A torn meniscus, a knee that recovered, and a tear that is almost certainly still there. Both facts are true, and the gap between them is where most of the confusion about this injury lives.
A man in his thirties tears his meniscus skiing. The scan confirms it. He is told, reasonably and in good faith, that the meniscus cannot heal itself.
Two years later he is skiing again, running, squatting, living an ordinary life without pain. The tear, if anyone rescanned him, is almost certainly still there.
Both of those things are true at once, and the gap between them is where most of the confusion about this injury lives.
This is the reasoning error underneath almost every conversation about a meniscal tear, and once you separate them it becomes much easier to think clearly.
Pain resolving does not mean the tear healed.
The tear not healing does not mean surgery was needed.
Those are separate claims. People routinely treat them as one, in both directions. Someone whose knee settles concludes the tear must have mended. Someone told the tear is permanent concludes that surgery is therefore inevitable.
A structurally persistent tear can become entirely asymptomatic. That is the ordinary course of events, not an unusual outcome. The scan still shows the tear. The joint stops caring.
The meniscus is not a uniform tissue, and this is the part almost nobody gets told.
The outer third — the red-red zone — has a blood supply. Tears there, particularly longitudinal tears near the edge, genuinely can heal, and can be repaired surgically with a reasonable chance of the repair taking.
The inner two-thirds — the white-white zone — is largely avascular. It receives nutrition by diffusion from joint fluid rather than from blood vessels. Tears there do not heal in the sense of the tissue knitting back together.
So "the meniscus cannot repair itself" is reasonable shorthand for an inner-zone radial or complex tear, and simply wrong for a peripheral one.
Most people are never told which they have. If you take one practical thing from this article, make it this: find out which zone, and which tear pattern. That single piece of information changes what "it can’t heal" means for you, and it is sitting in a report somebody has already written.
Two separate bodies of evidence, on two different populations, and they are often quoted as though they were one.
Degenerative tears — the wear-related tears common from middle age, often found alongside early osteoarthritis. Sihvonen and colleagues published a sham-controlled trial in the New England Journal of Medicine in 2013, in which the comparison group had a genuine operation performed up to the point of doing anything to the meniscus. METEOR, Kise’s trial in the BMJ, the ESCAPE trial with five-year follow-up and the OMEX cohort at ten years all point the same way: structured exercise therapy performs about as well as arthroscopic partial meniscectomy.
Traumatic tears in younger, active knees — the skiing and football population, where the received wisdom has always been that surgery is different. Van der Graaff and colleagues randomised 100 patients aged 18 to 45 in BJSM, following them for 24 months. DREAM reported similarly. Direction of effect: early surgery was not superior.
Notably, 41% of the exercise group in the van der Graaff trial crossed over to delayed surgery. Which is worth reading carefully — it is not a failure of the trial, it is the finding. A period of rehabilitation first does not close the surgical door, and roughly six in ten never needed to walk through it.
I am not going to present that evidence without the argument against it.
In 2018, Hohmann and colleagues assessed the trials of arthroscopic partial meniscectomy and judged all six to be at high risk of bias. Blinding is genuinely difficult in surgical trials. Crossover rates were substantial. Some outcome measures were not designed for these populations.
That critique does not overturn the findings, and the consistency of direction across quite different trials is itself meaningful. But anyone quoting these trials without mentioning the methodological objection is selling you a cleaner picture than exists — and you should be more suspicious of the clean version than the messy one.
There is one presentation where the conservative argument does not apply, and it needs stating plainly because getting it wrong causes harm.
True mechanical locking. A knee that physically will not straighten. A hard block, not stiffness and not pain-limited movement. Genuine giving way with a sense of the joint coming apart. This is the bucket-handle picture, where a fragment has displaced into the joint.
The ESCAPE trial excluded locked knees. This is also the presentation where repair rather than removal is often indicated — preserving the meniscus rather than trimming it.
If that describes your knee, this article is not about you. Go back to your surgeon.
One more piece of context that reframes the whole thing.
Culvenor and colleagues pooled 63 studies in BJSM in 2018 and found meniscal tears in around 19% of asymptomatic adults over 40 — people with no knee pain at all, walking around with a torn meniscus and no idea.
Which means finding a tear on a scan of a painful knee does not establish that the tear is causing the pain. It might be. It might also have been there for a decade.
That cuts both ways, and I want to be careful here. It is not a reason to dismiss anyone’s pain — the pain is real regardless of what the scan shows. It is a reason to be sceptical of the confident causal story, in either direction.
If you have been given a diagnosis and are trying to think it through, these are the ones that actually change the picture:
None of those are challenges. They are the questions a good surgeon will be pleased you asked.
I do not tell anyone whether to have surgery. That decision belongs to you and your surgeon, and a practitioner who nudges you either way on the basis of trial evidence they have read is overstepping badly.
What I can do is the part that nobody covers. Quadriceps and hamstring capacity deficits after a knee injury persist for years, long after the pain has settled and long after anyone stopped measuring. Limb symmetry, single-leg control, calf capacity, tolerance of loaded knee flexion — these are measurable, they are modifiable, and they are almost always invisible because the person stopped hurting and everyone moved on.
Whether or not you have the operation, that work still needs doing. It is the same on both sides of the decision.
This article is general information and not medical advice. Decisions about surgery belong with you and your surgeon. If your knee locks, gives way, or will not fully straighten, seek assessment.
Sources
Sihvonen R, et al. Arthroscopic partial meniscectomy versus sham surgery for a degenerative meniscal tear. New England Journal of Medicine 2013;369:2515-2524.
Katz JN, et al. Surgery versus physical therapy for a meniscal tear and osteoarthritis (METEOR). NEJM 2013;368:1675-1684.
Kise NJ, et al. Exercise therapy versus arthroscopic partial meniscectomy for degenerative meniscal tear in middle aged patients. BMJ 2016;354:i3740.
Noorduyn JCA, et al. Effect of physical therapy vs arthroscopic partial meniscectomy in people with degenerative meniscal tears: five-year follow-up of the ESCAPE randomized clinical trial. JAMA Network Open 2022.
van der Graaff SJA, et al. Arthroscopic partial meniscectomy versus physical therapy for traumatic meniscal tears in a young study population. British Journal of Sports Medicine 2022.
Hohmann E, et al. Arthroscopic partial meniscectomy versus physical therapy for degenerative meniscus lesions: how robust is the current evidence? Arthroscopy 2018.
Culvenor AG, et al. Prevalence of knee osteoarthritis features on magnetic resonance imaging in asymptomatic uninjured adults. British Journal of Sports Medicine 2018.
Quadriceps and hamstring deficits persist for years after a knee injury, on both sides of the surgical decision. They are measurable.
Ask AIdan → See the TDG Programme