Risk factors for and possible causes of systemic sclerosis (scleroderma)

INTRODUCTION — The pathogenesis of systemic sclerosis (SSc; scleroderma) involves vascular, fibrotic, inflammatory, and immunologic processes. Although a great deal is known about the abnormalities in immune, endothelial, and fibroblast cell function in SSc, it is unclear how these pathogenic pathways are initiated.

The etiology of SSc is unknown. The remarkable complexity of its pathogenesis suggests that no single gene or environmental trigger is, by itself, likely to be responsible for the development of SSc. Nevertheless, genetic factors appear to influence disease susceptibility as well as patterns of disease expression and disease-associated autoantibodies. The occurrence of SSc-like disease in response to certain environmental agents and the possible relationship of SSc to environmental exposures suggest that certain triggers lead to variable disease expression in genetically susceptible hosts.

A review of the risk factors for and possible causes of SSc is presented here. The pathogenesis of SSc, including vascular and endothelial changes, fibroblast activity, and the role of different cytokines, is discussed separately. (See "Pathogenesis of systemic sclerosis (scleroderma)".)

RISK FACTORS — Exposure to viruses or to certain environmental toxins and drugs appears to trigger the development of disease in genetically susceptible hosts.

GENETIC FACTORS — Evidence for genetic factors contributing to disease susceptibility in systemic sclerosis (SSc) comes from a variety of observations [1]. Studies of the Choctaw Native Americans in Oklahoma showed that the prevalence of SSc was 20 times higher than in the general population [2,3]. Furthermore, disease expression in this cohort was remarkably homogeneous, with diffuse skin disease, lung involvement, and anti-topoisomerase I (Scl-70) antibodies detected in almost all affected individuals. The risk of SSc in first-degree relatives of individuals with SSc is markedly increased.

Candidate genes — The strongest genetic associations linked to SSc are with the major histocompatibility complex (MHC). In the Choctaw Native American population, the presence of anti-topoisomerase I antibodies is strongly linked to the human leukocyte antigen (HLA) haplotype DQ7, DR2 (DRB1*1602) [3,4]. Other risk-associated alleles are DQA1*0501 and DQB1*0301 [3]. Subsequent studies showed a link between the fibrillin-1 gene, implicated in a putative mouse model of inherited scleroderma [5], and SSc in this Native American population [6], but not in White French or Italian populations [7]. Neither the MHC nor the fibrillin-1 gene shows a perfect association with the disease; there are affected individuals without the suspect genes and unaffected individuals with these genes.

Genes in the MHC influence disease susceptibility [8-12]. As an example, the HLA-DQA1*0501 allele was found in 42 percent of White men with diffuse cutaneous SSc but in only 29 percent of healthy men from the same geographic area (odds ratio 3) [8]. Risk associations have also been suggested to exist between SSc and HLA-DR5/11 and DR3 haplotypes in White individuals, DR2 haplotypes in Japanese, and, as noted above, DR2 (DRB1*1602) in the Choctaw population [3].

There is also an association between HLA and certain SSc-specific autoantibodies. These include antibodies directed against fibrillarin [13], ribonucleic acid (RNA) polymerase [14], anticentromere [3], anti-PM-Scl [3], and topoisomerase I [3,15,16].

Genetic polymorphisms in non-MHC genes associated with SSc include angiotensin converting enzyme (ACE), endothelial nitric oxide synthase (eNOS), the B cell surface protein associated with the antigen receptor (CD19), and PTPN22 [17-20]. However, the putative association between ACE and eNOS polymorphisms and SSc could not be independently confirmed [19].

Genome-wide association studies — Genome-wide screening in a case-control study in the Choctaw population confirmed earlier candidate gene regions (eg, MHC and the fibrillin gene) and identified other sites dispersed on several chromosomes that were associated with disease [21].

The largest genome-wide association study (GWAS) examined 2753 patients with SSc (mostly of European ancestry) and 5171 controls [22]. Several susceptibility loci were identified, including the HLA locus and gene regions for CD247, interferon regulatory factor 5 (IRF5), and signal transducer and activator of transcription 4 (STAT4). Each of these SSc-susceptibility loci is involved in immune recognition, in innate immunity, or in immune cell signaling. Interestingly, a majority of gene variants associated with SSc have also been linked to other autoimmune diseases, particularly systemic lupus erythematosus (SLE) [23].

Familial clustering — Further support for a genetic basis of SSc is provided by familial clustering of SSc or autoimmune disorders. One study of 710 families with one member having SSc reported a 1.4 percent incidence in another family member [24]. Similar frequencies have been observed in other studies (1.5 to 2.5 percent) [25,26]. The increased risk is similar for siblings and other first-degree relatives of affected patients (15- and 13-fold higher, respectively) [3]. Familial studies of autoantibodies in SSc are consistent with the hypothesis that genetic factors might determine whether disease develops and which phenotype is expressed. Although monozygotic twins are only modestly concordant for SSc (4.7 percent), they are 90 percent concordant for antinuclear antibody (ANA) positivity (titer of ≥1:40) [3,27]. Similarly, family members of patients with SSc are much more likely to have ANA than are healthy controls [28,29].

Non-HLA gene associations with systemic sclerosis

PTPN22 gene — A polymorphism of the PTPN22 gene was associated with the finding of anticentromere and anti-topoisomerase antibodies in SSc patients [20,30]. This gene has also been reported to be important in a number of other autoimmune conditions, including diabetes mellitus-type I, rheumatoid arthritis, SLE, Grave disease, and juvenile idiopathic arthritis (JIA) [31].

IRF8 gene — A polymorphism in the IRF8 gene was found by GWAS to be protective for SSc disease susceptibility [32].

TNFAIP3 (A20) gene — A polymorphism of the TNFAIP3 gene was found to be associated with diffuse cutaneous SSc and was further associated with interstitial lung disease and pulmonary hypertension [33]. Multiple polymorphic variants in this gene, encoding the immunomodulatory enzyme A20, have been associated with SLE and other autoimmune and inflammatory diseases.

IFN regulatory factor 5 gene — In a study of 427 patients with SSc, an association was noted between a functional polymorphism of the IRF5 gene and the risk of scleroderma (odds ratio 1.59, 95% CI 1.18-2.11) [34]. Carriage of the risk allele was also associated with an increased risk of fibrosing alveolitis in White European ancestry. International GWAS have identified IRF5 as a susceptibility locus for SSc in a European population [22].

Other — Other recurrent genes associated with SSc or specific SSc subsets include:

●STAT4, the signal transducer and activator of transcription 4 gene [35,36], is a transcription factor implicated in inflammatory signaling. Several studies have indicated an association of polymorphisms in STAT4 with SSc. Moreover, polymorphisms in STAT4 have additive effects with polymorphisms in IRF5 on SSc-associated interstitial lung disease [36].

●BANK1, a substrate of tyrosine kinase expressed on B cells, with a stronger association of the combination of BANK1, STAT4, and IRF5 alleles with diffuse cutaneous systemic sclerosis (dcSSc) [37]. These gene associations, together with another B-cell related gene BLK, implicate B cells in SSc pathogenesis.

●CD247 gene – Polymorphisms of the CD247 gene implicated in T-cell activation were found to be associated with SSc by GWAS [22].

●TBX21 gene [38].

●HLA-DPB1 and -DPB2 [39].

●Interleukin (IL)-23 receptor polymorphisms associated with pulmonary hypertension and anti-topoisomerase-I [40].

●TNIP-1 gene – Polymorphisms in the TNIP-1 gene (encoding NTFAIP3-interacting protein) were associated with SSc in GWAS [41]. Skin biopsies from patients with SSc showed reduced expression of this protein.

INFECTIOUS AGENTS — Genetic factors alone cannot fully explain disease expression. Thus, one hypothesis is that infection, particularly with a virus, may trigger a cascade of events in a genetically susceptible host leading to systemic sclerosis (SSc) and associated autoimmunity.

One possibility is that viruses harboring amino acid sequences similar to those found within host proteins may initiate the disease process in a host HLA haplotype-restricted manner. This general phenomenon, not unique to SSc, is called molecular mimicry. One study described a common epitope shared between topoisomerase I, a key target autoantigen in SSc, and certain retroviruses [42].

Latent viral infection may also accelerate or promote disease in the susceptible host. In particular, cytomegalovirus may induce the vascular, fibrotic, and immunologic features of SSc via effects on various cell types:

●Infection of endothelial cells results in vascular and immune injury [43].

●The virus can subvert immune responses.

●Fibrosis is promoted by stimulating production of transforming growth factor (TGF)-beta, connective tissue growth factor, and related cytokines [44].

Studies have suggested that SSc-specific autoantibodies react with the UL94 human cytomegalovirus late protein. These autoantibodies were able to induce endothelial cell apoptosis [44] and also to activate cultured human fibroblasts [45].

Studies from Europe suggest that localized scleroderma (morphea) may be associated with Borrelia burgdorferi infection [46,47]. A similar association has not been seen in the United States. (See "Clinical manifestations and diagnosis of systemic sclerosis (scleroderma) in adults", section on 'Differential diagnosis'.)

NONINFECTIOUS ENVIRONMENTAL FACTORS — A number of environmental agents have been implicated in systemic sclerosis (SSc), based upon case clustering or upon formal epidemiological studies. One of the earliest and strongest such associations, with silica dust, was discovered because of the high frequency of SSc among stone masons [48]. Similar clustering has been observed among gold miners in South Africa and among coal miners in the United States [49,50]. However, a case-control study from the United Kingdom did not support an association of SSc with occupational silica exposure [51]. Since the use of vibrating machinery is associated with the Raynaud phenomenon [52] and since silica may cause interstitial lung disease (silicosis), some cases described as SSc may instead have had the confluence of these two disorders.

A meta-analysis of 16 observational studies, which examined the risk of occupational silica exposure, found that such exposure was associated with significant risk for the development of SSc in males (combined estimator of relative risk [CERR] 3, 95% CI 1.2-7.4), but not in females (CERR 1, 95% CI 0.7-1.4) [53]. The risk estimate in cohort studies was higher than the estimate in case-control studies (CERR 15.4 versus 2.2).

Despite concern about a possible relationship between silicone exposure, primarily in breast implant recipients, and SSc or SSc-like disease, multiple large epidemiologic studies and a meta-analysis failed to provide evidence supporting a relationship. In one study, the prevalence rates for SSc and other autoimmune disorders were compared among 1576 women with breast implants (of whom 1112 had received silicone gel-filled implants) and 726 control individuals [54]. The incidences of typical or atypical connective tissue disease, as determined by a rheumatologist, were similar between the two groups.

Similarly, in a meta-analysis of nine cohort studies, nine case-control studies, and two cross-sectional studies, breast implants in general and silicone-filled breast implants in particular were not associated with an enhanced risk of SSc or other connective tissue disease [55]. This topic is further addressed elsewhere.

Petroleum-based products — A relationship between certain toxic exposures and SSc or SSc-like syndromes is supported by some studies. Examples include the following:

●Among 178 patients with SSc and 200 control individuals, men with the disorder had higher cumulative exposure to organic solvents than did controls [56].

●Among 660 women with SSc and 2227 controls, exposure to paint removers and paint thinners (but not to trichloroethylene [TCE]) was significantly more likely in the SSc group [57].

There are reports of linkage of SSc with certain industrial solvents. Trichloroethane (TCA), TCE, toluene, and xylene are among those most frequently cited. TCE has a similar structure to vinyl chloride. In addition, an association between solvent-associated hobbies and the development of SSc with anti-topoisomerase I antibodies has been reported [58].

An association between exposure to petroleum distillates and undifferentiated connective tissue disease was suggested by an epidemiologic study [59]. Potentially implicated compounds included hydrocarbons such as paint thinners, benzene, and chlorinated solvents such as TCA and TCE.

Vinyl chloride — An SSc-like disease has been reported in individuals exposed to vinyl chloride monomer [60]. This exposure occurs in the manufacturing process of vinyl chloride polymer and not in connection with any finished product. Affected individuals had SSc-like vascular features including Raynaud phenomenon, telangiectasia, and nailfold capillary changes but did not display characteristic skin changes.

Contaminated rapeseed oil — Ingestion of contaminated rapeseed oil was associated with a self-limited epidemic illness called toxic oil syndrome in Spain in the early 1980s. The syndrome was characterized by acute myalgia, fever, neuropathy, SSc-like skin disease, and pulmonary hypertension [61]. The disease was temporally closely related to the rapeseed oil exposure.

L-tryptophan — Another exposure linked to a toxic-epidemic SSc-like illness is eosinophilia myalgia syndrome (EMS), which occurred in the United States in the late 1980s. This illness, characterized by fasciitis and dermal induration, occurred in individuals consuming the nutritional supplement L-tryptophan [62-64]. Affected individuals showed abrupt onset of chronic diffuse skin induration, as well as neuropathy and myopathic features, but, in marked distinction to SSc, they did not have the Raynaud phenomenon, SSc-specific autoantibodies, or visceral disease [64].

The syndrome showed clinical similarities to the toxic oil syndrome described above and was linked to the consumption of L-tryptophan, particularly specific batches originating from a single manufacturer in Japan.

A well-documented post-epidemic case of EMS was described in an individual who used L-tryptophan subsequent to the 2005 lifting of the US Food and Drug Administration (FDA) ban on L-tryptophan in the United States imposed following the identification of the syndrome and its link to the specific contaminant [65]. This case highlights the continued occurrence of L-tryptophan-associated EMS. Sporadic cases meeting surveillance criteria for EMS but lacking L-tryptophan exposure have also been reported [66].

In addition to a significant positive association with L-tryptophan dose and increased age, both increased risk of EMS and protection following exposure have been associated with certain class II histocompatibility gene polymorphisms [67].

Another dietary supplement, 5-hydroxytryptophan (5-HTP) has been associated with an EMS-like disorder [68].

DRUGS — Several drugs have been associated with the development of systemic sclerosis (SSc)-like syndromes. Best studied of these is bleomycin, which induces skin and lung fibrosis in animal models [69]. Several cases of SSc with the Raynaud phenomenon have been reported among cancer patients undergoing chemotherapy with bleomycin [70]. (See "Cutaneous adverse effects of conventional chemotherapy agents".)

A mechanism potentially underlying the possible role of bleomycin in SSc-like conditions is the induction of chromosome breaks, a phenomenon that is enhanced with bleomycin ingestion and is seen in patients with SSc and their family members, independent of drug exposure [71,72]. Chromosome breaks may occur in response to oxidative stress, which may preferentially lead to the formation and release of unique autoantigens [73,74].

Other drugs potentially implicated in SSc-like syndromes include pentazocine, cocaine, and the taxanes docetaxel and paclitaxel [75].

MICROCHIMERISM — Healthy women frequently harbor viable immunologic stem cells of fetal origin many years following pregnancy [76]. The number of circulating fetal cells present in women with systemic sclerosis (SSc) who have been pregnant is elevated compared with healthy control women with a history of pregnancy [77,78]. Male cells were found in at least one organ obtained at autopsy from women with SSc who had male offspring [79].

The persistence of fetal cells may be linked to the pathogenesis of SSc via several potential mechanisms:

●Fetal (the graft) cells may mount an immune response against the mother (the host) in a graft-versus-host-like reaction.

●A maternal response to the fetal cells may subsequently be redirected against "self," leading to autoimmunity. It is noteworthy that patients with SSc are more likely to be closely human leukocyte antigen (HLA)-matched to the paternal HLA carried by their children [79,80].

●Maternal cells that cross the placenta and that are carried by the fetus may play an immunomodulatory role in men with SSc or in women with SSc who have never been pregnant [79,80].

However, not all studies have confirmed the increase in microchimerism in SSc. Furthermore, one's immunogenetic background may determine the development of microchimerism; as an example, among individuals with HLA-DQA1*0501, microchimerism was increased both in patients with SSc and in healthy controls [81]. Therefore, the significance of fetal microchimerism in the pathogenesis of SSc remains uncertain.

SUMMARY

●Genetic factors – Familial clustering of disease and gene associations suggest a role for genetic influences in systemic sclerosis (SSc; scleroderma).

•MHC class II – The genetic findings include associations with certain major histocompatibility antigen (MHC) class II alleles and other genes involved in immune recognition or signaling.

•HLA haplotypes – There is also an association between human leukocyte antigen (HLA) haplotypes and certain SSc-specific autoantibodies. (See 'Candidate genes' above.)

•Other genetic factors – Non-HLA genes linked to SSc susceptibility include TNFAIP3, CD247, and PTPN22 genes as well as genes for interferon regulatory factor 5 (IRF5). (See 'Non-HLA gene associations with systemic sclerosis' above.)

●Environmental factors – A number of environmental exposures have been linked to SSc based upon case clustering or formal epidemiological studies.

•Silica – Occupational silica exposure in males is associated with a significant increase in SSc risk. However, there is no evidence of a relationship between silicone exposure from breast implants and the development of SSc or SSc-like disease. (See 'Noninfectious environmental factors' above.)

•Infection – Infection may play a role in SSc pathogenesis. Potential mechanisms include molecular mimicry or more direct cellular effects of latent viral infection, such as cytomegalovirus, which may induce the vascular, fibrotic, and immunologic features of SSc. (See 'Infectious agents' above.)

•Toxins – A relationship between certain toxic exposures and SSc or SSc-like syndromes has been observed for several substances. (See 'Noninfectious environmental factors' above.)

-Exposure to organic solvents has been associated with the development of SSc.

-SSc-like disorders may occur in patients exposed to vinyl chloride, bleomycin, and several other medications.

-Local epidemics of SSc-like disorders have occurred in association with contaminated rapeseed oil (ie, toxic oil syndrome) and contaminated L-tryptophan (ie, eosinophilia myalgia syndrome [EMS]).

●Microchimerism – Some studies have suggested a potential role for increased microchimerism (ie, circulating fetal deoxyribonucleic acid [DNA] in the parent or parental DNA in the offspring) in SSc. (See 'Microchimerism' above.)