4Cartilage Shield and Palisade Tympanoplasty

FWhat a cartilage shield is and why it exists

A cartilage shieldis a plate of the patient’s own auricular cartilage — usually taken with one layer of perichondrium — laid across the drum to reinforce it. The word shield is well chosen: the cartilage stands between a fragile or absent tympanic membrane and the two forces that destroy reconstructions, namely the negative middle-ear pressure of a failing Eustachian tube and the point load of an alloplastic prosthesis. Where a thin fascia graft retracts, atrophies and eventually re-perforates or collapses onto the promontory, a cartilage shield is stiff enough to hold its shape. That single property — rigidity — is the reason the technique exists.

The idea is not new. Eviatar described the tragal perichondrium-cartilage composite graft in a large clinical experience in the 1970s, showing it could rebuild a drum even where the remnant was minimal or the annulus was missing [1978]. What has changed since is the evidence base. We now know from a randomised trial and a systematic review that cartilage out-performs temporalis fascia on graft take and needs revision less often, with hearing that is, in aggregate, no worse[2010, 2012]. In other words, the cartilage shield is not a niche salvage trick but a mainstream choice whenever the ear is at risk — atelectatic, revision, poorly aerated, or destined to carry a prosthesis. The chart below summarises the take-rate advantage across three representative series.

A practical caveat anchors the enthusiasm. Cartilage is heavier and stiffer than the native drum, so a careless, thick plate can damp the high frequencies. Almost everything else in this module is about how to keep the protection while giving back the sound: choosing the right donor cartilage, thinning it, and — when it helps — segmenting it into a palisade.

FDonor sites: tragus and concha

Two donor sites supply almost all shield cartilage, and the choice is largely a matter of approach and amount. Tragal cartilageis reached through a small incision on the medial surface of the tragus, leaves no visible scar, and yields a flat, easily handled plate roughly 1–1.5 cm across. It is the default for transcanal and endaural work and for most type 1 reconstructions. Conchal (cymba) cartilage is harvested through, or in continuity with, a postauricular incision; it is slightly more curved but gives a larger area, which matters when the whole pars tensa must be rebuilt or several palisade strips are needed. Both sites carry trivial donor morbidity and both regenerate a perichondrial framework.

Three biological facts make auricular cartilage well suited to the middle ear. It is avascularand survives by diffusion, so it tolerates the moist, sometimes inflamed cleft and resists resorption far better than bone. It is thin enoughat harvest — native tragal and conchal cartilage is about 0.7–1.0 mm — that it needs only modest thinning to reach an acoustically friendly target[2000]. And the attached perichondriumaids take and handling and can be left as a skirt to anchor an island graft. Dornhoffer’s thousand-patient experience leaned on exactly these properties to rebuild drums across cholesteatoma, recurrent perforation and atelectasis with around 96% closure [2003].

TShield versus palisade: trading stiffness for sound

Once the cartilage is harvested there are three broad ways to build the drum from it, and they sit on a single spectrum from maximum rigidity to maximum sound transfer. A solid shield is one full-thickness plate, often notched to straddle the malleus handle; it is the most retraction-proof construct and the natural choice for a wet, atelectatic or revision ear, at the cost of the most high-frequency damping. A thinned shield is the same one-piece plate reduced toward 0.5 mm. A palisadereplaces the single plate with several thin strips laid side by side, parallel to the malleus handle.

The acoustic logic is straightforward and is now well grounded in temporal-bone work. Sound transfer through a cartilage plate is degraded by its mass and stiffness, and both rise steeply with thickness. Laser-Doppler vibrometry showed that thinner and segmented reconstructions preserve high-frequency transmission better than a continuous full-thickness plate, which behaves as one heavy, stiff mass [2002]. The optimum thickness for a one-piece plate is about 0.5 mm— the point of least acoustic transfer loss that still resists deformation under the pressure swings of a working middle ear [2000]. Segmenting into a palisade lowers the effective stiffness further, which is why palisades edge out solid plates acoustically while keeping much of the retraction resistance. The explorer below lets you move along this spectrum and read off the trade-off.

Tos catalogued these variants — palisades and strips, foils and plates, perichondrium-cartilage island grafts and total pars-tensa composite grafts — into a six-group classification that gives a shared vocabulary for what would otherwise be a confusing family of techniques [2008]. The unifying rule that emerges is simple: use the thinnest, least massive cartilage construct that will still resist retraction in this ear. A stable, aerated ear can take a thin palisade; a wet, atelectatic, revision ear may need a sturdier solid shield, accepting a little high-frequency cost for the security of take.

TShaping and seating the graft

Good shield surgery is mostly careful carpentry. After harvest, the cartilage is laid on a cutting block and thinned with a scalpel or a dedicated cartilage cutter toward the 0.5 mm target, keeping one perichondrial surface intact for take and handling. The plate is then sized to the drum so it seats medial to the annulus all the way round, and a notch or groove is cut to accommodate the malleus handle so the graft lies flush rather than tenting laterally over the manubrium. For a palisade, the plate is sliced into strips that are arranged parallel to the handle; for an island graft, a central cartilage disc is left within a wider skirt of perichondrium that drapes over and anchors to the drum remnant.

Seating follows the same principles as any underlay graft, with two cartilage-specific points. First, the construct must rest on a well-vascularised bed— a freshened perforation margin, a de-epithelialised drum remnant, and a middle ear packed enough to support the graft but not so tightly that it lateralises. Second, the graft must not over-ride the round-window niche or wedge against the ossicular chain in a way that loads it; the aim is a membrane that vibrates, not a strut that splints. When these are respected, take is excellent: a full-thickness shield in type III tympanoplasty achieved graft take in every ear in one series, with mean hearing gains around 11 dB and most ears closing the air-bone gap to within 25 dB[2007].

CProtecting the prosthesis interface

The shield earns its keep most clearly where a prosthesis meets the drum. A rigid alloplastic head plate resting on a bare or thin membrane concentrates its load on a tiny area and, over months, erodes through and extrudes. Interposing cartilage between the head plate and the drum spreads that load over a wider area, gives the prosthesis a stable platform that holds its angle and length, and resists the retraction that would otherwise drag the construct medially. This is why titanium and other alloplastic heads are routinely wrapped against a cartilage shield rather than the drum itself, and why cartilage interposition is the standard answer when the malleus cannot be reached and the prosthesis must couple directly beneath the membrane[2003].

Two refinements matter at this interface. Use a perichondrium-cartilage island where you can: the perichondrial skirt anchors the graft to the remnant while the cartilage centre takes the head-plate load, combining stability with take. And keep the cartilage at the interface as thinas the situation allows, because the same mass-and-stiffness penalty that damps a drum graft also blunts the prosthesis’s delivery of energy to the oval window. The clinical pay-off of getting this right is durability: across a systematic review, cartilage reconstructions needed revision roughly half as often as fascia (about 10% versus 19%), which in a prosthesis-bearing ear is the difference between a result that lasts and one that slowly retracts or extrudes [2012].

CChoosing the construct for the ear in front of you

The decision collapses to two questions. First, how hostile is the ear? Negative middle-ear pressure, a history of retraction or atelectasis, revision surgery, a wet or poorly aerated cleft, and the presence of a prosthesis all push toward more cartilage and more rigidity. A dry, primary, well-aerated ear with good Eustachian function tolerates a thin palisade or even fascia. Second, how much does this patient’s high-frequency hearing matter, and how much margin do you have? Where the ear is stable enough to allow it, thin toward 0.5 mm and segment into a palisade to give back the high frequencies; where retraction risk dominates, accept a solid shield and the small acoustic premium that comes with it.

Put together, a workable plan reads: harvest tragal cartilage for transcanal work or conchal cartilage when more area is needed; thin toward 0.5 mm with one perichondrial surface preserved; choose a solid shieldfor the hostile, retraction-prone or prosthesis-bearing ear and a palisade or island for the stable ear that needs high-frequency preservation [2008, 2000]; notch the graft for the malleus, seat it on a vascular bed without splinting the chain, and interpose it against any prosthesis head. The randomised and pooled evidence is reassuring on the trade-off you are making: better take and fewer revisions, with hearing that holds its own — an insurance policy paid in a few decibels of high frequency that careful thinning largely buys back [2010, 2012].