Meniscal allograft transplantation

Once a meniscus is gone, the cartilage underneath is carrying a load it was never shaped to carry. Transplantation puts a size-matched donor meniscus in its place and anchors both horns to bone.

What the graft restores, and on what condition

The meniscus works as a tension ring. Load pressing down on a wedge-shaped meniscus tries to squeeze it outwards, and the circumferential fibres running from one horn to the other resist that — which is how a soft structure spreads load across cartilage instead of being extruded from between the bones. Take the meniscus out and the same load lands on a smaller patch of cartilage at higher pressure, which is the mechanism behind everything on this shelf.

A transplant is an attempt to put that ring back, and its fixation is not a detail. Both horns are anchored to bone, usually with bone plugs or a bone bridge, because a ring that is not anchored at both ends cannot carry tension at all. A transplanted meniscus is a hoop, not a spacer.

Measuring what that achieves inside a living knee is harder than it sounds, and the in-vivo evidence is thin. Two knees measured on the operating table with a surgical navigation system showed anteroposterior laxity and internal-external rotation both falling 25% to 50% once the graft was seated, at 30 and 90 degrees of flexion. Two patients, under anaesthesia, with no follow-up — it shows the graft changes the knee's mechanics immediately, and nothing about whether that persists.

How patients do, and for how long

Every systematic review that has measured patient-reported outcomes found them improved over baseline, most often on the Lysholm score. That finding is consistent and it is the least contested thing about this operation.

Durability is better established than the pooled-failure figures suggest. A single-surgeon series of 109 consecutive Dutch patients — who had each already had an average of 2.8 operations on that knee, one of them fourteen — gives a mean allograft survival of 16.1 years (95% CI 14.8 to 17.5). Age is what moves it: every additional year of age at surgery raised the hazard of failure by 19%.

Failure itself is not one number. Pooled across seventeen studies and 2,184 patients, it runs at 10.9% by five years and 22.7% by ten. Individual studies span 3.3% to 81%, and that spread is not noise about the same quantity — it is studies defining failure differently. The one risk factor that pools cleanly is the state of the cartilage before the graft goes in: damage worse than ICRS grade 3a raised the odds of failure more than fivefold. Patient sex and which compartment the graft went into did not separate.

Complications are common without usually being catastrophic. In a series of 93 knees, 21 (23%) needed at least one further operation, most often debridement of a torn graft.

What the survivorship numbers depend on

That same series reports 100% graft survival among the patients it followed, and then says plainly why the figure cannot be taken at face value. Thirty of its 91 patients did not take part, and institutional records show six of those thirty had already gone on to a knee replacement. The authors put the corrected success rate nearer 85%.

This is worth reading carefully before trusting any survivorship figure for this operation, including the better ones above. Patients whose knee has been replaced have less reason to return for a follow-up questionnaire about the graft it replaced, and a series that counts only the people who came back will report a survival rate that is too high by exactly the amount it failed to trace.

Does it protect the cartilage?

This is the claim the operation is really sold on, and it is the weakest part of the evidence.

The most direct thing that can be said about the literature is what an umbrella review found when it graded it. Forty-one systematic reviews and meta-analyses of transplantation were assessed against AMSTAR-2: 56.1% came out at critically low confidence, 26.8% at low, and 14.6% at high. The recurring faults were missing protocol registration, inadequate risk-of-bias assessment, and unreported conflicts of interest. Its conclusion on the question this section asks is one sentence long — evidence that transplantation delays the onset of osteoarthritis remains limited.

That is not the same as evidence it does not work. It is an accurate statement that the studies asked to demonstrate chondroprotection are, as a body, not built to answer it: no randomised trial of the operation exists, and the radiographic series that report joint space over time measure each knee against its own baseline rather than against a knee that did not get a graft.

Reading this before an operation

The mechanical rationale is clear and the symptom case is consistent and long — improvement over baseline in every review that has looked, and a mean graft survival past sixteen years in the best-followed series. Roughly one graft in nine has failed by five years and one in four by ten, and the strongest thing known about who fails is that it tracks the cartilage damage already present when the graft goes in.

The structural case — the one that justifies operating on a knee to protect it rather than to relieve it — is the part to hold lightly. It rests on a literature whose own appraisers rate most of it critically low.

What would move it

A randomised comparison against continued non-operative care or against meniscectomy alone, with a cartilage-specific imaging endpoint, is the study this field has never had. Short of that, the cheapest available improvement is agreement on what counts as a failed graft: a range from 3.3% to 81% across seventeen studies is mostly a definitional problem, and it is the reason no pooled figure from this literature can be quoted to a patient with a straight face. And any future series should report what happened to the patients it could not trace, because at least one has shown that the untraced are where the arthroplasties are.