How Organ Manifestations Beyond the Bone May Guide Risk Stratification and Treatment Decisions in Chronic Nonbacterial Osteomyelitis

Abstract

Chronic nonbacterial osteomyelitis is an autoinflammatory bone disease characterized by pronounced dysregulation of proinflammatory and anti-inflammatory cytokine production and release. Through innate immune dysregulation, patients with chronic nonbacterial osteomyelitis develop bone inflammation, pathological bone remodeling, and associated pain. Notably, secondary activation of adaptive immune mechanisms may contribute to phenotypic variability and the onset of additional tissue and organ involvement (joints, skin, gut). Currently used treatments are empiric, targeting inflammation and/or osteoclasts. While deemed efficacious across clinical cohorts internationally, in the absence of randomized controlled trials, they lack regulatory approval. While chronic nonbacterial osteomyelitis limited to bones frequently responds to anti-inflammatory treatment with naproxen or, in more severe cases, tumor necrosis factor inhibitors and/or osteoclast inhibition with bisphosphonates, patients with additional symptoms may require additional/different treatments targeting effector T cells and associated cytokines (such as interleukin-17 blockers, Janus kinase inhibitors). Thus, changing molecular, cellular, and phenotypic patterns throughout the disease course offer an opportunity for personalized treatment and suggest a ‘window of opportunity’ preventing the onset of additional organ involvement. However, ‘tools’ to guide personalized treatment approaches are currently lacking, and treatment is guided by clinical features, following a trial-and-error approach. This article reviews the available literature, deciphering shared immune mechanisms between chronic nonbacterial osteomyelitis and associated diseases that may guide individualized care. It focuses on shared and disease-specific molecular and cellular patterns of chronic nonbacterial osteomyelitis and associated diseases, delivering arguments for disease stage and phenotype specific treatments, including cytokine blockers and synthetic small-molecule inhibitors.

Key Points

Chronic nonbacterial osteomyelitis is characterized by an imbalanced cytokine release that results in systemic inflammation and pathological bone remodeling.

Chronic disease activity may result in secondary activation of adaptive immune responses, contributing to amplified inflammation and additional organ involvement.

Understanding evolving immunological patterns in relation to disease phenotypes and organ involvement will allow for target-directed and individualized care.

Introduction

Chronic nonbacterial osteomyelitis (CNO) is an autoinflammatory bone disease that, most commonly, affects children and adolescents [1, 2]. Likely because of improved awareness, the reported incidence of pediatric CNO has steadily increased, rising from historical estimates of 0.4 per 100,000 children annually (German Surveillance Study, 2006–8) [3], 0.65 per 100,000 children in the UK and the Republic of Ireland [4] to 2.3 per 100,000 children in the USA [5]. Nevertheless, clinical observations suggest these figures to represent a significant underestimation, and a German single-center cohort suggests that CNO may be as common as infectious osteomyelitis [6].

Chronic nonbacterial osteomyelitis is a clinically variable disease that may affect single bones, sometimes following self-limited courses over several months [1, 7]. However, most patients (> 75%) experience chronically active or recurrent multifocal disease, also referred to as ‘chronic recurrent multifocal osteomyelitis’. In patients with chronic recurrent multifocal osteomyelitis, inflammatory bone lesions typically involve the meta/epiphyses of long bones and frequently show a symmetric distribution [1, 8, 9]. Notably, in addition to bone inflammation, a proportion of patients with CNO develop extra-osseous inflammatory manifestations that can affect several organ systems. According to the Eurofever registry [10], arthritis complicates the clinical picture in a minimum of 29% of cases. Additionally, mucocutaneous involvement is observed in 19% of patients, presenting as acne (5%), palmoplantar pustulosis [PPP] (5%), psoriasis (4%), papulopustular (3%) or maculopapular (1%) rashes, urticaria (2%), and aphthous stomatitis (1%). Systemic inflammation may also affect the gastrointestinal tract (8%), lymphoid tissues (3%), eyes (2%), and the cardiovascular system (1%). Moreover, these associated inflammatory conditions are also enriched in the family members [11, 12].

In the following, current knowledge on the pathophysiology CNO and associated extra-osseous manifestations will be summarized. Special attention will be given to the question of how additional organ involvement may guide treatment decisions. Notably, while national societies have recently published treatment guidance for bone inflammation in CNO [13, 14], no shared recommendations exist between rheumatology and subspecialties treating associated diseases (such as dermatology, gastroenterology). This narrative review is based on a comprehensive literature search accessing PubMed followed by appraisal and data extraction by the authors.

Pathomechanisms of Chronic Nonbacterial Osteomyelitis (CNO) and Associated Diseases

Innate Immune Dysregulation

While its exact pathophysiology is complex and remains somewhat unclear, CNO is recognized as a chronic autoinflammatory bone disease, driven primarily by dysregulation of the innate immune system [1, 10]. Several mechanisms have been proposed linking CNO with pathologically increased expression [15–18] and activation of the NOD-like receptor pyrin domain-containing protein 3 (NLRP3) inflammasome [15–20], alongside imbalanced production of proinflammatory and anti-inflammatory cytokines [15, 16]. Most recent genetic analyses have identified rare variants in the P2X purinoceptor 7 gene (P2RX7), encoding for the ATP-dependent P2X7 transmembrane channel, in up to 40% of patients with CNO from a large cohort [19, 21]. This observation is of special interest because P2X7 promotes NLRP3 inflammasome assembly through potassium efflux. This offers molecular explanations for disease expression in a small proportion of patients with damaging/high-impact variants (in < 6% of patients), while suggesting P2RX7 variation as a risk factor for CNO in up to one third of patients [19, 21]. Notably, carriers of rare P2RX7 variants more commonly presented with extra-skeletal manifestations, including abdominal symptoms, and lymphadenopathy, suggesting more widespread symptoms in this patient cohort [19, 21]. Moreover, P2RX7 variants offer explanations for the bone phenotype of CNO because resulting imbalanced cytokine expression may disrupt bone homeostasis through enhanced osteoclast differentiation [19, 21], pathological promotion of osteoclast-mediated bone remodeling, and resorption [1]. Notably, a disease-associated rare variant in P2RX7 had previously been reported in a single patient with adult CNO who responded to treatment with the recombinant interleukin (IL)−1 receptor antagonist anakinra [22]. However, validation in additional unrelated multi-ancestral cohorts is necessary to fully understand the contribution of genetic variation in P2RX7 to disease expression and phenotypes.

Adaptive Immune Dysregulation

Reports on the development of additional organ manifestations in some patients suggest that CNO may progress from an autoinflammatory pathology to a ‘mixed pattern’ disease, showing aspects of innate and adaptive immune dysregulation [16, 21, 23]. Innate and adaptive immune mechanisms do not exist ‘in a vacuum’ but show close interactions that have, for example, been established in CNO-associated psoriasis [23, 24]. Indeed, strong associations of CNO with several other inflammatory disorders (including psoriasis) suggest shared immune mechanisms and common pathophysiologic features. Aspects of immune dysregulation may change over time, resulting in disease stage-dependent dynamic organ involvement [12, 16] (Fig. 1).

Fig. 1.

Molecular pathomechanisms in chronic nonbacterial osteomyelitis and the development of associated organ involvement. (A) In chronic non-bacterial osteomyelitis, innate immune dysregulation, including the pathological activation of NOD-like receptor pyrin domain-containing protein 3 inflammasomes, results in an imbalance in the release of proinflammatory and anti-inflammatory cytokines [65]. Pathological immune activation and imbalanced cytokine expression contribute to enhanced osteoclast differentiation and activation, promoting pathological bone remodeling and osteoinflammation. Proinflammatory signaling, over time, may result in subsequent activation of adaptive immune responses, differentiation and activation of effector CD4⁺ T-cell populations, and effector cytokine release. This enhances bone inflammation and remodeling through additional mechanisms as interleukin (IL)−17 also promotes osteoclasts. As the IL-23/IL-17 axis is centrally involved also in the development of arthritis (B), skin (C), and gut (D) pathology, over time, this may result in an extended phenotype that follows the spondylarthritis/psoriasis/inflammatory bowel disease spectrum. Possible treatments for chronic nonbacterial osteomyelitis that may prevent progression towards widespread organ involvement are indicated in red. CD4+ T cell cluster of differentiation 4-positive T cells, JAK Janus kinase, P2X7 P2X purinoceptor 7, RANKL receptor activator of nuclear factor kappa-B ligand, TGF transforming growth factor, Th T helper cell, TNF tumor necrosis factor

Cutaneous and joint involvement are the two most frequently reported additional organ manifestations in CNO that may co-occur [10, 25]. Notably, adolescent patients or adults with CNO, more frequently than younger children, exhibit ‘synovitis, acne, pustulosis, hyperostosis and osteitis (SAPHO)’ [now referred to as ‘adult CNO’]. This extended disease phenotype may be the result of secondary adaptive immune activation and effector T-cell engagement in a ‘mixed pattern disease’ [21, 26]. An ‘early’ hallmark of the skin involvement in this disease spectrum are innate inflammatory infiltrates involving neutrophils and monocytes, underscoring the central role of innate immune dysregulation [27]. Neutrophilic granulocytes from patients with SAPHO display enhanced migratory capacity, likely facilitating skin infiltration and inflammation [28]. Initial neutrophil-rich infiltrates may evolve over time toward a lymphocyte-dominated pattern, sometimes alongside non-caseating granulomas and reactive changes [29, 30]. This heterogeneity likely reflects dynamic immune dysregulation in CNO and associated skin disease [23, 24, 31] (Fig. 1C). Thus, in patients with long-standing (potentially even subclinical) bone inflammation, activation of tumor necrosis factor (TNF), IL-1β, and the IL-17/IL-23 axis may shift CNO towards a ‘mixed pattern disease’ in which T helper cells promote skin, joint and/or additional organ involvement (e.g., gut inflammation) [13, 32]. Early clinical reports from France (from the early 2000 s) suggested that insufficient disease control in CNO enables the development of arthritis, resulting in a spondylarthritis-like clinical picture in adulthood [33]. This may also link to differing bone inflammation patterns between patients with pediatric versus adult CNO (or SAPHO) with a shift from primary involvement of long bones of the extremities to more axial involvement at a later age [16, 34]. In this context, it is of interest that, similar to spondylarthritis and psoriatic arthritis, patients with adult CNO/SAPHO exhibit elevated numbers of circulating T helper cell (Th) 17 cell levels, underscoring activation of the IL-23/IL-17 axis and associated Th17 polarization [32, 35–37]. Moreover, Th17-derived IL-17 may promote osteoclastogenesis and bone remodeling [21, 37] through the induction of receptor activator of nuclear factor kappa-Β ligand expression [21, 38] (Figs. 1A, B, and 2).

Fig. 2.

Biphasic model of joint inflammation in chronic nonbacterial osteomyelitis. Similar to previously suggested for Still’s disease [269], joint inflammation in chronic nonbacterial osteomyelitis may also follow a bi-phasic course with (A) predominantly interleukin (IL)−1 driven early inflammation mediated by innate immune cells and cytokines, followed by (B) secondary adaptive immune activation, and effector T-cell and cytokine-driven inflammation at later stages, This may allow stage-specific and targeted therapeutic interventions using biological DMARDs and/or synthetic small-molecule inhibitors (red). CD4+ T cell cluster of differentiation 4-positive T cells, JAK Janus kinase, RANKL receptor activator of nuclear factor kappa-B ligand, Th T helper cell, TNF tumor necrosis factor

Disease Mechanisms Associated with Additional Organ Involvement

Enhanced NLRP3 inflammasome activation and release of inflammasome-associated cytokines (IL-1β, IL-18), expression of IL-6, and activation of the IL-23/−17 axis in CNO may extend the clinical phenotype to clinical manifestations along the spondylarthritis/psoriasis/inflammatory bowel disease (IBD) spectrum [39, 40] (Fig. 1D). Inflammatory bowel diseases have been reported in up to 10% of patients with CNO [6, 41–43]. Notably, patients with CNO, even in the absence of clinically manifest IBD, may exhibit signs of bowel inflammation, such as (asymptomatic) increase of fecal calprotectin [41]. This is similar to observations in patients with spondylarthritis and/or enthesitis-related arthritis (ERA) where fecal calprotectin may even correlate with the activity and progression of distant skeletal lesions [44–47]. Notably, treatment with non-steroidal anti-inflammatory drugs (NSAIDs) should be considered as a potential confounder, as it may contribute to increased fecal calprotectin in individual patients [48]. Another shared feature between CNO, spondylarthritis, and IBD is the impaired expression of immune-regulatory cytokines IL-10 and IL-19 [15, 21, 25, 49]. Reduced IL-10 production and/or signaling have been implicated in the pathogenesis of early-onset IBD, and single nucleotide polymorphisms within the IL-10 promoter alter serum cytokine levels, increasing an individual’s susceptibility to IBD, spondylarthritis, and other inflammatory musculoskeletal diseases [50–53]. Intestinal inflammation in CNO, such as in IBD, may be associated with an impaired epithelial barrier, microbial dysbiosis, and subsequent inflammatory responses driving disease expression and phenotypes [54] (Figs. 1D, 3B). Indeed, altered composition of microbiota and ‘epithelial breach’ have been proposed in CNO [1, 54–56] and in related diseases psoriasis, spondylarthritis, and IBD [24, 57, 58]. Propionibacterium acnes, a microorganism with low virulence, has been long suspected as a trigger of CNO [59] and SAPHO [60]. Detection of this species in bone biopsies from patients with CNO, however, most likely represents contaminants from the skin [61]. Indeed, while adult patients with CNO/SAPHO experienced symptom relief (osteoarticular and cutaneous disease) during a 16-week course of antimicrobial therapy with azithromycin, responses did not persist beyond the discontinuation of antimicrobial treatment [62]. Thus, effects may have been linked to immunomodulatory effects of azithromycin, which was suggested also by others in pediatric CNO where treatment-associated benefit also vanishes after its discontinuation [34, 62]. Nonetheless, the presence of P. acnes may be interesting in the context of understanding disease mechanisms and disease progression as it may promote effector Th17 polarization and IL-17 expression in patients with acne [63], which is typically present in SAPHO. A genetic predisposition leading to impaired bacterial clearance has also been hypothesized in people affected by adult CNO/SAPHO, further supporting a model in which host/microbiota interactions drive a dysregulated inflammatory response [64].

Fig. 3.

Possible mechanisms and treatment targets in chronic nonbacterial osteomyelitis with skin and gut involvement. (A) In skin disease, effector cytokine (interleukin [IL]−23) and T helper cell (Th) 17 cell-mediated events on keratinocytes and immune cells create and amplify inflammation. Innate cytokine expression, including IL-23 from dendritic cells have broad effects on adaptive immune activation and effector T-cell differentiation/activation. The effector cytokine IL-17 induces epidermal hyperplasia through IL-19 and IL-36, further immune cell infiltration (through CXCL chemokines), amplification of Th17 responses (CCL-20), and the transcriptional expression of innate proinflammatory signals (IL-1β, IL-6, and IL-8). These events may be preventable or could be terminated by a number of biological disease-modifying antirheumatic drugs or synthetic small-molecule inhibitors already used in chronic nonbacterial osteomyelitis and/or psoriasis treatment (red) [270]. (B) In inflammatory bowel diseases, microbial dysbiosis and the disruption of mucus layers result in failed epithelial integrity with increased permeability and antigen exposure. Activated macrophages produce proinflammatory cytokines tumor necrosis factor (TNF), IL-1, IL-6, IL-12, and IL-23. Neutrophils invade and release inflammatory signals, including IL-17A. Antigen-bearing dendritic cells in Peyer’s patches migrate to the mesenteric lymph nodes, where they present antigens to naive T cells that differentiate into effector Th17 or Th1 cells producing their lineage-defining cytokine patterns. Therapeutic approaches (red) include biological disease-modifying antirheumatic drugs or small-molecule inhibitors that inactivate cytokines or block lymphocyte trafficking to the intestine (vedolizumab) [271]. CCL CC motif chemokine ligand, CXCL CXC motif chemokine ligand, JAK Janus kinase, P2X7 P2X purinoceptor 7, RANKL receptor activator of nuclear factor kappa-B ligand, TGF transforming growth factor

Treatment of Bone Inflammation in CNO

In the absence of randomized controlled trials (RCTs), treatment of CNO is empiric and based on observational cohort studies, registry data, and personal experience. To date, there are no licensed treatments available. As a result, all patients with CNO are treated ‘off label’. Patients need to be informed about lacking approval and potential consequences, and reimbursement of treatment-associated costs may have to be discussed with regulators/insurances. To standardize care, the Childhood Arthritis and Rheumatology Research Alliance, supported by international colleagues, developed Consensus Treatment Plans in 2018 [14], and the German Society for Pediatric and Adolescent Rheumatology recently proposed guidelines [13] (Table 1).

Table 1.

Medications routinely used ‘off label’ in chronic nonbacterial osteomyelitis [13, 65]

Class effects	Agent	Dose and frequency	Efficacya with highest evidence level [66]	References
NSAIDs: prostaglandin inhibition with effects on nociception and osteoclast activation, ‘weak’ inhibitor of the NLRP3 inflammasome [13, 16, 67]	Usually naproxen	Naproxen: (10-)15 mg/kg/day in 2 doses Ibuprofen: 30 mg/kg/day in 3 doses	Good-to-moderate (except axial involvement) [LoE 2b]	[7, 20, 30, 33, 68–99]
Conventional DMARDs: suppress immune cell activation and proliferation, modulate cytokine expression by decreasing proinflammatory and increasing anti-inflammatory mediators, inhibit osteoclasts [13, 16]	Sulfasalazine	30–50(−60) mg/kg/day in 2 doses (maximum 2000 mg/day)	Limited [LoE 2b]	[7, 20, 30, 73–76, 78, 79, 81, 82, 86, 89, 90, 92–94, 96, 97, 100, 101]
	Methotrexate	10–15 mg/m2 once a week, (maximum 20–25 mg per week)	Moderate-to-limited [LoE 2b]	[7, 69, 72–76, 78–90, 92–98, 100, 101]
TNFi: (partially) restore balance between proinflammatory cytokines and the immunomodulatory cytokines IL-10 and IL-19 [67], reduce osteoclast differentiation and activation [67, 102]	Etanercept	0.8 mg/kg sc, (maximum 50 mg per dose), once a week	For individual TNFi: insufficient [LoE 4] TNFi as a therapeutic class: good-to-moderate [LoE 2b]	[7, 73, 74, 76, 78, 81, 85, 86, 88–93, 95, 100, 103–105]
	Adalimumab	15–30 kg: 20 mg sc every 2 weeks > 30 kg: 40 mg sc every 2 weeks	For individual TNFi: insufficient [LoE 4] TNFi as a therapeutic class: good-to-moderate [LoE 2b]	[7, 74, 78, 79, 81, 82, 86, 88, 90, 91, 95, 100, 101, 103–106]
	Golimumab	< 40 kg: 30 mg/m2 (maximum 40 mg) sc every month > 40 kg: 50 mg sc every month	For individual TNFi: insufficient [LoE 4] TNFi as a therapeutic class: good-to-moderate [LoE 2b]	[107–109]
	Infliximab	5–10 mg/kg i.v. at weeks 0, 2, and 6 and then every 4–8 weeks	For individual TNFi: insufficient [LoE 4] TNFi as a therapeutic class: good-to-moderate [LoE 2b]	[72–74, 78, 81, 83, 86–89, 91, 92, 95, 98, 101, 103–106]
Bisphosphonates: inhibit osteoclasts, reduce bone pain and bone inflammation [13, 110]	Pamidronate	1 mg/kg/day i.v. (maximum 60 mg/dose) for 3 consecutive days (first dose of first cycle 0.5 mg/kg), repeated every 3 months (for a total of 3–4 cycles) 1 mg/kg/day i.v. (maximum 60 mg/dose) as a single dose every month (4 weeks), continued for 9–12 months	Good [LoE 2b]	[7, 69–73, 81–83, 85–89, 92, 94, 95, 101, 105, 111–119]
	Zoledronic acid	Initial dose 0.0125–0.025 mg/kg every 6 months. May increase dose to 0.05 mg/kg/dose (maximum 4 mg/dose) [14]	Insufficient [LoE 4]	[98, 101]

DMARDs disease-modifying antirheumatic drugs, IL interleukin, i.v. intravenously, LoE level of evidence [66], NLRP3 NOD-like receptor pyrin domain-containing protein 3, NSAIDs non-steroidal anti-inflammatory drugs, sc subcutaneously, TNF tumor necrosis factor, TNFi tumor necrosis factor inhibitor

^aEfficacy definition: Very good: Consistent efficacy across multiple cohorts/registries; standard of care; Good: reproducible efficacy in several studies/cohorts; Moderate: some evidence of efficacy, but inconsistent results; Limited: small case series or heterogeneous results; Insufficient: anecdotal evidence only

As the majority of treatment response data available originate from retrospective cohort studies or case series in which heterogeneous definitions of ‘treatment response’ were used, their interpretation is complex. Response rates in individual studies may reflect pain improvement, a reduction in serum inflammation markers, changes on radiographic imaging, or combinations thereof. As a result, comparisons between individual studies are difficult and should be interpreted with caution.

Non-Steroidal Anti-Inflammatory Drugs (NSAIDs)

Based on their effects on nociception, anti-inflammatory effects (including ‘mild’ inflammasome inhibition), and the involvement of prostaglandins in osteoclast activation, NSAIDs have been used in CNO [67]. Both Childhood Arthritis and Rheumatology Research Alliance Consensus Treatment Plans and German Society for Pediatric and Adolescent Rheumatology guidelines recommend NSAIDs to be part of first-line therapy, considering possible contraindications [13, 14]. While NSAIDs provide rapid pain control and can induce remission in up to 75% of patients in the first year [65], long-term benefits are limited. Retrospective data suggest that 50% of patients with CNO experience disease flares after a median of 29 months, and more later [7]. Moreover, NSAIDs are not considered ‘potent’ enough for the treatment of patients with vertebral involvement as disease may progress [20, 86]. In patients with vertebral involvement, even in the absence of structural damage, or in those with insufficient response or secondary treatment failure, escalation to second-line therapies is recommended. Additional (off-label) treatment options include conventional synthetic disease-modifying antirheumatic drugs (csDMARDs), biologic disease-modifying antirheumatic drugs (DMARDs), and/or bisphosphonates.

Conventional Synthetic Disease-Modifying Antirheumatic Drugs (csDMARDs)

Among csDMARDs, methotrexate and sulfasalazine are the most used options, which is particularly true for concurrent IBD and/or arthritis [65]. Conventional DMARDs have multiple incompletely understood effects, including correcting the inflammatory imbalance in CNO by attenuating proinflammatory cytokine release while boosting immune regulatory factors [67]. Based on reports from uncontrolled cohort studies, clinical efficacy in CNO is variable. The efficacy of methotrexate ranges between 16 and 66%, while 40–60% respond to sulfasalazine in small cohorts [65]. Data from the Eurofever registry confirm modest effects of csDMARDs, reporting complete response rates of 38% for sulfasalazine and 22% for methotrexate [10].

Tumor Necrosis Factor (TNF) Inhibition

For severe or otherwise treatment-refractory presentations of CNO, including vertebral involvement, bisphosphonates and/or TNF inhibitors (TNFi) currently have the largest observational evidence supporting their use. It remains to be mentioned here that vertebral involvement, while holding the risk of vertebral compression fractures, may be clinically asymptomatic, thereby representing a significant diagnostic challenge [6, 21]. Thus, to prevent future complications through missed diagnosis and treatment, whole body imaging (usually magnetic resonance imaging) is generally agreed and recommended for all patients with CNO, at least at the time of diagnosis and treatment initiation [13, 21].

Inhibition of TNF may (partially) correct the balance between proinflammatory and anti-inflammatory cytokines [67]. Moreover, TNF plays a key role in the differentiation and activation of osteoclasts [102], which are major contributors to bone inflammation and pathological bone remodeling in CNO [67].

Biologic treatment with TNFi (e.g., adalimumab, infliximab, etanercept) have been introduced to the treatment of CNO, aiming to correct inflammation and pathological bone remodeling [67]. Dosing of TNFi in CNO is based on established experience and approvals in juvenile idiopathic arthritis (JIA) [67, 120]. Notably, across uncontrolled studies, TNFi associate with high response rates among patients with CNO (up to 91%) and may be preferentially utilized in patients with co-existing psoriasis, arthritis, and/or IBD [65].

Bisphosphonates

Bisphosphonates inhibit osteoclasts through the induction of apoptosis and have incompletely understood effects on the balance between proinflammatory and anti-inflammatory cytokines [67]. Bisphosphonates, usually pamidronate, exhibit high overall efficacy of up to 75% [65, 105, 111, 114]. However, relapses are not uncommon after treatment withdrawal [7, 116]. Pamidronate remains the best-studied bisphosphonate in pediatric CNO. Recent reports suggest that zoledronic acid [98, 121] or neridronate [122] may also be effective in the treatment of CNO. However, reports remain preliminary and no head-to-head comparative studies are currently available.

Recent retrospective cohort data from the UK and Germany found pamidronate and TNFi to be effective in children with CNO [105]. Notably, while patients treated with pamidronate achieved clinical and radiographic remission more quickly when compared with TNFi, although not reaching a statistical significant level, fewer relapses were observed among patients with CNO treated with TNFi [105]. Notably, treatment duration may have influenced the reported relapse rates because treatment duration and cumulative treatment exposure were not adjusted [105]. In routine clinical practice, pamidronate is usually administered over approximately 9–12 months (monthly or as three to four treatment cycles with three doses on consecutive days, 3 months apart) [16, 21]. Across centers involved in the study, TNFi are usually used for a minimum of 12 months; longer treatment courses may be used in patients with remaining disease activity. Therefore, RCTs are urgently needed to define the comparative effectiveness of pamidronate and TNFi [26]. Notably, the risk of pamidronate failure is associated with female sex, widespread skeletal involvement (on imaging), and marked elevations in C-reactive protein and erythrocyte sedimentation rate. Thus, identifying these risk factors at baseline may, in the future, provide clinicians with tools to tailor individualized therapeutic strategies [105].

However, while data on commonly used treatments are encouraging, in the absence of RCTs, they lack evidence and licensing [13, 65]. Moreover, a subset of patients with CNO will not respond to any of these therapeutic agents and require additional treatment options, especially in cases with co-existing associated diseases. Several additional potential treatments for CNO have been reported in single cases, small cohorts, or case series and may represent options in otherwise treatment-refractory cases.

Interleukin (IL)−1 Inhibition

Based on the pathological expression and activation of NLRP3 inflammasomes reported in CNO by several groups, blockade of IL-1β and/or IL-18 are promising targets [15–20]. Published clinical evidence supporting IL-1 blockade in sporadic CNO remains limited. Nine children from Italy with otherwise treatment-refractory CNO received the recombinant IL-1 receptor antagonist anakinra, a post-translational regulator of IL-1 signaling. After 6 months, over half the patients demonstrated improvement in disease activity alongside normalization of inflammatory markers. However, new, predominantly asymptomatic bone lesions developed in 78%, suggesting that radiological disease control may remain incomplete [123]. Importantly, this cohort represented a treatment-refractory population and anakinra was administered at a relatively low dose of 2 mg/kg/day, both of which may have negatively affected treatment efficacy. More recently, a retrospective Turkish cohort of children with refractory CNO reported favorable outcomes in response to IL-1-targeted therapies, including anakinra and the inactivating IL-1β antibody canakinumab [124, 125]. Similarly, a small cohort of adults with refractory SAPHO experienced a clinical improvement of osteoarticular manifestations in response to treatment with anakinra [126]. Nevertheless, compared with TNFi, the evidence base for IL-1 blocking strategies currently remains limited, and neither prospective cohort studies nor RCTs are available. Nonetheless, IL-1 inhibition may be considered in selected patients with otherwise treatment-refractory CNO, particularly when prominent systemic inflammation, recurrent fever episodes, neutrophilic dermatoses, and/or other autoinflammatory features are present [127–129].

IL-6 Inhibition

The pleiotropic proinflammatory cytokine IL-6 plays a key role in innate and adaptive immune responses, osteoclast differentiation, bone resorption, and chronic synovial inflammation. Elevated IL-6 levels have been reported in patients with CNO, and IL-6 is recognized as a central mediator in several autoinflammatory diseases [16, 130]. Blockade of IL-6 with tocilizumab is an established and effective treatment for JIA, particularly in systemic and/or polyarticular disease [131, 132]. However, clinical evidence supporting IL-6 blockade in CNO remains limited and is restricted to isolated case reports and small case series. Clinical improvement in response to tocilizumab has been described in a few selected patients with otherwise treatment refractory CNO, but consistent efficacy has not been demonstrated [85, 124, 133].

Treatment of CNO-Associated Diseases

CNO and Arthritis

Arthritis represents the most common extra-osseous manifestation in CNO, although its presentation is heterogeneous (Figs. 1 and 2). Data from the Eurofever registry indicate that approximately 29% of patients with CNO develop arthritis, most commonly presenting as monarthritis (15%), followed by oligoarthritis (12%) and, less frequently, polyarthritis (2%) [10]. The anatomical relationship between joint involvement and inflammatory bone lesions remains insufficiently characterized, but a cohort study from Germany and data from the German National Pediatric Rheumatology Database (NPRD) suggest that CNO-associated arthritis may most frequently occur adjacent to bone lesions (two thirds vs one third with ‘distal’ lesions) [6, 90]. However, differences in the frequency, extent, and pattern of joint involvement may be present and relate to ancestry. A Turkish cohort displayed markedly higher proportions of CNO-associated axial joint involvement, with sacroiliitis in 37% of patients [124]. Notably, a substantial proportion of patients with CNO in this cohort had concomitant familial Mediterranean fever (FMF) (13/80; 16%) and/or ERA (9/80; 11%), both of which are associated with musculoskeletal and axial inflammation and may contribute to the increased frequency of sacroiliac involvement observed in this cohort [124].

As mentioned above, insufficiently treated CNO may evolve into spondyloarthropathies in a subset of patients (up to 40% of adult patients during long-term follow-up), supporting the concept of a disease continuum between CNO and spondylarthritis [33]. Similarly, approximately 20% of children with CNO may developed features consistent with spondyloarthropathy over a 5-year period [134]. Indeed, overlapping features with psoriasis/axial spondyloarthropathy and/or disease progression have been described by several groups. In the NPRD cohort, ERA was diagnosed in 2.4% of patients with CNO, while psoriatic arthritis was reported in approximately 1%. Notably, while no strong association was seen between HLA-B*27 positivity and CNO per se, an association between HLA-B*27 positivity and co-existing ERA (14.6% vs 2.9%), and foot or ankle inflammation (54.6% vs 27.7%) was present [90]. In a large European cohort, HLA-B*27 positivity was associated with a two-fold increased risk of developing CNO (odds ratio = 2.2), particularly in boys. While not reaching a significance level, also in this study, HLA-B*27 was associated with an increased frequency of arthritis, psoriasis, and/or IBD [135]. This suggests that, in CNO, HLA-B*27 may define a subgroup of patients with clinical features overlapping with ERA and/or spondylarthritis, further supporting the pathophysiological continuum between CNO and spondyloarthropathies, offering additional therapeutic considerations such as blockade of TNF and/or the IL-17/IL-23 axis [1, 14, 16] (Fig. 2) [26].

csDMARDs

Conventional synthetic DMARDs, such as methotrexate or sulfasalazine, represent a cornerstone in the management of ‘peripheral’ JIA and may therefore be considered in patients with predominantly peripheral arthritis and ‘mild’ CNO activity (Table 2). However, their efficacy in axial arthritis as well as CNO is limited. Across several CNO cohort studies, complete clinical remission in response to methotrexate ranges between 20 and 44%, while remission rates for sulfasalazine are even lower (18–54%) [10, 136]. In pediatric ERA cohorts, only 20–30% of patients achieve clinically inactive disease with methotrexate, and remission is uncommon [137–139]. Accordingly, a substantial proportion of pediatric patients with ERA (50–70%) require treatment escalation, usually with TNFi [140].

Table 2.

Treatment considerations in chronic nonbacterial osteomyelitis with associated inflammatory arthritis

Arthritis	Conventional DMARDs	TNF inhibitors	IL-17 inhibitors	IL-12/IL-23 inhibitors	JAK inhibitors	IL-6 inhibitors
jPsA	Methotrexate: moderate efficacy in peripheral arthritis (derived from studies in JIA rather than disease-specific ERA or jPsA randomized trials) [LoE: 1a] Sulfasalazine: limited [LoE: 4]	Good [LoE: 1b]	Good [LoE: 1b]	Limited [LoE: 4] Limited pediatric evidence; extrapolated from adult PsA and pediatric psoriasis studies	Moderate [LoE: 1b] Evidence derived from the phase III randomized placebo-controlled tofacitinib JIA trial; efficacy data for jPsA and ERA are based on exploratory subgroup analyses	Insufficient [LoE: 5] Very limited evidence
ERA	Methotrexate: limited efficacy in axial arthritis [LoE: 1a] (derived from studies in JIA rather than disease-specific ERA or jPsA randomized trials) Sulfasalazine: moderate in peripheral arthritis [LoE: 1b]	Good efficacy in peripheral arthritis, enthesitis, and sacroiliitis [LoE: 1a] (often used in combination with methotrexate)	Good [LoE: 1b]	Insufficient [LoE: 4]		Insufficient [LoE: 5] Very limited evidence
Juvenile polyarthritis	Methotrexate: good [LoE: 1a] Sulfasalazine: Limited [LoE: 1b-2b]	Very good (often used in combination with methotrexate) [LoE: 1a]	Insufficient [LoE: 5]	Insufficient [LoE: 5]	Good [LoE: 1b]	Good [LoE: 1b]
References	[137, 141]	[142, 143]	[144]	[132]	[131, 145]

DMARDs disease-modifying antirheumatic drugs, ERA enthesitis-related arthritis, IL interleukin, JAK Janus kinase, JIA juvenile idiopathic arthritis, jPsA juvenile psoriatic arthritis, LoE level of evidence [66], PsA psoriatic arthritis, TNF tumor necrosis factor

Efficacy definition: Very good: consistent efficacy across multiple cohorts/registries; standard of care; Good: reproducible efficacy in several studies/cohorts; Moderate: some evidence of efficacy, but inconsistent results; Limited: small case series or heterogeneous results; Insufficient: anecdotal evidence only

TNF Inhibition

Tumor necrosis factor inhibition currently represents the treatment with the most evidence for efficacy in treating patients along the pediatric (and adult) ERA/psoriatic JIA/CNO spectrum (Table 2). Cohort studies and registry data consistently demonstrate efficacy of TNFi in JIA, including axial disease, and are successfully used in CNO [7, 14, 82, 86, 103, 105, 124]. In patients with JIA with ERA, TNFi are among the most effective therapies, with high efficacy in peripheral arthritis, enthesitis, and sacroiliitis [139, 140, 146]. Because of their rapid anti-inflammatory effects and corticosteroid-sparing potential, TNFi are particularly advantageous in patients with high systemic inflammatory activity. Notably, most recent treatment guidelines (2026) do not recommend preferential use of one TNFi over another in patients with axial spondyloarthritis [147]. However, in patients with concomitant IBD, monoclonal anti-TNF antibodies (e.g., adalimumab or infliximab) may be preferred because of their efficacy in controlling both musculoskeletal and intestinal inflammation [148–153] (below). In patients with CNO and arthritis, TNFi may enable ‘dual control’ of bone and joint inflammation [14]. While concomitant methotrexate may reduce the immunogenicity of monoclonal TNFi in some inflammatory diseases, evidence supporting this strategy in CNO is currently lacking. Therefore, routine combination therapy to prevent anti-drug antibody formation cannot be generally recommended, although it may be considered on an individual basis. Concerns raised by some authors about the development of psoriasis in patients with CNO treated with TNFi will be discussed below (see Sect. 4.2).

IL-17 Inhibition

Blockade of IL-17 has proven effective in adult spondyloarthritis [154, 155] and represents an effective therapeutic option in patients with pediatric inflammatory arthritis. In a recent RCT in patients with JIA with ERA or juvenile psoriatic arthritis, the IL-17A inhibitor secukinumab prolonged the time to disease flare when compared with placebo (27% vs 55%; p < 0.001) and demonstrated marked improvements in disease activity within 12 weeks [144] (Table 2). Currently, the role of IL-17 inhibition in CNO remains less clearly defined. However, because of the central role of IL-17 in spondylarthritis and the involvement of IL-17 in osteoclast differentiation and activation [65], IL-17 blockade may represent a rational option in selected patients with CNO, although evidence currently remains limited. It may be particularly efficacious in patients with CNO with axial disease, arthritis, psoriasis and/or PPP [35, 156]. Caution is required in patients with coexisting IBD, as IL-17 inhibition has been associated with worsening intestinal inflammation [157].

Janus Kinase (JAK) Inhibition

A newer class of targeted therapies represent Janus kinase (JAK) [158] inhibitors with increasing evidence in TNFi-refractory JIA, including ERA. A recent RCT using tofacitinib in JIA (82% polyarticular JIA, 9% psoriatic arthritis, and 9% ERA) delivered ACR50 response rates between 65 and 70% [145]. However, data specifically for ERA remain limited, and their use is currently only considered in otherwise treatment-refractory cases (including failed biologic therapies). Only a few cases are documented using JAK inhibitors in CNO [159, 160], predominantly in highly selected patients with severe refractory disease, where they delivered mixed results. Consequently, the current evidence is insufficient to define their role in CNO treatment, and prospective studies are needed to determine whether JAK inhibitors may be effective ‘earlier’ in the treatment of CNO.

IL-12/IL-23 Inhibition

Evidence for IL-12/IL-23 blockade with ustekinumab remains limited in pediatric inflammatory arthritis. In juvenile psoriatic arthritis, the available evidence for the use of ustekinumab is based on the extrapolation of adult data from RCTs and cohort data in pediatric psoriasis, which led to US Food and Drug Administration approval of ustekinumab in children aged > 6 years [161]. Targeting the shared p40 subunit of IL-12 and IL-23, ustekinumab suppresses Th1- and Th17-mediated inflammatory responses [162]. This mechanism may provide a pathophysiological rationale for its use in CNO/SAPHO with prominent skin, gastrointestinal, and/or arthritis-related manifestations (Figs. 1 and 2). In a small cohort of patients with adult CNO/SAPHO, ustekinumab was associated with improvement of cutaneous manifestations, whereas osteoarticular symptoms showed limited response [36]. Consequently, current evidence is insufficient to support the routine use of ustekinumab in patients with CNO. However, it may be considered in selected refractory cases, particularly in the presence of concomitant psoriasis or IBD (below).

IL-1 and IL-6 Inhibition

Blockade of innate inflammatory cytokines IL-1 and IL-6 has been used in selected patients with CNO (above). Evidence supporting the efficacy of IL-1 inhibitors in ERA or juvenile psoriatic arthritis is sparse, and IL-1 blockade is not considered a standard therapeutic approach for these conditions alone and/or in association with CNO. Likewise, evidence supporting the use of the IL-6 blocker tocilizumab in ERA or juvenile psoriatic arthritis remains scarce and/or the combination with CNO. Taken together, while no published recommendations exist for the treatment of CNO and associated arthritis, in the view of the authors, it should follow a phenotype-driven approach (Table 2). In patients with ‘mild’ CNO and predominantly peripheral arthritis, NSAIDs and/or methotrexate may be sufficient. In cases with active osteitis combined with axial ERA, early initiation of TNFi therapy should be considered. In cases of inadequate response, particularly in the presence of concomitant psoriasis or persistent axial inflammation, IL-17 inhibition may be considered, whereas in patients with coexisting IBD, monoclonal TNF antibodies should be preferred. However, in the absence of evidence from RCTs, well-planned prospective studies are urgently needed to establish standardized treatment within licensing.

CNO with Inflammatory Skin Disease

Dermatologic manifestations represent one of the most common and clinically relevant extra-osseous features of CNO, occurring in approximately 18–21% of patients [10, 21, 163] (Fig. 3A). The spectrum of skin involvement is heterogeneous and includes psoriasis [164–167], PPP [168–170], severe acne, hidradenitis suppurativa, pyoderma gangrenosum [171–173], and other neutrophilic dermatoses, reflecting a shared inflammatory background across disease phenotypes [163]. While cutaneous involvement in CNO is more frequent among adults [10, 174], they can occur in all age groups. Skin manifestations may be present prior to bone symptoms or develop over time [10, 91, 163]. Thus, pediatric CNO and adult CNO/SAPHO are considered to be part of a nosological continuum [31].

Skin involvement in pediatric CNO represents a clinically heterogeneous condition. Distinct phenotypes, differing in age at onset, sex distribution, skeletal involvement, and treatment response are present.

Psoriasis and Palmoplantar Pustulosis (PPP)

Palmoplantar pustulosis is most frequently observed in prepubertal girls and typically associates with predominant peripheral bone involvement and a favorable therapeutic response. In contrast, acne- or hidradenitis suppurativa-associated disease more commonly affect pubertal boys, with a higher prevalence of axial involvement and more treatment-resistant skin manifestations [175–179]. While reports are preliminary, observations may inform future patient stratification towards individualized care with more or less potent agents at early disease stages.

Treatment of CNO-associated skin disease can be complex, and not all agents used for the treatment of CNO affect skin inflammation [180]. While pamidronate is effective in achieving improvement of osteoarticular manifestations in CNO/SAPHO, its efficacy on cutaneous involvement is often limited [159]. Nevertheless, contrasting data have been reported by Wu et al., who observed considerable efficacy of bisphosphonate therapy in improving PPP in adult patients with CNO/SAPHO [181].

Notably, recent data from Italy suggest that CNO patients with skin manifestations, when compared with those without skin involvement, more frequently require second-line treatment with biologic agents [176]. Nonetheless, authors of the same study reported that patients with CNO with PPP-psoriasis vulgaris manifestations generally showed good response to topical corticosteroids and/or csDMARDs [176] (Table 3).

Table 3.

Efficacy of treatments on cutaneous manifestations associated with chronic nonbacterial osteomyelitis/SAPHO

Skin manifestation	Conventional DMARDs and corticosteroids	TNFi	IL-12/IL-23 inhibitors	IL-17 inhibitors	JAK inhibitors
Psoriasis and palmoplantar pustulosis	Effective in mild cases; can be associated with topical corticosteroids [LoE: 4]	Effective, but relapses are common [LoE: 4]	Limited efficacy for psoriasis and palmoplantar pustulosis, preferred if cutaneous involvement is predominant and/or associated IBD [LoE: 4]	Effective for extensive skin disease with arthritis/axial involvement [LoE: 4]	High efficacy, limited evidence on pediatric SAPHO [LoE: 4]
Severe acne/hidradenitis suppurativa	Often used in combination with biologic agents [LoE: 4]	Effective, often requires combined approach TNFi + csDMARDs [LoE: 4]	No specific evidence reported	No specific evidence reported	Potential therapeutic role suggested by case reports LoE: 4]
Pyoderma gangrenosum	Mild localized: topical corticosteroids and calcineurin inhibitors Progressive: systemic corticosteroids or cyclosporine [LoE: 4]	Effective but limited evidence [LoE: 4]	Limited evidence [LoE: 5]	Limited evidence [LoE: 5]	Limited evidence [LoE: 5]
References	[176, 182]	[34, 183–190]	[187]	[156, 191–196]	[160, 182, 197, 198]

csDMARDs conventional synthetic disease-modifying anti-rheumatic drugs, DMARDs disease-modifying anti-rheumatic drugs, IBD inflammatory bowel disease, IL interleukin, JAK Janus kinase, LoE level of evidence [66], SAPHO synovitis, acne, pustulosis, hyperostosis and osteitis, TNF tumor necrosis factor, TNFi tumor necrosis factor inhibitor

Efficacy definition: Very good: consistent efficacy across multiple cohorts/registries; standard of care; Good: reproducible efficacy in several studies/cohorts; Moderate: some evidence of efficacy, but inconsistent results; Limited: small case series or heterogeneous results; Insufficient: anecdotal evidence only

TNF Inhibition

In otherwise treatment-refractory cases, TNFi have demonstrated efficacy on dermatologic manifestations [176]. They are considered particularly effective in psoriasis, PPP, and severe acne [34, 187], but relapses are not uncommon [183, 189]. Paradoxically, while psoriasis usually responds to TNFi treatment, the development of psoriasis has been reported in patients receiving TNFi for variable other inflammatory diseases. A large retrospective cohort (N = 4111) including children with IBD, JIA, or CNO showed increased incidence of psoriasis when compared with the general population. The highest rates were observed in patients with TNFi exposure [109]. However, as psoriasis is also associated with CNO/chronic recurrent multifocal osteomyelitis independently of TNFi therapy, and switching within the TNFi may improve psoriasis manifestations [199], the extent to which TNFi induce psoriasis remains unclear. Indeed, pediatric patients with CNO, IBD, or JIA possess an inherently elevated risk for psoriasis (standardized incidence ratio: 18), and the incidence of psoriasis in patients with CNO remains high even in the absence of TNFi exposure (38.9/1000 person-years), closely mirroring the rates seen in those receiving the biologic therapy (33.5/1000 person-years) [109]. Lastly, in a cohort of patients with CNO from Philadelphia, a larger proportion of patients with CNO exposed to treatments other than TNFi developed psoriasis when compared with patients receiving TNFi (resulting in a higher incidence ratio for the development of psoriasis in patients with CNO not treated with TNFi). Therefore, the development of psoriasis in patients with CNO treated with TNFi may, in at least a subset of patients, be a consequence of inadequate disease control [27] (Table 3).

IL-17/IL-23 Inhibition

In patients with CNO with skin involvement, not responding to csDMARDS and/or TNFi, inhibition of the IL-17/IL-23 axis may represent a therapeutic option [35]. Both IL-23 and IL-17 inhibitors are approved for the treatment of psoriasis. Notably, in clinical practice, patients with psoriasis and predominant cutaneous involvement in the absence of axial disease may derive greater benefit from IL-23 blockade, whereas IL-17 inhibitors may be more efficacious in cases with extensive skin disease associated with arthritis and/or axial disease [187]. In patients with PPP, therapeutic responses are more heterogeneous. While IL-12/−23 blockade with ustekinumab has limited efficacy, IL-17 inhibitors have shown encouraging effects in CNO-associated PPP and SAPHO [156, 192–196, 200]. Nevertheless, responses to IL-12/IL-23/IL-17 inhibition are not uniform, and paradoxical psoriasis as well as treatment discontinuation have been described in association with secukinumab and Ustekinumab [36], which may result from insufficient disease control. Notably, as mentioned above, IL-17 inhibitors are generally avoided in patients with IBD [157, 201]. Conversely, IL-23 inhibition has demonstrated efficacy in inducing remission in this setting [157, 187, 201] (Table 3).

JAK Inhibition

An alternative therapeutic approach to biologics targeting of IL-12/IL-23 or IL-17 are synthetic small-molecule JAK [158] inhibitors. Unlike cytokine-specific biologics, JAK inhibitors exert broader immunomodulatory effects through interference with multiple cytokine signaling cascades [187]. While case reports and small case series, mostly including adult patients with CNO/SAPHO [197, 202–205], suggest a role for JAK inhibition in the treatment of CNO with skin involvement otherwise refractory to treatment, data remain limited and preclude final recommendations [160, 182, 197] (Table 3).

Acne and Hidradenitis Suppurativa

A therapeutically complex group representing patients with CNO and acne-hidradenitis suppurativa spectrum disorders often require systemic corticosteroids and/or biologic therapies. Complete cutaneous remission may only be achieved in a subset of patients, frequently necessitating combination approaches [176]. Currently the TNFi adalimumab is the only approved biologic agent for the treatment of moderate-to-severe hidradenitis suppurativa [185, 188]. In adult patients with CNO/SAPHO with concomitant hidradenitis suppurativa, combined therapeutic approaches with adalimumab and csDMARDs, such as methotrexate, may be effective in controlling cutaneous and osteoarticular disease [184, 186, 190]. Lastly, one case report described the efficacy of the JAK inhibitor upadacitinib, suggesting a potential therapeutic role, warranting further investigation [198] (Table 3).

Pyoderma Gangrenosum

Pyoderma gangrenosum has been reported in patients with CNO/SAPHO and related autoinflammatory diseases [10, 171]. Topical treatments, including corticosteroids and calcineurin inhibitors, are used for mild localized presentations [191]. However, systemic treatment with corticosteroids or cyclosporine is required in most cases owing to the progressive nature of the disease [191]. Biologic therapies or small-molecule inhibitors may be used as second-line options, particularly TNFi, alongside inhibitors of IL-1, IL-17, IL-23, and the JAK pathway [172, 173, 206–211].

A recent review on the use of biologics in SAPHO reported N = 385 cases of conventional or biologic DMARD use, including ten pediatric patients [182]. Tumor necrosis factor inhibitors resolved skin lesions completely in 72% of cases, and excellent responses were similarly observed with IL-23 inhibitors (80%). Conversely, IL-17 inhibitors yielded a lower skin response rate, achieving complete resolution in 41% of patients. Notably, JAK inhibitors were associated with the highest efficacy, achieving complete cutaneous responses in 94% of patients (alongside 89% skeletal response). Collectively, these findings underscore that targeted therapies with biologic DMARDs or JAK inhibitors may be options for the management of co-existing skin and bone manifestations in CNO/SAPHO. However, data are largely derived from retrospective reports and small case series and paradoxical reactions have been reported [191, 212], warranting future research (Table 3).

CNO and Inflammatory Bowel Disease (IBD)

The association between CNO and IBD (either Crohn’s disease [CD] or ulcerative colitis [UC]) has first been described in the 1990 s [172, 213, 214] (Fig. 3B). A proportion of patients with CNO (estimated 10%) may exhibit asymptomatic or mildly symptomatic bowel inflammation that does not require specific treatment [41]. However, a minimum of 3–10% of patients with CNO will also experience ‘full-blown’ IBD [6, 41, 42, 215, 216]. Indeed, extra-intestinal manifestations occur in both adult and pediatric IBD [217] and may include inflammatory arthritis, sacroiliitis and/or ankylosing spondylitis (which all may also occur in patients with CNO). Further non-inflammatory bone diseases associated with IBD include osteopenia, osteoporosis, osteonecrosis, and compression fractures [218]. The association between CNO and IBD may be under-reported, especially among the relatively large proportion of patients with IBD (30–50%) with joint and/or bone pain [219, 220].

Recently, a large cohort study from Sweden, comparing patients with IBD alone with those experiencing IBD and CNO, concluded that the association between IBD and CNO may define a subgroup of patients with more severe inflammation overall [221]. Reporting N = 8244 patients with childhood-onset IBD (and N = 8244 matched controls) with 21 years of follow-up, authors reported five-fold higher incidences for CNO among patients with IBD when compared with otherwise healthy participants. Among patients with IBD, those with co-existing CNO developed IBD at a younger age, while the age at onset of CNO did not differ between patients with CNO with versus without IBD. Patients with a diagnosis of both IBD and CNO were more likely to develop additional extra-intestinal manifestations when compared with those with IBD alone. Thus, unsurprisingly, patients with IBD and CNO were more likely to receive biologic DMARDs and required more frequent outpatient appointments and hospitalizations [221].

NSAIDs

Data on the treatment and care of patients with CNO and IBD are limited. While NSAIDs are used as first-line treatment for ‘isolated’ CNO [14, 65, 222], their use in patients with CNO and IBD is controversial, especially in children with active colitis [223, 224]. The use of NSAIDs may associate with an increased risk of toxic megacolon [225] and exacerbation of IBD symptoms [223]. While a meta-analysis found no association between NSAIDs or selective COX inhibitors and exacerbation of IBD [224], European Society for Gastroenterology, Hepatology and Nutrition (ESPGHAN) consensus guidelines (2025) specifically argue against the use of NSAIDs in patients with acute severe colitis [223, 224, 226].

csDMARDs

The role of conventional synthetic DMARDs in IBD-associated musculoskeletal manifestations is nuanced. A recent review of CNO-associated IBD demonstrated comparable overall use of csDMARDs (methotrexate, sulfasalazine) between CNO-associated IBD and ‘isolated’ CNO cohorts. Sulfasalazine was specifically initiated after the establishment of an IBD diagnosis in the CNO-associated IBD group [41]. Although indicated as maintenance treatment for CD [153], methotrexate is not effective in axial arthritis. Thus, methotrexate may have a role specifically in CNO and IBD with peripheral arthritis [227, 228].

TNF Inhibition

Tumor necrosis factor inhibitors are used to treat either UC or CD in children and have remained a recommended maintenance therapy since they were first studied in pediatric patients with IBD approximately 20 years ago [229–232]. Accessing TNFi to treat CNO is more difficult and varies internationally, and the diagnosis of co-existing IBD may allow treatment escalation [65]. Moreover, patients with concomitant CNO and IBD may be a subgroup of patients in whom TNFi as first-line therapy is most appropriate (to reduce the risk of gastrointestinal flare with NSAID use) [233].

IL-12/IL-23 Inhibition

The inactivating monoclonal IL-12/−23-directed antibody ustekinumab was recently approved for the treatment of children with active CD and is an established therapy for UC or CD [234, 235]. Limited evidence (including four reported patients with IBD and CNO) exists regarding its efficacy in CNO-associated IBD [42, 43, 236]. Based on these preliminary reports, ustekinumab may be a future option in CNO patients with IBD who, for example, develop psoriasis under treatment with TNFi or require additional treatment because of high disease activity.

Notably, though the spondylarthritis/psoriasis/IBD spectrum centrally involves the IL-23/IL-17 axis, based on current reports, IL-17 blockade should not be considered in patients with CNO and co-existing IBD [39, 40]. Indeed, IL-17 inhibition has been associated with worsening intestinal inflammation in CD [157] and UC [237, 238].

Vedolizumab

Vedolizumab is a gut-specific monoclonal antibody directed against α4β7 integrin, regulating lymphocyte trafficking [158]. Vedolizumab effectively maintains remission in adult patients with CD or UC [239, 240]. Results of RCTs in children are awaited, with initial results suggesting clinical remission by 1 year in only 50% [241]. Because of its mode of action, vedolizumab is not used in patients with isolated CNO [65]. Few reports are available for the use of vedolizumab to treat IBD in patients with CNO [43, 242] in whom CNO was treated ‘separately.’

JAK Inhibition

Janus kinase inhibitors are emerging as effective therapeutic options for IBD. Tofacitinib is approved for the treatment of moderate-to-severe UC [243]. Notably, while tofacitinib successfully controls intestinal inflammation, its benefits on concomitant joint manifestations typically require extended treatment periods [244] after which JAK inhibitors control IBD-associated spondylarthritis [245]. For patients with CNO/SAPHO and IBD, data are limited and derived from case reports with mixed outcomes. Lanis et al. reported on four patients treated with tofacitinib, including one with CD. Although the patient’s bowel and skin manifestations improved, magnetic resonance imaging showed progression of CNO in the spine and sacrum, leading the authors to suggest that tofacitinib might be better suited for ‘milder’ CNO cases [160]. In contrast, induction of remission was documented in an adult patient with refractory SAPHO and UC who received tofacitinib [246]. Consequently, in a review of IBD-associated SAPHO, the authors proposed JAK inhibitors as a therapeutic option for patients not responding to anti-TNF therapy [247].

CNO-Associated Takayasu Arteritis (TAK)

Takayasu arteritis (TAK) is a rare granulomatous large-vessel vasculitis that has occasionally been reported in coexistence with CNO [248–250], and, in two cases, with CNO and IBD [248, 250]. Because of the relative rarity of reported cases, the question of whether this reflects a ‘true’ association or a coincidental finding, remains somewhat unclear. Takayasu arteritis predominantly affects the aorta, its major branches, and/or pulmonary arteries [251]. Vascular inflammation causes vessel wall thickening, stenosis, occlusion, and/or thrombosis, whereas destruction of the vessel wall may result in aneurysm formation and dissection [252]. Although the coexistence of TAK and CNO is rare, it supports the hypothesis that CNO may represent part of a broader systemic autoinflammatory disease spectrum, sharing key immunopathogenic mechanisms. In TAK, enhanced expression of IL-6, pathological activation of the NLRP3 inflammasome, and increased release of inflammasome-associated cytokines such as IL-1β and IL-18 contribute to vascular inflammation, granulomatous infiltration of the vessel wall, and progressive arterial remodeling [253, 254]. Notably, partially overlapping with CNO and associated bone inflammation, in TAK, elevated levels of circulating IL-6, IL-18, TNF, and the regulation upon activation normal T expressed and secreted chemokine (RANTES) correlate with disease activity, further highlighting the central role of these mediators in disease pathogenesis [254].

Overlapping pathological cytokine profiles in TAK and CNO provide a rationale for the use of targeted biologic DMARD therapies. In TAK refractory to corticosteroids and/or methotrexate, blockade of TNF or IL-6 signaling with tocilizumab are important therapeutic options [255–259]. While, because of its rarity, data on treatment response in pediatric TAK remain limited [260, 261], European Single Hub and Access Point in Pediatric Rheumatology (SHARE) recommendations emphasize the importance of timely diagnosis and immunomodulating therapy initiation to prevent irreversible vascular damage [262]. In this context, the occurrence of vasculitic manifestations in children with CNO, including arterial hypertension, blood pressure discrepancies between limbs, and persistently elevated inflammatory markers, should be regarded as a potential indicator of severe systemic inflammation warranting comprehensive vascular imaging and early initiation of targeted therapy.

CNO and Familial Mediterranean Fever (FMF)

Familial Mediterranean fever is a hereditary autoinflammatory disease caused by mono- or bi-allelic mutations in the Mediterranean fever gene (MEFV), encoding for the innate immune-associated pyrin protein. While several pathomechanistic models have been discussed over the years, spontaneous formation of pathogenic MEFV-variant associated pyrin inflammasomes and subsequently enhanced IL-1 signaling remain as the most likely mechanism underpinning FMF [263]. Clinically, FMF is characterized by skin (erythema), joint (arthralgia and/or arthritis) and abdominal (serositis) involvement. Ongoing inflammation in patients with undiagnosed and/or uncontrolled disease can result in amyloidosis, a potentially life-threatening complication [263, 264]. Recently, cohorts from Türkiye reported co-existing CNO in a subset of patients with FMF [264, 265], especially in patients with homozygous or compound heterozygous MEFV p.M694V variants [265] that are associated with ‘aggressive’ disease courses overall [266]. Previously, another group had reported patients with CNO exhibiting gene variants in MEFV, only some of which also fulfilled criteria for FMF [267]. Additionally in this cohort, an over-representation of the MEFV p.M694V variant was observed. Because FMF is caused by pathological activation of pyrin inflammasomes, resulting in IL-1-mediated inflammation, it appears unsurprising that bone inflammation can be present, at least in a subset of patients. As FMF is a genetically defined autoinflammatory disease with a relatively variable clinical picture, in the view of the authors, it remains to be decided whether patients with FMF with bone disease should be classed as having ‘FMF with associated CNO.’

Nonetheless, treatment of FMF with bone inflammation, even in cases with predominant bone disease, should follow internationally agreed guidance for the treatment of FMF and include colchicine (because of FMF-associated amyloidosis risk) [263, 268]. Based on the very limited literature available, biological agents, particularly IL-1 inhibitors, may be effective and safe options in patients with FMF-associated inflammatory bone disease and colchicine-resistant FMF [263, 268].

Conclusions

The autoinflammatory bone disease CNO may, over time, evolve into a ‘mixed-pattern’ disease, associated with additional organ involvement. This likely results in ‘moving therapeutic targets,’ complicating the treatment of individual features but, at the same time, offering an opportunity for stage- and phenotype-specific approaches to care. Based on the individual patient, variable organ involvement in CNO may require the assembly of interdisciplinary teams experienced in the care of CNO and associated disease, such as inflammatory arthritis, skin or bowel disease, and/or systemic vasculitis. Especially for rare associations, such as TAK or FMF, international collaboration is required to assemble clinical cohorts for knowledge generation and exchange. Lastly, clinical trials are urgently needed in ‘isolated’ CNO and beyond to generate evidence needed to achieve licensing of treatments for this underserved patient population.

Funding

Open access funding provided by Università degli Studi di Firenze within the CRUI-CARE Agreement. While the authors received no funding specific for the preparation of this article, Christian M. Hedrich received support from the NIHR (NIHR GOSH BRC, NIHR Alder Hey CRF, NIHR HTA), Arthritis UK, Funding Autoimmune Research (FAIR), Alder Hey Children’s Charity, The US Department of Defense, and the Alder Hey Kidney Fund.

Declarations

Conflict of Interest

A Schnabel, Giulia Inguscio, and E Roberts, have no conflicts of interest to disclose. Christian M. Hedrich receives unrestricted research funding from Merck (MISP programme) to study lupus nephritis.

Ethics Approval

Not applicable.

Consent to Participate

Not applicable.

Consent for Publication

Not applicable.

Availability of Data and Material

Not applicable.

Code Availability

Not applicable.

Author Contributions

AS, GI, ER, and CMH wrote the first draft of this manuscript, revised the manuscript, added to the discussion, and agreed to the final version.