Background Chronic non-bacterial osteomyelitis (CNO) is a rare characterized by inflammatory bone lesions with no detectable infectious agents. It may be unifocal (U-CNO) or multifocal (M-CNO) and have a monophasic, relapsing or persistent course. The typical presentation of CNO is local bone pain with or without fever or other local signs of inflammation (1). Imaging evaluation (bone scintiscan and MRI) is very important to identify and evaluate lesions but, to date, no validated outcome measure has been identified (2).
Objectives to define the role of MRI in detecting disease activity in patients with CNO.
Methods Patients with CNO, lasting longer than 6 month, were prospectively followed every 3-4 months. MRI were performed at diagnosis, at the time of clinical relapse or at least yearly in all patients. Clinically, active disease was considered when patient complained bone pain and needed analgesic or second line therapy to control symptoms. Active bone lesions on MRI were those in which BME was associated with soft tissue inflammation (STI, edema and/or perilesional effusion). As resulted in sports medicine studies, BME alone was not considered a parameter of activity (3). A single radiologist, blinded on patients’ clinical status, reviewed all MRI. Sensitivity, specificity and predictive value for disease activity detection of MRI were calculated by comparing radiological evaluation with clinical status of the patients at different time points.
Results 13 CNO patients entered the study. 8 had UF-CNO, 5 had MF-CNO, mean age at disease onset 10,8 years (range 2,33-18,5), 54% were female. Disease duration at diagnosis was longer in patients with UF-CNO (14,2 vs 10,1 months). Localized bone pain was the leading symptom at onset in all patients; systemic symptoms, such as fever and fatigue, were more frequent in MF-CNO. 38,5% presented associated skin disease and 61,5% positive family history for autoimmune disease. At onset WBC was normal, CRP and ESR were elevated in 69,2%, especially in MF-CNO. After median 3 years follow up, 46% of patients had no symptoms and were off-therapy. At disease onset, all patients were evaluated by MRI then, during the follow-up, 12 patients repeated MRI once (T1) and 6 twice (T2). At disease onset, all 13 MRI showed BME with STI. Of the 18 follow up MRI, 4 (22%) completely normalized, 6 (33%) showed only BME, 8 (44%) showed pathological changes.
Sensitivity of the MRI ranged between 0.83 and 1.00, specificity 1.00, positive predictive value was 1,00 while negative predictive value ranged between 0.86 and 1.00.
Conclusions Despite the small sample, MRI resulted to be a sensitive, specific and predictive tool for monitoring CNO. Once validated in a larger cohort of patients, these preliminary findings will allow including MRI among the most reliable outcome measures for CNO.
Girschick HJ, et al. Chronic non-bacterial osteomyelitis in children. Ann Rheum Dis. 2005; 64:279.
G. Khanna, et al. Imaging of Chronic Recurrent Multifocal Osteomyelitis. Radiogaphics 2009; 29:1159.
Kornaat PR, et al. Bone marrow edema-like signal in the athlete. Eur J Radiol. 2008;67:49.
Disclosure of Interest None Declared