How the diagnosis gets made
A pediatric rheumatologist puts together the pattern of bone pain, the appearance and distribution of lesions on MRI, blood work, and the absence of infection or malignancy. Several formal criteria sets exist, including the Jansson and Bristol criteria.
New since 2025
EULAR and the American College of Rheumatology published the first joint classification criteria for pediatric CNO in 2025, developed across a dozen countries and validated in a separate cohort of 514 children, where they performed at 82 percent sensitivity and 98 percent specificity.
The important caveat: classification criteria are not diagnostic criteria. They exist to define consistent study populations so that drug trials in CNO become possible, and they are not intended for deciding whether your individual child has the disease. A child can fall outside them and still have CNO. If a clinician cites them as a reason your child does not have CNO, that is a misuse of the tool and a fair thing to question.
They matter to families anyway, because the absence of an agreed case definition has been one of the reasons no treatment has ever been approved for this disease.
Median time to diagnosis is commonly reported at around 15 months, and plenty of families describe two years or more of being told it was growing pains, a sports injury, or an infection. That delay is a feature of the disease being rare and having no defining blood test, not a failure on your part.
Whole-body MRI and dedicated MRI are different tools
This comes up constantly, and misunderstanding it leads to real problems. The two scans answer different questions.
If a whole-body MRI came back clean but the pain is real
A negative whole-body MRI does not close the question on a symptomatic area. It is reasonable to ask for a dedicated MRI of that region, and it is reasonable to ask again if time has passed, because CNO lesions wax and wane and a scan is a snapshot of one day. Asking for a targeted scan of a painful area is a normal request, not a difficult one.
What each test contributes
- Whole-body MRI
- The current standard for mapping disease at diagnosis. Active lesions show as bone marrow edema, bright on fluid-sensitive sequences such as STIR. No radiation. Often done without contrast.
- Dedicated MRI
- Higher detail for a single region. The right tool for the spine, for confusing pain, and for assessing whether a specific lesion is improving.
- Plain X-ray
- Often the first imaging done. Can show sclerosis, lysis, or hyperostosis, but is frequently normal early in the disease. A normal X-ray does not rule out CNO.
- Bone biopsy
- Not required to diagnose CNO, and not able to confirm it. Used to exclude tumor or infection, most importantly when there is a single lesion or the imaging looks atypical. See the section below.
- ESR and CRP
- Inflammatory markers. Often mildly elevated, but commonly normal even during active disease. Normal labs do not mean the disease is quiet.
- Bone scan
- An older approach, largely replaced by whole-body MRI, which gives better detail without radiation.
Do they always do a biopsy?
No, and this is worth understanding properly, because the answer is neither "always" nor "never" and both of those get repeated in family groups.
Biopsy is not required to diagnose CNO. Consensus recommendations state this directly. CNO remains a diagnosis built from clinical findings, labs and imaging, with other causes excluded. Around a third of the children in the EULAR/ACR validation cohort had never had a biopsy. If your child has a characteristic multifocal picture at typical sites, a biopsy may add nothing.
A biopsy cannot confirm CNO either. There is no finding that says CNO. What a biopsy shows is inflammation, sometimes fibrosis, and negative cultures. That is consistent with CNO but it is also the point: the job of the biopsy is to rule things out, not to rule CNO in.
Why the diagnosis can still be uncertain after a biopsy
This is frustrating but medically possible. A biopsy can make infection or malignancy less likely without proving that CNO is the answer. The rheumatologist still has to put the pathology together with the lesion's MRI appearance, its location, the clinical course, cultures and laboratory results, and whether other lesions are present.
This matters most with an apparently isolated lesion, such as one in the jaw. If the diagnosis stays uncertain, a useful question is whether whole-body MRI has been done, not only MRI of the painful bone. CNO lesions can be clinically silent, so finding additional ones in characteristic locations can turn an apparently unifocal presentation into a multifocal pattern and materially strengthen the diagnosis.
When consensus does call for one
Published recommendations name specific situations, and in these a biopsy is the appropriate step rather than an overreach:
- A single lesion, other than in the clavicle or mandible. This is the big one. A solitary bone lesion is the presentation that overlaps with bone tumors such as osteosarcoma and with infection, and distinguishing them matters enormously.
- Atypical imaging: abscess formation, fistula, involvement of the skull, or a diffuse pattern that does not look like CNO.
- No response to adequate second-line treatment, where the question becomes whether the original diagnosis was right.
The purpose in each case is excluding primary or secondary bone tumors and infection. Guidance also recommends the decision be made jointly across pediatric rheumatology, radiology, orthopedics or surgery, oncology and infectious disease, rather than by one clinician alone.
Before deciding a biopsy is unnecessary
If your child has one lesion and no whole-body imaging yet, that is the situation where a biopsy is most defensible, not least. Declining it on the basis that biopsy is not routine in CNO risks delaying a diagnosis that is not CNO at all.
What can reduce the need for one
Whole-body MRI, done first. If it finds additional lesions at characteristic sites, the picture becomes typical of CNO and the argument for biopsy weakens considerably. The classification criteria reflect this: a single lesion counts against a CNO classification, while a multifocal pattern counts strongly toward it.
So the productive question is usually not "does he really need a biopsy" but "has he had a whole-body MRI yet, and what did it show?" That is the study that may answer the question without an invasive procedure, and it is reasonable to ask whether it should come first.
If a biopsy is going ahead
Ask that the microbiology be thorough. Consensus recommendations specify extended bacterial culture of at least 10 days, molecular testing for bacteria, and testing for typical and atypical mycobacteria. A short standard culture can miss a slow-growing infection, and a falsely negative result sends everyone down the CNO path with an infection still untreated. This is a specific, reasonable thing to request.
- What exactly are you trying to rule out with this biopsy?
- Has he had whole-body imaging, and would that change whether we need it?
- Is this a single lesion, and is it in a site where biopsy is usually recommended?
- Who is involved in this decision besides rheumatology?
- What cultures and molecular tests will be sent, and for how long will they be held?
- What happens if the result is inflammation with negative cultures, which is the expected CNO result?
Preparing for an MRI
A whole-body MRI is a long scan, commonly 45 to 90 minutes, and the child has to stay still for it. This is the part of CNO care families ask about most and get told about least.
- Sedation or awake
- Younger children, roughly under six or seven, usually need sedation or general anesthesia. Older children often manage awake, especially with preparation. Ask which is planned, because sedation means fasting instructions and a longer day.
- Child life specialists
- Most pediatric centers have them, and they are the single most useful resource here. They offer mock scanners, practice sessions, and distraction techniques that let many children avoid sedation entirely. Ask for one rather than waiting to be offered.
- Noise and comfort
- MRI scanners are loud. Earplugs or headphones are standard, and many centers can play music or a movie. Ask what is available at yours, and bring a comfort item.
- Contrast
- CNO lesions are usually identified on non-contrast fluid-sensitive sequences, so contrast is not always needed. Worth asking whether it is planned and why, particularly if an IV is a significant stressor for your child.
- Practical setup
- No metal, comfortable clothing without zips or snaps if the center allows own clothes, and feet-first positioning where possible if your child is anxious in enclosed spaces.
- Afterwards
- Ask when results will be available, how you will get them, and request a copy of the written report rather than only a verbal summary. Those reports accumulate into the record that supports insurance appeals and second opinions.
Scan anxiety is worth treating as a real problem
Children with CNO have repeat imaging for years. A child who develops genuine dread of scans is facing that many more times, so it is worth investing in preparation early rather than getting through each one. Child life involvement, a consistent routine, and honesty about what will happen do more than reassurance does. If dread builds anyway, see fear of scans and procedures.
Who is on the team
A pediatric rheumatologist coordinates care and is the person who makes the treatment decisions. Depending on how your child presents, others get drawn in:
- Pediatric radiologist
- Ideally one who reads CNO imaging regularly, since the patterns are distinctive and easy to misread as infection or tumor by someone who rarely sees them.
- Orthopedics
- For structural complications: pathological fracture, vertebral collapse, limb length differences, deformity.
- Gastroenterology
- CNO is associated with inflammatory bowel disease. New or persistent gut symptoms get referred here, and a coexisting diagnosis changes which drug is chosen.
- Dermatology
- For the psoriasis and other skin conditions that travel with CNO, and for the paradoxical psoriasis that TNF inhibitors can trigger.
- Ophthalmology
- Less commonly needed, but CNO can involve eye inflammation. Report new eye redness, pain, or vision change promptly rather than waiting, since eye inflammation can be quiet and still be doing damage.
- Pain management and pain psychology
- For pain that persists beyond what disease control explains, and for the skills side of living with a body that hurts. See managing pain.
- Dentistry and orthodontics
- For jaw lesions, and required before starting a bone-directed drug.
- Infectious disease
- Usually involved early, to confirm infection has been properly excluded rather than to treat anything.
Dedicated CNO programs exist at a small number of centers and run this as a single multidisciplinary clinic rather than as a series of separate referrals you chase yourself. If you are coordinating all of this alone, that is a reason to ask about a second opinion at such a center.
Questions worth asking at the first visits
- How many CNO patients do you currently treat?
- How many lesions did imaging show, and exactly where are they?
- Is there any spinal involvement, and how will we monitor it?
- Is any lesion near a growth plate, and what does that mean for growth?
- What is the treatment plan, and what would make us change it?
- What symptoms should make me call you rather than wait for the next visit?
- What activity is safe right now, and what is not?
- Are there registries or clinical trials we should join?
- How often will we scan, and what are we looking for each time?
- How will you measure whether the disease is active, and is there anything you want me tracking between visits?
Second opinions
Expertise in CNO is concentrated at a relatively small number of pediatric centers. A second opinion from a rheumatologist who sees CNO regularly is reasonable in almost any case, and is worth pursuing if the diagnosis was made outside a major pediatric center, if treatment is not working, or if there is spinal involvement. Asking for one is not a criticism of your current doctor and most are used to it.
Is it genetic?
For the large majority of patients, no single gene explains it. CNO has a complex genetic background involving the IL-1 pathway among others, and it is not inherited in a simple pattern. Siblings carry a somewhat elevated risk but the absolute risk stays low. Mention it to your rheumatologist if a sibling develops unexplained bone pain.
A small minority of cases are caused by identified single-gene variants. These usually present differently, often much earlier or with other organ involvement, and identifying them can change treatment.
Most families will never need this table. It is here for the ones who have been told genetic testing is relevant to their child.
What genetic testing can and cannot tell you
Some children are offered a broad primary immunodeficiency or autoinflammatory gene panel. This is more likely when the presentation is unusually early, includes recurrent or unexplained high fevers, unusual infections, significant skin or gastrointestinal disease, a strong family history, or other features that make the team wonder whether a rare single-gene inflammatory disorder could be producing a CNO-like picture.
These panels can include hundreds of genes. They are not a test for a single "CRMO gene," because no such diagnostic gene exists for typical CNO. Their value is in looking for a different or overlapping genetic immune disorder that might help explain the presentation and, in a small number of cases, change treatment.
A negative panel does not rule out CNO and does not mean genetics play no role. Current panels only test genes and variant types that are known and technically detectable today.
A variant of uncertain significance, usually written VUS, is also not the same thing as finding the cause. It means a genetic difference was found but there is not enough evidence to know whether it causes disease. Most VUS findings should not be used by themselves to diagnose a condition or choose treatment. Ask whether the laboratory plans to reclassify the variant as new evidence becomes available, and whether a genetics consultation would add anything.
What about HLA-B27?
HLA-B27 is sometimes described in family groups as a "CRMO gene," but that is not accurate. It is a genetic marker associated most strongly with the spondyloarthritis spectrum. It may be clinically relevant when a child with CNO also has sacroiliitis, enthesitis, inflammatory arthritis, or other features suggesting juvenile spondyloarthritis or an overlap phenotype.
Being HLA-B27 positive does not diagnose CNO, and being negative does not rule it out.
Conditions that get involved
CNO travels with other inflammatory conditions often enough that your team will watch for them: inflammatory bowel disease, psoriasis, inflammatory arthritis, and less commonly eye inflammation. Report new gut symptoms, skin changes, joint swelling, or eye redness and pain. These associations also shape treatment, since a drug that treats both conditions is usually preferred.
Involvement outside the bones
CNO is a bone disease but it does not always stay there. Recent work identifying distinct clinical clusters in both children and adults found that extraosseous involvement, meaning disease outside the skeleton, marks a more severe course and is associated with needing biologic treatment sooner.
Pulmonary involvement in particular has been flagged as a marker of greater severity. This is worth knowing because breathing symptoms in a child with CNO are easy to attribute to something else, and because chest pain and difficulty taking a deep breath are also classic symptoms of thoracic spine lesions. The two are different problems with different workups, and neither is a reason to dismiss the other.
Practically: report persistent cough, shortness of breath, or chest symptoms to the rheumatology team rather than only to a pediatrician, so they are interpreted in the context of the CNO rather than as an unrelated respiratory issue.
Jaw involvement
Mandibular lesions are one of the more characteristic sites in CNO and they bring specific concerns: jaw pain, swelling, difficulty chewing or opening the mouth, and dental problems. Tell your dentist about the diagnosis, and tell the rheumatology team about jaw symptoms.
The connection that catches families out is bisphosphonates, and it runs in both directions.
They are used to treat jaw lesions, not merely tolerated despite them. In a case series of patients with CNO of the jaw treated with zoledronic acid, average pain scores fell from about 7.7 to 2.5 out of 10, with roughly 84 percent reporting partial or complete relief. Worth knowing alongside that: about two thirds had pain return, usually within the first year, so a repeat course is common rather than a sign of failure. Imaging and pain did not track together reliably, which is the same pattern as everywhere else in this disease.
The osteonecrosis concern is real but small, and it is mostly about dental procedures. Osteonecrosis of the jaw is a known risk of bone-directed drugs, and it is rare in children. The jaw CNO series above reported none from zoledronic acid. What remains standard is a dental check before starting, and scheduling extractions or oral surgery around treatment rather than during it. Tell the dentist your child is on or starting a bisphosphonate before any procedure, and tell the rheumatology team about jaw symptoms.