OP0342 IDENTIFYING CANDIDATE ITEMS TOWARDS THE DEVELOPMENT OF CLASSIFICATION CRITERIA FOR CHRONIC NONBACTERIAL OSTEOMYELITIS (CNO) AND CHRONIC RECURRENT MULTIFOCAL OSTEOMYELITIS (CRMO)

Background Chronic nonbacterial osteomyelitis (CNO) is a severe and occult autoinflammatory bone disease of unknown cause. Early diagnosis is challenging, and CNO may debilitate affected children when left untreated. Currently, evidence-based and validated diagnostic and classification criteria for CRMO/CNO are lacking. The insidious disease course, increasing disease incidence, and significant delay in diagnosis highlight the need for the development of classification criteria that leads to more precise and early selection of patients for clinical trials1,2. Objectives To identify candidate items towards developing classification criteria for CNO using anonymous survey and nominal group technique. Methods An international collaborative effort was formed within the pediatric and adult rheumatology communities to conduct the following phases: 1) to generate candidate criteria items by a Delphi survey among international rheumatologists; 2) to reduce candidate criteria items through consensus processes involving physicians managing CNO and patients or caregivers of children with CNO. This study was approved by Seattle Children’s Hospital Institutional Review Board. Results In Phase 1, 259 pediatric rheumatologists (30%, N=865) participated in an online questionnaire about features most relevant to the classification of CNO. Of those, 77 (30%) practiced in Europe, 132 (51%) in North America, and 50 (19%) in other continents. A total of 138 (53%) responders had >10 years of practicing experience and 108 (42%) had managed >10 CNO patients. There were 33 candidate criteria items initially identified. In Phase 2, candidate items were presented to 39 rheumatologists and 7 parents and items were refined or eliminated through item reduction techniques. Seventy-seven (94%, N=82) workgroup members then participated in a second survey to rank the remaining items by their distinguishing power of CNO from mimicking conditions. Figure 1 shows the mean score for the remaining 31 candidate criteria. Multifocal lesions, ruling out malignancy and infection and typical location on imaging had the greatest means. CRP and/or ESR greater than 3x the normal upper limit had the greatest negative means. Discriminatory Score +3/-3 (increases/decreases the likelihood of CRMO the most) +2/-2 (increases/decreases the likelihood of CRMO moderately) +1/-1 (increases/decreases the likelihood of CRMO slightly) 0 (no difference) Conclusion Through surveys and consensus technique, candidate items towards developing classification criteria for CNO were identified. This list of items will guide the design of a feasible patient data collection form towards weighting of each item in the classification criteria. References [1] Roderick MR, Shah R, Rogers V, Finn A, Ramanan AV. Chronic recurrent multifocal osteomyelitis (CRMO)–advancing the diagnosis. Pediatric Rheumatology. 2016 Dec;14(1):47. [2] Jansson A, Renner ED, Ramser J, Mayer A, Haban M, Meindl A, et al. Classification of non-bacterial osteitis: retrospective study of clinical, immunological and genetic aspects in 89 patients. Rheumatology. 2006 Jun 17;46(1):154-60. Acknowledgement CNO/CRMO Work Group, Childhood Arthritis and Rheumatology Research Alliance Disclosure of Interests Melissa Oliver: None declared, Eveline Wu: None declared, Raymond Naden Speakers bureau: Was a speaker at conferences paid by pharmaceutical companies several times in the past, but not in the last 7 years., Matthew Hollander: None declared, Polly Ferguson: None declared, Fatma Dedeoglu Consultant for: Attended a scientific meeting for Novartis in 2017. Overall monetary amount was less than $5000., Seza Ozen Consultant for: Seza Ozen is receiving consultancy fees from Novartis, Speakers bureau: Roche, Yongdong Zhao Grant/research support from: I have grant support from Bristol-Myer Squibb