Spondyloarthritis (SpA) represents a heterogeneous group of chronic inflammatory diseases characterized by varying degrees of shared pathophysiological, genetic, clinical and radiographic features (1, 2). A distinguishing hallmark across all subtypes of SpA is the primary involvement of the enthesis, a fibrocartilaginous transitional tissue that anchors tendons and ligaments to bone, conferring resistance to biomechanical stress (3).

Axial spondyloarthritis (axSpA) is a subset of SpA, predominantly (though not exclusively) affecting the spine the spine and sacroiliac joints. Its characteristic progression towards bony ankylosis contributes to profound disability, pain and functional limitations in advanced disease stages (1). According to the 2009 ASAS classification criteria (4), axSpA encompasses both radiographic and non-radiographic forms, distinguished by the presence or absence of structural damage detectable via conventional radiography at the time of assessment. The historical entity of ankylosing spondylitis (AS), previously defined by the New York criteria (5), now aligns with the classification of radiographic axSpA (r-axSpA). The updated classification for non-radiographic disease, enables earlier diagnosis and timely therapeutic intervention. The global prevalence of axSpA is estimated to range between 0.1% to 1.4%, with a male-to-female ratio of approximately 2-3:1 (6).

Within the SpA spectrum, psoriatic arthritis (PsA) constitutes a multifaceted immune-mediated disorder marked by varying incidences of peripheral arthritis, enthesitis, dactylitis and spondylitis, frequently co-occurring with personal or familiar histories of psoriasis. PsA is classified using the CASPAR (7) and typically presents in middle-aged females (8), with an estimated global prevalence of 0.1-1% (9). Axial involvement has garnered increasing attention in recent years. This subset of the disease is more frequently observed in males (8), with prevalence estimates varying widely from 25% to 70% depending on the criteria applied (10, 11). These findings support the recognition of axial PsA as a distinct clinical phenotype. In individuals aged over 60–65 years, this phenotype is associated with more severe clinical and radiographic manifestations and poorer outcomes (12).

Bone involvement in both axSpA and PsA arises from complex interactions between anabolic and catabolic processes, variably contributing to bone loss and new bone formation (13).

Bone loss may manifest as systemic reductions in bone mineral density, thereby increasing fracture risk (14), or as local bone erosions (15), which are more prevalent in peripheral PsA but may also affect the sacroiliac joints and vertebrae in axSpA. Conversely, new bone formation can lead to progressive ankylosis and bridging between adjacent bones, substantially exacerbating disease burden (16).

It is esteemed that up to 66% of patients with axSpA experience intense fatigue, stiffness, limitations in daily activities (including self-care and low-effort tasks), diminished quality of life, higher incidences of nocturnal awakenings, and even severe insomnia (17).

The functional limitations resulting from bone remodeling are directly implicated in occupational impairment (18), with nearly 40% of affected individuals unable to work. Among these, 24% are compelled to take early retirement, while 45% transition to less demanding roles (19).

This review seeks to elucidate the intricate topic of bone metabolism in spondyloarthritis, highlighting shared features and key differences between axSpA and PsA. To this purpose, the latest insights into the radiographic characteristics and underlying molecular pathways of each condition are examined, with the aim of advancing our understanding of the bone-centric pathogenetic mechanisms. Such knowledge may provide clinicians with improved tools to refine therapeutic strategies and enhance disease management.

The delicate balance of bone tissue homeostasis is maintained through a continuous process of resorption and new bone formation, primarily mediated by osteoclasts and osteoblasts, respectively. The key molecular pathways and therapeutic targets involved in bone remodeling in axSpA and PsA are depicted in Table 1 . Osteoclastogenesis refers to the process by which osteoblasts (20) directly and indirectly influence the differentiation of macrophage/monocyte precursor cells into active bone- resorbing osteoclasts by producing key signaling molecules (21). The two main cytokines driving this cascade is macrophage colony-stimulating factor (M-CSF) (22) and receptor activator of nuclear factor-kappa B (NF-κB) ligand (RANKL) (23).

Osteoclast precursor express RANK on their surface, rendering them responsive to RANKL produced by osteoblast and other cell types, such as osteocytes. This interaction activates multiple intracellular pathways, including NF-kB and MAPK (24).

Recently, significant attention has been given to immunoreceptor tyrosine-based activation motif (ITAM) proteins, which have been shown to play a critical role in inducing calcium signaling within the mineralized matrix. This function is absent in the RANK/RANKL pathway (25). Ultimately, these interactions converge to activate NFATc1, a key transcription factor for osteoclastogenesis. The complexity of this cytokine interplay is further modulated by NFATc1-inhibiting molecules that counteract RANKL activity (26).

RANKL- and RANK-deficient murine models, exhibit marked osteopetrosis due to impaired osteoclast differentiation (27). An important inhibitor of RANK/RANKL pathway is osteoprotegerin (OPG), a soluble decoy receptor for RANKL produced by osteoblasts. OPG suppresses osteoclastogenesis, thereby increasing bone mineral density (28). Engineered mice lacking OPG demonstrate severe osteoporosis (29). Building on these findings, denosumab, anti-RANKL antibody that mimics OPG natural activity, has been developed and is now widely employed in the treatment of osteoporosis and cancer-related bone disorders (30).

Once activated, osteoclasts adhere to the bone surface and initiate the dissolution of hydroxyapatite by creating an acid environment through the secretion of intracytoplasmic hydrogen ions (31). Additionally, the organic components of the bone matrix are remodeled by proteolytic enzymes, such as cathepsin K (27).

Osteoblasts, on the other hand, differentiate from mesenchymal stem cells (MSCs) (32) under the influence of various transcription factors, most notably Wnt proteins and bone morphogenetic proteins (BMPs) (33). Wnt/Wntβ-catenin signaling pathway plays a pivotal regulatory role in new bone formation, which is particularly relevant in inflammatory contexts such as axial spondyloarthritis (34). The equilibrium in bone formation is maintained by two key inhibitors of Wnt pathway: Dickkopf1 (DKK1) (35) and sclerostin (36).

Sclerostin, an antagonist of Wntβ-catenin signaling, is produced by osteocytes (37). Reduced sclerostin levels are associated with increased bone mass (38), making the anti-sclerostin antibody romosozumab, a therapeutic option for diseases characterized by systemic bone loss (39). Recent studies have suggested that cardiotrophin-1 (CT-1), secreted by osteoclasts, suppresses sclerostin production by osteocytes, providing a functional link between bone resorption and new bone formation (40).

Moreover, it is crucial to highlight that sclerostin expression has been shown to be decrease under mechanical stress (41) and in the presence of elevated inflammatory mediators such as interleukin-6 (IL-6) (42) and prostaglandin-E2 (PGE2) (43).

Axial involvement typically originates in the sacroiliac joints (SIJ), making this the primary focus of diagnostic imaging (44). Figure 1 (Panel 1.a, 1.b) summarizes the most common radiographic findings observed throughout the disease course. Historically, plain radiography has been the gold standard for assessing bone changes associated with axSpA, with radiographic findings serving as key features in classification criteria for AS. Joint erosion typically occurs in the cartilaginous compartment of the joint while the ligamentous retroarticular space is usually spared. These changes manifest as blurring of subchondral bone on opposing articular surfaces, followed by erosions, joint space widening, and, over time, new bone formation, which appears as articular sclerosis and, eventually, total ankylosis of the joints (45).

However, these radiographic features typically indicate an advanced disease stage, often developing over years after symptoms onset. Consequently, negative radiography will be followed by magnetic resonance imaging (MRI) as the preferred modality for early detection. Although radiography has historically been a cornerstone in diagnosing axSpA, its inability to detect active inflammation in joints and entheses significantly limits its utility in early disease stages. In these instances, MRI offers superior sensitivity and specificity (46).

MRI can reveal active synovitis, periarticular inflammation, tenosynovitis and BMO in axial and peripheral joints as hyperintense signals on T1-weighted images with gadolinium enhancement or short tau inversion recovery (STIR) sequences (46).

On MRI, active inflammation is indicated by bone marrow oedema (BMO), while capsulitis and enthesitis are less specific signs of inflammation. BMO adjacent to the cartilaginous joint surface appears as a high-signal lesion on fat-suppressed T2-weighted (T2w) sequences and a corresponding signal reduction on T1-weighted (T1w) sequences. These findings define osteitis or, when affecting SIJ, sacroiliitis, as outlined in the 2019 update of the ASAS definitions for MRI lesions (47).

The presence of sacroiliitis is a critical criterion for the diagnosis and follow-up of axSpA and is included in the most recent ASAS classification criteria (4).

Bone erosion is another MRI sign of ongoing articular inflammation. It manifests as morphological alterations in the cartilaginous part of the joint, discernible on T1w sequences with and without fat-suppression, and progresses to apparent joint space widening (48). As active inflammation subsides, reparative processes emerge. Fat lesions, visible as bright signals on T1w images, replace areas of BMO. Erosions may become filled with fat metaplasia (backfill), and in advanced stages, bone budding can lead to SIJ narrowing and eventual ankylosis (49) ( Figure 2 ).

A similar pathological sequence occurs in the spine. Active osteitis of the vertebrae is followed by fat metaplasia, corner erosions, and reactive sclerosis, also known as “shiny corner” (45, 50). New bone formation is indicated by the presence of vertebral body squaring or syndesmophytes (51), which are bony outgrowths originating from the attachment sites of the anulus fibrosus at vertebral corners. Syndesmophytes may extend vertically as fine “hair- shaped” projections along the anterior vertebral margin. They may remain non-bridging (52) or, in chronic cases, fuse adjacent vertebrae (bridging-syndesmophytes), often accompanied by similar fusion processes in the facet joints, culminating in the classic “bamboo spine” appareance (51, 53). Inflammation of the vertebral endplates (so-called Andersson lesions) may additionally result in transdiscal ankylosis, best depicted by fat signal inside the disc space on MRI ( Figure 3 ).

PsA is characterized by the involvement of bone tissues in both axial and peripheral joints. A key radiographic hallmark of the disease is the pronounced new bone formation, often resulting in a “fuzzy” appearance of the juxta-articular bone (90), which is incorporated into the CASPAR classification criteria (7).

In peripheral joints, notable radiographic features include metaphyseal periostitis, which manifests as periosteal new bone layering or irregular cortical thickening. In some cases, this may culminate in the so-called “ivory phalanx” sign, where an entire phalanx displays increased radiodensity. Conversely, ongoing marginal erosions and bone resorption may lead to substantial alterations in joint integrity, including subluxations and the characteristic “pencil- in-cup” deformity (91). These two distinct aspects of bone pathology, new bone formation and bone resorption, often coexist, contributing to the overall burden of the disease and negatively affecting the patient’ s clinical progression and treatment outcomes ( Figure 5 ). In axial PsA, sacroiliitis is a hallmark feature, sharing similarities with axSpA. It is defined by the progressive development of erosions, sclerosis, bony bridges, and ankylosis, as per the New York criteria (5). The rest of the axial skeleton may be the initial site of disease manifestation, independent of prior SIJ inflammation ( Figure 6 ). Axial involvement is typified by neo-appositional changes such as ligamentous ossifications, ossification of anterior vertebral ligaments, squaring of vertebral bodies, and irregular, bulky syndesmophytes (para-syndesmophytes). Simultaneously, erosive lesions of vertebral surfaces are commonly observed (92). Figure 1 (panel 2.A) illustrates para-syndesmophytes, along with the two phases of bone inflammation, and unilateral sacroiliac joint involvement with progressive changes. Panel 2.B shows distal interphalangeal joint alterations, while panel 2.C depicts Achilles tendon involvement, both characteristic of PsA.

AxSpA and PsA share a fundamental characteristic: both involve local inflammation of joints and periarticular tissue. The cytokines driving this inflammation exert extensive effects on bone remodeling. Both conditions are characterized by systemic and local bone loss alongside a pro-ankylosing tendency, though the patterns of these changes differ significantly between the two diseases (45–53, 90–92).

Key mediators such as TNFα, IL-17, and IL-6 play critical roles in driving the distinctive bone lesions that are hallmarks of axSpA and PsA. While the relative contributions of these cytokines to the pathological processes are not yet fully elucidated, it is evident that their interplay underpins the complex regulation of tissue changes, including erosions and new bone formation (56–72, 93–118).

Despite sharing a common cytokine milieu involved in bone remodeling, axSpA and PsA exhibit profound differences in their morphological, pathogenetic, and topographic characteristics. AxSpA is marked by axial bony neo-apposition, which ca culminate in complete joint ankylosis. In the early phases erosive lesions predominantly affect axial skeleton, representing scars of intense inflammation targeting the joints. These erosions lead to subtle alterations in bone structure, which in turn disrupt mechanical force distribution and activate a feedback loop involving osteoblasts, osteocytes and osteoclasts. New bone formation in axSpA appears to be a reparative response to inflammatory or mechanical injury, aiming to restore an impaired weight-bearing system (88, 89).

Mechanical stress may also act as a potent activator of γδ-Tc within spinal entheseal and peri- entheseal tissues, driving IL-23-independent IL-17A production, osteoblast activation, and subsequent new bone formation in forms such as fine hair-shaped syndesmophytes, bone buds, and eventual ankylosis (66–73). Furthermore, syndesmophyte formation contributes to trabecular bone unloading, which suppresses osteocyte activity and exacerbates bone loss, perpetuating a vicious cycle (50–84).

In contrast, PsA primarily shows a peripheral erosive pattern, likely influenced by the close interplay between joint and skin inflammation. Psoriatic plaques, through skin resident γδ-Tc, ILCs and keratinocytes activated in the inflammatory milieu, promote enhanced osteoclastogenesis and local and systemic bone loss (107, 113).

A key distinction in PsA lies in the uncoupling of osteoblasts and osteoclasts activity. Unlike axSpA, erosive lesions and new bone formation in PsA are not chronologically linked and may occur simultaneously. Moreover, axial bone neo-apposition in PsA differs significantly from that in axSpA, producing bulkier, asymmetrical para-syndesmophytes, indicative of a more chaotic new bone formation process. This distinction likely reflects differences in the cytokine’s networks driving disease processes (91–121).

AxSpA appears to be predominantly influenced by TNFα, as suggested by the “TNFα brake hypothesis”, which posits a more controlled balance between osteoblast and osteoclast activity, leading to a more structured bone formation process (58). In contrast, axial bone remodeling in PsA shows reduced sensitivity to TNFα effects and greater reliance on IL-22, IL-23, and IL-17 activity, resulting in a less regulated remodeling process. Myeloid cells resident in peripheral entheses of axPsA patients exhibit increased IL-23 production (127), further highlighting this molecular disparity.

A comparative overview of bone remodeling mechanisms, structural features, and therapeutic effects in axSpA and PsA is provided in Table 2 . This divergence in cytokine activity is clinically underscored by the differential efficacy of anti-IL23 agents. Phase III trials of IL-23 inhibitors such as ustekinumab (128), risankizumab (129), and guselkumab (130) failed to demonstrate efficacy in axSpA patients. Conversely, these agents demonstrated significant benefits in PsA cohorts, including positive effects on axial radiographic progression (131–133).

While significant advancements have been made in understanding the molecular pathways underlying bone remodeling in axSpA and PsA, the intricate interplay of inflammatory and mechanical factors remains incompletely elucidated. This review not only highlights the shared cytokine-driven processes of bone erosion and neoformation but also underscores the distinct pathogenetic and phenotypic differences between these two diseases. These distinctions call for a more nuanced differential definition, enhanced assessment tools, and tailored, disease- specific therapeutic strategies. One particularly compelling observation is the differential impact of cytokine networks, with a relative predominance of TNFα in axSpA and a greater reliance on IL-17, IL-22, and IL-23 pathways in PsA. This divergence is reflected in the clinical efficacy of targeted treatments, as IL-23 inhibitors have shown promising results in PsA but have failed to demonstrate efficacy in axSpA. These findings challenge the notion of a “one- size-fits-all” approach to managing spondyloarthtitis and emphasize the need for context- specific therapeutic paradigms. Moreover, the apparent dichotomy in bone remodeling, ankylosis and erosive lesions in axSpA versus the uncoupled yet concurrent processes of erosion and neoformation in PsA, warrants further exploration. A deeper understanding of how these processes is regulated and interconnected at the molecular and cellular levels could elucidate the mechanisms underlying disease progression and clinical heterogeneity. Future research priorities should include:

Such efforts hold the potential to advance truly personalized medicine, equipping rheumatologists with the tools to refine therapeutic strategies that not only mitigate inflammation but also address the fundamental bone remodeling abnormalities driving disease burden, disability, and societal costs. Ultimately, addressing these critical gaps in knowledge will be essential to revolutionize the management of axSpA and PsA. By transitioning from reactive to proactive approaches, we can aim to preserve bone integrity, improve patient quality of life and reduce the significant socioeconomic impact of these chronic conditions.