Priorities in Chronic nonbacterial osteomyelitis (CNO) – results from an international survey and roundtable discussions

Two questionnaires were designed to collate experience with diagnosis and treatment, as well as opinions on research priorities among CNO patients and families, health care professionals and CNO researchers. Questionnaires were developed by the Experimental Arthritis Treatment Centre for Children (EACT4Children), involving paediatric rheumatologists from the UK (CMH, CP) and the USA (PJF), the EATC4Children’s manager (JFG) and administrator (LW), a former CNO patient and paediatric trainee (ER), the EATC4Children’s Public Involvement and Engagement (PPIE) manager (JP), and the scientific lead of the CRMO Foundation USA (AT) (Supplements 1 & 2).

A 24-item questionnaire (Supplement 1), designed in 2020, was circulated in 2021 among 27 expert clinicians and clinical academics using Google Forms. Proportional representation of North America, the UK, and continental Europe was targeted during the selection of experts which included members of the Childhood Arthritis and Rheumatology Research Association (CARRA), the Paediatric Rheumatology European Society (PReS), the British Society of Rheumatology (Paediatric and Adolescent members) (BSR), and the German Society of Paediatric and Adolescent Rheumatology (GKJR). Proportional representation of North America, the UK, and continental Europe were targeted during the selection of experts. All experts were consultant-grade clinicians who diagnose and treat a minimum of 10 patients with CNO per year and have clinical experience with CNO for > 5 years. All have published work on CNO's clinical and/or pathophysiological aspects with an impact factor > 4 within the past five years (2016–2021). Experts who responded proportionally represented geographic regions as follows: seven from North America, seven from the UK, and seven from continental Europe.

A second 20-item questionnaire was designed using lay language to explore the experience and priorities of CNO patients and family members (Supplement 2). The survey was circulated internationally (Google Forms) in 2021 via the EATC4Children’s social media pages (Facebook: 50 followers, Twitter: 618 followers), the UK CRMO Facebook group (486 members), the CRMO Foundation USA (Facebook: 1300 followers, Twitter: 284 followers), and individuals at European Network for Children with Arthritis (ENCA) and Rare Autoinflammatory Conditions Community (RAAC-UK) who distributed it to their patient contacts/groups.

Both questionnaires included qualitative and quantitative questions. For analysis, quantitative data were grouped, and graphics were produced to summarise differences and similarities between clinicians’ versus patients’ experiences and priorities. A thematic analysis of free-text data was conducted by two investigators (MM and ER) to collate the information and highlight key areas of concern and research priorities. The efforts of the two investigators were then compared and moderated (CMH and PJF) to confirm themes.

Results from the questionnaires, as outlined in this report, were used to assign topics for four moderated roundtable discussions at the “International Conference on CNO and autoinflammatory bone disease” in Liverpool, May 25^th-26^th, 2022. Facilitated discussion topics included: i) Clinical Trials (facilitators: CP, ER), ii) Pathophysiology and Molecular Mechanisms (PF, AT), iii) Mental and Emotional Well-being (MH, MM), and iv) (to allow tracing and comparison of future studies) Nomenclature (HJG, CA). Roundtable discussion participants included patients and families (N ~ 15), academics, and clinical staff (consultants, medical subspecialty trainees, specialist nurses) (N ~ 30). Moderation by facilitators was minimal to allow for free and uninterrupted discussion amongst parties. It included agreeing a framework for discussion (e.g. discussing Population, Intervention, Control, and Outcomes in the “clinical trials” discussion), reminding participants of timing of session, note-keeping and encouragement of contribution from all parties. At the end of each discussion session, facilitators moved from table to table, so all participants discussed all topics (20 min each). Results were recorded and presented to the groups after the subsequent conference sessions.

There were 93 responses to the patient questionnaire (Fig. 1A). Of these, 23% (21/93) were from children and young people (CYP) affected by CNO (Age range: 8–31 years, median age: 20 years), and 77% (72/93) were from parents or carers on behalf of the young person they care for (on behalf of children and young people aged 5–20 years, median: 12 years) (S1Q2, Q3) (Supplement 3A). Of patient/carer responses, 58% (54/93) were from North America, 21% (19/93) were from the UK and Ireland, 11% (10/93) were from continental Europe, and 10% (10/93) were from other regions (including Australia, New Zealand, Central and South America) (S1Q4).

A total of 21 expert clinicians and clinical academics responded to the questionnaire (21/27: 78%, Fig. 1B). Of these, 71% (15/21) had been practising as a specialist rheumatologist for > 10 years (S2Q5). All respondents treated paediatric patients, with 19% (4/21) also treating adults (S2Q6). Responses from clinicians/clinical academics were equally distributed between North America, the UK, and continental Europe (all 7/21, 33.3%) (S2Q3) (Supplement 3B).

The patient/carer questionnaire asked about their experience with time to (correct) diagnosis, initial incorrect diagnoses, and clinical phenotypes and treatment, and the clinicians/clinical academics were asked about their experience in practice with the aforementioned aspects of CNO (Fig. 2). Patients and carers reported a median time to diagnosis of approximately 6 months, with a mean of 14 months (S1Q8). Clinicians/clinical academics estimated a median and mean of 6 months (S2Q11) (Fig. 2A). As many as 56% of CYP/carers reported they had received an alternative diagnosis before they were diagnosed with CNO (S1Q9). The most common initial diagnoses included osteoarticular infections (28%), biomechanical joint pain (23%), “inflammation” (13%), and cancer (13%) (S1Q10) (Fig. 2B). Among expert clinicians/clinical academics, 100% reported that, in their experience, CNO patients receive an incorrect working diagnosis before CNO was diagnosed (S2Q13), with the most common being infectious causes (61%) (S2Q13) (Fig. 2B).

Because CNO can be associated with inflammatory involvement of extraosseous structures that can aid in diagnosing CNO, CYP/families and clinicians were asked about their experience with these. Of all CYP/carers responding, 46% reported inflammatory conditions alongside CNO (S1Q12); 41% of these patients experienced psoriasis, 21% had inflammatory bowel disease (IBD), and 18% had arthritis (S1Q13) (Fig. 2C). All clinicians had treated patients with an additional inflammatory condition (S2Q14), the most common being arthritis (39%), followed by psoriasis (35%) and IBD (26%) (S2Q14) (Fig. 2C).

Treatment of CNO is empiric and not standardised across centres, and therefore is largely based on the personal experience of the treating clinician, case series, and expert opinion [8]. Therefore, Patients/carers and clinicians were surveyed about their experience with treatment choices. Among CYP/carers responding to the survey, the most common medications taken were nonsteroidal anti-inflammatory drugs (NSAIDs) (34%), followed by bisphosphonates (28%), biological (15%), and conventional (14%) disease-modifying antirheumatic drugs (DMARDs) (S1Q14). Clinicians were asked to rank medications, with one being the medication they most frequently prescribed for their patients with CNO. Responses were in close agreement with patient/carer experience (Fig. 3) (S2Q16).

Lastly, patients/carers and clinicians were asked to rank research topics/areas of interest (namely: pathophysiology, medication trials, outcome measures, diagnosis and classification criteria, mental and emotional well-being) to allow future prioritisation of joint efforts. While both groups agreed on the highest priority of deciphering the pathophysiology of CNO, clinicians did not come to an agreement on how to rank closely related medication trials, outcome measures, and diagnosis and classification criteria (Fig. 4A). Both groups ranked impacts on emotional and mental well-being as their lowest research priority (S1Q15, S2Q17). Notably, this question had to be re-shared with a subset of respondents (both patients/carers and clinicians/academics) with additional clarification. The question was worded in a way that a subset of respondents responded in a different manner than intended. Several respondents isolated each topic and rated them each out of 5, resulting in incomparable data. Thus, we redistributed this question to those who did not answer as intended, hence giving a lower number of respondents to this specific item (Fig. 1).

In addition to the ranking exercise, open questions were asked, allowing free text responses. Patients/carers were asked what their biggest concerns were regarding the impact of CNO on their well-being. Most commonly occurring themes were ‘long-term outcomes and future quality of life when living with CNO’ (42%), followed by both ‘treatment choice and safety including side effects’ (16%) and ‘disease management monitoring of response and remission’ (16%) (Fig. 4B) (S1Q17).

From aforementioned responses to questionnaires, final themes for roundtable discussions, including free-text responses. Notably, research priorities aligned closely with the five topics previously ranked, including ‘Investigations into the underlying causes of CNO (Pathophysiology)’ (patients/carers: 45%, clinicians: 34%), ‘Medication trials testing drugs’ (patients/carers: 18%, clinicians: 33%), ‘Defining the ways we diagnose and classify CNO’ (patients/carers: 11%, clinicians: 14%), ‘Finding outcome measures that help us to treat patients more effectively in clinic, and research studies’ (patients/carers: 3%, clinicians: 14%), and ‘Studies into how the disease affects your mental and emotional wellbeing’ (patients/carers: 1%, clinicians: N/A), (Fig. 4C). The top 3 themes were chosen for roundtable discussions.

Considering variable nomenclature for CNO across the existing literature, an additional topic ‘nomenclature’ was added by the organizers.

The four aforementioned topics were discussed at a face-to-face meeting in Liverpool, involving expert clinicians, clinical academics, patients and carers. Discussions were moderated and recorded, and outcomes are outlined below (Fig. 5).

The discussion around pathophysiology and molecular mechanisms of CNO was extensive and detailed, with themes extending from individual predisposition/susceptibility to associated long-term outcomes. Equally, patients, carers and experts recognised the vast need for research into this field. Parents/carers wish for a greater understanding of modifiable factors that may allow for disease prevention or early and “mild” interventions, i.e., dietary interventions, smoking, geographic variability, etc. The urgent need for collaboration was recognized, especially considering logistical issues around sharing comparable samples, i.e., sample handling, shipment, variability in reagents used.

Discussions around designing clinical trials for patients with CNO naturally considered the ‘PICO’ framework, discussing Population, Intervention, Controls and Outcome measures. Most participants felt that trials should, ideally, be accessible to all CNO patients and not limited to sub-cohorts, e.g., only those with multifocal bone disease. However, it was agreed that some CNO patients may be harder to study - i.e., those with vertebral fractures, where treating with NSAIDs alone as a comparator may be considered unethical. Consideration was given to classification of CNO patients and eligibility criteria, reflecting diagnostic challenges and lack of criteria used for diagnosis in clinical practice. Soon to be published suggestions for EULAR/ACR criteria were agreed as a valuable tool [13]. Discussion around appropriate interventions was broad, likely reflecting the lack of clinical trials in this field. Interestingly there were differing opinions between CYP, parents and clinicians regarding controls. Families favoured the generation of evidence for the efficacy of current and future treatments in placebo-controlled trials, whilst clinicians expressed their preference for head-to-head comparisons. This was based on their concerns around ongoing pain and damage accrual in the placebo group. There was a unanimous understanding that the measured outcomes, including physical exam, biomarkers and imaging, were required to be patient centred. It was recognised that outcome measures important to patients and families (particularly pain, sleep, health economics, school attendance) are difficult to quantify and validate reliably. Thus, additional objective tools must be included. Acknowledging the current lack of prospectively and independently validated scoring tools is important for future research in the field of CNO and in paediatrics in general.

This round table was an opportunity for all delegates to share their experiences of living with or caring for people with CNO. Many common issues were shared. Families emphasized uncertainties of diagnosis, many referring to frightening conversations where the differential diagnosis of malignancy was discussed. The insidious and relapsing/remitting pattern of the disease was also acknowledged, with many patients/carers expressing frustrations around the unpredictable course of symptoms and having to explain to others why their levels of functional ability may vary depending on whether they are experiencing flares or remission. Other themes included helping children/patients to understand their diagnosis, and the implications for the wider family, including emotional and financial impacts.

Varying nomenclature was a concern equally for clinicians/scientists and patients and their families. Clinicians recognised the importance of nomenclature for classification, both for diagnostic purposes and for future laboratory and clinical research. For example, the acronym CRMO may disadvantage patients with monofocal disease in their access to certain/future treatments. Parents/patients felt the use of the word ‘chronic’ sounded “scary”, and took away hope for recovery, commenting that many other chronic diseases do not reference chronicity in name, e.g., asthma or epilepsy. Overall, there was consensus that CNO may be a good "umbrella term” and that sub-type nomenclature may be of use, especially because some sub-groups of CNO require more “aggressive” management plans, e.g., CNO of the jaw. Because of its inclusivity, there was an overall preference for CNO. Because families and some charities use CRMO as a “brand”, the combination of “CNO/CRMO” may be useful when publishing data.