Meniscus repair

Repair puts sutures across a tear and leaves the meniscus in the knee. The mechanical argument for it is direct: the meniscus carries load as a tension ring, and how far contact stress rises tracks roughly linearly with how much tissue is taken out — about 100% after a medial meniscectomy, 200% to 350% after a lateral one. Sutures leave the ring in place; resection shortens it.

The European consensus on traumatic tears puts the same reasoning as a recommendation: preservation should be the first line of treatment wherever it is possible, because the long-term clinical and radiological outcomes are worse after partial meniscectomy than after preservation. That is expert consensus graded as such, not a trial, and the trial it would take does not exist.

Which tears are even repairable

Blood supply decides, and it recedes with age. A developing meniscus is supplied across its whole width; by adulthood only the peripheral 10% to 30% still has a blood supply and everything inside that depends on diffusion. That vascular border is the thick, convex, capsule-attached rim surgeons call the red zone, and it is where the red-red, red-white and white-white classification comes from. It is why a peripheral vertical tear in a young knee gets offered sutures and a central horizontal flap in a sixty-year-old does not.

The trade, quantified

A repair fails often enough that it is the main thing to weigh. Pooled across thirteen studies with at least five years of follow-up, the failure rate — reoperation or clinical failure counted together — was 23.1%, or 131 of 566 repairs. That rate barely moves with technique, with which meniscus was repaired, or with whether the cruciate was intact or deficient, and late failures more than two years out made up 29% of all the failures counted. So it is not a rate that front-loads and then settles.

Push the follow-up past seven years and the picture blurs rather than sharpening. Twelve retrospective case series report failure anywhere from 5% to 48% — a spread that says as much about how each study counts a failure as about the operation. What that review could rule out is that technique explains it: open, arthroscopic inside-out and all-inside repair with flexible implants did not differ from each other, though two individual implants did, at 22.3% against 48%. Failure was significantly higher in children and adolescents than in adults, which is the opposite of what the age-and-healing argument predicts. Six of the twelve reported minor radiographic change differing little from the opposite knee, and clinical scores at final follow-up were good to very good.

Repair is followed by less arthritis. A meta-analysis of twenty studies and 31,783 patients with acute tears found advanced knee osteoarthritis about half as likely after repair at four years (OR 0.51), and reported progression to knee arthroplasty at the same figure. In the same analysis, mean joint space width did not separate, IKDC did not separate, and the resection arm actually had the higher Lysholm score — by four points, which the authors say is below the threshold a patient would notice. Those results point different ways, and the authors report all of them.

Root tears, where the gap is widest

A posterior root tear is a radial tear within a centimetre of the root insertion, or an avulsion of the insertion itself. Mechanically it is closer to removing the meniscus than to tearing it: the meniscus stops converting axial load into hoop stress, and the cartilage degeneration that follows is described in the review literature as comparable to what follows a total meniscectomy. All the tissue is still there and it has stopped doing its job.

Extrusion is the visible sign. The medial meniscus slides out of the joint after both partial and complete root tears, and extrusion beyond three millimetres tracks with more cartilage degeneration and more osteophyte formation.

Pooling thirteen studies, Kellgren-Lawrence grade worsened in 22.2% of knees after medial posterior root repair and 48.3% after resection, and 4 of 95 repaired knees were reoperated against 30 of 92 resected ones. The reoperations were not the same operation, either — a second pullout repair on one side, an osteotomy or a new joint on the other. Lysholm improved more after repair; the Tegner activity score did not separate.

The reviews are careful about how far that goes. Knee function improves after root repair and is reported consistently. Whether the operation prevents osteoarthritis is reported inconsistently, and the reviews say so rather than resolving it.

The comparison nobody has run

Every number above comes from pooled cohorts and case series. Repair against resection has never been randomised, in any tear pattern — and an umbrella review of level 1 evidence across the ten commonest elective orthopaedic operations records meniscal repair as one of only two on its list with no randomised comparison against non-operative care either.

That matters more here than it usually does, because the selection is not random in a knee-shaped way. Surgeons repair peripheral vertical tears in younger, more active patients with better cartilage, and resect degenerative flaps in older knees that are already changing. Both arms of every comparison above were chosen, and the difference between the arms includes the difference between the people in them. The signature of that problem is visible in the results: the structural gaps are large and the patient-reported gaps mostly are not, which fits a real chondroprotective effect and also fits repair being offered to better joints.

One boundary worth being explicit about. The case for preservation is built on traumatic tears, and the consensus that states it says so: it distinguishes traumatic from degenerative injury by cause and is scoped to the first. So "resect less" and "repair instead" are not the same recommendation, and the second does not follow from the first in the degenerative tears that dominate middle-aged practice.

What would settle it

A randomised trial of repair against resection in comparable tears is the missing study in this section, and everything on this page would change meaning if one existed. Two smaller gaps sit behind it. The long-term failure literature reports a range from 5% to 48% and cannot say why, because failure is defined differently in every series in it — an agreed definition would be the cheapest improvement available here. And the root-tear result with the widest structural gap is medial-only, so whether it carries to lateral roots is untested. In the meantime preservation has the better mechanical argument, the better structural numbers, and the weaker study designs.