Muscle, joint & pain

Platelet Injections for Stubborn Tendons, Trial by Trial

Save

Trial by trial, the history of tendon and joint treatment is a history of things everyone was sure about. A painkiller nobody doubted. An operation performed everywhere. An injection that made perfect sense. Each was eventually tested against a fair comparator, and each came back smaller than advertised. That record is the context any new injection walks into.

Last updated: July 2026

Talk to a clinician

Gale can help you find a clinician in your state and request a visit.

Find care →

Why an injection that promises healing is such an easy sell

Because the alternative on offer sounds like nothing. When a rotator cuff tendon becomes painful and inflamed, the standard menu is rest, anti-inflammatories, physical therapy, and injections 1. Read that list at 2am with a shoulder you cannot lift and it reads as a shrug. Then someone offers to repair the tissue instead of quieting it. Of course that lands.

Tendinopathy is a tendon that has turned painful and intolerant of load; shoulder pain of this kind is often written up as impingement or rotator cuff tendinitis on the referral 1. The names change. The experience does not.

The pitch works on a grammatical trick. Relief is a verb about you. Repair is a verb about the tissue. One promises to change how the shoulder feels this month; the other promises to change what the shoulder is. The second is a far larger claim, and larger claims need larger evidence rather than more enthusiasm.

It is also a claim that travels well. A platelet preparation is offered for tendons, for joints, and — under the heading platelet-rich plasma skin — for faces. A mechanism that elastic might be a genuinely general repair signal, or a general-purpose story. Only the trials tell you which.

The most important tendon trial here is about doing less

It compared three treatments for a torn rotator cuff and found no winner. Adults with nontraumatic supraspinatus tears were randomized to physiotherapy alone, to acromioplasty plus physiotherapy, or to rotator cuff repair plus physiotherapy. At two years, with clinical and imaging follow-up, there was no significant clinical difference between the three groups 2.

Sit with the size of that. One group had the tendon surgically reattached. One group had bone shaved from the underside of the acromion. One group did exercises. Two years later the trial could not tell them apart on clinical outcome, and its conclusion was that conservative care is a reasonable initial option for these tears 2.

A torn tendon and a tendon that needs repairing are not automatically the same object. That is the finding that should reframe the platelet question. If surgically reattaching the tendon did not beat exercise at two years, then the premise underneath every repair-flavoured injection — that the tear is the problem and closing it is the solution — is not a premise this trial supports. The word nontraumatic in those inclusion criteria is load-bearing, and the last section of this page comes back to it.

A painkiller everyone believed in, tested

Paracetamol — the most ordinary painkiller there is. A systematic review of placebo-controlled trials measured it. For low back pain it was ineffective. For hip and knee osteoarthritis, the effect on pain and disability was small and not clinically important 3. Not weaker than expected — effectively absent.

That finding is here for one reason: it is the cleanest demonstration that widespread use is not evidence. Paracetamol was not a fringe product pushed by a clinic with a centrifuge. It was the safest, dullest, most consensus-backed thing on the shelf, and when the placebo-controlled trials were run and pooled, the effect was not there 3.

If that can happen to paracetamol, it can happen to anything. The useful question about a treatment is never how many people use it, or how sensible the mechanism sounds. It is what happened when somebody compared it with a convincing fake.

A procedure everyone performed, tested

Arthroscopic surgery for the degenerative knee. A systematic review and meta-analysis of its benefits and harms found at most a small and short-lived benefit for pain, no benefit for function, and a set of harms that came along with it — a package the authors judged unsupported for middle-aged and older patients 4. This was not an obscure operation.

Look at the exact shape of that result, because it is a shape you will meet again: small, and short-lived 4. Not zero. A small, real, temporary improvement is the easiest result for a treatment to produce and the easiest to mistake for a repair. It is what you get from attention, from the ritual of a procedure, from the natural settling of a flare, and from a genuine but modest treatment effect — and from the outside, all four look identical.

Small and short-lived is the default result in this field. A treatment claiming more is claiming to be unusual, and unusual claims are the ones worth checking hardest.

Another injection with a measured ceiling

Epidural corticosteroid for sciatica, which has been meta-analysed properly. It produces small, short-term relief of leg pain and disability. It produces no meaningful long-term benefit. And — this is the endpoint that matters — it does not reduce the rate of subsequent surgery 5. The injection changes the month. It does not appear to change the trajectory.

That third endpoint is the one worth stealing. Pain scores are soft, self-reported, and they move for a dozen reasons that have nothing to do with the treatment. Whether people ended up having the operation anyway is hard. It either happened or it did not, and it is very difficult for a hopeful trial design to massage 5.

So the question to put to any tendon injection is not whether the pain score improved. It is whether fewer of these people ended up needing the thing the injection was supposed to prevent. Other injections have their own measured ceilings — hyaluronic acid injection for the knee is its own question with its own page — and the pattern across this page's sources, small and short and not trajectory-changing 45, is the sensible starting expectation. None of that is a verdict on platelets. It is a description of what the injection literature has looked like every previous time somebody measured carefully.

What the platelet trials would have to show

A bar, now that the field's habits are visible. This page cannot tell you whether PRP works for a tendon: the platelet trials are not among its sources, and a verdict with nothing behind it is worth nothing. But the sections above describe what a convincing platelet trial would have to do, which is more useful than a borrowed opinion.

  • A sham arm, not a comparison with nothing. The paracetamol reversal only happened because someone ran placebo-controlled trials 3.
  • Follow-up past the point where a small, short-lived effect would have faded 4.
  • A hard endpoint. Did fewer people go on to have the operation — the thing an epidural steroid failed to change 5?
  • A comparator that is real care. Beating rest is not the question. The rotator cuff trial's physiotherapy arm, the one that matched surgery at two years 2, is the arm to beat — and the one most likely to be quietly skipped.

There is a documented reason to be careful rather than merely cynical. Low-value care for low back pain — unnecessary imaging, opioids, injections, and surgery — is widespread across the world, and the call in the literature is to reduce it 6. That finding is about backs, not tendons. What it establishes is that widespread use and good evidence come apart in this field.

The trial-by-trial work lives elsewhere in the library: the PRP RCT summary reads the platelet literature directly, prp vs cortisone sets the two needles against each other, and prp for knee arthritis takes the joint rather than the tendon. What prp injection cost adds is its own question, better answered before the decision than after.

Where an operation on a tendon is clearly the right call

The rotator cuff trial answers this by what it excluded. Its patients had nontraumatic tears — tendons that had frayed over years, not tendons torn by an event 2. That single word is the boundary of the finding. A trial of degenerative tears says nothing about the tendon that ruptured when someone caught a falling ladder, and was never designed to.

Read the inclusion criteria. A trial's conclusion stops exactly where its population stopped. That is the most useful habit on this page, and it cuts in both directions. It is why conservative care first is well supported for a degenerative cuff tear 2. It is equally why nothing in this article is an argument against operating on a tendon that a surgeon judges needs operating on.

The honest limit is that this page cannot hand you the full list of situations where surgery is clearly indicated. Those criteria are tendon-specific, and they belong to the guideline for that tendon and to a surgeon who has examined it. What can be said is that the list is real, and that its decisions are made tendon by tendon — whether a gluteal tendon repair beats an injection for a painful lateral hip, whether a biceps tenodesis is the right call rather than a tenotomy. Each is its own question with its own evidence.

And the standard menu is not a placeholder. Rest, anti-inflammatories, physical therapy, and injections are what orthopaedic guidance lists for a painful rotator cuff 1 — injections included, not grudgingly. The order of care exists because the cheap and reversible things sometimes work, not because anyone is withholding the good stuff. When they do not, the next step is still there.

Common questions

Not a question this page can answer, because the platelet trials are not among its sources and a verdict without one is worth nothing. What it can give you is the bar: a sham arm rather than a comparison with nothing, follow-up long enough that a small short-lived effect would have faded, a hard endpoint such as whether people still had surgery, and a comparator of real physiotherapy.

The trial that asked randomized adults with nontraumatic supraspinatus tears to physiotherapy alone, acromioplasty plus physiotherapy, or cuff repair plus physiotherapy. At two years there was no significant clinical difference between them, and the authors concluded conservative care is a reasonable initial option. Note nontraumatic: a tendon torn by an event was not studied, and that conclusion does not reach it.

Because they are the field's cautionary tales, and they are the reason to be careful rather than cynical. Paracetamol was the most ordinary painkiller there is, and placebo-controlled trials found it ineffective for back pain and not clinically important for hip and knee arthritis. Knee arthroscopy was performed everywhere, and the meta-analysis found at most a small, short-lived pain benefit alongside harms. Consensus is not evidence.

It is a point about the source material, not about the effect. Where a substance comes from speaks to its safety profile and to the appeal of the story. It does not predict whether the injection changes a tendon — that is a separate question, and the only thing that answers it is a trial with a sham arm and a long enough follow-up. Natural is not an endpoint.

No, and reading it that way would be a genuine mistake. The rotator cuff trial studied nontraumatic tears only; a tendon torn by an event sits outside its conclusion entirely. Orthopaedic guidance lists injections among the standard options for a painful cuff. The argument here is about order and evidence, not about refusing treatment, and the criteria for surgery belong to a surgeon who has examined the tendon.

Related

Say it back

How would you explain this to someone you love?

Two or three sentences, just as you’d say it. Gale reflects back what you focused on — a mirror, not a quiz.

Talk to a clinician

Gale can help you find a clinician in your state and request a visit.

Find care →

When a tendon problem needs seeing quickly

  • A sudden pop or snap in the tendon during effort, followed by an inability to use the limb normally — no rising onto the toes, no lifting the arm at all, no straightening the knee.
  • A tendon or joint that is hot, red, and swollen along with a fever or spreading redness, particularly in the days after an injection.
  • Sudden weakness in the limb rather than pain — an arm that will not lift even once the pain is dulled, or a foot that slaps against the floor when walking.
  • Pain that wakes you from sleep and will not settle with a change of position, or pain alongside unexplained weight loss or a history of cancer.

A hot, swollen, painful tendon or joint with a fever — above all within days of an injection into it — belongs in an emergency department the same day, because an infection there is treated in hours. A sudden snap with loss of function needs urgent assessment rather than a wait for the next appointment.

Gale's library explains what the trials measured and how to read them. It cannot examine your tendon and does not know your history. Nothing here is a treatment recommendation for you, and what belongs in your own sequence of care is a conversation with a clinician who can assess the tendon itself.

References

  1. 1.American Academy of Orthopaedic Surgeons (OrthoInfo) (2024). Shoulder Impingement / Rotator Cuff Tendinitis. OrthoInfo — AAOS. linkLay-education account of shoulder impingement and rotator cuff tendinitis, and of its nonsurgical management — rest, NSAIDs, physical therapy, and injections — used for the article's description of the standard menu and its point that injections are part of that endorsed menu rather than excluded from it.
  2. 2.Kukkonen J, Joukainen A, Lehtinen J, et al. (2015). Treatment of Nontraumatic Rotator Cuff Tears: A Randomized Controlled Trial with Two Years of Clinical and Imaging Follow-up. Journal of Bone and Joint Surgery (American). doi:10.2106/JBJS.N.01051Physiotherapy alone, acromioplasty plus physiotherapy, and rotator cuff repair plus physiotherapy produced no significant clinical difference at two years for nontraumatic supraspinatus tears, and conservative care is a reasonable initial option — the article's central tendon finding, and, via the trial's nontraumatic inclusion criterion, the explicit boundary of that conclusion.
  3. 3.Machado GC, Maher CG, Ferreira PH, et al. (2015). Efficacy and safety of paracetamol for spinal pain and osteoarthritis: systematic review and meta-analysis of randomised placebo controlled trials. BMJ. doi:10.1136/bmj.h1225Paracetamol is ineffective for low back pain and gives only a small, not clinically important effect on pain and disability in hip and knee osteoarthritis — the article's demonstration that widespread use is not evidence, and that a placebo-controlled comparison is what settles the question.
  4. 4.Thorlund JB, Juhl CB, Roos EM, Lohmander LS (2015). Arthroscopic surgery for degenerative knee: systematic review and meta-analysis of benefits and harms. BMJ. doi:10.1136/bmj.h2747Arthroscopic surgery for degenerative knee disease provides at most a small, short-lived benefit in pain and no benefit on function while carrying harms, and is not supported for middle-aged and older patients — the article's demonstration that a widely performed procedure can still lose to its comparator, and the source of its 'small and short-lived' pattern.
  5. 5.Pinto RZ, Maher CG, Ferreira ML, et al. (2012). Epidural Corticosteroid Injections in the Management of Sciatica: A Systematic Review and Meta-analysis. Annals of Internal Medicine. doi:10.7326/0003-4819-157-12-201212180-00564Epidural corticosteroid injections give small, short-term relief of leg pain and disability in sciatica but no meaningful long-term benefit and no reduction in subsequent surgery — the article's example of an injection with a measured ceiling, and its argument for judging an injection on a hard endpoint rather than a pain score.
  6. 6.Buchbinder R, van Tulder M, Öberg B, et al. (2018). Low back pain: a call for action. The Lancet. doi:10.1016/S0140-6736(18)30488-4Low-value care for low back pain — unnecessary imaging, opioids, injections, and surgery — is widespread globally and should be reduced; used only for the over-treatment claim that widespread use and good evidence come apart in this field, and explicitly scoped to back pain rather than tendons.

6 sources, numbered by first appearance. General health information, not medical advice. AI-assisted editorial content — citations link their sources. Editorial policy