11Prosthesis Length, Angulation, and Tension

FLength, vector, tension: three trims, one columella

Choosing the family of prosthesis — PORP or TORP, titanium or hydroxyapatite — is only the opening move. The reconstruction that actually restores hearing is made in the last few minutes at the microscope, when the surgeon trims the strut to length, sets its angulation so it sits perpendicular, and seats it under the right tension. These three geometric decisions, not the catalogue choice, decide whether the rebuilt columella vibrates like the chain it replaces.

The goal is to recreate a columella: a single, near-vertical strut that carries vibration from the tympanic membrane or malleus handle straight down onto the stapes head or footplate. The native footplate moves predominantly as a piston, rocking only slightly, across the whole audible range[1986]. A good reconstruction therefore drives the footplate along that piston axis with as little wasted motion as possible. Three properties of the trim govern how well it does so:

• Length— how far the medial end seats, which sets the preload (tension) on the chain.
• Angulation (vector)— how nearly perpendicular the shaft sits to the footplate, which sets how much force is delivered along the piston axis rather than wasted as lateral shear.
• Tension— the resting load between the lateral and medial ends, dominated by the stapedial annular ligament, which supplies most of the chain’s compliance.

Length and angulation are the levers the surgeon pulls; tension is the mechanical state they produce. Get them right and the construct is stable, perpendicular, and lightly sprung; get them wrong by fractions of a millimetre or a few degrees and the same components transmit poorly or slip [1994].

FLength sets tension

The most common beginner’s error is to think of length as merely “bridging the gap.” In fact length is the dial that sets tension, and tension is the single most decisive mechanical variable the surgeon controls. A strut even 0.1–0.2 mm too long drives the foot medially, distending the annular ligament and preloading the entire chain; one slightly too short loses firm contact and leaves the construct loose to the point of instability.

The structure that absorbs and resents this preload is the stapedial annular ligament, the compliant suspension that lets the footplate move. Most of the reconstructed chain’s compliance lives there, and most of the system’s damping sits at that footplate–cochlea interface[1986]. Distend the ligament with an over-long prosthesis and you stiffen the whole transmission line; unload it with an over-short one and the foot rattles instead of pushing. Because length and tension are inseparable, the surgeon sizes the strut not to a ruler but to the behaviourof the chain as it seats — firm contact without tenting the drum or splinting the stapes.

This is why measured length tolerances in ossiculoplasty are so tight, and why adjustable-length and trimmable designs exist: the difference between a snug fit and an over-tight one can be a single trim of the shaft. The practical lesson for the trainee is to stop chasing a number and start reading the chain.

TThe Goldilocks rule of tension

Tension fails in two opposite directions, and the failures are not symmetric. An over-tensioned(too-long) prosthesis stiffens the annular ligament, raising cochlear input impedance. Because low frequencies are stiffness-controlled, it worsens the low and midfrequencies first, leaving a residual air–bone gap, and it risks splinting, subluxating, or even fracturing the footplate, with vertigo or perilymph fistula. An under-tensioned (too-short) prosthesis suffers intermittent contact, acoustic leakage, displacement, and fluctuating hearing.

Temporal-bone experiments resolve the trade-off cleanly. Measuring stapes velocity by laser-Doppler vibrometry across deliberately varied prosthesis lengths, shorter, looser reconstructions gave the best stapes vibration, particularly at low frequencies, while a snug best fit gave the best broadband result and a too-tight fit was worst at both ends of the spectrum [2004]. The same pattern holds when the prosthesis runs from the malleus: low tension was optimal, and the advantage of looseness was greatest at low frequencies [2004].

From this comes the governing maxim of ossicular reconstruction, often credited to Bance and colleagues: aim for the loosest configuration that remains positionally stable. Many surgeons deliberately err to the loose side, because postoperative healing, fibrosis, and retraction tend to tighten the construct over weeks to months, quietly converting a perfect intraoperative fit into an over-tensioned one. A small amount of shaft flexibility helps for the same reason: a slightly flexible strut conforms to the conical drum, accommodates middle-ear pressure swings, and forgives minor sizing errors, whereas a rigidly stiff column transmits every micron of preload straight to the footplate.

TAngulation and the force vector

Correct length is necessary but not sufficient: a perfectly sized strut still transmits poorly if it lies at the wrong angle. Because the footplate works as a piston, only the component of force perpendicular to the footplate plane does useful work; force applied obliquely is partly wasted as lateral shear and tends to slide the head off its seat. Biomechanical and temporal-bone studies converge on a favourable window of roughly 45–90°between the shaft and the footplate or capitulum axis. Angles above 90° disperse force laterally; acute angles below 45° couple markedly worse and destabilise the construct.

The commonest cause of an unfavourable angle is a medialised or foreshortened malleus that drags the lateral end of the shaft toward the promontory, tilting it acute and pushing the head contact eccentrically toward the annulus, where drum vibration is small. The classic remedy is to lateralise the malleus by partial sectioning of the tensor tympani tendon, which swings the manubrium laterally, restores a near-vertical shaft within the favourable window, and re-centres the head contact[1994]. Where the malleus cannot be recruited at a workable angle, a centred drum contact under cartilage is the fallback — but the surgeon should reach for vector correction before reaching for more length or more tension.

TWhere the ends sit: head and foot

Length and angle position the two ends of the columella, and where each end sits has its own consequences. At the lateral (head) end, vibratory amplitude is greatest centrally, near the umbo and manubrium, and falls toward the annulus, so a centrally placed head captures more energy. Coupling to the malleus is the most powerful single choice: a malleus-anchored assembly recruits the natural lever and a stable central contact, and in cadaveric testing it transmitted vibration to the footplate more efficiently than a tympanic-membrane-only assembly, across tension levels [2004]. Clinically, the presence of the malleus handle improved the mean postoperative air–bone gap (11.6 dB with the malleus versus 16.9 dB without) [2001].

Because a bare alloplastic head against the drum tends to extrude, a thin cartilage interposition is interposed for protection — but it must be kept small and thin. Bench work covering a prosthesis head with materials of differing size showed that smaller cartilage covers transmitted best and that cartilage of the size used clinically had little overall effect, whereas large, thick discs add mass and rigidity and steal the high frequencies [2004].

At the medial (foot) end, a PORP rests on the stapes capitulum, which self-centres the load; a TORP rests directly on the footplate and is far less forgiving of length and angle, being prone to displacement or subluxation. Where on the footplate the foot sits also matters: temporal-bone analysis comparing anterior, central, and posterior contact found that the contact site measurably affects transmission, supporting a stable, well-centred seat for the medial end [1999]. A footplate shoe or cartilage shoe can stabilise the foot and preserve its angle when the superstructure is absent.

CTrimming to fit at the microscope

These principles collapse into a short sequence of operative habits that reliably separate good reconstructions from disappointing ones:

• Measure the gap, then size for the loosest stable fit.Trial-seat the strut: it should make firm contact without tenting the drum or splinting the stapes. If in doubt, trim a fraction shorter — over-tensioning is the commoner and more punishing error because healing tightens the construct further[2004, 2004].
• Set the vector before the length. Confirm the shaft sits near-perpendicular to the footplate. If a medialised malleus forces an acute angle, lateralise it by partial tensor tympani sectioning rather than compensating with extra length [1994].
• Recruit the malleus and keep the head central.A malleus-coupled, centrally seated head buys both a better lever and a more stable contact point, narrowing the air–bone gap[2004, 2001].
• Cap thinly. Use just enough cartilage to prevent extrusion; a bulky disc adds mass and rigidity and quietly degrades the high frequencies [2004].
• Seat the foot squarely. Centre the medial end on the capitulum or footplate and stabilise it with a shoe if the superstructure is gone, so the angle and tension you set survive closure[1999].

The unifying idea is that length, angulation, and tension are not three separate adjustments but one act of shaping a columella to the anatomy in front of you. Trim it to the loosest stable length, swing it perpendicular, centre both ends, and the rebuilt strut will drive the footplate along its native piston axis — reproducing the mechanics of the chain it replaces far more faithfully than any catalogue choice ever could.