Pathogenesis of spondyloarthritis

INTRODUCTION — The term spondyloarthritis (SpA, formerly spondyloarthropathy) refers to a group of disorders that includes ankylosing spondylitis (AS), nonradiographic axial SpA (nr-axSpA), undifferentiated spondyloarthritis (USpA), reactive arthritis, and the arthritis and spondylitis that may accompany psoriasis and inflammatory bowel diseases (IBD). SpA can also be differentiated into axial and peripheral SpA, depending upon the predominant regions of involvement. Axial SpA includes both AS and nr-axSpA, based upon the presence or absence, respectively, of abnormalities of the sacroiliac joints on plain radiography.

This topic review will focus primarily on the pathogenesis of AS, regarding which the most is known. The pathogenesis of each of the other members of the SpA family, especially nr-axSpA, is probably closely related to that of AS [1]. The clinical manifestations, diagnosis, and treatment of AS are presented separately. (See "Clinical manifestations of axial spondyloarthritis (ankylosing spondylitis and nonradiographic axial spondyloarthritis) in adults" and "Diagnosis and differential diagnosis of axial spondyloarthritis (ankylosing spondylitis and nonradiographic axial spondyloarthritis) in adults" and "Treatment of axial spondyloarthritis (ankylosing spondylitis and nonradiographic axial spondyloarthritis) in adults".)

The clinical aspects of the other types of SpA are also presented in detail elsewhere, as is SpA in children. (See "Clinical manifestations of axial spondyloarthritis (ankylosing spondylitis and nonradiographic axial spondyloarthritis) in adults" and "Clinical manifestations and diagnosis of peripheral spondyloarthritis in adults" and "Reactive arthritis" and "Clinical manifestations and diagnosis of psoriatic arthritis" and "Clinical manifestations and diagnosis of arthritis associated with inflammatory bowel disease and other gastrointestinal diseases" and "Spondyloarthritis in children".)

OVERVIEW OF PATHOGENESIS — Several elements are important in the pathogenesis of spondyloarthritis (SpA), a group of diseases with diverse clinical manifestations, which involve several different structures (figure 1). These elements include interactions in the context of a particular genetic background between the gut microbiome, various immune cells, and mechanical stress at the anatomic structures that are disease targets. Those structures include, for axial SpA, the entheses along the axial skeleton, the sacroiliac and facet joints, and for peripheral SpA, the peripheral entheses and the peripheral joints. At the sites of pathology, the major known mediators are tumor necrosis factor (TNF) and interleukin 17A (IL-17A). (See 'Proinflammatory mediators validated by clinical observations' below and 'The gut mucosa, gut microbiome, and IL-17A' below.)

The largest single genetic contribution is from the gene for human leukocyte antigen (HLA)-B27, but the presence of HLA-B27 is not absolutely essential. Moreover, non-HLA genes and others are also involved. (See 'Genetic factors' below.)

A major challenge for investigators is that at the entheses, where ligaments are attached to the cartilage in the vertebrae, two different processes both occur that may seem paradoxical. One is inflammation, sometimes with destruction of bone (an osteoclastic process), while the other process is new bone formation leading to syndesmophytes (an osteoblastic process). At its worst, the new bone formation can convert the entire vertebral column into a rigid bamboo spine, the hallmark of severe ankylosing spondylitis (AS). A comprehensive hypothesis of SpA pathogenesis needs to address both the inflammatory osteoclastic and the osteoblastic processes. (See 'Coexisting bone erosion and new bone formation' below.)

PROINFLAMMATORY MEDIATORS VALIDATED BY CLINICAL OBSERVATIONS — Among the variety of targeted therapies tested in spondyloarthritis (SpA) clinical trials, the only ones considered effective are aimed at one of four targets: cyclooxygenase (COX), tumor necrosis factor (TNF), interleukin 17A (IL-17A), and Janus kinases (JAKs) [2,3]:

●Cyclooxygenase – Inhibition of COX by nonsteroidal antiinflammatory drugs (NSAIDs) is very effective in controlling disease activity in some patients with SpA. This is probably because the COX enzymes are required for the generation of the proinflammatory compounds, the prostaglandins [4]. (See "NSAIDs (including aspirin): Pharmacology and mechanism of action", section on 'Mechanisms of analgesia and antiinflammatory effects'.)

●Tumor necrosis factor – TNF is a pleiotropic cytokine released among other cells by macrophages, neutrophils, and lymphocytes. It is a strong inducer of a host of inflammatory mediators [5]. The targets of therapeutic TNF blockers are the innate immune response system, the prostaglandin system, macrophages, and the nuclear factor kappa-light-chain-enhancer of activated B cells (NF-kB) pathway, as well as pathways downstream of NF-kB [6].

●Interleukin 17A – There are six members in the IL-17 family, named IL-17A through IL-17F [7]. IL-17A is the best characterized and often referred to as IL-17. However, except for the highly similar IL-17A and F, the IL-17 cytokine family is diverse and the other IL-17 family members are produced by different cell types and have different functions. IL-17 cytokines are dimeric molecules. IL-17A and IL-17F may form IL-17AA and IL-17FF homodimers and IL-17AF heterodimers. All three variants bind to the same receptor expressed on stromal cells, which activates a cascade of pathways and the transcription factors NFkB, AP-1, and C/EBP. These induce activation of genes encoding multiple inflammatory cytokines, chemokines, and metalloproteinases. Hence, IL-17A, like TNF, is also an orchestrator for a family of mediators [8,9]. IL-17A also acts in a positive feedback loop to enhance the production and effects of IL-17A. As a result of these interactions, IL-17A has a clinical effect that is as potent as that of TNF [10-12]. (See 'Overview of pathogenesis' above.)

●Janus kinases – Multiple cytokines signal through the JAK/signal transducer and activator of transcription (STAT) signaling pathways. Extracellular ligand binding leads to cytoplasmic activation of JAKs. JAK activation results in phosphorylation of STAT proteins, which then dimerize and translocate to the nucleus, where they initiate proinflammatory gene transcription programs. The JAK/STAT system is highly heterogeneous, involving four different JAKs (JAK1, JAK2, JAK3, TYK2) and seven STATs (STAT1, STAT2, STAT3, STAT4, STAT5a, STAT5b, STAT6) in humans [13].

Given that an ultimate goal of studying pathogenesis is to provide better therapy, the observations from clinical trials that these four classes of target-specific therapeutic agents are effective in at least some patients with SpA have been important contributions to the identification of critical elements in disease pathogenesis. Neither TNF nor IL-17A signal through the JAK/STAT system, suggesting that other cytokines must play a significant role in the disease process. Inhibition of the JAK/STAT cytokines IL-6 or IL-23 using monoclonal antibodies with proven efficacy in other inflammatory diseases was not successful in ankylosing spondylitis (AS) clinical trials [14-17]. The identity of the relevant JAK/STAT cytokines in AS remains to be identified.

MAJOR ELEMENTS IN THE PATHOGENESIS OF SPONDYLOARTHRITIS — The events that contribute to the pathogenesis of spondyloarthritis (SpA) are not a linearly linked series of cascades. It is a complex network of interactions involving the following:

●The axial skeleton and its entheses

●The peripheral entheses and joints

●Genetic influences

●The innate and perhaps also the adaptive immune systems

●The bowel

There are substantial variations among these elements, accounting for the enormous diversity of clinical presentations in SpA patients. As an example, these processes are thought to begin in the bowel, at least in patients with compromised bowel integrity. (See 'The gut mucosa, gut microbiome, and IL-17A' below.)

THE GUT MUCOSA, GUT MICROBIOME, AND IL-17A

Presence of mucosal lesions — The gut microbiome has been incriminated as having a role in the pathogenesis of several types of inflammatory arthritis, including spondyloarthritis (SpA), in which defects in the gut mucosal barriers have been implicated in this process. Normally, the microbiome is separated from the host by a gut epithelial barrier and a gut vascular barrier [18]. However, when the integrity of the barriers is compromised, the microbes become capable of initiating a systemic immune response [19]. Lesions in the gut mucosa in patients with SpA were first demonstrated in the 1990s [20]. Evidence supporting a role for the gut and gut microbiota in SpA pathogenesis includes [21-24]:

●The demonstration by ileocolonoscopy of the presence of subclinical acute or chronic intestinal inflammation in about two-thirds of patients with SpA, with functional breaches in the gut epithelium in SpA

●The correlation between chronic gut inflammation and magnetic resonance imaging (MRI)-demonstrable SpA disease activity

●The sharing of genes between ankylosing spondylitis (AS) and Crohn disease

●The presence of AS in 4 to 10 percent of patients with ulcerative colitis and Crohn disease

●Findings of sacroiliitis on plain radiographs in a significant proportion of patients with Crohn disease, many of whom do not have symptoms of inflammatory back pain

SpA-specific gut microbiome and damage of intestinal mucosal barriers — The composition of the gut microbiome, which is influenced by genetic and other factors, differs in patients with SpA from healthy individuals. Evidence in both humans and more so in animal models supports the view that the gut microbiome plays a vanguard role in the cascade of events leading to inflammation in many patients with SpA.

The strongest evidence that the gut microbiome is necessary for the development of SpA comes from animal models [25]. In these models, rats and mice develop SpA-like clinical and pathologic features when housed in their usual laboratory environment; however, they fail to develop such features when raised in a germ-free environment, although the SpA-like features do appear when the animals are brought out of the germ-free environment.

Data in humans also argue for a role of the gut microbiome in disease pathogenesis. The enormous diversity in the trillions of commensal microbial cells in the human gut is affected by multiple factors, including geography, ethnicity, major histocompatability complex (MHC) genes, therapy, age, sex, and diet. A description, or "census," of the microbiota in each individual can allow for the characterization of that individual's microbiome by sequencing and analytic techniques [26,27]. Using this approach, a core microbiome set can be demonstrated, which is unique for each person at a given point in time and has the capacity to distinguish between different individuals [28].

A large number of studies have compared the gut microbiome between SpA and healthy subjects, consistently demonstrating differences in the microbiome between SpA and healthy subjects. However, findings concerning the particular SpA-specific microbial species vary between studies. What appears to be more consistent is that the SpA microbiome is more enriched in species that are mucolytic and potentially able to degrade the gut barrier. Breakdowns were demonstrated by immunohistologic findings of the downregulation of junctional proteins and evidence of increased serum levels of bacterial lipopolysaccharides [29-34].

IL-17A and gut microbial invasion — The loss of architectural and functional integrity in the intestinal epithelium allows for passage of microbiota or their metabolites into the submucosa and the systemic circulation (see 'SpA-specific gut microbiome and damage of intestinal mucosal barriers' above). The first line of cytokine defense is IL-17A. Acting in synergy with IL-22, it guards the integrity of the gut epithelium against microbes by inducing the generation of anti-microbial peptides [35]. The IL-17A inhibitor secukinumab failed in a clinical trial in Crohn disease, and intestinal inflammation appeared to worsen in a subset of patients [36]. However, while new-onset inflammatory bowel disease (IBD) or exacerbation of preexisting gut inflammation in SpA patients treated with IL-17A inhibitors has been reported, the incidence of such events is low [37,38].

There are multiple factors that can drive a variety of cell types to generate IL-17A [8,35].

Although IL-17A-producing cells activated by IL-23 were at one time considered the most likely disease-causing candidate cells in SpA, randomized controlled trials have failed to demonstrate a significant clinical benefit of two different anti-IL-23-blocking antibodies in AS. This is despite the efficacy of both drugs for psoriasis, psoriatic arthritis, and Crohn disease [14,15,39]. Hence, the IL-17A-producing cells in AS are probably predominantly IL-23 independent [40].

Potential roles of several types of IL-17A-positive cells — Besides Th17 cells and mast cells, several types of innate-like lymphocytes in the intestine are capable of producing IL-17A, including type 3 innate lymphoid cells (ILC3), mucosal-associated invariant T cells (MAIT), gamma-delta T cells, and innate-like invariant natural killer T cells (iNKT) [12,41]. Innate-like lymphocytes can rapidly produce a large quantity of cytokines. Some of them can recognize microbial products presented by nonclassical MHC molecules [32,40,42,43].

ILC3 cells provide an example of how innate immune cells might travel from the gut to the entheses and joints (figure 1) [44-46]. Upon becoming activated in the gut, these ILC3 express the alpha-4/beta-7 integrin, which can function as a homing receptor. The ligand for this integrin is mucosal vascular addressin cell adhesion molecule 1 (MADCAM1), which, in patients with AS, is strongly expressed in the high endothelial venules (HEV) of the gut and bone marrow [45]. Although not directly demonstrated, these gut-derived ILC3 cells in patients with AS are postulated to migrate into the systemic circulation and to home via this integrin-ligand interaction into the bone marrow, the peripheral joints, and the entheses [47-49].

Alternately, such IL-17A-positive cells might originate in locations other than the gut or be tissue-resident, ie, they may originally reside and become activated in the target tissue [50,51].

Entheses and the role of mechanical stress — It is thought that the major targets of the disease process in patients with SpA are the entheses, where tendons and ligaments are attached to bone [52]. Because of the mechanical load, entheses are highly susceptible to micro-injury. The attachment of the Achilles tendon to the calcaneus, for example, is subjected to mechanical stress of 3 to 10 times the body weight during activities. Such micro-injuries can activate resident immune cells that include the ILC, gamma delta T cells, natural killer cells, and conventional T cells. Together with the neutrophils, they also release chemoattractants for circulatory proinflammatory cells. All these cells orchestrate a TNF- and IL-17-dependent inflammatory reaction even in subjects with no arthritis. It is postulated that SpA is caused by a faulty fine-tuning of this local reactivity [53-55].

GENETIC FACTORS — Genetic factors have overwhelming importance in susceptibility to ankylosing spondylitis (AS). First-, second-, and third-degree relatives of patients with AS have markedly increased risks of developing the disease (relative risks of 94, 25, and 4, respectively) [56]. The mode of inheritance is polygenic with multiplicative interaction among loci [57,58]. The association with the human leukocyte antigen B27 (HLA-B27) gene was recognized in 1973 and has the strongest association with the disease (table 1) [59,60]. The overall contribution to AS heritability by HLA-B27 is estimated at approximately 20 to 30 percent [61]. The contribution of the major histocompatibility complex (MHC) region is 40 to 50 percent [40]. (See 'Role of HLA-B27' below.)

Role of HLA-B27 — Because HLA-B27 is present in approximately 80 to 95 percent of patients with AS in most ethnic groups, compared, for example, with approximately 6 percent of the general population in the United States [59,60,62], it is assumed to play a major role in the pathogenesis of AS. HLA-B27 itself is highly polymorphic and more than 250 protein-coding variants have been identified; small differences in the amino acid sequence between these subtypes determine whether they are associated with spondyloarthritis (SpA), including AS [63-65]. The most frequent subtypes, HLA-B2705 and HLA-B2704, are strongly associated with disease [63,66,67]. Only two subtypes, HLA-B2706 and HLA-B2709, are considered not to be associated with SpA [56,68,69]. Many other subtypes are rare and their association with AS has not been studied in detail.

At this point, it is still unclear what role HLA-B27 plays in the pathogenesis of SpA [70]. Besides its role in cell-mediated immunity, studies in healthy HLA-B27 positive individuals have shown that there is an association of the HLA-B27 genotype with the overall gut microbial composition, which is distinct from the composition of the gut microbiome that is observed in subjects without HLA-B27 [71,72]. These findings may relate to the proposed role of the gut microbiome in the pathogenesis of SpA, and will be a focus of further research.

Most research on the role of HLA-B27 in the pathogenesis of SpA has historically focused on how different structures of HLA-B27 mediate inflammatory processes. Although no fully satisfactory explanations have been established, it is clear that the entire set of intracellular processes of formation of the HLA-B27 molecule need to be considered (figure 2).

Classical (canonical) structure of HLA-B27 — Several features distinguish HLA-B27 from most other HLA class I molecules; these features may be relevant to disease susceptibility according to some hypotheses. The classical (canonical) structure is shared with other HLA class I molecules [73-76]:

●The HLA class I molecule is composed of a 45 kD polymorphic heavy chain, noncovalently complexed with a 12 kD monomorphic unit, beta-2-microglobulin.

●The heavy chain is composed of three domains. The first two domains (alpha-1 and alpha-2) together form two antiparallel helices resting on a platform of an eight-stranded beta-pleated sheet, which itself rests on two barrel-shaped structures derived from the complex of the third domain (alpha-3) and the beta-2-microglobulin (figure 3).

●An antigenic peptide that is usually 8 to 11 amino acids in length rests inside the platform. These peptides are derived from endogenous proteins and from proteins of viruses and bacteria that have invaded the cells (figure 4).

The features that distinguish HLA-B27 from most other HLA class I molecules include:

●Most antigenic peptides associated with HLA-B27 have arginine as the second residue [73-75].

●The presence of an unpaired cysteine at residue 67 (Cys67) – This unique feature allows for the formation of homodimers and oligomers of free heavy chains. (See 'HLA-B27 misfolding and autophagy' below.)

Since the physiologic function of HLA-B alleles is to present cytosolic peptides to the T-cell receptors on CD8+ T cells, the favorite hypothesis for SpA is the "arthritogenic peptide hypothesis." It postulates that there are certain microbial peptides that are very similar to self-peptides from the point of view of the T-cell receptors of certain HLA-B27-restricted CD8+ T lymphocytes (cytotoxic T lymphocytes). The reactivity of these T lymphocytes with these HLA-B27-peptide complexes would then lead to autoreactivity and autoimmune disease [77-79]. There is indeed an enrichment of such bacterial peptides in the stool of patients with AS [80]. Most importantly, this hypothesis is supported by multiple studies of T cell profiling to identify both the responsible T-cell receptors as well as the cross-reactive self and microbial peptides [81,82]. Intriguingly, an HLA-B27-positive patient with refractory AS has been reported whose symptoms responded favorably to immunotherapy with repeated infusions of a monoclonal antibody depleting a limited set of TCRVb9+ T cells [83].

HLA-B27 as free heavy chains — HLA-B27 can also exist as a dimer of two heavy chains without the beta-2-microglobulin [84-86]; these proteins are present in the gut and synovium of SpA patients and may contribute to the pathogenesis of SpA. HLA-B27 homodimers interact with the killer immunoglobulin-like receptor KIR3DL2 expressed on NK cells and other lymphocytes. HLA-B27 homodimers have been shown to stimulate KIR3DL2-positive CD4+ T cells to produce IL-17A [87].

HLA-B27 misfolding and autophagy — HLA-B27 folds more slowly than other HLA molecules into the canonical MHC class I structure inside the endoplasmic reticulum. The "misfolding hypothesis" explains the HLA-B27 association with SpA protein through the accumulation of misfolded HLA-B27 heavy chains in the endoplasmic reticulum leading to the activation of secondary effector mechanisms [88,89]. The HLA-B27 misfolding hypothesis can be summarized as follows (figure 5):

●Mature HLA-B27 has a quaternary structure with three different components (HLA-B27 heavy chain, b2 microglobulin and an 8-11 aa peptide). It is assembled and folded from a linear structure in the endoplasmic reticulum, a cellular compartment.

●For several reasons, including the cysteine residue at position 67, the folding process of HLA-B27 is slower than that for other HLA alleles.

●Improperly folded HLA-B27 proteins (ie, those not yet in the canonical mature class I conformation) accumulate in the endoplasmic reticulum.

●This can lead to a misfolding process, which activates autophagy and activation of the IL-23/IL-17A pathway. One type of cells that undergo autophagy in AS are the Paneth cells in the intestinal epithelial lining [47,90].

●Alternately, misfolding can induce a process termed the unfolded protein response (UPR), a cellular stress response that can also activate the IL-23/IL-17A pathway (figure 6). (See 'Coexisting bone erosion and new bone formation' below.)

Non-HLA genes — The search for non-HLA-B27 genes that may be important in SpA pathogenesis has focused almost exclusively on patients with AS and has suggested that non-MHC genes are also important in disease susceptibility. Since 2007, several large genome-wide association studies (GWAS) have been carried out in several populations of European [91-95] and Han Chinese descent [96]. These studies have identified a total of at least 114 genetic variants. The MHC, especially HLA-B27, is a major contributory factor. Approximately 7 percent of the heritable risk is from non-MHC variants (table 1) [61,94,95,97-102]. The apparent need for an HLA-B27-positive individual to also carry some of these non-MHC genes in order to develop AS may explain, at least in part, why only 1 to 5 percent of HLA-B27-positive individuals develop AS.

Even though the total contribution of all these non-MHC genes to AS heritability is relatively small, these associations provide clues about the pathogenesis of AS. In addition, their significance can be amplified by gene-gene interactions [103]. These non-MHC AS risk genes can be grouped into several functional categories:

●ERAP1 and ERAP2 – The two endoplasmic reticulum aminopeptidases that encode genes related to AS are termed ERAP1 and ERAP2; each of the genes has variants that may increase the risk of AS and variants that are protective. These two enzymes are responsible for the generation as well as trimming and destruction of peptides in the endoplasmic reticulum to achieve the correct length for loading into HLA class I molecules, such as HLA-B27, for antigen presentation [104]. The list of AS risk loci contains additional genes encoding enzymes involved in peptide trimming in the endoplasmic reticulum. The precise mechanistic role of ERAP1, ERAP2, and these other peptidases in SpA pathogenesis remains unclear [87,105].

●Tumor necrosis factor receptor gene family – The presence of an additional group of genes, including those for the lymphotoxin beta receptor (LTBR) and tumor necrosis factor receptor 1 (TNFRSF1A), provides additional support for the role of tumor necrosis factor (TNF) in axSpA [61]. (See 'Proinflammatory mediators validated by clinical observations' above.)

●Interleukin 23/interleukin 17A axis – GWAS have identified a number of genes that are associated with IL-23 receptor signaling [94].

●T lymphocyte activation and differentiation – The association of AS with genes modulating activation and differentiation of either CD4+ or CD8+ T lymphocytes is consistent with potential involvement of these cells with disease pathways.

●Genes associated with radiographic progression in AS – The genes listed above are associated with susceptibility of disease. In AS, spinal damage demonstrable on radiographs is the most disabling factor. Genomic data on determinants of radiographic damage are limited. Using GWAS in 444 AS patients, 1 study identified a gene known as ryanodine receptor 3 (RYR3), which was associated with severity of radiographic spinal damages [94].

●A gene associated with radiographic progress in AS – The genes listed above are associated with susceptibility of disease. In AS, spinal damage is the most disabling factor. Using GWAS in 444 AS patients, 1 study identified a gene known as ryanodine receptor 3 (RYR3), which was associated with severity of radiographic spinal damages [106]. RYR3 encodes a channel release protein that regulates intracellular calcium homeostasis. It is expressed in musculoskeletal tissues.

How these various genes are activated to express their respective proteins and how their expression is modulated in flares and relapses of AS remain to be clarified.

More advanced nucleotide-based technologies might be able to answer these questions. These include whole genome sequencing, studies of microribonucleic acid (microRNA), methylation, and histone acetylation [107].

COEXISTING BONE EROSION AND NEW BONE FORMATION — The mechanisms by which bone inflammation and erosions can occur in patients with spondyloarthritis (SpA) along with new bone formation have not been fully elucidated. These findings may appear paradoxical and there is controversy regarding the interpretation of the available data. The findings in SpA are in contrast with those in rheumatoid arthritis (RA), where only bone erosion is observed.

One clue is provided by longitudinal studies of the spine in the same patients using MRI and radiography. Using two different sequences, MRI can visualize both inflammation and repair processes, identified respectively as bone marrow edema and fat metaplasia. New bone formation (for example, syndesmophytes) is best visualized by plain radiography or computed tomography (CT) imaging. These studies suggest that the initial change is inflammation in which cytokines such as tumor necrosis factor (TNF) and IL-17A directly or indirectly activate osteoclast precursor cells. This is then followed to a certain extent by a reparative process at the vertebral corners and other sites of the axial skeleton, and finally by new bone formation (such as syndesmophytes) (figure 7) [108]. Longitudinal imaging studies suggest that new bone formation is more likely to take place in locations where there has been fat metaplasia and that new bone formation appears after the inflammation has subsided [109]. Osteoblast differentiation is driven by bone morphogenic proteins and the Wnt pathway, and it is thought that these signals also play a critical role in the pathological bone formation in AS [52]. [110]

New bone formation in AS does not occur in the vertebral body, where the inflammation has destroyed the microarchitecture, leading to osteoporosis; rather, syndesmophytes form at entheses. Stromal cells such as bone marrow mesenchymal cells (BM-MSC) and fibroblast-like synovial cells (FLS) are thought to play an important role in this process. Normally, the BM-MSC reside in the bone marrow, but they are capable of migrating through pores into the entheses. There, mesenchymal cells are driven into osteogenesis by interleukin 22 (IL-22), IL-17A, and TNF. These mesenchymal cells also secrete a chemokine to augment the reactivity [12,55,111]. The FLS on the other hand can develop into osteoblasts even independent of an inflammatory environment [41].

Other observations suggest that activation of the unfolded protein response by misfolded HLA-B27 in mesenchymal cells of the spinal entheses drives syndesmophyte formation in AS via induction of tissue-nonspecific alkaline phosphatase (TNAP). In that study, serum levels of bone-specific TNAP correlated with radiographic progression [112]. Another group found that macrophage migration inhibitory factor (MIF) was raised in the serum of patients with AS and can predict radiographic progression [113,114]. MIF was subsequently shown to be a critical mediator of inflammation in a mouse model of spondyloarthritis [115].

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Beyond the Basics topics (see "Patient education: Axial spondyloarthritis, including ankylosing spondylitis (Beyond the Basics)")

SUMMARY

●Targets of inflammation – The major areas of involvement in ankylosing spondylitis (AS) and the other forms of spondyloarthritis (SpA) are at the articulations of the axial skeleton, at the interface between ligaments/tendons/bone (entheses). (See 'Overview of pathogenesis' above.)

●Mediators of inflammation – Mediators that are current targets of therapy are the cyclooxygenases (COX), tumor necrosis factor (TNF), interleukin 17A (IL-17A), and cytokines signaling through the Janus kinase (JAK)/signal transducer and activator of transcription (STAT) signaling pathway, whose identity remains to be clarified. Results from clinical trials document the significance of these enzymes and cytokines to processes important in the pathogenesis of SpA. (See 'Proinflammatory mediators validated by clinical observations' above.)

●Enthesitis pathogenesis

•Microbiome – For patients with microscopic bowel lesions, the disease processes probably start in the gut, where several types of rapidly responding cells such as the innate lymphoid cells (ILCs) that produce IL-17A (a proinflammatory cytokine) and IL-22 are activated by SpA-specific gut microbiota (figure 1). (See 'The gut mucosa, gut microbiome, and IL-17A' above.)

Immune cells activated in the gut may migrate to the entheses and the joints, causing an inflammatory process in which TNF and IL-17A participate. IL-17A causes a bone-erosive process. (See 'IL-17A and gut microbial invasion' above and 'Coexisting bone erosion and new bone formation' above.)

•Mechanical stress – The entheses themselves also contain several types of these cells, which can be activated via mechanical stress to generate IL-17A. (See 'Entheses and the role of mechanical stress' above.)

•Genetics – The pathogenic events in SpA take place in a complex genetic background. The major gene is human leukocyte antigen (HLA) B27, which probably participates in several of the key pathogenic processes. Additionally, a large number of non-HLA genes have also been identified as having associations with AS. They interact and reinforce several of the disease-causing pathways. (See 'Genetic factors' above.)

●Syndesmophyte pathogenesis – Initial changes in bone result from inflammation, followed by a reparative process involving mesenchymal cells at the entheses. The ultimate consequence is new bone formation, such as the formation of syndesmophytes (figure 7). (See 'Coexisting bone erosion and new bone formation' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges David Yu, MD, who contributed to earlier versions of this topic review.