Introduction

Systemic sclerosis (SSc) is a systemic inflammatory disorder affecting the skin and other organs [1–4]. The condition is characterized by three histopathologic features: (i) both structural and functional vascular lesions, associated with progressive endothelial damage, reduction in the number of capillaries, thickening of arterial walls, and obliterative vasculopathy; (ii) perivascular and tissue infiltration of mononuclear inflammatory cells at early stages of SSc [5–8] (histological analyses of skin biopsy specimens have shown prominent CD4+ cellular infiltration with overexpression of cellular adhesion molecules in both vessels and interstitium); and (iii) increased synthesis and excessive deposition of extracellular matrix, resulting in fibrotic destruction of internal organs during the course of SSc [5–8].

The pathogenesis of SSc still remains unclear, although it is increasingly thought to result from interactions between environmental factors and epigenetic features leading to the onset and progression of SSc in genetically susceptible patients. To date, it is established that epigenetic modifications are characterized by both stable and heritable changes in gene expression without modifications to the original DNA sequence [5, 9–11]; thus, three main epigenetic features are thought to determine the epigenome in SSc, i.e.,

  • DNA methylation, leading to dysregulation of many cells, including principally (1) CD4+ T cells which seem to be hypomethlylated. CD4+ T cell hypomethylation has been shown to be associated with reduced expression of methylation-regulating genes: DNA methyltransferase-1 (DNMT1), methyl-CpG-binding domain (MBD) 3, and MBD4 [6, 9, 12]. Furthermore, CD40L gene promoter has been shown to be hypomethylated and transcriptionally active in CD4+ T cells [6, 9, 13]. In female SSc patients, the presence of demethylated CD40L regulatory elements on the inactive X chromosome leads to CD40L overexpression; these latter findings may, in part, explain the predominance of SSc in females [6, 9, 13]; (2) endothelial cells from SSc patients have been found to show CpG hypermethylation of the bone morphogenetic protein receptor factor type 2 (BMPR2), which may contribute to higher vulnerability of endothelial cells to apoptosis and oxidation damage [14]. Additionally, CD70 is part of the tumor necrosis factor (TNF) α superfamily and is restricted to T and B lymphocytes and mature dendritic cells [6, 9, 10]; in SSc patients, the promoter of CD70 is hypomethylated, leading to the overexpression of CD70 by CD4+ T cells; and (3) fibroblast global DNA hypermethylation has been found in dermal fibroblasts from SSc patients; histological analyses of skin biopsy specimens of SSc patients showed higher levels of methylation regulatory genes DNMT1, MBD1, and methyl-CpG-binding protein (MECP) 2 than controls [6].

  • Histone modifications, contributing to impairment in (i) transcription of genes and (ii) modulation of gene expression [6, 9]. Transforming growth factor β (TGF-β)-induced transcription of fibrotic genes is markedly dependent on histone acetylation catalyzed by acetyltransferases, including p300 [15, 16]. In experimental models of skin/lung fibrosis and in SSc skin biopsy samples and explanted fibroblasts, analyses have shown elevated lesional tissue p300 levels, promoting histone H4 hyperacetylation and collagen transcription [16]. Furthermore, the histone deacetylase sirtuin 1, which exerts antifibrotic effects by blocking Smad-dependent transcription, has been found to be reduced in human and mouse fibroblasts in culture as well as murine model of fibrosis [16].

  • MicroRNA profile, resulting in a pro-fibrotic process [5, 6, 9, 10, 17]. Previous investigators have shown that SSc patients, compared with healthy subjects, had (1) ex vivo lower levels of miR-29 (which has a master antifibrotic action on collagen I and III and extracellular matrix components via inhibition of expression of COL1-A1, COL1-A2, and COL1-A), which suggests an association with TGF-β and platelet-derived growth factor β (PDGFβ) pathways [6, 10, 18, 19]; (2) higher values of miR-92a in dermal fibroblasts and in sera, resulting in reduced metalloprotease 1 [6]; (3) decreased levels of miR-129-5p in dermal fibroblasts, leading to increased expression of collagen I and connective tissue growth factor (CTGF) [20]; and (4) higher serum levels of miR-143-3p, which were predictive of SSc severity [21].

To date, there is a growing evidence to underscore the effect of environmental factors on alterations and/or modulation of epigenetic determinants (DNA methylation, histone modifications, and microRNA profiles), resulting in SSc onset and progression [5, 6, 9, 21]. Therefore, many environmental factors have been incriminated to play a role in SSc pathogenesis, which can be divided into (1) chemical agents, including notably occupational chemical agents; (2) physical agents, including ionizing radiation, ultraviolet radiation, and electric and magnetic fields; and (3) biological agents, such as infections, foods, and dietary contaminants. The aim of the present article is thus to review the association between these environmental factors and both SSc onset and progression.

Chemical agents

In our review, the association between SSc risk and the following chemical agents has been evaluated: silica, asbestos, organic solvents, other industrial chemicals, pesticides and persistent organic pollutants, air pollutants, smoking and drug abuse, as well as personal care.

Silica

Freshly fractured particulate silica (crystalline silica, quartz) is released in gold mining, sandblasting, rock drilling, sand factory work, granite cutting, construction work, brick laying, sail tilling, cement work, and scouring powder factory work (Table 1) [2, 4].

Table 1 Occupations associated with exposure to crystalline silica
Full size table

Bramwell [22] was the first to describe SSc in Scottish stonemasons. In 1957, Erasmus [23] has reported a higher incidence of SSc in gold miners exposed to crystalline silica, compared with the general population (2/1000 vs. 0.35/1000). Previous case–control series have found a correlation between crystalline silica and SSc [23–28]. Haustein et al. [29] have mentioned that SSc risk, for crystalline silica-exposed subjects, was 25 to 50 times that of unexposed subjects. In a case–control study (80 patients, 160 controls), the authors have found an association between silica exposure and SSc (OR 5.57, 95 % CI 1.60 to 18.37); the risk was elevated in patients with high level (>1) of crystalline silica exposure (OR 3.74, 95 % CI 1.06 to 13.18) [25]. Moreover, a 2012 meta-analysis of 16 series (total 1101 cases in 3 cohorts and 9 case–control series) demonstrated that SSc patients were more often exposed to crystalline silica (OR 3.20, 95 % CI 1.89 to 5.43); the risk appeared to be higher in men (OR 3.02, 95 % CI 1.24 to 7.35) than in women (OR 2.03, 95 % CI 0.74 to 1.44) [30]. Finally, in a 2014 case–control study, a marked association between SSc and crystalline silica exposure (OR 5.32, 95 % CI 2.25–13.09) has been reported; although, there was a significant association between SSc and crystalline silica exposure only in male patients (OR 8.30, 95 % CI 2.58–29.60) [2]. Additionally, the risk was higher in male SSc patients with high level (cumulative exposure score >1) of crystalline silica exposure (OR 9.63, 95 % CI 2.49 to 43.01); the risk was further stronger in patients with very high level (cumulative exposure score >5) of crystalline silica exposure in overall SSc patients (OR 9.68, 95 % CI 2.14–59.43) [2]. In this study, most of SSc patients exposed to silica were foundry workers, building workers, and roofers [2]. Furthermore, a case report of limited cutaneous SSc has also been described in a winegrower, who commonly filtrated wine using calcined diatomaceous earth (i.e., diatoms are microscopic plants extracting silica from lakes/seas), containing 90 % pure crystalline silica [31].

Interestingly, occupational exposure to crystalline silica has also been found to be a predictive parameter of SSc severity [4, 25]. In a previous series, the authors reported that SSc patients, who exhibited exposure to crystalline silica, compared to non-exposed SSc patients, developed more frequently (i) diffuse cutaneous SSc (P = 0.02), higher value of modified Rodnan score (P = 0.001), digital ulcers (P = 0.05), lower median value of left ventricular ejection fraction (P = 0.006), and (ii) interstitial lung disease (ILD) (P = 0.0004) [4]. ILD was also more severe in the group of silica-exposed patients with SSc who exhibited lower median values of forced vital capacity (FVC) (P = 0.004), vital capacity (VC) (P = 0.001), and diffusing capacity of the lungs for carbon monoxide (DLCO) (P = 0.01) on pulmonary function tests and more commonly honeycombing (P = 0.045) pattern on HRCT scan [4].

The pathogenic mechanisms of crystalline silica remain unclear in the development of SSc. Because silica is a strong T cell adjuvant, dysregulated immune response triggered by crystalline silica in genetically predisposed subjects could lead to tissue damage and systemic inflammatory response [32]. Experimental series have indicated that silica, administered intratracheally or intravenously, may interact with alveolar macrophages leading to immune dysfunction and notably to (1) activation of T and B lymphocytes, resulting in cytokine synthesis (TNF, interleukin-1b, TGF-β) and autoantibody production; (2) autoimmunity related to apoptosis with autoantigen alterations [33] (when neutrophils undergo apoptosis, primary granule constituents translocate to the cell surface, leading to both generation of self antigens and tolerance breakdown [34]); and (3) activation of fibroblast proliferation, with increased synthesis of collagen and other components of extracellular matrix (fibronectin, glycosaminoglycans) [2, 30]. In addition, chronic exposure to silica has been reported to activate both responder T cells and regulatory T cells [32, 35, 36]. These activated responder T cells enter the peripheral CD4+25+ fraction and activated regulatory T cells, which may result in reduced inhibitory function of regulatory T cells [32, 35, 36]. Furthermore, persistently activated responder T cells may also express Fas-mediated apoptosis inhibitory molecules such as soluble Fas and survive longer, which may contribute to progress of SSc [32, 35, 36].

Asbestos

Asbestos is composed of mineral fibers, i.e., (1) serpentine (chrysotile) and (2) amphibole (actinolite, amosite, anthophyllite, crocidolite, tremolite) [37]. Asbestos is used due to heat resistance, occurring in mining and construction [37].

Previous series have shown that subjects exposed to amphibole more often developed high levels of antinuclear antibodies than non-exposed controls [37, 38]. Noonan et al. [39] have, in fact, found that workers in an asbestos-contaminated vermiculite mining company were at twice the risk of either rheumatoid arthritis, systemic lupus erythematosus, or SSc compared with the unexposed population. Nevertheless, in this group of exposed patients, because asbestos was concurrent with crystalline silica exposure (contaminated vermiculite), it is difficult to assess the role of each material separately. Moreover, based on examination of US death certificates of SSc patients (n = 5642), Gold et al. [40] have also observed an association between occupational asbestos exposure and SSc (OR 1.2, 95 % CI 1.1–1.3).

In experimental models, which are C57BL/6 mice and rats, intratracheal amphibole asbestos has been shown to induce (1) synthesis of antinuclear antibodies [41–43] and (2) modifications of serum cytokine profile, such as marked increase in the concentration of serum IL-17. This Th17 response was characterized by high levels of IL-17 triggered or maintained by other cytokines (e.g., IL-6, IL-23, TGF-β) [44]. Interestingly, Th17 responses have been implicated in the pathogenesis of SSc [45, 46].

Taken together, the limited number of epidemiological series exploring a causal association between asbestos exposure and SSc makes it difficult to draw definite conclusions. Further investigations are therefore warranted to assess the accurate relationship between SSc and occupational exposure to asbestos.

Organic solvents

Organic solvents are compounds that can be dichotomized into aliphatic-chain compounds (n-hexane), aromatic compounds (benzene, xylene), and chlorinated compounds. Common uses for organic solvents are for dry cleaning (tetrachloroethylene), paint thinners (toluene), nail polish removers and glue solvents (acetone, methyl acetate, ethyl acetate), and spot removers (hexane) [2]. To date, the applications of organic solvents are increasingly diversified in both developed and developing countries (Table 2).

Table 2 Occupations associated with exposure to organic solvents
Full size table

The association between SSc and organic solvent exposure was first described by Reinl [47]. Previous authors have also evaluated the association between organic solvent exposure and increased risk of SSc [25, 48–51]. In a case–control study (178 patients, 200 controls), Nietert et al. [52] have observed a correlation between solvent exposure and SSc (OR 2, 95 % CI 1.51–2.50). In another case–control series (21 SSc, 42 controls), the association between exposure to organic solvents and SSc was observed only in male patients with SSc [48]. A 2012 meta-analysis included 11 case–control series (1291 cases and 3435 controls), the meta-analytic risk estimate for occupational exposure to organic solvents (ever vs. never) being 2.4 (95 % CI 1.7–3.4) in SSc; the risk of SSc associated with solvent exposure was higher in men (OR 3.0, 95 % CI 1.9–4.6) than in women (OR 1.8, 95 % CI 1.5–2.1) [30]. Furthermore, two series have shown elevated relative risk estimate for high cumulative exposure scores [25, 52]. To date, only few studies have evaluated the association between specific organic solvents and SSc [25, 48, 49, 51, 52]. Some series have reported elevated risks associated with such specific organic solvents—which are trichloroethylene, chlorinated solvents, aromatic solvents, toluene, xylene, and white spirit—with various results [25, 49, 52]. Nietert et al. [52] have observed a correlation between trichloroethylene exposure and SSc (OR 4.4, 95 % CI 1.3–15); by contrast, Garabrant et al. [49] failed to find a significant association between trichloroethylene exposure and SSc (OR 2.29, 95 % CI 0.95–5.71). In a meta-analysis of three series, the risk estimates for occupational trichloroethylene exposure were 2.5 (95 % CI 1.1–5.4) for men and 1.2 (95 % CI 0.6–2.6) for women [53]. In another case–control series (80 patients, 160 controls), a significant association was observed between SSc and aromatic solvents (OR 2.67, 95 % CI 1.06–6.75), chlorinated solvents (OR 2.61, 95 % CI 1.20–5.66), trichloroethylene (OR 2.39, 95 % CI 1.04–5.22), toluene (OR 3.44, 95 % CI 1.09–10.90), white spirit (OR 3.46, 95 % CI 1.48–8.11), ketones (OR 8.78, 95 % CI 1.82–42.38), and any type of solvent (OR 2.66, 95 % CI 1.35–5.23) [25].

Additionally, a larger case–control study also showed the marked association between SSc and organic solvents for white spirit (OR 7.69, 95 % CI 4.11–14.7), aromatic solvents (OR 8.17, 95 % CI 2.29–36.5), chlorinated solvents (OR 2.46, 95 % CI 1.12–5.32), trichloroethylene (OR 2.26, 95 % CI 0.95–5.26), and ketones (OR 3.37, 95 % CI 1.51–7.53) [2]. However, in this series, exposure to other specific organic solvents was different in both male and female patients [2]. In male patients with SSc, the authors found increased ORs for white spirit (OR 10.29, 95 % CI 3.10–38.90), chlorinated solvents (OR 4.01, 95 % CI 1.2–13.37), and trichloroethylene (OR 2.77, 95 % CI 0.80–9.35); there was also an elevated risk associated with a high final cumulative score (>1) in male patients with SSc compared with male controls for white spirit (OR 5.59, 95 % CI 1.54–21.44), chlorinated solvents (OR 14.09, 95 % CI 2.38–152.20), trichloroethylene (OR 4.49, 95 % CI 1.11–18.89), and ketones (OR 5.92, 95 % CI 1.03–41.95) [2]. Furthermore, in female patients, the authors found increased ORs for white spirit (OR 9.77, 95 % CI 4.06–25.51), aromatic solvents (OR 26.4, 95 % CI 3.45–1183), and ketones (OR 5.52, 95 % CI 1.76–19.37); there was an elevated risk associated with a higher final cumulative score (>1) in female patients with SSc compared with female controls for white spirit (OR 6.22, 95 % CI 2.02–21.28), aromatic solvents (P = 0.0008), any type of solvent (OR 4.95, 95 % CI 1.15–17.52), and ketones (OR 7.91, 95 % CI 1.26–84.93) [2]. These authors reported that (1) building trade and painting were more commonly found in patients exposed to white spirit; (2) exposure to aromatic solvents was principally observed in patients in chemistry-related occupations and in workers exposed to petrol; and (3) exposure to chlorinated solvents and trichloroethylene was found in patients working in metal industry and automotive industry and in sheet metal workers, paint/varnish/adhesive manufacturing, and manufacture of electronic components [2]. Altogether, the differences between the occupational exposure to specific organic solvents of male and female SSc patients were explained, in part, by the differences in occupations.

In addition, occupational exposure to organic solvents has been found to be a predictive parameter of SSc severity [4]. Recently, SSc patients who were exposed to organic solvents, compared to non-exposed patients, have been found to exhibit more frequently (i) diffuse cutaneous SSc (P = 0.001); (ii) severe microangiopathy as shown by higher prevalence of digital ulcers (P = 0.01) and lower median value of left ventricular ejection fraction (P = 0.04); (iii) ILD (P = 0.02), which was more severe in exposed patients with lower median values of FVC (P = 0.005), VC (P = 0.002), and DLCO (P = 0.02) on pulmonary function tests; (iv) cancer (P = 0.003); and (v) positivity of anti-Scl70 antibody (P = 0.04) [4]. Interestingly, in this latter series, the authors have found that patients who were exposed to chlorinated solvents, compared to those exposed to white spirit, aromatic solvents, or ketones, had higher median values of Rodnan score (P = 0.037) and lower median values of left ventricular ejection fraction (P = 0.04), VC (P = 0.04), FVC (P = 0.03), and DLCO (P = 0.009) [4].

The pathogenic mechanisms of organic solvents remain unclear in the onset of SSc. Previous authors have speculated that organic solvents may link with nucleic acids and proteins, resulting in immune disruptors and increased risk of SSc [30]. MRL+/+ mice, when exposed to trichloroethylene, develop higher levels of overall serum IgG and increased prevalence of antinuclear antibodies [54]. Other experimental models have shown that exposure to trichloroethylene led to (1) protein modifications [55–59]; (2) synthesis of higher levels of IL-17 and IL-21 by splenocytes in trichloroethylene-treated mice [60]; (3) production of reactive oxygen species and increased nitric oxide by nitric oxide synthase in cultured human epidermal keratinocytes [61]; (4) Th1 T cell activation in MRL+/+ mice [62]; (5) inhibition of cellular apoptosis of naïve CD4+ and CD8+ T cells in MRL+/+ mice [63]; and (6) DNA hypermethylation on rat cardiac myoblasts [64]. Furthermore, other experimental studies have shown that exposure to benzene resulted in (1) decreased T cell population in wild/captive American kestrel birds [65]; (2) reduced number of CD4+/CD8+ T cells and B cells in exposed workers [66]; (3) increased reactive oxygen species and induction of DNA fragmentation in pump workers [67]; and (4) changes in gene transcription involved in apoptosis, oxidative stress, cellular cycle, and cytokine production in benzene-treated mice [68].

Other industrial chemicals

The association between SSc and other industrial agents related to occupational exposure has more uncommonly been described.

Previous investigators have reported a 2.5-fold higher risk of SSc associated with vinyl chloride, epoxy resins, and/or formaldehyde (OR 2.5, 95 % CI 0.8–8) that was statistically significant in men (OR 18.6, 95 % CI 1.4–251) [24]. In a case–control study (56 SSc patients), the authors failed to find a relationship between epoxy resin exposure and SSc [23]. In another series (80 SSc patients, 160 controls), significantly increased OR for SSc was observed for epoxy resins (OR 4.24, 95 % CI 1.03–17.44) [6]. However, Silman et al. [15] failed to find an association between SSc and epoxy resin exposure (OR 1.7, 95 % CI 0.4–7.3). More recently, in a larger study (100 SSc patients, 300 controls), there was no significant difference between epoxy resin exposure and SSc (OR 3.03, CI 95 % 0.02 to 12.6) [2].

Additionally, one case–control series (80 patients, 160 controls) has shown significant increased OR in SSc for welding fumes (OR 3.74, 95 % CI 1.06–13.18); the risk associated with high cumulative exposure (>1) was increased in SSc (OR 6.41, 95 % CI 1.26–32.49) [25]. Another larger case–control study (100 SSc patients, 300 controls) demonstrated a significant association between SSc and occupational exposure to welding fumes (OR 2.60, 95 % CI 1.15 to 5.81) [2]. This latter study has shown that the risk of SSc was increased in patients with a high cumulative score (>1) (OR 4.70, 95 % CI 1.09–23.14) for welding fumes [2]. Additionally, there was also an elevated risk in male patients with SSc compared with male controls (OR 8.62, 95 % CI 2.64–32.04) [2]. SSc patients exposed to welding fumes were welders, boilermakers, and electromechanical workers [2].

Finally, there has been no epidemiologic research on risks of SSc associated with relatively widespread metals and synthetic chemical exposures such as plasticizers, phthalates, and bisphenol A.

Pesticides and persistent organic pollutants

Few authors have mentioned that pesticides may contribute to increased risk of rheumatoid arthritis and systemic lupus erythematosus [69, 70]. A recent case–control study (100 SSc patients, 300 controls) failed to find an association between SSc and occupational exposure to pesticides (OR 3.06, 95 % CI 0.04 to 241.94); these latter data may be explained, in part, by the fact that SSc patients and controls included few agricultural workers/farmers (2 vs. 1 %) [2]. Further investigations are thus required to assess accurately the risk of SSc associated with specific pesticide agents.

Furthermore, based on published data, there are also insufficient results showing that exposure to persistent halogenated organic pollutants that are resistant to environmental degradation plays a role in the onset of SSc, including organochlorine pesticides, polychlorinated biphenyls, dioxins, and furans.

Air pollutants

Particulate emission from diesel exhaust engines has been associated with fibrosis [71, 72]. Diesel exhaust nanoparticles have been shown to be internalized by monocyte-derived macrophages and keratinocytes from healthy subjects, leading to oxidative, pro-fibrotic, and pro-inflammatory processes in normal human skin [71, 72]. A recent series has demonstrated that diesel exhaust nanoparticles induced, in vitro, (1) higher mRNA gene expression of metalloproteases 2, 7, 9, and 12; collagen I and III; and vascular endothelial growth factor (VEGF) in SSc fibroblasts than in controls and (2) activation of keratinocytes of patients with diffuse cutaneous SSc with increase of pro-fibrotic cytokines: IL-1α, IL-8, and IL-6 mRNA levels [73]. IL-1α has been reported to be greatly increased in SSc epidermis, and it is speculated to initiate keratinocyte–fibroblast interactions, leading to fibroblast activation [74].

These results suggest that traffic-derived pollution may play a role in SSc genesis. However, no definite conclusion can be drawn from these data, and further investigations are warranted to confirm these preliminary results.

Smoking and drug abuse

Smoking does not confer a risk for SSc onset. In a case–control study (621 SSc, 1228 controls), the authors failed to find an association between SSc and smoking status dichotomized into (1) ever smoking (43 vs. 42.5 %) and (2) never smoking (57 vs. 57.5 %) [75]. However, it is established that cigarette smoking results in significant negative effects on vascular and pulmonary outcomes in SSc [76, 77]. Furthermore, in 606 SSc patients, a correlation has been observed between cigarette smoking and a higher modified Rodnan skin score (P = 0.0029) [78].

Furthermore, cocaine is a vasoconstrictive agent that has also been reported to be associated with SSc-like syndromes [79–82].

Personal care

Silicone

Silicone is one of the major constituents of the Earth’s crust and an important trace mineral in bone formation and mineralization [83]. In fact, the silicones used in medical products are heterogeneous in polymer length, side-chain substitutions, and fillers, leading to a great variation in the biological and physical properties of these chemicals [84]. The synthetic polymer may be formed into a gel by lengthening of polymer chains or can be modified into a rubber-like material (elastomer, by cross-linking the polymer chains) [84].

Case reports of women with silicone breast implants who developed SSc were published. Several case–control series and prospective studies of connective tissue diseases (including SSc) and breast implants failed to find an increased risk of SSc associated with breast implants. A meta-analysis (9 cohort studies, 9 case–control studies, and 2 cross-sectional studies) found no association between silicone breast implants and SSc (OR 1.01, 95 % CI 0.59–1.72) [85]. In a 2014 case–control study, the authors did not find any significant difference between SSc patients (n = 100) and controls (n = 300) regarding previous history of cosmetic surgery using silicone breast implants [2].

Prosthesis

In a 2014 case–control study, the authors did not find any significant difference between SSc patients (n = 100) and age- and sex-matched controls (n = 300) regarding previous history of prosthesis (OR 0.83, 95 % CI 0.29–2.05) as well as contact lenses (OR 0.74, 95 % CI 0.13–2.83) [2].

Hair dye

In a case–control study, the authors did not find any significant difference between SSc patients (n = 100) and age- and sex-matched controls (n = 300) regarding previous history of hair dye (OR 0.94, 95 % CI 0.57–1.58) [2].

Drugs

Case reports have identified clusters of SSc in patients receiving therapy, i.e.,

  • Antimitotics: bleomycin, taxanes, gemcitabin, tegafur-uracile, interferon α, and aldesleukine [54, 81, 86–92]. Moreover, SSc deterioration related to cisplatin and bleomycin therapy has been reported [90]. Taxane-associated SSc may induce increased synthesis of IL-4 and IL-6 and growth factors, leading to fibroblast activation [86].

  • Anorexigens, including diethylpropion chlorhydrate, mazindol, dexamphetamine-metaqualone, fenproporex, and fenfluramine [79, 80, 88, 93–95].

  • Tryptophan [88, 96–98].

In a 2014 case–control study, the authors did not find any significant difference between SSc patients (n = 100) and age- and sex-matched controls (n = 300) regarding the use of the following drugs: anorexigens (OR 1.39, 95 % CI 0.54–3.37), pentazocine (P = 0.441), bromocriptine (OR 0.94, 95 % CI 0.22–3.12), and taxanes (OR 3.01, 95 % CI 0.04–237.39) [2].

Skouby et al. [99] have also reported a case of scleroderma-like picture following a single serum injection during vaccination.

Thus, based on our data and literature review, there is insufficient evidence to show that personal care (silicone, prosthesis, hair dye, drugs) exposure plays a causative role in the development of SSc.

Physical agents

Physical agents include ionizing radiation, ultraviolet radiation, and electric and magnetic fields. To date, there is growing evidence for (1) an association between ionizing radiation and risk of autoimmune thyroiditis and Graves’ disease and (2) an inverse relationship between ultraviolet exposure and risk of multiple sclerosis [30]. However, no investigators have yet assessed the risk of SSc in patients exposed to physical agents, including ionizing radiation, ultraviolet radiation, and electric and magnetic fields.

Based on current published data, there is thus insufficient evidence that exposure to ionizing radiation, ultraviolet radiation, and electric and magnetic fields is a causative factor of SSc.

Biological agents

Biological agents can be divided into infections and diet, foods, and dietary contaminants.

Infections

Infections have been incriminated to play a role in the SSc pathogenesis. Several mechanisms of infections have been suggested, including endothelial cell damage, molecular mimicry, and self-reactive antibodies [100].

  • Parvovirus B19: Parvovirus B19 has been speculated as a causative agent in SSc because (1) serum parvovirus B19 viremia has been more often detected in SSc patients than in healthy subjects (4 vs. 0.6 %) [101, 102] and (2) the presence of the parvovirus B19 has been detected in bone marrow biopsy specimens of 57 % of SSc patients [103], suggesting that the bone marrow may represent a reservoir from which the parvovirus B19 virus spreads to SSc tissues [104]. Endothelial injury in patients with parvovirus B19 has been suggested to reflect a combination of direct viral cytotoxicity and humoral immunity [105]. It has also been found that parvovirus B19 exerts a cytotoxic effect on infected cells through a non-structural protein NS-1 [106]. The ability of parvovirus B19 to persistently infect SSc fibroblasts might be responsible for marked cell alterations [107].

  • Cytomegalovirus: Cytomegalovirus may play a role in SSc onset due to its ability to infect both endothelial and monocyte/macrophage cells and through the upregulation of fibrogenic cytokines and induction of immune dysregulation [108, 109]; the onset of SSc shortly after an acute episode of viral infection suggested cytomegalovirus as a possible trigger for SSc. Infection of endothelial cells alters the expression of different integrins and may induce the expression of fibrogenic cytokines and dysregulation of different antibodies, especially anti-Scl70 antibody [108, 110]; anti-Scl70 antibody could cross-react with a peptide sequence of the UL-70 protein of cytomegalovirus [111].

  • Helicobacter pylori: Previous authors have reported increased prevalence of H. pylori in SSc patients, compared with healthy subjects [112–114]. In 42 SSc patients, H. pylori infection has also been found to be associated with SSc activity, especially with higher values of Rodnan score (P < 0.0001) [115]. It has been speculated that H. pylori sheds extracellular products and notably heat shock protein 60 (HSP 60) eliciting local and systemic immune responses and leading subsequently to tissue damage [116]. However, there are other conflicting and controversial data regarding the association of H. pylori infection with SSc.

Altogether, based on current published data, there is insufficient evidence to show that infections play a causative role in the development of SSc.

Diet, foods, and dietary contaminants

Regarding dietary factors in autoimmune diseases, the gold standard is the causation of celiac disease by gluten ingestion [30]. Furthermore, there is evidence for association between outbreaks of (1) toxic oil syndrome and ingestion of contaminated rapeseed oil containing 1,2-di-oleyl ester and oleic anilide [117–119] and (2) eosinophilia–myalgia syndrome and l-tryptophan supplements [96, 98, 120].

Nutritional sources may provide the methyl donors and co-factors (folic acid, vitamin B12, and pyridoxal phosphate) essential for DNA and histone methylation. There are reports of diet-induced epigenetic changes in the adult state [121]. It has been proposed that epigenetic links between nutrition and autoimmunity may contribute to several autoimmune diseases.

However, to date, based on current published data, there is insufficient evidence to show that diet, foods, and dietary contaminants play a causative role in the development of SSc, including (1) breast-feeding and the age at which certain foods are introduced into an infant’s diet; (2) exposure to complex foods early in infancy; (3) low antioxidant vitamin intake, low fruit or fiber intake, and high sweets or fat intake; (4) low alcohol consumption; and (5) consumption of food chemicals, dyes or additives, as well as nitrates, nitrites, nitrosamines.

Conclusion

As a mirror image of the Roman god Janus Bifrons, the environment has a hidden face. To highlight this hidden face of the environment in the field of autoimmune diseases and particularly in SSc will allow to see responsible agents emerging in the very near future. To date, there is, in fact, a growing scientific evidence that environmental factors have a crucial impact on both alterations and modulation of epigenetic determinants, resulting in SSc onset and progression. Interestingly, a marked correlation has been found between SSc onset and occupational exposure to crystalline silica and the following organic solvents: white spirit, aromatic solvents, chlorinated solvents, trichloroethylene, and ketones; the risk associated with high cumulative exposure to silica and organic solvents further appears to be strongly increased in SSc. Altogether, occupational exposure should be systematically checked in all SSc patients at diagnosis, as (1) exposed patients seem to develop more severe forms of SSc, especially diffuse cutaneous SSc, severe microangiopathy, and ILD, and (2) the identification of the occupational toxic agent will allow its interruption, which may lead to potential improvement of SSc outcome [2, 4, 121]. Therefore, further investigations are still warranted to identify other environmental factors that may be associated with SSc onset and progression.