*Axial SpA (axSpA)*

Pathologically, axSpA is characterized by a three-stage process, resulting in a paradoxical bone destruction and formation. It has been postulated that the initial inflammation (in which TNFα is the principal cytokine involved [71]) causes erosions in cartilage and bone; these lesions are then filled in by fibrous tissue, that is later ossified leading to abnormal bony outgrowth (syndesmophytes, bone bridges and complete ankyloses) [9,68,72]. The Wnt signaling and its inhibitors, such as Dkk-1 and sclerostin, might play a role in the third phase of axSpA, due to their role in bone homeostasis [9,73].

In AS the primary site of inflammation is located at the enthesis or subchondral bone marrow, with bone marrow edema, lymphocytic infiltrates, increased osteoclast density, and increased microvessel density as typical findings in acute inflammation [35]. TNFα is a primary cytokine in axSpa [73], not only for its pro-inflammatory effect [71,74], but also as a stimulator of osteoclastogenesis [67] and for its effect on the Wnt signaling (enhancing the expression of Wnt inhibitors, Dkk-1 and sclerostin [10,21,23,63,75]).

Treatment with TNFα inhibitors has greatly improved the clinical outcome of axSpA patients, reducing the signs and symptoms improving the health-related quality of life. However, it is still unclear whether they slow the structural damage and radiographic progression or not [63,67,68]. In this context, several studies focused on the role of Dkk-1 in axSpA, in relation to bone formation and treatment with TNFα inhibitors [33,34,76–78].

Studies have shown that Dkk-1 is involved in the pathogenesis of structural damage and in molecular mechanisms of syndesmophyte formation and new bone formation in SpA, but most data are conflicting [33,34]. Various studies showed that Dkk-1 circulating levels respond in a different way to TNFα blockade in patients with AS and in patients with RA. It has been shown that serum levels of Dkk-1 were significantly higher in patients with AS compared to patients with RA and healthy controls and were further increased in AS patient receiving anti-TNFα treatment [34]; the authors hypothesized that the higher levels of Dkk-1 may reflect a counter-balancing mechanism to attenuate Wnt signaling, which is turned on after resolution of inflammation following the treatment. Conversely, another study revealed that circulating levels of Dkk-1 were lower in patient with AS compared to healthy controls and did not change after anti-TNFα treatment [33]. It could be therefore hypothesized that in AS the circulating levels of Dkk-1 are unable to suppress Wnt-mediated bone formation, and thus the inability of TNFα to induce Dkk-1 in SpA may be one explanation why anti-TNFα therapy might not block new bone formation [63].

Levels of functional Dkk-1 are increased in AS patients with no syndesmophyte growth compared to patients with syndesmophyte growth, suggesting that low levels of serum DKK-1 are necessary to develop syndesmophytes. Therefore, inhibition of Wnt signaling could protect from the development of syndesmophytes through an inhibition of bone formation processes [76]. In an animal model of transgenic mice overexpressing TNF, which develops bilateral sacroiliitis, Dkk-1 blockade promotes the expression of type X collagen and the formation of hypertrophic chondrocytes, promoting the ankylosis of sacroiliac joints, without affecting inflammatory processes [77]. Nevertheless, a prospective study on SpA patients showed that treatment with TNFα blockers induced a significant decrease of circulating levels of Dkk-1, which correlated with the reduction of MRI bone marrow edema of sacroiliac joint and spine, suggesting that the reduction of Dkk-1 by anti-TNFα agents could be related to the inhibitory effects of these drugs on new bone formation in SpA [78].

The hypothesis that the Wnt pathway plays an important role in the pathogenesis of AS is supported by studies that focused on another Wnt signaling inhibitor, sclerostin. It was observed that sclerostin expression was nearly absent in osteocytes in the periarticular bone of AS patients, suggesting a specific alteration of osteocyte function in this disease [79]. Moreover, serum sclerostin levels were found lower in AS patients than in healthy controls [79,80]. Interestingly, radiographic progression was associated with significantly lower sclerostin levels, suggesting that a low sclerostin level in AS could increase susceptibility for syndesmophyte formation [79]. Furthermore, although increasing after anti-TNFα treatment, sclerostin levels were still lower than in healthy subjects after 12 months of therapy [80]. Moreover, since AS patients with higher baseline sclerostin levels showed a rapid reduction of inflammatory parameters following the treatment and a concomitantly gradual increase in this Wnt inhibitor, the authors hypothesized that low sclerostin serum levels could be a possible marker of persistent inflammation in AS patients under anti-TNF therapy.

These data confirm the potential role of Wnt signaling in the pathogenesis of structural bone changes in SpA. Several physiopathogenetic hypothesis have been proposed to explain the conflicting available data on Wnt involvement in bone changes observed in SpA patients. There is a difference between the total levels of Dkk-1 and receptor-bound and thus functional Dkk-1, showing a minor capacity of Dkk-1 in axSpA patients to bind the LRP coreceptor [22,34,76]. Another explanation could be that Dkk-1 serum levels in patients with axSpA could be related to disease duration [72], as Dkk-1 serum levels were found higher in patients with early axSpA compared with patients with established disease.

This explanation may also clarify the link between TNFα and Dkk-1, and could contribute to explain the relationship between inflammation and osteoproliferation. The lower levels of Dkk-1 in patients with established axSpA may explain why TNFα blockers in this stage of the disease did not lead to a change in Dkk-1 levels in the study conducted by Kwon et al. [33] and is not able to avoid the development of syndesmophytes [68]. On the contrary, in patients with early axSpA, anti-TNFα treatment could lead to a decrease in Dkk-1 serum levels [75,78] and also to a better radiographic outcome, as shown by a longitudinal cohort study conducted by Haroon et al. [81]. In this study, radiographic progression was slower in AS patients who underwent TNFα-blocker therapy earlier in the course of disease than in patients in whom treatment was delayed. Indeed, in the early stages of the disease, inflammation and osteoproliferation may be coupled. Considering the three-stage pathological process of axSpA (inflammation, erosion, osteoproliferation), targeting the inflammation with TNFα inhibitors in patients at an early stage of the disease, may avoid the development of the next stages and thus the new bone formation processes. In other terms, if early inflammatory lesions resolve without undergoing chronic changes, the sequelae of new bone formation following inflammation might be halted [68,72]. Therefore, anti-TNFα treatment in the early stages of the disease may prevent development of osteogenesis, while at later stages osteogenesis and inflammation appear to be two separate therapeutic targets [63,68,75].

The clinical application of this theory is to establish the therapeutic time window to avoid the new bone formation process and improving the radiographic outcome through TNFα inhibitors, before the disconnection between inflammation and osteoproliferation [75]. This window might be characterized by higher serum levels of Dkk-1 in axSpA patients than in controls.

#### **4. Conclusions**

The Wnt signaling is involved in pathological aspects of rheumatic diseases, with several evidences supporting its role especially in bone complications, bone loss and new bone formation.

Higher serum levels of Wnt inhibitor Dkk-1 found in RA patients than in controls and its decrease after biological treatment (TNFα inhibitors, IL-1Ra, anti-IL-6 antibody), suggest the strong connection between the dysregulation of Wnt signaling leading to bone loss—both locally and systemically—in RA and the pro-inflammatory state typical of the disease. Interestingly, Dkk-1 concentration correlates with disease activity, elevated acute-phase reactants and more severe bone damage, suggesting its role as a possible biomarker of structural damage and new therapeutic target in RA. Moreover, in RA patients, higher levels of another Wnt inhibitors, sclerostin, correlates with disease activity scores and inflammation markers; in addition, in RA mice models anti-sclerostin antibodies arrested the bone damage progression and in combination with anti-TNF led to regression of cortical bone erosions.

In axSpA, the role of Wnt signaling has not yet been fully understood. There is evidence of a critical time window in the early stages of the disease when treatment with TNFα inhibitors might lead to two otherwise separate therapeutic targets: reducing signs and symptoms, and slowing the structural damage and radiographic progression. Further exploration of Dkk-1's role and the connection between bone formation and inflammatory response is needed in order to achieve a better knowing of the biologic background of the disease; this would subsequently lead to better management of current therapeutic options.

**Author Contributions:** Conceptualization, A.C. and F.P.C.; literature collection and preparation, D.C. and A.C.; writing—original draft preparation, D.C. and A.C.; writing—review and editing, D.C., A.C. and C.R.; supervision, F.P.C.

**Funding:** This work received no external funding.

**Conflicts of Interest:** The authors declare no conflict of interest.
