Introduction

Chronic nonbacterial osteomyelitis (CNO) is the most common autoinflammatory bone disorder in childhood (1). Diagnostic information is provided by TC-99 bone scintigraphy (BS) and/or whole body MRI. Non-steroidal anti-inflammatory drugs (NSAIDs), glucocorticoids, bisphosphonates and tumour necrosis factor inhibitors have been used until now with variable response (2).

Objectives

To describe clinical, biological and radiological response to treatment with anakinra in patients with CNO refractory to NSAIDs and bisphosphonates.

Materials and methods

Seven patients (4 females and 3 males) with refractory CNO were treated with anakinra for at least 6 months in our institution. Response to treatment was evaluated assessing clinical manifestations (pain, local swelling, functional impairment), laboratory findings (C-reactive protein (CRP)), erythrocyte sedimentation rate (ESR)) and serum amyloid A level (SAA))) and number of bone lesions on TC-99 BS at the start of treatment and at 6 months.

Results

The median age at diagnosis and before starting anakinra was 9.7 years (IQR 7.8-14.7) and 13.3 years (IQR 8.0-15.9) respectively. All were treated with NSAIDs and bisphosphonates as first-line therapy. Glucocorticoid therapy was required in one patients with concomitant recurrent fever and pleural effusion. These patients did not respond satisfactorily and anakinra (2 mg/kg/day) was started. At the start of treatment 7/7 patients (100%) had pain, 3/7 (43%) local swelling and 5/7 (71%) functional impairment; at 6 months of follow up 6/7 patients (86%) were completely asymptomatic, with one patient complaining of arthralgia. Before starting anakinra the median CRP, ESR and SAA were 2.7 mg/dl (IQR 1.7-4.9) 26 mm/h (IQR 12-46) and 53 mg/dl (IQR 27-112); at 6 months 5/7 patients (71%) normalized CRP, ESR and SAA; 2/7 had a decrease in inflammatory markers. Before anakinra 59 bone lesions were detected on TC-99 BS. After 6 months of therapy 24/59 lesions (40%) had completely resolved, 1/59 lesions (2%) had partially improved and 29/59 lesions (49%) remained stable. In two patients with persistent high biological inflammatory markers, new lesions (14) developed during treatment.

Conclusion

Our data suggest that anakinra appears effective in CNO in controlling symptoms and laboratory findings; subclinical bone inflammation was still detectable by BS after 6 months of treatment. Long-term follow-up studies with a larger number of patients are needed.