What's Actually Going On
Sarcoidosis makes the immune system build small clumps of cells, called granulomas, in whatever tissue they settle in. In the chest that shows up as swollen lymph nodes on a plain x-ray. In the musculoskeletal system it produces four separate patterns rather than one: Löfgren syndrome (acute, ankles, shin lumps, chest nodes, often fever), chronic sarcoid arthritis, bone involvement, and muscle involvement. Across 58 musculoskeletal cases at one centre, Löfgren syndrome was 46.6%, bone 25.9%, chronic arthritis 24.1% and muscle 6.9%.
The part that matters most for a physical therapist is that the acute ankle presentation is usually not a joint problem. Le Bras 2014 scanned 36 consecutive patients with high-resolution ultrasound and blood-flow imaging. 26 of 36 (72.2%) had no ultrasound signs of active arthritis at all. Fluid inside the joint appeared in only 9 of 36, and was mild in 8 of those 9. Meanwhile 23 of 25 (92%) had extensive swelling in the tissue around the joint, and 14 (38.8%) had inflamed tendon sheaths. Mañá 1996 went one step further and biopsied the tissue around four of these ankles: no granulomas in any of them.
Muscle involvement is common and mostly silent. Uddenfeldt 1983 biopsied the calf muscle of 20 unselected patients and found granulomas in 9 of them, of whom only 3 had any muscle ache at all. A granuloma in a muscle is not the same thing as a muscle problem.
How to Identify It
This condition has exactly one published diagnostic instrument, and it is a pattern you score from the history and by looking. It is not a manoeuvre you perform on someone.
- Both ankles affected, symmetrically
- Symptoms present for less than two months
- Age under 40
- Tender red or purple lumps on the shins
Three or more of those four is a positive result. catches 93 in 100 rules out 99 in 100 NPV 99.7% Derived from 579 consecutive patients newly referred with recent-onset joint swelling, 55 of whom had this condition.
The catch, and it is the important part: those figures come from a specialist clinic where roughly 1 in 10 such referrals turned out to have this. In a general practice or a sports clinic the odds going in are far lower, and the "if positive, you have it" side of the test weakens with them while the "if negative, you don't" side holds up. In practice this is a rule-out instrument wearing rule-in numbers, and a positive score means get a chest x-ray, not you have sarcoidosis.
Other findings: movement is characteristically preserved despite dramatic swelling, which is unusual and useful. Onset clusters in spring. Fever in about half. Smoking is negatively associated with this presentation, which is an epidemiological curiosity and is not advice in any direction. No orthopaedic ankle test has any published accuracy here, because those tests examine structures this disease is not damaging.
The Differential
The differential is unusually easy at the pattern level and unusually hard at the tissue level, and that gap is the whole problem.
Two ankle sprains at once, with no incident, is not a thing. That single observation closes the commonest wrong answer. Gout and septic arthritis are single-joint problems, and single-joint involvement here is pooled at 1%, so they rarely compete. Septic arthritis is the one that cannot wait: one hot joint, cannot weight-bear at all, systemically unwell, same-day assessment.
Reactive arthritis follows a gut or urinary infection. Rheumatoid arthritis takes the small joints of the hands and feet over months, and the long-running form of sarcoid arthritis is the harder call against it. Painless pitting swelling in both legs is a heart, kidney or vein problem and a different pathway entirely. And shin lumps have other causes, including throat infections, some medicines, and bowel disease. The lumps do not make the diagnosis. The lumps plus the ankles plus the chest film do.
The Debate
Is it even arthritis?
The whole literature, and the name itself
The condition is called sarcoid arthritis, pooled at 19% of sarcoidosis patients.
vs
Le Bras 2014, Mañá 1996
72.2% have no ultrasound evidence of arthritis. 92% have swelling around the joint instead. 4 of 4 biopsies of that tissue found no granulomas.
Follow the imaging. The name predates anyone looking inside the joint, and it was never revised.
Does it always clear up?
Standard teaching, repeated in current reviews
Löfgren syndrome is self-limiting, with over 90% resolving within two years.
vs
Grunewald & Eklund 2009, 275 patients followed past 2 years
Almost every HLA-DRB1*03-positive patient resolved. 49% of DRB1*03-negative patients did not.
Follow the stratified figure. The reassuring population number is carried entirely by the carrier majority, and the other half is close to a coin toss.
The drug that treats it, and appears to cause it
Smedslund 2022
TNF inhibitors are used in refractory sarcoidosis, recorded in 0 to 100% of patients across studies.
vs
Donzella 2024
43 of 46 published cases of new-onset sarcoidosis in seronegative inflammatory arthritis arose during anti-TNF therapy.
Both are true. Case reports have no denominator, so this is a reason to watch and to tell the prescriber, not a reason to stop a drug. Nobody should stop a biologic because of a web page.
Honest Limitations
1. There is no treatment evidence, and the systematic review says so itself
What the research shows: across eleven studies, steroids were used in 23-100% of patients, anti-inflammatories in 0-100%, hydroxychloroquine in 5-100%, methotrexate in 12-100%, TNF inhibitors in 0-100%.
The gap: those are not doses and they are not effects. They are the spread of who received what across studies with nothing to compare against. None of the eleven had a control group, and non-drug treatment was recorded in ten patients in a single study.
The adjustment: read every treatment on this page as convention, never as evidence. If you ask what the research says about physical therapy here, the honest answer is that there isn't any.
2. The diagnostic criteria do not travel to a general caseload
What the research shows: catches 93 in 100, rules out 99 in 100, with a positive predictive value of 75%.
The gap: 55 of 579 patients in that clinic had the diagnosis, roughly 1 in 10. In a general or sports setting the starting odds are a fraction of that, and the rule-in half weakens accordingly while the rule-out half holds.
The adjustment: use the pattern to decide whether to refer, not to make the diagnosis.
3. The study that reclassifies the whole condition is small and unrepeated
What the research shows: 72.2% with no active arthritis on standardised ultrasound.
The gap: one centre, 36 patients, no comparison group, and not repeated anywhere in the 193 papers reviewed here.
The adjustment: the finding survives its small size because it is structural rather than statistical. A joint either shows capsule distension and active blood flow or it does not. Treat it as the best available description while knowing it rests on one study.
The Nuance
The hardest thing about this condition is that the examination actively supports the wrong answer. Both ankles are swollen, warm and tender. The tissue involved is soft tissue rather than joint. Movement is preserved. Put those four findings in front of any clinician and they describe a sprain, because that is what a sprain is.
Nothing you can do with your hands separates them. What separates them is three questions: were there two of them, did they come up at the same time, and was there any story at all. If the answers are yes, yes and no, the next step is a chest x-ray and not a treatment plan.
There is a second nuance worth carrying, and it is about the word "resolved". Most people are told this settles within two years, and for the genetic majority that is accurate. But roughly half of the non-carrier group is still dealing with it past two years, and nobody knows in advance which group a given patient is in until the results are back. That is a very different conversation from "it'll be fine in a few months", and it is worth having honestly the first time rather than correcting later.
