An international team has developed and validated the first global classification criteria for pediatric chronic nonbacterial osteomyelitis (CNO), offering a standardized tool to improve research on the rare disease.
CNO, an autoinflammatory bone disease with no known cause, has an estimated global incidence of 0.4 to 2.3 cases per 100,000 children and adolescents annually, according to a paper describing the guidelines and the process of devising them.1 CNO causes bone inflammation, leading to pain and swelling, plus such potential long-term complications as deformity or disability. Diagnosing the disorder is difficult because of its gradual onset and varying symptoms, which often overlap with other conditions.
The CNO classification criteria resulted from an effort jointly supported by the ACR and EULAR. Involvement and testing by a global group of researchers allows generalization of results beyond a single center or country, says the criteria’s first author, Yongdong Zhao, MD, PhD, director of the CNO clinical program at Seattle Children’s Hospital.
The criteria are intended primarily to improve research on the condition and will be used in future clinical drug trials for CNO, Dr. Zhao adds. He points out that clinicians and patients have no medication approved by the U.S. Food & Drug Administration for the condition, while the estimated number of newly diagnosed CNO patients worldwide has increased from 8 per million children to 23 per million each year.1
Dr. Seza Ozen
The criteria development process “raised awareness for this disease on both sides of the Atlantic,” says criteria senior author Seza Ozen, MD, MSc, head of the Department of Pediatric Rheumatology at Hacettepe University, Ankara, Turkey, and president of the Pediatric Rheumatology European Society (https://www.pres.eu). “I’m sure this classification system is going to help in the future studies.”
The Criteria
The criteria were developed in four phases by researchers from a dozen countries in North and South America, Europe, Australia and New Zealand. The phases involved generating candidate criteria items, reducing and defining those items, weighting criteria and identifying a threshold for CNO classification, and refinement and validation.
First, pediatric rheumatologists identified relevant disease features via surveys. The researchers refined and weighted the identified features in a structured decision-making process. The researchers then tested the criteria in a development cohort of 441 patients and validated them in a separate cohort of 514 children worldwide. They found the new criteria had a sensitivity of 82% and specificity of 98% in the validation cohort.
To be classified as having CNO, a child must be younger than 18 years old and have bone pain or musculoskeletal limitations for at least six weeks, the final criteria say. Imaging must show abnormal bone findings, such as inflammation visible on radiography or magnetic resonance imaging (MRI). Additionally, the classification process involves excluding mimicker conditions including malignancies, infections, vitamin C deficiency and the genetic disorder hypophosphatasia.
Using the criteria involves evaluating clinical findings and laboratory and pathology findings in 10 domains and assigning points to them. The clinical domains include bone lesion sites, bone lesion patterns, patient age at disease onset, coexisting conditions and fever. Laboratory and pathology domains include anemia, C-reactive protein (CRP) level, erythrocyte sedimentation rate (ESR) and bone biopsy findings, although the criteria can be applied without a bone biopsy. A minimum of 55 points is required to meet the classification threshold.
The paper and Dr. Zhao describe the domains included in the point system:
- Symptom onset before age 18—The non-genetic form of CNO is rare in children younger than 3, so clinicians and researchers must exercise extreme caution when classifying a patient younger than 3 years old.
- Bone lesion pattern and sites—The mandible and clavicle get the most points because these sites are more common in CNO than in mimicker conditions. Sites other than clavicle, mandible, skull or hand get fewer points. Affected bones, such as the femur and tibia, are equally affected in mimicker conditions and do not carry extra weight toward CNO classification.
- Distribution of bone lesions based on imaging—Multifocal lesions get the most points because previous research shows most patients with CNO have multiple lesions at diagnosis or eventually develop them. Symmetrical lesion patterns draw points because they are more common in CNO than in mimicker conditions.
- Coexisting conditions prior to CNO diagnosis—Inflammatory bowel disease and specific skin diseases, such as psoriasis and pustulosis, are highly associated with CNO and draw the most points. Axial arthritis, which primarily affects the spine and sacroiliac joints, gets fewer points if it does not involve skin symptoms.
- Anemia—The study showed a significant difference between the distribution of patients with CNO and mimicker conditions, and hemoglobin under 10 g/dL.
- Fever—Fever is more associated with infectious osteomyelitis and malignancy, but may occur in 10–20% of patients with CNO. Therefore, fever at presentation is less favorable toward classification of CNO and receives no points.
- ESR below 60—ESR of 60 mm/h or less is associated more with infectious osteomyelitis and malignancy and receives no points. ESR above this threshold garners points.
- CRP—A CRP of 30 mg/L or greater is more associated with infectious osteomyelitis and malignancy, so it does not get any points. A CRP below that threshold gets points. However, A CRP of 30 mg/L or greater can also be present in 10–20% of patients with CNO.
Biopsy
Although the criteria do not require bone biopsy, the criteria do include it in scoring. Dr. Zhao explains that inflammation within bones refers to the presence of immune cells, such as neutrophils, lymphocytes and plasma cells. The latter two are more commonly seen in bone biopsy in CNO. In the study, fibrosis indicated subacute and chronic repairing process. Neither is specific for CNO, but chronic inflammation and fibrosis are more common in CNO than in mimicker conditions, according to Dr. Zhao.
Although the criteria are primarily intended to distinguish CNO from mimicker conditions for research purposes, not diagnosis, some clinicians “may have different thresholds of making the diagnosis based on known diagnostic criteria” and choose biopsy, although biopsy is not always required for clinical diagnosis, Dr. Zhao says. Neither is biopsy required in classification criteria. However, in patients with features of mimicker conditions, it may be indicated for clinical diagnosis, he adds.
Dr. Sandrine Lacassagne
A bone biopsy makes classification easier, says Sandrine Lacassagne, MD, MSc, a consultant in pediatric rheumatology at Great Ormond Street Hospital for Children, London. She points out that up to 17 points out of a total of 55 can potentially come from biopsy results.
The bone biopsy domain may be subject to debate because not all patients have a bone biopsy and the criteria do not guide on the decision to perform a bone biopsy, Dr. Zhao notes.
Another criterion that may spur debate is an exclusion criterion calling for complete and sustained response to antimicrobials alone. This criterion does not suggest every patient should receive a trial of antibiotics, according to Dr. Zhao.
“Vitamin C deficiency, a mimicker condition, can be more common than we expect in western countries, and it is important to exclude it before applying the criteria because all other features can be very similar to CNO,” he adds.
Looking for Better Treatments
An important area of CNO research focuses on better therapies. Currently, the most used drugs to treat CNO include non-steroidal anti-inflammatory drugs (NSAIDs), cytokine-blocking agents, bisphosphonates and Janus kinase (JAK) inhibitors. At least half of CNO patients continue to need medications for CNO during adulthood.
Dr. Zhao is collecting clinical data to characterize the long-term outcomes of several drugs. In an ongoing study to establish the Chronic Nonbacterial Osteomyelitis International Registry (CHOIR), he aims to build a prospective disease registry for CNO to investigate the responses of patients to different clinical management over 10 years and better define the natural history of the disease. Drugs included in the study include methotrexate, sulfasalazine, leflunomide, pamidronate, zoledronic acid, etanercept, adalimumab, certolizumab, infliximab, golimumab and NSAIDs, according to clinicaltrials.gov. Currently, the study has 600 patients at 17 sites, Dr Zhao says.
His study will use the new criteria. According to Dr. Ozen, studies in Turkey are also using the criteria. The studies include efforts to examine how bone parameters change within the disease and to validate the classification criteria she and Dr. Zhao established.
Dr. Lacassagne, who notes the U.K. does not yet have a clinical trial of a CNO treatment, praises the criteria’s high sensitivity and specificity. “These classifications tend to try to identify the more typical cases” for the purposes of research so investigators can better compare patient responses to treatment, she explains. “This is an important international effort with strong methodology for a rare disease to support future research and improve care.”
With the recently published classification criteria, “we can capture main features of CNO and recognize that patients with CNO do not present in a single pattern,” adds Dr. Zhao.
The criteria, while intended for research, can serve as a reminder to clinicians that “a combination of different features may still classify a patient as [having] CNO,” he adds. “Keep mimicker conditions in mind and make sure that adequate laboratory and imaging data are collected during the diagnostic process. Prioritize non-invasive tests before an invasive test unless a clear indication is present.”
Deborah Levenson is a writer and editor based in College Park, Md.
Disclosures
Dr. Zhao made the following disclosures: ACR, EULAR, CARRA (grants to institution, finance committee), Bristol Myers Squibb (grants to institution), Rheumatology Research Foundation (grants to institution), CTIP (grants to institution), UpToDate CRMO/CNO (royalties/license), Novartis (consulting fees to self), U.S. Provisional Application No. 63/356977 (patent issued), Seattle Children’s CRMO Warriors Guild (leadership/fiduciary role), American Board of Pediatrics, Pediatric Rheumatology Subboard (leadership/fiduciary role).
Dr. Ozen reports consulting fees from Novartis and Sobi.
Dr. Lacassagne reports no conflicts of interest.
Reference
- Zhao Y, Oliver MS, Schnabel A, et al. EULAR/American College of Rheumatology Classification Criteria for Pediatric Chronic Nonbacterial Osteomyelitis. Arthritis Rheumatol. 2025 May 8.
