Introduction

The recent characterization of the canonical WNT pathway in the regulation of bone modeling and remodeling provided important insights in our understanding of the pathophysiology of bone involvement in chronic arthritis [1,2,3].

The WNT/β-catenin signaling (canonical WNT pathway) regulates osteoblast proliferation, maturation and differentiation and, along with bone morphogenetic proteins, acts as the master regulator of osteogenesis. The WNT/β-catenin signaling plays a key role in bone tissues by inducing the differentiation of stem cells into mature osteoblasts rather than into chondrocytes or adipocytes. Its regulation is driven by the production of two WNT signaling antagonists, sclerostin and Dickkopf-related protein 1 (Dkk-1) [4]. On one hand, the WNT pathway boosts bone formation by fostering osteoblast activity, while on the other it inhibits osteoclastic bone resorption by regulating the production of RANKL and thus by interfering with the osteoprotegerin/RANKL ratio [5, 6].

In a previous study we showed that, in patients with rheumatoid arthritis (RA), Dkk-1 is significantly increased and associated with a lower BMD and with the presence of the typical erosions [7]. The positive correlation found between circulating Dkk-1 and serum CTX-I (a marker of bone resorption) suggested that the over-production of Dkk-1 (and the resulting inhibition of the WNT pathway) may contribute both to the locally increased bone resorption and to the impaired bone repair, typical features of the more erosive forms of RA [7]. This phenomenon may also justify the development of RA-associated osteoporosis [8].

Ankylosing spondylitis (AS) is a chronic rheumatic disease characterized by the inflammation and extensive remodeling of spine and joints. In contrast to RA, erosive changes are limited in AS, while extensive bone neoformation leads to the development of spinal syndesmophytes and extra-articular enthesophytes and therefore to joint or spine ankyloses [9]. Previous studies already reported that both circulating sclerostin and Dkk-1 levels are significantly lower in AS patients when compared to matched controls [10,11,12]. The observation of lower circulating levels of WNT inhibitors suggests that WNT overexpression could play a relevant role in the focally exuberant bone formation in AS.

Currently, the WNT pathway is thought to be heavily involved in the pathogenesis of bone damage during RA and SA with different and opposite profiles, the inhibition of WNT pathway in RA and its overexpression in SA. The influence of these inflammatory diseases over the WNT pathway seems to be driven by the dysregulation of Dkk-1 and sclerostin. In previous studies, we already documented PTH to be a major determinant of Dkk-1 serum levels not only in patients with arthritis but also in subjects affected by bone metabolic diseases such as primary hyperparathyroidism [7, 13, 14]. Psoriatic arthritis (PsA) is a chronic, systemic inflammatory disease that affects peripheral joints, connective tissues, and the axial skeleton and it is associated with psoriasis of the skin and nails [15, 16]. In PsA, we can find bone erosive damage often associated with an exuberant bone formation especially in enthesial sites [17].

At this point, we asked ourselves in which manner the WNT pathway is involved in these different conditions. Is PsA more similar to RA or to AS? We performed this study in order to investigate the WNT-pathway regulators along with BTM and parathyroid hormone (PTH) in three different groups, one group of patients affected by PsA, one group of patients affected by RA, and a healthy controls (HC) group.

Materials and methods

Study design

This is a cross-sectional study conducted in the Rheumatology Unit of the University of Verona on patients with a recent diagnosis of PsA or RA. The study sample included 33 female patients with PsA classified with the CASPAR criteria [18], 28 female patients with RA classified with the ACR/EULAR 2010 criteria [19], and 35 female healthy controls. The patients were enrolled consecutively in our PsA outpatient clinic and in our RA outpatient clinic between September 2015 and March 2016. The criteria of inclusion were the female sex, stable treatment with parenteral methotrexate in the last 6 months (ranging from 10 to 15 weekly), time since the diagnosis ≤3 years, for patients affected by RA, a 28-joints disease activity score (DAS28) ≤3.2 and for patients affected by PsA presented a disease activity index for psoriatic arthritis (DAPSA) ≤14. The healthy controls were enrolled from the hospital personnel.

The exclusion criteria for all the groups were further systemic inflammatory diseases, active infections, neoplasms, metabolic bone diseases, pregnancy, current use of biologic therapy, and current use of drugs known to affect bone metabolism and use of corticosteroids in the last 6 months. All subjects received vitamin D3 supplementation during the course of the study (7500 IU/week).

Biochemical assessment

Blood samples were collected and stored at −50 °C and finally assayed for intact N-propeptide of type I collagen (PINP), C-terminal telopeptide of type I collagen (CTX-I), sclerostin, and Dkk-1. All the samples were processed in the laboratory of the Rheumatology Unit of the University of Verona. Bone turnover markers (PINP and CTX-I), 25OH-vitamin D, and intact-PTH were measured by the IDS-ISYS Multi-Discipline automated analyzer (Immunodiagnostic System, Boldon, UK) based on chemiluminescence technology. The coefficients of variation (CV) intra-assay measured in our laboratory were 4% for intact-PINP (inter-assay CV 6%), 3% for CTX-I (inter-assay CV 7%), 6% for 25OHVITD (inter-assay CV 9%), and 2.7% for intact-PTH (inter-assay CV 5.5%).

Serum Dkk-1 and sclerostin were measured by ELISA (Biomedica Medizinprodukte, Vienna, Austria) with a sensitivity of 1.7 and 3.2 pmol/l and intra-assay CV of 7 and 5% (inter-assay CV 8.2 and 6.9%), respectively.

Statistical analysis

All statistical analyses were performed per protocol by SPSS software, version 22 (SPSS, Inc., Chicago, IL, USA).

We compared mean levels of all parameters for the three groups with three-way ANOVA. Differences between groups were confirmed by t test for independent samples.

Correlations were calculated using linear regression.

Two-sided p values of 0.05 or less were considered significant. Data are presented as mean ± SD.

Results

The three groups were comparable for age (58.8 ± 8.8, 56.2 ± 8, 60.8 ± 6.3 years old, respectively for PsA, RA, and HC). There was a statistically significant difference only concerning the weight, with a higher mean weight for the PsA group (74 ± 17, 60.2 ± 9.2, 61.2 ± 9.0 Kg, respectively for PsA, RA, and HC).

The levels of 25OH-vitamin D were comparable in the three groups (28.5 ± 15.6, 30.5 ± 8.4, 29.0 ± 11.0 ng/ml, respectively for PsA, RA, and HC).

The values of bone turnover markers (CTX-I, PINP), Dkk-1, and sclerostin of PsA and RA patients and in healthy controls are shown in Table 1. PINP, sclerostin, and 25OH-vitamin D did not show any significant difference in the two groups of patients (Table 1).

Table 1 Values of bone turnover markers (CTX-I, PINP), Dkk-1, and sclerostin of PsA, RA patients, and control group (mean ± SD)
Full size table

Dkk-1 was found to be significantly lower in PsA patients when compared both with the RA and the control groups (ANOVA p < 0.01 and p < 0.05 both vs RA and vs HC) (Table 1, Fig. 1). A similar trend was documented also for PTH (ANOVA p < 0.001). A statistically significant difference was observed when we compared the PsA vs the RA group and also when we compared the PsA vs the HC and the RA vs the HC groups (Table 1, Fig. 1).

Fig. 1
figure 1

Dkk-1 and PTH levels in the three groups (mean ± SD)

Full size image

We found a positive correlation between the levels of PTH and Dkk-1 when we examined the study population altogether (R 2 = 0.227, p < 0.001). However, when the different groups were tested singularly, this correlation remained significant only in the RA group (R 2 = 0.224, p < 0.01).

Finally, we found a significantly lower CTX-I in the PsA group vs both the RA and control groups.

Discussion

The aim of the study was to document a possible difference of Dkk-1, sclerostin, and BTM serum levels in PsA patients when compared to HC or the ones affected by RA.

In order to homogenize as much as possible the samples of RA and PsA patients from the disease activity perspective, we chose to select them on the basis of the current cutoff for low disease activity defined by the most used clinical indexes DAS28 and DAPSA, respectively [20, 21].

What emerged at the analysis was that the levels of Dkk-1 were significantly lower for PsA subjects when compared with HC or the RA group.

It should be noted that our remark is not consistent with what found in the study of Dalbeth et al., in which Dkk-1 in the PsA group resulted higher than HC [22]. These conflicting results do not present an easy explanation, even though in that particular study the healthy control group was very limited in size (only 12 patients) [22]. In addition, neither the PTH levels nor the vitamin D status of those patients was reported. Since psoriatic patients are significantly at risk of hypovitaminosis D [23], the influence on the concentration of Dkk-1, driven by a possible secondary hyperparathyroidism, may not be excluded.

It is very interesting to speculate why Dkk-1 levels might be lower in patients with PsA, as it has been demonstrated that Dkk-1 blockade leads to an enhanced bone formation and ankyloses of peripheral joints [24], and also that WNT agonists (such as R-spondin) induce bone spur formation in inflamed joints [25]. Furthermore, Heiland et al., in a previous study [17], postulated that high levels of functional Dkk-1 could have a protective action towards the syndesmophytes formation in AS. As we know, PsA shares with AS (even if with some differences) the tendency to develop bone “reparative” proliferative lesions. This important aspect differentiates PsA from RA, in which the bone involvement is typically limited to erosive phenomena. Therefore it is not surprising to find a suppressed Dkk-1 in PsA, similarly to what has already been documented in AS [10, 26].

Consistently with our previous experience [7], Dkk-1 and PTH resulted significantly higher in the RA group. Indeed, in that study we observed that these biomarkers were associated with an increased risk of bone erosions in RA, and an increased risk of osteoporosis and vertebral fractures in both RA and AS [7, 10, 27].

The Dkk-1 levels in PsA resulted about halved when compared to those seen in RA (Fig. 1), but this remark was apparently not accompanied by a concomitant difference on bone formation (PINP is similar in the three groups of patients, Table 1). Possible explanations for this observation could be the limited sensibility of the marker in revealing such difference or even the fact that the effects of serum Dkk-1 could represent the expression of a localized and not of a systemic phenomenon. On the contrary, CTX-I seems to be lower in the PsA group, as we could expect as a consequence of a low Dkk-1 [5,6,7].

To our knowledge, this is the first study that evaluated the levels of PTH and compared them between PsA, RA, and HC. It is important to underline that the levels of 25OH-vitamin D were comparable in the three groups, thus the levels of this hormone might explain the different pathophysiology of RA and PsA.

The significant correlation observed between PTH and Dkk-1 confirms what we already found in our previous study in patients with RA [7] and may contribute to suggest the possible relevant role of PTH in the pathogenesis of both systemic and local bone loss in RA.

On the contrary, lower serum levels of PTH, Dkk-1, and bone resorption marker (CTX-I) were observed in our study in PsA patients, in whom a typical concomitant excess in bone formation is common. Interestingly, in PsA, the current evidences about the presence of systemic bone loss are not consistent [28]. Very recently, Odgie A. et al. [29] showed that in PsA, the risk for all the fractures appears to be slightly increased, while, when taking into account only vertebral fractures, the increase is not significant [29].

Our study has several limitations. First of all, the sample size. Overall, the total number of cases was quite limited. Second, no stratification based on the disease severity, the radiographic damage, or the disease duration was performed. It is conceivable that differences could be present, in example, between patients with mild vs severe radiographic involvement or patients with a prevalently erosive vs proliferative lesion. In addition, it should be noted that there are several other conditions, besides arthritis, able to interfere with Dkk-1 levels, as its expression is not limited to the bone tissue. Dkk-1 has been extensively studied in oncology, as elevated levels of Dkk-1 in bone marrow plasma and peripheral blood is associated with the presence of osteolytic bone lesions in patients with multiple myeloma [30]. Furthermore, the upregulation of Dkk-1 expression takes place also in prostate cancer and hepatocellular carcinoma [31, 32]. Overexpression of the protein is detected in tissue as well as serum from hepatocellular carcinoma patients [33].

Further studies with greater numbers are warranted, ideally with an adequate evaluation of all these features.

In conclusion, our study showed that Dkk-1 and PTH serum levels may differ in PsA patients, RA patients, and HC. These results might also contribute to explain the different kind of bone involvement in RA and PsA. However, further studies are needed to determine whether these preliminary results really have an actual clinical relevance.