A discussion of the literature, intended primarily for healthcare professionals.
A child with CNO may have pain at just one site and several other bone lesions without symptoms. This is the most useful finding of a study published on 25 August 2026 in the European Journal of Pediatrics: 32 patients had 144 lesions on MRI, with a median of 4.5 lesions per child. The number of symptomatic sites did not correlate with the number of MRI lesions. [1]
The 22 patients assessed with whole-body MRI (WB-MRI) had a median of 5.5 lesions, compared with 2.0 in the ten patients who underwent regional MRI (p=0.003). The difference is clear, but needs to be interpreted in context: a scan covering the whole skeleton has more opportunity to detect silent lesions than an examination limited to the painful area. This comparison does not establish that WB-MRI improves prognosis, or that every patient should undergo the same examination at the same intervals. [1]
Symptoms do not reliably describe the extent of CNO
Chronic non-bacterial osteomyelitis (CNO), also called chronic recurrent multifocal osteomyelitis (CRMO) in its chronic, multifocal forms, is a sterile inflammatory bone disease that mainly affects children and adolescents. It may cause pain, swelling or a limp and can involve one or more sites. Diagnosis remains based on clinical and imaging findings, with particular attention to excluding infection, malignancy and other conditions that can produce similar appearances. [2,3]
In the series by Galvis and colleagues, the most frequently involved sites were the femur (65.6%), tibia (56.3%), clavicle (37.5%) and sacrum (31.3%). Clavicular involvement was therefore common, but did not occur in the majority of patients. Pelvic and spinal lesions were also found in children who did not necessarily report symptoms at those sites. [1]
This is where the study becomes clinically useful. Apparently unifocal pain does not mean that the disease itself is unifocal. In this series, the correlation between the number of symptomatic sites and the total number of MRI lesions was effectively absent (ρ = -0.03; p=0.879). [1]
Earlier literature had already pointed in the same direction. WB-MRI is now considered the most comprehensive method for assessing the distribution and activity of CNO without exposure to ionising radiation, particularly when clinically silent lesions need to be identified. A review focused on imaging also explains that a typical distribution can help with the differential diagnosis and, in characteristic multifocal presentations, may avoid unnecessary biopsies. [3]
Whole-body MRI is useful, but this study does not establish everything
The authors interpret their findings as supporting routine WB-MRI at the initial assessment. This is a reasonable position, consistent with much of the recent literature, but this study alone is not enough to establish it.
Patients were not randomised to regional or whole-body MRI. The choice depended on the clinical indication and the clinician's preference; protocols also changed over the ten years studied. The WB-MRI and regional MRI groups therefore cannot be treated as directly comparable experimental populations. [1]
There is a further, simpler limitation: finding more lesions does not automatically improve a clinical outcome. That would require evidence that detecting silent sites early changes treatment and reduces chronic pain, deformity, fractures, vertebral damage or other complications. The study by Galvis was not designed to answer that question. [1]
Some silent sites do, however, matter more than others. Vertebral involvement deserves particular attention because CNO can be associated with vertebral deformity and collapse; lesions near the growth plates can also affect growth. The CARRA consensus treatment plans recognise active spinal lesions as a factor that changes the treatment pathway and generally favour WB-MRI for imaging assessment, while explicitly stating that these consensus plans are not clinical guidelines. [4]
Biopsy still has a role
Thirteen of the 32 patients in the study, or 41%, had undergone a bone biopsy. Indications included persistent symptoms, diagnostic uncertainty and an inability to confidently exclude infection or malignancy. Cultures were negative and histology showed no malignancy. [1]
This finding helps avoid another oversimplification. WB-MRI can make a typical multifocal pattern much easier to recognise, but it does not turn CNO into an automatic imaging diagnosis. Biopsy may still be necessary in atypical or unifocal cases, when imaging appears aggressive, or when the clinical history and laboratory findings do not fit. Recent reviews retain this distinction. [2,3]
For the orthopaedic surgeon, the problem often arises before anyone mentions CNO: persistent bone pain, a limp, swelling, a non-specific radiograph, or a lesion initially thought to represent osteomyelitis or a tumour. At that point, it is useful to remember that the painful site may tell only part of the story.
The new study therefore supports whole-body MRI for defining the extent of CNO, but does not yet establish how this greater sensitivity should translate into fixed imaging protocols or changes in treatment. The next question matters more than simply counting lesions: which silent lesions actually change the child's management?
References
[1] Galvis I, Jaramillo D, Imundo LF, Delgado J, Kvist O. The quiet flame: whole-body MRI and the underestimated burden of pediatric chronic non-bacterial osteomyelitis (CNO). Eur J Pediatr. 2026;185:691. Published August 25, 2026. doi:10.1007/s00431-026-07311-9. PMID: 42642677.
[2] Triaille C, De Bruycker JJ, Miron MC, Lecouvet F, Girschick H, Wouters C. Update on the diagnosis and treatment of CNO in children: a clinician’s perspective. Eur J Pediatr. 2025;184:48. Published online November 27, 2024. doi:10.1007/s00431-024-05823-w. PMID: 39604722. PMCID: PMC11602790.
[3] Nico MAC, Araújo FF, Guimarães JB, et al. Chronic nonbacterial osteomyelitis: the role of whole-body MRI. Insights Imaging. 2022;13:149. doi:10.1186/s13244-022-01288-3. PMID: 36114435. PMCID: PMC9481810.
[4] Zhao Y, Wu EY, Oliver MS, et al. Consensus Treatment Plans for Chronic Nonbacterial Osteomyelitis Refractory to Nonsteroidal Antiinflammatory Drugs and/or With Active Spinal Lesions. Arthritis Care Res (Hoboken). 2018;70(8):1228-1237. doi:10.1002/acr.23462. PMID: 29112802. PMCID: PMC5938153.
Disclaimer: This article provides general information and does not replace an individual medical assessment.
