1983]. Austin, who codified ossicular reconstruction, framed the problem in terms of what is missing and what must be rebuilt, leaving the surgeon to choose a material that reproduces the needed geometry; cortical bone is one durable, biologically safe way to do that [1971].

The strut must obey the same acoustic rules as any prosthesis: it should be light enough not to load the chain with mass (which damps high frequencies), stiff enough to transmit without flexing, and stableenough not to tilt or migrate. Cortical bone scores well on rigidity but is heavier than a thin-walled titanium strut, so the surgeon’s task is to sculpt away every unnecessary milligram while keeping the cross-section that gives the strut its stiffness. That tension — lighten, but keep it rigid — runs through every step of the carving.

FHarvest and the sculpting sequence

The harvest is opportunistic: dense cortical bone is lifted from the mastoid cortex or the bony canal wall that the approach has already exposed, so there is no second donor site and no added morbidity. The one inviolable rule is to take bone from a clean, disease-free part of the field. This matters most in cholesteatoma surgery, where the temptation to reuse a convenient but matrix-engulfed ossicle is exactly the temptation to avoid: reimplanting epithelial rests risks seeding recurrent disease, and a fresh cortical chip sidesteps that hazard entirely [1983].

Sculpting is unhurried microsurgery with a diamond or cutting burr under continuous irrigation to avoid thermal injury to the bone and to nearby structures. The strut is roughed out slightly long, then a shallow cup or notch is drilled in its medial end so it seats and self-centres on the stapes head; a captured foot resists tilting without crimping. It is then thinned to shed mass while retaining rigidity, and finally trimmed to a just-contacting length and seated, ideally with a cartilage shield interposed at the drum. The step-through below walks the sequence.

Two practical caveats deserve emphasis. First, sculpting cortical bone is technically demanding and time-consumingunder the microscope; this labour, and the variability of graft availability and integrity, is the chief everyday argument against autografts and in favour of off-the-shelf prostheses. Second, the lateral interface — where the bone meets the drum — should almost always be protected by a thin cartilage shield, which spreads the point load and curbs erosion, the same principle that governs any rigid strut abutting a membrane.

TGeometry: reproducing the columella

A cortical bone strut can reproduce any of the standard columellar geometries, and the choice follows Austin’s defect groups [1971]. When the stapes superstructure is intact and mobile, a short columellaruns from the stapes head to the drum or malleus handle — the most forgiving and durable arrangement, because the load path is short and the strut is well captured on the capitulum. When the superstructure is absentbut the footplate is mobile, a longer strut must reach all the way to the footplate — a long columella that is inherently less stable, more prone to tilt and displacement, and which benefits from any retained malleus to steady it. Series of cortical bone columellas confirm the expected gradient: ears with an intact stapes superstructure achieve better hearing than those reconstructing to a bare footplate [1983].

Whatever the length, three geometric goals are constant. The strut should sit on the mobile part of the stapes, aligned so its force vector points into the oval window rather than skewing across it; it should make light, even contact laterally without tenting the drum or over-loading the footplate; and it should be captured— notched onto the stapes head and shielded laterally — so it cannot migrate. A cortical strut that is well captured and well aligned behaves, mechanically, much like the sculpted incus it replaces, with the added advantage that there is no incus-to-prosthesis junction to loosen [2011].

TThe biology: survival, creeping substitution and resorption

It is tempting to picture a bone strut as an inert peg, but a cortical autograft is biologically active. Like cortical grafts in orthopaedic surgery, it is incorporated by creeping substitution: host capillaries and osteoprogenitor cells slowly invade the dense graft, osteoclasts resorb devitalised bone, and osteoblasts lay down new bone in its place. Cortical bone revascularises far more slowly than cancellous bone — its dense structure is a barrier to vascular ingrowth — so this remodelling unfolds over months, and at any moment the graft is a mixture of living and dead bone. When cortical bone columellas were explanted and examined histologically, that is precisely what was found: 78% were a mixture of living and dead bone, 9% were mainly vital bone, and only 13% showed no surviving bone at all [1987].

This biology is the root of the material’s one real liability: resorption. If the balance of creeping substitution tips toward net resorption — as it tends to in a poorly aerated, inflamed or chronically drainingear — the strut can thin, shorten and lose contact, causing a delayed-onset conductive loss months or years after an initially good result. The corollary is reassuring and clinically actionable: in a healthy, well-aerated ear the same graft behaves well. A long-term mastoid-cortical-bone series reported no extrusion, necrosis or resorption over years of follow-up, and a remarkable stapedectomy cohort followed for two to three decades saw the columella deteriorate by under one decibel per year, with the late drift attributed mainly to age-related sensorineural decline rather than failure of the bone strut itself [2014, 2011]. Bone resorbs when the ear is hostile and survives when the ear is healthy.

CHearing outcomes and how bone compares

Hearing results with cortical bone columellas are solidly in the range expected of any good partial reconstruction. In a long-term mastoid-cortical-bone series the mean air-bone gap fell from 31.6 dB to 20.3 dBand stayed there over three to six years — with canal-wall-up ears closing to about 19.9 dB and canal-wall-down ears to about 21.0 dB — and, notably, no extrusion, necrosis or resorption [2014]. Older comparative work found that cortical bone columellas gave somewhat greater air-bone-gap improvement than ossicle grafts, recommending bone specifically when the patient’s own ossicles are diseased and unusable [1983]. The chart contrasts these bony results with a benchmark sculpted-incus series.

For perspective, sculpted incus interposition— the classic autograft yardstick — closes the air-bone gap from about 26.8 dB to 18.6 dB, brings roughly two-thirds of ears to within 20 dB, and produces essentially no extrusions [2005]. Cortical bone lands in the same territory: a little behind incus in absolute gap closure in some series, ahead of it in others, and broadly comparable to modern alloplastic prostheses when the ear environment is favourable. The honest summary is that material is not destiny. Statistical staging of ossiculoplasty outcomes repeatedly shows that the middle-ear environment— mucosal health, aeration, drainage, prior surgery, stapes status — outweighs the strut material chosen [2001]. A cortical bone columella in a dry, aerated, superstructure-intact ear will usually outperform any prosthesis in a wet, atelectatic, footplate-only ear.

CChoosing wisely: when bone is the right strut

Cortical bone autograft is at its best in a defined niche. Favour itwhen the native ossicle is diseased or unusable but the ear is otherwise healthy — a dry, well-aerated cavity with sound mucosa and, ideally, an intact mobile stapes superstructure; when a separate, immunologically safe, low-extrusion biological strut is wanted at no implant cost; and when prostheses are unavailable or not preferred. It is especially apt in cholesteatoma surgery as a fresh strut that avoids reimplanting a matrix-laden ossicle [1983, 2014].

Be cautiousin the ears where its biology works against it: poorly aerated, atelectatic, chronically draining or heavily revised middle ears, where resorption is more likely and a non-resorbing titanium strut (cartilage-shielded) or a staged reconstruction may serve better. And respect its costs — the demanding intra-operative sculpting and the variable availability of healthy graft bone. The practical workflow falls out cleanly: assess the environment and the stapes; if the ear is healthy and the superstructure intact, harvest clean cortical bone, sculpt a captured, lightened columella of the right length, shield the lateral interface with cartilage, and protect aeration. Done so, a chip of the patient’s own mastoid becomes a rigid, biocompatible columella that can carry hearing for decades [2011, 2001].