Systemic sclerosis (SSc) is a multisystem auto-immune disease. The two main subtypes of SSc (limited and diffuse) typically have differing courses and prognoses. New classification criteria have been proposed to identify SSc in the earliest stages, before skin involvement. Over the past three decades, there has been an apparent increase in the incidence of SSc to approximately 20 per million, possibly due to improved diagnosis. The most extensively studied environmental associations of SSc are organic solvents and silica but no single risk factor has emerged. Recent genetic studies have identified new susceptibility factors including human leucocyte antigen (HLA) haplotypes and polymorphisms in immune regulatory genes. Despite earlier disease recognition and effective treatment for some of its complications, SSc still carries a high mortality, particularly due to cardiorespiratory complications. Although some predictors of organ involvement and outcomes have been identified, novel biomarkers are greatly needed. Due to low disease prevalence, large multicentre research collaborations are required.
Systemic sclerosis (SSc) is a multisystem connective tissue disease. Its clinical features include Raynaud’s phenomenon (RP) and scleroderma as well as internal organ involvement. This article gives an overview of the classification of SSc followed by a review of disease frequency, including geographic variations and time trends. We discuss the burden of disease in terms of morbidity and mortality and provide an update on the risk factors for disease occurrence and prognostic factors for its outcomes.
Case definition and classification of SSc
The two main subtypes of SSc are defined by the extent of skin involvement: limited cutaneous systemic sclerosis (lcSSc) and diffuse cutaneous systemic sclerosis (dcSSc). In lcSSc, skin involvement is confined to the face, neck and the area distal to elbows and knees. In dcSSc, skin involvement extends proximally to involve upper arms, thighs and/or trunk . The American College of Rheumatology (ACR) classification criteria for SSc , listed in Table 1 , are insensitive for lcSSc . Therefore, criteria for ‘early’ disease have been proposed ( Table 1 ) . In 2009, the European League Against Rheumatism (EULAR) Scleroderma Trial and Research Group (EUSTAR) proposed another set of criteria for the ‘very early’ diagnosis of SSc ( Table 1 ). These criteria have yet to be validated.
| ACR criteria for SSc a | Early SSc b | Very early SSc c | |
|---|---|---|---|
| Major | Scleroderma proximal to MCPs | Raynaud’s phenomenon | Raynaud’s phenomenon |
| Antibodies (ANA, ACA, anti-topo1) | |||
| Diagnostic nailfold capillaroscopy | |||
| Minor | Sclerodactyly | SSc-type nailfold capillary pattern | Calcinosis |
| Digital pitting or pulp atrophy | SSc-selective autoantibodies d | Puffy fingers | |
| Bibasilar pulmonary fibrosis | Digital ulcers | ||
| Dysfunction of the oesophageal sphincter | |||
| Telangiectasiae | |||
| Ground glass on high-resolution CT scan of the chest |
a Presence of the major criterion or 2 of the 3 minor criteria indicates SSc.
b Objective documentation of Raynaud’s phenomenon and one minor criterion or subjective evidence of Raynaud’s phenomenon and both minor criteria indicates early SSc.
c Presence of all three major or two major and one minor criterion indicates very early SSc.
d ACA, anti-topo 1, ARA I or III, anti-fibrillarin or anti-PM-Scl.
The two main subtypes of SSc have different natural histories and prognoses. DcSSc is associated with rapid early progression of skin thickening and a high incidence of early internal organ involvement whilst lcSSc tends to have a long period of RP prior to diagnosis and a later peak in mortality from pulmonary arterial hypertension (PAH). Recognition of these clinical subtypes and their differing outcomes has been one of the major advances in SSc over the past three decades. Of note, a subgroup of patients with lcSSc has no skin thickening (SSc sine scleroderma). In addition, patients with either lcSSc or dcSSc may have overlap syndromes, wherein the features of SSc overlap with those of systemic lupus erythematosus (SLE), polymyositis, Sjogren’s syndrome or rheumatoid arthritis.
SSc-specific autoantibodies tend to be associated with certain clinical manifestations . For example, anti-topoisomerase I antibody (anti-topo 1) is associated with dcSSc and an increased risk of pulmonary fibrosis. Anti-centromere antibody (ACA) is associated with lcSSc and an increased risk of severe digital ischaemia. Anti-RNA polymerase (I, II and III) antibodies (ARA) are associated with dcSSc and an increased risk of renal involvement. Anti-U1 ribonucleoprotein (U1RNP) antibodies are associated with overlap syndromes, while anti-polymyositis (PM)-Scl antibodies are associated with myositis.
Incidence of SSc
Varying prevalence and incidence rates of SSc have been reported due to the differences in the geographic area surveyed, the definition of disease and the method of case ascertainment . In Table 2 , we have summarised prevalence and incidence rate data from more recent epidemiological studies of SSc. In three studies in which ‘incidence’ was measured, the rates were very similar (22.8, 19.3 and 23 new cases/million/year in South Australia, Detroit and Northwestern Spain, respectively) . In the Spanish study, when diagnosis was based on the ACR criteria alone, the incidence rate was 12/million/year. When diagnosis was based on either the ACR or the LeRoy and Medsger ‘early’ criteria, the incidence rate was 23/million/year, highlighting the impact of accurate disease definition on measurement of disease frequency. A recent study of the incidence of childhood SSc in the UK and Ireland reported a very low rate of 0.27 cases/million children/year (95% confidence interval (CI): 0.1, 0.5), despite comprehensive case ascertainment over a 25-month surveillance period.
| Geographical area | Period of study a | Number of cases | Incidence rate b (95% CI d ) | Prevalence rate c (95% CI d ) |
|---|---|---|---|---|
| Iceland | 1975–1990 | 18 | 3.8 | 71 |
| Greece | 1981–2002 | 109 | 11 | 154 |
| Japan | 1987 | 629 | 7.2 | 38–53 |
| Northwestern Spain | 1988–2006 | 78 | 23 (16, 25) | 277 (211, 358) g |
| Detroit (US) | 1989–91 | 706 | 19.3 (12.4, 30.2) | 242 (213, 274) |
| South Australia | 1993–99 | 348 | 22.8 (16, 32) e | 233 (210, 260) |
| Northeast England | 2000 | 80 | Not assessed | 88 (68, 108) |
| Northeastern Paris | 2001 | 173.2 | Not assessed | 158 (129,187) |
| Quebec, Canada | 2003 | Not stated | Not assessed | 443 (411,476) |
| UK and Ireland f | 2005–2007 | 7 | 0.27 (0.1, 0.5) | Not assessed |
b New cases per million per year.
f Incidence in children under 16 years of age.
Incidence of SSc
Varying prevalence and incidence rates of SSc have been reported due to the differences in the geographic area surveyed, the definition of disease and the method of case ascertainment . In Table 2 , we have summarised prevalence and incidence rate data from more recent epidemiological studies of SSc. In three studies in which ‘incidence’ was measured, the rates were very similar (22.8, 19.3 and 23 new cases/million/year in South Australia, Detroit and Northwestern Spain, respectively) . In the Spanish study, when diagnosis was based on the ACR criteria alone, the incidence rate was 12/million/year. When diagnosis was based on either the ACR or the LeRoy and Medsger ‘early’ criteria, the incidence rate was 23/million/year, highlighting the impact of accurate disease definition on measurement of disease frequency. A recent study of the incidence of childhood SSc in the UK and Ireland reported a very low rate of 0.27 cases/million children/year (95% confidence interval (CI): 0.1, 0.5), despite comprehensive case ascertainment over a 25-month surveillance period.
| Geographical area | Period of study a | Number of cases | Incidence rate b (95% CI d ) | Prevalence rate c (95% CI d ) |
|---|---|---|---|---|
| Iceland | 1975–1990 | 18 | 3.8 | 71 |
| Greece | 1981–2002 | 109 | 11 | 154 |
| Japan | 1987 | 629 | 7.2 | 38–53 |
| Northwestern Spain | 1988–2006 | 78 | 23 (16, 25) | 277 (211, 358) g |
| Detroit (US) | 1989–91 | 706 | 19.3 (12.4, 30.2) | 242 (213, 274) |
| South Australia | 1993–99 | 348 | 22.8 (16, 32) e | 233 (210, 260) |
| Northeast England | 2000 | 80 | Not assessed | 88 (68, 108) |
| Northeastern Paris | 2001 | 173.2 | Not assessed | 158 (129,187) |
| Quebec, Canada | 2003 | Not stated | Not assessed | 443 (411,476) |
| UK and Ireland f | 2005–2007 | 7 | 0.27 (0.1, 0.5) | Not assessed |
b New cases per million per year.
f Incidence in children under 16 years of age.
Prevalence of SSc
Improved survival has contributed to the higher prevalence of SSc reported in recent studies. The prevalence of SSc in the South Australian, Detroit and Spanish studies was broadly similar (233, 242 and 277 cases/million, respectively) ( Table 2 ). Possible reasons for the higher prevalence of 443 cases/million reported in Quebec include case ascertainment methods (which included physician billing) and the long ascertainment period (10 years) .
Geographic variation in SSc disease frequency
Allowing for variation in case definition, the reported prevalence of SSc is consistently higher in USA and Australia than in Japan and Europe. In Europe, a north–south gradient has been observed with lower rates reported in northern European countries ( Table 2 ) .
A number of geographical clusters of SSc have been reported in several countries. Silman reported a high prevalence (150 cases/million) in three areas close to two major airports in the UK . A cluster of SSc and scleroderma-like syndromes has been reported in a rural province of Rome, wherein the inhabitants have a high frequency of human leucocyte antigen (HLA)B51 and DR2 haplotypes . Englert et al. reported a clustering of cases of SSc in Edenhope, Australia, where the estimated prevalence of SSc was 610 cases/million. Many of the cases were male farm workers and the authors postulated that dust-storm silica inhalation might be an aetiological factor .
To date, the highest prevalence of SSc has been reported among a group of Choctaw Indians living in Oklahoma, USA. Based on 14 cases, the prevalence of SSc in full-blooded Choctaws was 469 cases/100,000 and in all Choctaws, 66 cases/100,000 . This is significantly higher than in other native Americans in Oklahoma (9.5 cases/100,000). The majority of the cases have dcSSc with a high prevalence of interstitial lung disease (ILD) and anti-topo 1. Anti-topo 1 was strongly linked to HLA haplotypes DQ7 and DR2 (DRB1*1602).
Time trends in incidence of SSc
Studies from the USA suggest that the incidence of SSc has increased over the last 60 years. The incidence in Tennessee in 1947 was estimated to be 0.6 new cases/million/year whilst in Detroit in 1991, the incidence was 19 cases/million/year . Variations in study methodology, increased disease awareness and improvement in diagnosis may explain this apparent change. Using the same methods in the same region of Pennsylvania, Steen et al. demonstrated an increase in the mean incidence of SSc from 9.6 cases/million/year in the years 1963–1972 to 18.7 cases/million/year in the years 1973–1982 . The incidence increased most in women in whom there was a rise from 13 to 27.6 cases/million/year Subsequent studies in the USA have reported relatively stable incidences . In South Australia, a trend towards a rising incidence over the 10-year period from 1993 to 2002 (from 15.1 to 20.4 cases/million/year) was not statistically significant . In Europe, the incidence seems to be relatively stable in northern areas with rates of 3.7 cases/million/year in England in 1986 , 3.8 cases/million/year in Iceland in 1990 and 3.7 cases/million/year in Finland in 1990 . There are no earlier European studies available to compare rates over a longer time period.
Racial variation in SSc disease frequency
Several studies have reported a higher incidence of SSc in black populations and that these patients are more likely to have severe disease than white patients. In Michigan, the incidence of SSc in black women was 22.5 cases/million/year whilst in white women it was 12.8 cases/million/year . Fifty percent of black women with SSc had dcSSc compared with only 25% of white women with SSc . In a French multiethnic study, there was a trend towards a higher prevalence of SSc among non-Europeans (Africans, Asians and Caribbeans), with dcSSc also being significantly more common (34% cf. 17% in Europeans). ILD was also more common in the non-Europeans (53 cf. 33%) . These findings, along with the high prevalence of SSc among the Choctaw Indians, support the hypothesis that genetic background influences the development of SSc and its manifestations.
Risk factors for the occurrence of systemic sclerosis
The complexity of SSc pathogenesis along with its heterogeneous clinical manifestations suggests that no single genetic or environmental trigger is likely to be responsible.
Genetic factors
Support for a genetic basis of SSc is provided by familial clustering. Studies of families with one member having SSc reported a 1.4–2.5% incidence in another member . In population-based registers, family history of SSc is associated with a 13–14-fold increase in relative risk for a first-degree relative . However, monozygotic twins have only a 5% concordance for SSc, though they are up to 90% concordant for presence of autoantibodies, in particular, antinuclear antibody (ANA) .
Various HLA class II alleles influence disease susceptibility among people of different racial origins . In a recent, large, multiethnic study by Arnett et al. , the DRB1*1104, DQA1*0501, DQB1*0301 haplotype and DQB1 alleles encoding a non-leucine residue at position 26 ( DQB1 26 epi) were positively associated with SSc in whites and Hispanics. The HLA-DRB1*0701, DQA1*0201, DQB1*0202 and HLA-DRB1*1501, DQA1*0102, DQB1*0602 haplotypes were negatively associated with SSc suggesting that they confer a ‘protective’ effect. SSc in blacks was associated with DRB1*0804, DQA1*0501, DQB1*0301 alleles.
Certain HLA alleles are strongly associated with SSc-specific autoantibodies. For example, haplotypes of DRB1*11 have been associated with anti-topo 1 in Caucasians , although, in the study of Arnett et al. , DPB1*1301 had the highest odds ratio (OR) for anti-topo 1 (OR = 14). ACA has been associated with DQB1*0501 and DQB1 26 epi in this and other studies . ARA has been associated with DRB1*0404, DRB1*11 and DQB1*03 in whites and Hispanics and with DRB1*08 in blacks . DRB1*1302 and DQB1*0604 have been associated with the anti-fibrillarin antibody .
Candidate gene studies have identified single nucleotide polymorphisms (SNPs) associated with susceptibility for developing SSc , and some are correlated with certain autoantibodies and/or clinical phenotypes. Many of these genes are involved in regulating immune tolerance, and polymorphisms in these genes are associated with other auto-immune diseases, particularly SLE, suggesting that they are common auto-immune-disease-susceptibility genes. Interferon regulatory factor 5 (IRF5) regulates the expression of type I interferons (IFNs), and polymorphisms in its gene are associated with SLE. Overexpression of IFN-inducible genes is found in the skin and peripheral blood of SSc patients, implying a similar dysregulation of IFN signalling pathways in SSc . A functional SNP in intron 1 of IRF5 was reported in 427 Caucasian patients with SSc (OR for development of SSc: 1.59) and in patients with fibrosing alveolitis.
The excess accumulation of extracellular matrix (ECM) in SSc skin due to increased production of type I collagen by SSc fibroblasts involves various regulators, including connective tissue growth factor (CTGF) and transforming growth factor (TGF)-β. The expression of the CTGF gene (also known as CCN2 ), is upregulated in the skin of SSc patients . In a large group of British patients , the GG genotype at −945 in the CTGF gene promoter was more common in patients with SSc than in control subjects (OR = 2.2) and was associated with anti-topo 1 (OR = 3.3) and fibrosing alveolitis (OR = 3.1).
Other ECM proteins are overexpressed by SSc skin fibroblasts. Secreted protein, acidic and rich in cysteine (SPARC) regulates the deposition and assembly of ECM components . Homozygosity for the C allele at SPARC SNP +998 is significantly increased in SSc patients in Choctaw Indians and other ethnic groups. In addition, SNPs +1551 and +1922 correlate with RP and pulmonary fibrosis, respectively.
Fibrillin (FBN-1) is a regulator of TGF-β , and a defective FBN-1 gene is the basis of an animal model for SSc . Studies have shown a link between the FBN-1 gene and SSc in Choctaw Indians, but not in Caucasians .
Groups in the US and Europe recently completed the first phases of genome-wide association studies (GWASs) in SSc, with 1617 and 842 cases, respectively. These studies identified associations with HLA and with several genes important in regulating immune function including IRF5 .
Signal transducer and activators of transcription-4 (STAT4) is involved in signalling through the interleukin (IL)-23 receptor and, possibly, IFN receptors. Several studies have independently identified STAT4 as a susceptibility factor for SSc , a finding which was confirmed in the US GWAS . The association is greatest for lcSSc and ACA . Gene–gene interactions between polymorphisms in STAT4 and in the transcription factor genes, IRF5 and TXB1 , are associated with greater susceptibility to SSc and with certain phenotypes such as pulmonary fibrosis .
The roles of HLA and polymorphisms in IRF5 and STAT4 as risk factors were confirmed in a recent GWAS in 2296 SSc patients . This study also identified a new susceptibility locus at CD247. This gene encodes the T cell receptor (TCR) ζ chain (CD3Z), which plays a role in the assembly and signalling function of the TCR-CD3 complex. Reduced expression of CD3 ζ chain due to genetic variants has been associated with systemic autoimmunity.
The presence of auto-antibodies with specificity for different disease variants suggests that humoral immunity is involved in the pathogenesis of SSc. B-cell scaffold protein with ankyrin repeats (BANK1) links the B-cell receptor to kinases such as Lyn. BANK1 polymorphisms have been linked with dcSSc and anti-topo 1 .
Overall, genetic factors fail to explain the majority of the disease variance. Plotting the age at onset of SSc versus age-specific incidence in the South Australian Scleroderma Registry suggests that between five and eight random events occur during the development of the disease . It has been postulated that some of these events could be explained by genetic instability in a predisposed population, for example, acquired mutations in pivotal somatic genes. Alternatively, rare genetic variants arising from a common ancestor many generations earlier could be inherited by distantly related cases of SSc, which might explain the geographical clusters of SSc .
Female sex
The 7:1 female preponderance of SSc suggests hormonal or pregnancy-related factors may have a role in SSc pathogenesis. In a study of 472 women with SSc only published in abstract form, there was no evidence that the oral contraceptive pill, earlier age of menarche or reproductive history influenced the risk of developing SSc , although the same study showed a small but significant increase in the risk of SSc in post-menopausal women taking oestrogen replacement.
Environmental factors
Based on clusters of cases and on epidemiologic studies over many years, a number of environmental agents have been implicated in SSc but no single chemical has emerged as a leading contender. The most extensively studied environmental associations of SSc are occupational exposures to organic solvents and silica.
A possible relationship between occupational exposure to organic solvents and an increased risk of SSc was first described in the 1950s. In a meta-analysis of seven case-control studies and one cohort study published between 1989 and 1998 , the combined estimator of relative risk (CERR) for all studies was 2.91 (95% CI: 1.60, 5.30).
A potential role for silica has been suspected for many years, based on case series of SSc among stonemasons and gold miners . In a recent meta-analysis of 16 studies published between 1949 and 2009 , the CERR for silica exposure was 3.20 (95% CI: 1.89–5.43), mainly due to an increased risk in men. Other occupational exposures may have a role in women: teaching and textile work , medical diagnostic laboratories, professional cleaning, film developing and publishing . In a meta-analysis, there was no evidence of an association between breast implants and any of the connective-tissue diseases .
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
