Systemic sclerosis (SSc) is a connective tissue disease at the interface between inflammation and autoimmunity progressively leading to fibrosis of the skin and of multiple internal organs (1).

Calcinosis is a clinical complications occurring in 20-40% of SSc patients (2–5) being characterized by calcium salt deposition in skin and in subcutaneous tissues causing pain, ulceration, infection, and deformities causing severe functional limitations (6). The underlying mechanisms are still elusive, even if chronic inflammation, trauma, and vascular damage have been hypothesized (7). In the last years studies performed in unrelated diseases (e.g., vascular calcification, chronic kidney disease, ectopic mineralization on genetic basis) demonstrated that pathological calcification is a complex process in which molecular (e.g., fetuin A, osteopontin and BMP-2, -4) and/or different cellular mediators (e.g., smooth muscle cells, circulating vascular progenitor cells and fibroblasts) may play a key role (8, 9).

However, the role of calcification regulatory proteins has been poorly investigated in the pathogenesis of SSc-related calcinosis (10).

Furthermore, the mineral phase of calcinosis has not yet been clearly defined due to contradictory results either describing calcinotic particles preferentially composed of hydroxyapatite (HAP), or of carbonate apatite (11, 12), that has a different stability and solubility compared to HAP.

Therefore, the present observational study aimed to compare clinical and laboratory parameters in SSc patients with (C group) or without (NC group) cutaneous calcinosis. Moreover, circulating markers and in vitro serum potential of calcification were evaluated in the two subgroups of patients. Finally, in a few SSc patients, the composition of calcium deposits was also investigated.

Calcinosis is a manifestation of SSc characterized by the deposition of calcium minerals in the skin and in subcutaneous tissues (4).

Some Authors report an association between calcinosis either with dcSSc (22) or lcSSc (10), whereas other studies (23), including this one, did not found differences in the prevalence of calcinosis between SSc cutaneous subtypes. To explain these discrepancies, it has been suggested that different results can depend on ethnicity (e.g., Caucasians vs African-American) and/or on geographic regions (4, 24). Moreover, as already reported (25), calcinosis occurs after several years (i.e., 7-10 years) from SSc diagnosis.

Regarding internal organ involvement, patients with calcinosis develop more frequently arterial hypertension, gastro-esophageal reflux disease, intestinal involvement, and pulmonary arterial hypertension (PAH). This observation confirms previous data showing that calcinosis is associated with a more severe disease course (26). Interestingly, PAH is a serious complication of SSc affecting 8-15% of patients (27) and has been related to inflammation and calcification (28). In agreement with data from the literature, our patients show a significant association between PAH and calcinosis. It is worth noting that 20% of our SSc patients in the C group, but none in NC group, were complicated by PAH, although it cannot be excluded a possible contribution of longer disease duration characterizing this patient subgroup. Several studies have underlined some clinical characteristics and serological markers associated with the development of PAH in SSc patients, such as the presence of calcinosis, or telangiectasias, or lcSSc, and/or serum ACA (26, 29–32). In our study, both ACA seropositivity and telangiectasias are significantly more frequent in patients with calcinosis compared to those without.

In several pathological contexts, ectopic calcification has been shown to be associated with a systemic imbalance between pro- and anti-calcifying molecules (33). Since, scattered and sometimes contradictory data are present for SSc patients, we have investigated several circulating calcification-related molecules in both C and NC groups of patients.

Fetuin A is a systemic inhibitor of calcification and is known to protect tissues from inflammation-related damage. No difference was observed between C and NC patients, in contrast with data from Belloli and coworkers who found fetuin A to be significantly lower in serum of SSc patients with calcinosis compared to those without calcinosis (10). This discrepancy could be attributed to the smaller randomly selected sample size of patients analyzed in the previous study (10). Moreover, we cannot exclude that the levels of fetuin A might differ between the two groups in the early phases of SSc. For instance, in chronic kidney disease (CKD) patients, it has been suggested that fetuin A may have a protective role being upregulated in the early stages of the disease, whereas severe or prolonged exposure to a pro-inflammatory and/or pro-calcific environment may eventually lead to low levels due to decreased production and/or increased consumption (34). As mentioned above, calcinosis usually occurs after a prolonged disease course (i.e., 7-10 years after SSc diagnosis).

Indeed, several studies have highlighted an association between fetuin A levels and the severity of different pathologic conditions (e.g., chronic obstructive pulmonary disease, coronary calcification; CKD; calcific aortic valve diseases, aortic stiffness) (35–37). Therefore, the inverse correlation found between fetuin A levels and the global score in the C group of patients may suggest that lower fetuin A is associated with an increased number of clinical manifestations.

Osteopontin (OPN), a highly negatively charged secreted protein, regulates several processes including apoptosis, bone metabolism, inflammatory response, fibrosis. It has been also found in sites of ectopic calcification (e.g., renal stones, aortic stenosis) suggesting that it can interact with crystal surfaces, and regulate the mineral process (38). No changes were observed in the present study, although high levels of serum/plasma OPN were previously found in SSc patients or in SSc patients with interstitial lung disease, however, in these studies SSc patients were compared with healthy controls without considering the presence of calcinosis (39–41).

Osteoprotegerin (OPG) is a soluble decoy receptor for the receptor activator of nuclear factor-κB ligand (RANKL), which stimulates osteoclastic bone resorption (42). The literature on OPG levels in SSc is quite heterogeneous. For instance, one study reported higher OPG levels in lcSSc patients compared to healthy subjects (43), while Tayalan Ali and collaborators (44) found similar OPG levels in both SSc patients and healthy subjects. In the present study, OPG levels were similar in C and NC patients. To our knowledge, only one study showed higher OPG levels comparing SSc patients with and without calcinosis (45). Differences between our results and previous findings could be due to the technical procedures utilized, as suggested by the detection range, and/or by the fact that in the previous study males and females were not separated, despite the known sex-dependent variation in OPG levels (higher in females than in males) (45).

SPARC/osteonectin is a protein with collagen type I-binding domain and hydroxyapatite-binding sites (46, 47) that may be involved in both the initial and the progressive stages of the calcification process (48, 49). This molecule was previously found to be significantly increased in the dermis of lcSSc patients with calcinosis compared to those without calcinosis (50) and in the plasma/fibroblasts from SSc patients compared to healthy subjects (51, 52). The discrepancies between our findings showing no differences in the two groups of patients and those reported in the literature may be related to: i) different types of bio-samples analyzed (i.e., dermis/fibroblasts/plasma vs serum) and/or on ii) differences in the studied population, i.e., we firstly stratified SSc patients based on the presence or the absence of calcinosis, whereas in other studies, patients with or without calcinosis were selected within one SSc cutaneous subtype (i.e., lcSSc).

Bone morphogenic proteins –2 and -4 (BMP-2 and BMP-4) play a crucial role in vascular calcification by promoting osteogenic activation of vascular cells (53) and have also been implicated in cutaneous calcification (54) and inflammation (55).

Currently, there are no data in the literature on BMP-2 and BMP-4 in SSc patients. Our results indicate no changes for BMP-2, but higher BMP-4 levels in the C group. BMP-4 has been extensively investigated as a promoter of calcification either in the vascular system (53) and in the skin (54). Besides its role as pro-osteogenic signaling inducer, BMP-4 is also related to inflammation and vascular damage, inducing endothelial dysfunction through oxidative stress (56). Since overexpression of BMP-4 in endothelial cells enhances vascular remodeling in pulmonary hypertension (57) it cannot be excluded that higher serum levels of BMP-4 can contribute to the vascular abnormalities frequently observed in the C group (e.g., telangiectasia, PAH and arterial hypertension). This may, in turn, accelerate the need of remodeling and renewal of endothelial cells, potentially leading to a shift towards a pro-osteogenic phenotype (58) and underlines the potential role of BMP-4 in SSc calcinosis.

Under certain conditions, Ca, P and specific serum proteins (e.g., fetuin A and albumin) can aggregate to form amorphous, soluble Ca–P particles (CPP1). These particles can subsequently transform into larger and crystalline calciprotein particles (CPP2) and the half-maximal time of transformation from CPP1 to CPP2 is known as T50. A lower T50 indicates a faster conversion from CPP1 to CPP2, and this has been associated, for example, with cardiovascular mortality, and with progressive aortic stiffening (59, 60). In our study, we did not observe significant differences of T50 and CPP2 between C and NC groups. However, we revealed a correlation between T50 and the global score, as well as between CCP2 and circulating inhibitors of calcification.

Since its introduction in 2012 (16), T50 test has been evaluated in both healthy subjects and in various patient cohorts (e.g., diabetic, hemodialysis, Pseudoxanthoma elasticum patients) (61–65) demonstrating that lower T50 values could predict all-cause and cardiovascular mortality in hemodialyzed patients, in those with CKD or with diabetes. Furthermore, some studies have proposed the use of T50 value as a surrogate marker of arterial calcification or as biomarker of disease severity (65). Interestingly, a negative correlation between T50 and global score was found in the C group. Although further studies with a larger number of SSc patients are necessary to confirm this result, T50 could be an important biomarker in SSc, potentially useful for risk stratification and management of the SSc patients with calcinosis.

Interestingly, in the C group we found that fetuin A and albumin negatively correlated with CCP2, consistently with the inhibitory role of fetuin A (66, 67) and of albumin (68) (69, 70) in preventing the growth and the aggregation of CPP2 thereby hindering their precipitation.

In addition, in the C group, larger CPP2 radius correlated with higher serum OPG. These results are akin to those obtained in hemodialysis patients and in pre-dialysis CKD patients (69, 71). However, further studies are needed to clarify the role of OPG in influencing the size of CPP2.

Ectopic calcification is generally associated with deposition and progressive accumulation of insoluble HAP, however, several studies demonstrated that different type of crystals with a different solubility are present in mineralization sites (72). Present results highlight that calcinosis, even in the same SSc patient, is characterized by a mixture of carbonate and phosphate minerals including calcite and HAP, respectively. It is important to note that analyses were performed on calcinosis samples removed from patients and not on isolated crystals. Since X-ray analysis is not capable to detect crystals that constitute less than 5% of the sample’s weight (73), we cannot exclude the additional presence of other minerals further increasing the complexity of calcinosis samples. Previous studies described SSc calcification composed solely of HAP or of B type carbonated apatite (11, 12, 74, 75). These discrepancies can be explained by the fact that calcification is a complex, non-linear process, and that mineral characteristics (i.e., structure, composition and morphology) depend on several factors such as ion concentration, temperature, pH value and location (76–79). Some in vitro and in vivo studies report that a range of intermediate phosphates (e.g., carbonated apatite, octacalcium phosphate) are formed before transforming into HAP (80–82), even though Thomson and collaborators suggested that hydroxyapatite is formed directly on cholesterol-containing droplets suggesting the absence of intermediate phases (83). The demonstration that different types of mineral deposits are present in SSc calcinosis may lead to the possibility of selecting the most appropriate treatment to improve the solubility of mineral components and to decrease extraosseous calcification. The differences observed among various samples could be due to changes in the composition/organization of the organic matrix. It is well known that aging, oxygen availability, mechanical stress, for example, can modify the characteristics of the extracellular matrix favoring the development/progression of pathological mineralization (84–86). Therefore, a thorough study of both the mineral and the organic components will provide a better understanding of the mechanisms leading to the development of calcinosis.

While the present study indicates some molecules that can correlate with calcinosis in SSc patients, it comes with some limitations such as: i) the relatively small number of patients in both subgroups; ii) the lack of data in the early phases of the disease; iii) the absence of male patients; iv) the difficulty to recruit patients in a fairly homogeneous phase of the disease.

Although these potential biases should be kept in mind to draw general conclusions, in summary, present data allow: i) to find statistically significant differences in the prevalence of clinical manifestations (i.e., telangiectasias, pulmonary arterial hypertension, artery hypertension, gastro-oesophageal reflux disease, intestinal involvement) and laboratory tests (i.e., ACA and Scl70 autoantibodies) in SSc patient with calcinosis compared to those without calcinosis; ii) to highlight higher serum levels of BMP-4 in SSc patients with calcinosis; iii) to identify fetuin A as potential circulating prognostic marker in SSc patients with calcinosis; iv) to underline a higher global score in C patient group; v) to show negative correlation between T50 and global score in patients with calcinosis; vi) to demonstrate the heterogenous composition of the mineral component in calcinosis samples.

The present data may pave the way for future studies on larger cohorts of patients. Moreover, measuring pro- and anti-calcifying molecules at the time of SSc diagnosis and during the clinical long-term follow-up in patients who will do or do not develop calcinosis might clarify their pathogenetic role.