Chronic non-bacterial osteomyelitis (CNO) in children may present as single or multiple lesions and there may be a single episode or recurrent episodes. Chronic recurrent multifocal osteomyelitis (CRMO) is thought to be the childhood equivalent of the synovitis, acne, pustulosis, hyperostosis, and osteitis/osteomyelitis (SAPHO) syndrome that affects adults. Although there have been reports of bacterial growths from CRMO lesions (Staphylococcus aureus, Mycoplasma hominis, Propionibacteriumacnes, Bartonela henselae, and Coxiella burnetti have all been reported) most microbiological studies, including PCR testing, have given negative results. Antibiotic treatment does not work but anti-inflammatory agents do (non-steroidal anti-inflammatory drugs are effective and steroids, interferon α, interferon γ, biphosphonates, and tumour necrosis factor antagonists have been used). Little has been written about the long term course of the disease but a follow up study has recently been reported from Germany ( H J Girschick and colleagues. Annals of Rheumatic Diseases2005;64:279–85OpenUrlCrossRef).

They followed up 30 children (aged 2–17 years at symptom onset, mean 10.3 years; 21 girls) for at least 5 years (mean 5.6 years). The mean age at diagnosis was 11 years (range 2–21 years) and time from first symptoms to diagnosis 5–9 months. Nine children had a single lesion and no recurrence, three had a single lesion on more than one occasion, nine had multiple lesions with no recurrence, and nine had multiple recurrent lesions (CRMO). The most commonly affected bones were the clavicle and the calcaneum but altogether 59 different bones were affected. Only one patient had multiple lesions in any one bone. The clinical features were those of osteomyelitis. Twenty-four children had joint involvement; seven had cutaneous pustules.

Laboratory test results (haemoglobin, leucocyte count, erythrocyte sedimentation rate (ESR), serum ferritin, serum IgG) were generally normal. Serum IgM was moderately raised in four multifocal cases and serum IgA in three. Tests for antinuclear antibody were negative. Twenty-five children had biopsies of bone lesions and all showed the histology of chronic osteomyelitis usually with reparative changes (marrow fibrosis, trabecular osteoid apposition, and periosteal hyperostosis). Patients with CRMO were more likely to have granulocyte infiltration and hyperostosis. Microbiological investigations were negative.

Twenty-seven patients were treated with naproxen, which appeared to be effective in general. Children with a single episode needed less lengthy treatment than those with recurrent episodes (mean 13.4 months vs 25 months). One patient with non-relapsing multifocal disease and four with CRMO were treated with prednisone in addition to naproxen before achieving remission. Other treatments (sulfasalazine, biphosphate, hyperbaric oxygen, surgical decortication) were apparently ineffective. All lesions were quiescent 5 years after disease onset. Bone growth was not affected but one patient had difficulty eating because of unilateral mandibular hypertrophy.

There appears to be a spectrum of CNO ranging from a single non-recurrent lesion to recurrent multifocal lesions (CRMO). The cause is not known but the prognosis is good.