Background/Purpose: Chronic nonbacterial osteomyelitis (CNO) is an inflammatory bone disease that can result in bone destruction/deformity, persistent bone pain and pathological fractures. Due to the lack of randomized control trials, treatment is empirical and based on physician experience. Clinical studies to compare treatments require validated definitions of minimal disease activity (MDA) and disease flare. This study aims to develop and validate definitions of MDA and flare in patients with CNO.

Methods: A Delphi survey on priorities for defining MDA and flare was sent to CNO patient/caregivers and CARRA’s CNO Workgroup. Based on responses, key parameters from our multisite prospective registry (CHronic nonbacterial Osteomyelitis International Registry, CHOIR) were extracted to create 2 cohorts to define MDA and flare (Table 1). Missing variables were imputed. A global expert panel of 13 pediatric rheumatologists and 7 patient/caregivers voted on whether a clinical scenario was considered MDA (a single time point assessment) or flare (worsening disease compared to prior visit). Consensus was defined as ≥80% agreement using nominal group technique. A decision tree or ROC analysis were used to determine the cutoff points. The model was validated using 10-fold cross-validation to ensure generalizability.

Results:

Results: A total of 252 participants (172 patients/caregivers and 80 providers) completed the survey from 5 continents. Most providers (90%) had >5 years of experience treating children with CNO with ≥5 patients/year. Among patients, 91% were diagnosed within 10 years, 98% had taken medications for CNO and 85% had experienced a flare. Discrepancy was noted among items deemed important for MDA and flare between the 2 groups (Figure 1). Of the MDA cohort, 253 cases were classified by consensus as MDA, 324 as non-MDA, and 23 as inconclusive. Pain level (p< 0.001), physician global (p=0.012), patient global (p=0.005), clinical lesion count (p=0.001) and MRI lesion count (< 0.001) were significantly associated with MDA classification. Of the flare cohort, 404 were classified by consensus as flare, 159 as non-flare and 10 were inconclusive. Pain level (p< 0.001) and MRI lesion count (p=0.049) were associated with flare designation. Cases with presence of either sacroiliitis or spinal lesions on MRI or active clinical arthritis were consistently voted as flare and as non-MDA (p< 0.001). The median [IQR] clinical disease activity score (CDAS) was 0 [0,2] in cases with MDA compared to 10 [6, 12] in those without MDA (p< 0.001) with a threshold CDAS ≤ 3 for being categorized as MDA. The accuracy of the MDA model without CDAS and with CDAS was 0.77 and 0.89, respectively. The median [IQR] CDAS was 8 [4, 12] in cases with flare, compared to 0 [0, 2] in those without flare (p< 0.001) with a threshold CDAS ≥3 for being categorized as flare.The accuracy of the flare model without CDAS and with CDAS was 0.78 and 0.87 respectively. Decision pathways were developed (Figure 2).

Conclusion: Through Delphi survey and formal consensus techniques, we developed definitions and decision pathways of MDA and flare in patients with CNO proposed as useful targets for future clinical trials.

Provider and Patient/Caregiver Family Priorities in Defining Minimal Disease Activity and Flare in CNO Delphi Survey Results

Table 1. Characteristics of Cohorts Used to Develop Definitions of Minimal Disease Activity and Flare in CNO

Figure 2. Decision Pathways for Defining Minimal Disease Activity and Disease Flare in CNO

« Back to ACR Convergence 2025

ACR Meeting Abstracts - https://acrabstracts.org/abstract/development-and-validation-of-minimal-disease-activity-and-disease-flare-for-children-with-chronic-nonbacterial-osteomyelitis-using-a-consensus-and-data-driven-approach/