5Cartilage Interposition at the Prosthesis-Drum Interface

FThe cap that almost every prosthesis wears

Watch an experienced otologist finish an ossiculoplasty and you will almost always see the same final move: before the tympanomeatal flap is laid back down, a small disc of cartilage is slipped between the head of the prosthesis and the undersurface of the drum. This little cap is so routine that it is easy to overlook, yet it is one of the most consequential decisions in the whole reconstruction. Its job is deceptively simple — to stop the prosthesis from extrudingthrough the eardrum — while doing as little harm as possible to the sound coupling the operation was built to restore. The whole module is the story of that single trade-off: a buffer that buys durability at a small, manageable acoustic price.

The need arises from a mismatch of materials. The lateral end of a partial (PORP) or total (TORP) ossicular replacement prosthesis is a flat head plate— rigid, often titanium or a ceramic, and quite unlike the thin, living, three-layered membrane it has to push against. Press a hard plate against a thin membrane and leave it for months and one of two things happens: either the membrane tolerates it, or it slowly gives way and the prosthesis works its way out. Interposing cartilage tips the odds decisively toward tolerance. It is, in effect, a piece of the patient’s own tissue acting as a shock-absorbing, biocompatible washer between hardware and host. The same biological logic that makes cartilage the graft of choice for the difficult drum makes it the cap of choice at the prosthesis interface [2003].

FWhy a bare head plate extrudes

Extrusion is a mechanical and biological process, and understanding it explains why the cap works. A rigid head plate resting directly on the drum concentrates its load on a tiny area of the epithelial surface. That focal pressure, sustained over weeks, produces low-grade pressure necrosis: the keratinising squamous epithelium that lines the outer drum begins to migrate around and over the foreign body, the membrane thins beneath it, and the prosthesis is gradually walled off and pushed laterally toward the canal. Eventually the head plate erodes clean through, sits in the external canal, and the reconstruction fails. The drum’s natural tendency to retract in a poorly aerated ear accelerates the whole sequence by drawing the membrane down onto the unforgiving plate.

The histology of this has been shown directly. In a guinea-pig model, a synthetic prosthesis placed against the drum tended to protrude and extrude, whereas interposing a cartilage disc kept the rigid material separated from the epithelium and markedly reduced both protrusion and extrusion [2002]. The cartilage does two things at once: it physically holds the hard plate off the squamous layer, and — because it is firm but not point-sharp — it spreads the load over a broader, more forgiving footprint. The widget below contrasts the two interfaces and steps through what each looks like months later.

It is worth being precise about which prostheses need this protection. Highly biocompatible ceramics such as hydroxyapatite are, in principle, well tolerated against the drum and were originally marketed as not requiring a cap; in practice, retraction-driven extrusion still occurs and cartilage helps. The workhorse titanium head, although exceptionally biocompatible, is rigid and is conventionally wrapped or capped with cartilage at the drum interface for exactly the reasons above. Older porous polyethylene (Plastipore) prostheses extruded notoriously often without a buffer. Across all of them, the cap is the common answer to a common failure mode.

TWhat the cartilage cap does, and the evidence it works

The single most quotable piece of evidence comes from a clinical comparison of hydroxyapatite prostheses placed with and without an interposed autologous cartilage disc. In ears reconstructed without cartilage, the prosthesis extruded in 13.2%; in ears reconstructed with a cartilage cap, extrusion fell to 1.9%— a roughly sevenfold reduction — and, crucially, the hearing gain was no worse in the capped group [2002]. That last point is the whole argument in miniature: the cap removed most of the extrusion risk withoutsacrificing the acoustic result. More broadly, reported hydroxyapatite extrusion rates of roughly 5–22% with a bare interface fall to the low single digits once cartilage is interposed.

These numbers do not stand alone. They sit on top of a much larger body of work showing cartilage to be the most forgiving tissue in the middle ear: Dornhoffer’s series of more than 1,000 cartilage tympanoplasties documented around 96% closure and durable results even in cholesteatoma, atelectasis and revision ears — precisely the high-risk settings in which a bare prosthesis is most likely to extrude [2003]. And the reason the high-risk ear matters so much is captured by statistical staging of ossiculoplasty outcomes: the middle-ear environment— mucosal health, drainage, aeration — drives results more than the prosthesis material, so the cap earns its keep most in exactly the ears where the environment is poor[2001].

TMaterial, harvest and shaping

The cap is almost always autologous auricular cartilage, taken from the tragus or the concha. Both are within or adjacent to the operative field, are harvested with minimal morbidity, and have the mechanical properties the job needs: they are firm enough to hold a hard plate off the epithelium, yet thin and trimmable; they tolerate the moist, sometimes inflamed middle-ear environment; and, unlike fat or loose connective tissue, they resist resorption, so the protection lasts. Tragal cartilage yields a small, flat, conveniently sized plate; conchal cartilage is slightly more curved but larger. Whichever is used, the perichondriumis often left on one or both faces — it aids handling, improves graft take where the cap also helps reconstruct the drum, and lets the disc be fashioned into a perichondrium-cartilage island when extra anchorage is wanted.

The forms the cap can take have been catalogued. Tos grouped cartilage techniques into a six-part classification — palisades and strips, foils and plates, perichondrium-cartilage island grafts, full pars-tensa composite grafts, and special methods — which gives a shared vocabulary for what is, at the prosthesis interface, usually a simple disc or shield a few millimetres across [2008]. The practical shaping steps are consistent: harvest a piece comfortably larger than the head plate, trim it to a disc that covers the plate with a small margin, and — the part beginners forget — thin it. Native tragal or conchal cartilage is about 0.7–1.0 mm thick, which is more mass than the interface needs; thinning it toward roughly 0.5 mm preserves acoustic transfer while keeping enough stiffness to do its protective job [2000]. A small notch or a cupped centre can be cut to seat the head plate and stop the prosthesis from sliding.

CThe acoustic cost: sizing the cap

Cartilage is not acoustically free. Every disc adds mass and stiffness to the lateral end of the vibrating system, and added mass preferentially damps the high frequencies— the region where the ear’s own transformer contributes most. The clinical message is therefore not “more cartilage is safer,” but “use the smallest, thinnest cap that still protects the interface.” Temporal-bone work makes the cost concrete: across reconstruction variants studied with stroboscopic holography and laser-Doppler vibrometry, the largest high-frequency reductions in tympanic-membrane motion occurred precisely with a large oval of cartilage interposed between the drum and the prosthesis[2014]. The cap that over-protects is the cap that over-damps.

Two design levers control the cost. The first is thickness: thinning toward about 0.5 mm minimises acoustic transfer loss while retaining enough rigidity to resist deformation under the pressure swings of a working middle ear, which is why surgeons shave the native graft rather than use it at full thickness[2000]. The second is area: the disc should cover the head plate with only a small margin, not blanket the pars tensa. An oversized cap not only damps high frequencies but can creep onto the annulus or tent the drum, distorting the very geometry the reconstruction is trying to restore. The goal at the interface is a small, thin, well-seated disc — just enough cartilage to keep the hardware off the epithelium, and no more.

CWhen is the cap optional, and a practical routine

Is the cap truly mandatory? Not in every ear. With a very biocompatible head and favourable conditions — a dry, well-aerated middle ear, healthy mucosa, a thick well-vascularised drum — some series report acceptable extrusion rates with the titanium head placed directly against the undersurface of the membrane, sparing a step and a fraction of mass [2014]. But this is the exception that proves the rule: the ears that tolerate a bare interface are precisely the low-risk ears, and the moment the environment turns unfavourable — atelectasis, recurrent infection, revision surgery, a thin atrophic drum — the calculus swings firmly back toward interposing cartilage, because that is exactly where extrusion clusters[2001].

A workable routine falls straight out of the principles above. First, decide whether you need the cap: in any wet, atelectatic, revision or high-risk ear, and with any rigid head plate coupling directly to the drum, the answer is yes. Second, harvest and shape: take tragal or conchal cartilage, trim a disc just larger than the head plate, leave perichondrium where it helps, and thin toward 0.5 mm[2000, 2008]. Third, seat it: cup or notch the centre so the head plate sits without sliding, and confirm the disc covers the plate with only a small margin rather than blanketing the drum [2014]. Fourth, check the result: light, even contact, no tenting, and a prosthesis held off the epithelium. Done this way, the cap delivers what the evidence promises — extrusion driven down from low double digits to low single digits, with the hearing result preserved[2002, 2002].