Systemic sclerosis in Asians: Are there racial differences?

Abstract

Systemic sclerosis is a multisystemic autoimmune disease characterized by vasculopathy and fibrosis. Racial factors exert a significant influence on the epidemiology, clinical manifestations, antibody profile, mortality and genetic factors in systemic sclerosis. In this review, we examined Asian systemic sclerosis cohorts reported in Asia and multi-racial cohort studies to evaluate the disease characteristics and outcomes of systemic sclerosis in Asians. Asian patients have distinct genetic susceptibility to systemic sclerosis, younger age of systemic sclerosis onset, higher frequency of diffuse skin involvement, different autoantibody profiles such as higher frequency of anti-Scl70 and anti-U1-RNP antibodies, and more severe clinical phenotype. There was a suggestion of poorer survival among Asians that may be contributed by more severe disease, socioeconomic factors and differences in healthcare systems. Recognizing the influence of racial differences in systemic sclerosis disease course is important as it has implications for appropriate treatment, monitoring and prognostication.

Introduction

Systemic sclerosis (SSc) is a multisystemic autoimmune disease characterized by immune dysregulation, vasculopathy, inflammation and fibrosis. The heterogeneous nature of SSc in terms of epidemiology, disease manifestations, mortality, genetic factors and autoantibody profile is evidently influenced by racial and geographic variations. ¹

Studies of multi-racial cohorts, including Caucasian and Asian patients with SSc reported that Asians have a younger age of SSc disease onset, higher frequency of diffuse skin involvement, more severe clinical phenotype and different autoantibody profile such as higher frequency of anti-Scl70 antibody.^2–4 Within Asia, SSc cohort studies conducted in countries including Japan, Korea, China, Thailand, Singapore, Taiwan, India and Iran also suggest that race influences disease manifestation and the course of SSc.^5–14 Studies from China and Thailand have reported greater proportion of diffuse cutaneous SSc (DcSSc) cases, whereas Japan and Korea had more prevalent cases of limited cutaneous SSc (LcSSc).^5–7,9 Higher frequency of interstitial lung disease (ILD) was also reported in China and India compared to Singapore, Japan and Korea.^5,6,8,11,13

To date, studies examining the clinical characteristics and outcomes in Asians with SSc remain relatively few and limited to independent cohort studies. In this review focusing on SSc in Asians, we summarize the literature highlighting SSc cohorts reported in Asia, and studies of multi-racial cohorts that compare Asians with the Caucasian populations. Observed differences in demographics, clinical characteristics, autoantibody profile and outcomes are described and their implications discussed. Recognizing the influence of racial differences in SSc disease course is important as it has implications for appropriate treatment, monitoring and prognostication.

Incidence and prevalence

Geographical variations in epidemiology have been described. A recent meta-analysis showed that the overall pooled prevalence of SSc globally was 17.6 per 100,000, and the overall pooled incidence rate of SSc was 1.4 per 100,000 person-years. ¹⁵ SSc prevalence is higher in the United States and Australia (25 and 24 per 100,000, respectively) than in Europe (15 per 100,000) and Asia (6.8 per 100,000). ¹⁵ The incidence rate of SSc is also higher in the United States compared to Asia (1.5–2.0 vs 0.9 per 100,000 person-years). ¹⁵ In Asia, SSc prevalence varied from 5.3 per 100,000 in Japan to 8.0 per 100,000 in Taiwan.^16,17

Genetic polymorphisms

Genetic differences between different racial groups have shown to exert an effect on SSc. Human leukocyte antigen (HLA) and its associations with SSc susceptibility have been extensively examined. ¹⁸ Several studies have found that HLA class II alleles were associated with SSc and vary in different racial populations. ¹⁹ A genome-wide association study (GWAS) in a Korean SSc cohort revealed that the specific single nucleotide polymorphisms (SNPs) of HLA DPB1 conferred the strongest susceptibility to SSc, and HLA-DPB1*13:01 and DPB1*09:01 were the most susceptible alleles in Korean SSc. ²⁰ A Japanese SSc cohort supported only DPB1*09:01 in association with SSc. ²¹ Studies of Caucasian populations in the United States and the United Kingdom supported DPB1*13:01, but not DPB1*09:01, as a major susceptibility allele to SSc, and most prominent in SSc patients with anti-Scl70 antibody.^22,23 Han Chinese patients with SSc were more likely to have HLA-DPB1*03:01 and to develop pulmonary fibrosis. ²⁴ A strong association of DRB1*15 with SSc was also seen in Thai, Han Chinese and Korean patients, which suggests it is specific to Asian SSc populations.^25–27

Clinical characteristics and outcomes

Tables 1 and 2 summarize pertinent studies that describe the clinical characteristics and outcomes of Asian patients with SSc, details of which will be elaborated in the rest of this review. Table 1 includes studies conducted in Asia; Table 2 summarizes multi-racial cohort studies conducted in Western countries that make direct comparisons of Asian and Caucasian patients with SSc. A meta-cohort of four SSc registries comparing Western and Asian patients is also included here.

Table 1.

Studies conducted in Asia describing SSc clinical characteristics and mortality.

Cohort	Race	Female, %	Age at onset a /diagnosis b , years	Subtype	Antibody	RP, %	DU/ischaemia, %	Joints, %	ILD, %	Suspected PH/PAH on RHC, %	Cardiac, %	GI, %	SRC, %	Mortality	Association with mortality
Li et al.; 7 2002–2014; n = 201; single centre (China)	Han Chinese	91	41.6 a	D 50%, L 30%, O 19% (RA, SLE, DM)	Scl70 29%; ACA 13%; U1RNP 25%	86	27	30	74	19/ NA	28	25	2	Mortality rate 2.4/100 py SMR 2.2 10-year survival 95% COD: ILD, renal	(+) Pericardial effusion; digital ischaemia
Hu et al.; 8 2009–2015; n = 448; single centre (China)	Han Chinese 98%, Others 3%	90	39.0 a	D 43%, L 57%, O 8% (RA, SLE, DM)	Scl70 47%; ACA 17%; U1RNP 26%	97	33	42	85	17/ NA	61	70	1	10-year survival 88% COD: PAH, renal	(+) PAH, arrhythmia (−) MTX, TwHF
Hashimoto et al.; 5 1973–2008; n = 405; single centre (Japan)	Japanese	93	47.0 a	D 33%, L 67%, O 28% (SS, myositis, SLE, RA)	Scl70 23%; ACA 36%; U1RNP 24%	NA	NA	NA	50	16/ NA	20	46	3	SMR 2.8 10-year survival 88% COD: malignancy, infection	(+) ILD, cardiac, renal, myositis overlap (−) ACA, SS overlap
Moon et al.; 6 1986–2016; n = 751; multicentre (Korea)	Korean	87	46.0 a	D 35%, L 65%	Scl70 43%; ACA 26%; RNAP3 16%	NA	NA	NA	53	14/ NA	11	33	3	10-year survival 87%	(+) Scl70, cardiac, age > 50 years
Santosa et al.; 11 2008–2013; n = 349; multicentre (Singapore)	Chinese 78%, Malay 10%, Indian 7%, Others 5%	87	46.2 b	D 37%, L 34%, O 27% (SLE, RA, myositis)	Scl70 38%; ACA 15%	79	23	63	63	30/ 9	24	51	3	Mortality rate 4.7/100 py COD: ILD, PAH, GI	(+) Indian versus Chinese; smoking; overlap; renal; echo PASP ⩾ 40 mmHg; vasculopathy treatment; parenteral nutrition
Kuo et al.; 12 2002–2007; n = 1479; nationwide (Taiwan)	Taiwanese	78	51.3 b	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	Mortality rate 4.3/100 py 5-year survival 83% SMR 3.2	(+) male, older age, cancer, ESRD
Wangkaew et al.; 9 2010–2014; n = 115; single centre (Thailand, Chiang Mai)	Thai	59	52.5 a	D 79%, L 21%	Scl70 80%; ACA 8%	NA	7	52	73	4/ NA	50	40	0	Mortality rate 4.2/100 py 4-year survival 88% COD: cardiomyopathy	(+) high mRSS, Hb < 10, ESR ⩾ 40
Foocharoen et al.; 10 2013–2019; n = 566; single centre (Thailand, Khon Kaen)	Thai	63	50.0 a	D 73%, L 27%	Scl70 81%; ACA 16%	60	19	27	47	NA/ 18	2	41	2	Mortality rate 3.5/100 py 5-year survival 84% 10-year survival 70% COD: cardiac, ILD, infection	(+) age > 60 years, FC III, IV, edematous skin, early ILD onset, mRSS, multi-organ
Janardana et al.; 13 2008–2013; n = 327; single centre (India, Vellore)	Indian	92	34.7 a	D 95%, L 4%, sine scleroderma 1%	Scl70 75%; ACA 5%	77	23	52	88	8/ NA	2	43	1	NA	NA
Poormoghim et al.; 14 1998–2012; n = 220; single centre (Iran)	Middle Eastern	87	36.2 a	D 40%, L 60%,	Scl70 70%; ACA 8%	90	23	21	62	16/ NA	24	88	2	10-year survival 82% COD: ILD, PH, cardiac	(+) age > 50 years, advanced ILD, TFR, arthritis

SSc: systemic sclerosis; RP: Raynaud’s phenomenon; DU: digital ulcers; ILD: interstitial lung disease; PH: pulmonary hypertension; PAH: pulmonary arterial hypertension; RHC: right heart catheterization; GI: gastrointestinal; SRC: scleroderma renal crisis; D: diffuse; L: limited; O: overlap; RA: rheumatoid arthritis; SLE: systemic lupus erythematosus; DM: dermatomyositis; ACA: anti-centromere antibody; U1RNP: U1-ribonucleoprotein; py: patient-years; SMR: standardized mortality ratio; COD: cause of death; MTX: methotrexate; TwHF: Tripterygium wilfordii Hook F; SS: Sjogren’s syndrome; RNAP3: RNA polymerase 3; PASP: systolic pulmonary arterial pressure; ESRD: end-stage renal disease; mRSS: modified Rodnan skin score; Hb: haemoglobin; ESR: erythrocyte sedimentation rate; FC: functional class; TFR: tendon friction rub; (+): positive association; (−): negative association.

^aAge at onset (reported as mean or median).

^bAge at diagnosis (reported as mean or median).

Table 2.

Multi-racial cohort studies conducted in Western countries comparing Asian and Caucasian SSc clinical characteristics and mortality.

Race	Chung et al.; 2 2007–2016; n = 609; single centre (United States)		Jaeger et al.; 3 2004–2018; n = 9700; multicentre (Europe)		Al-Sheikh et al.; 28 1970–2017; n = 1005; single centre (Canada)			Low et al.; 29 1990–2007; n = 336; single centre (Canada)		Proudman et al.; 4 2004–2016; n = 17,838; multicentre (Australia, Canada, Europe and Singapore)
	Asian a 16% Caucasian 51% Others 33%		EUSTAR: Asian 4% Caucasian 94% Others 2%		South Asian b 7% East Asian c 8% Caucasian 74% Others 11%			Chinese 11% Caucasian 89%		ASCS: Asian 5%; Caucasian 92%; Others 3% CSRG: Asian 3%; Caucasian 79%; Black 1%; Others 19% EUSTAR: Asian 4%; Caucasian 89%; Others 8% SCORE: Asian 99%; Caucasian 0.6%; Others 0.4%
	Asian n = 96	Caucasian n = 310	Asian n = 341	Caucasian n = 9162	South Asian n = 70	East Asian n = 80	Caucasian n = 745	Chinese; n = 36	Caucasian; n = 300	SCORE n = 500	ASCS n = 1714	CSRG n = 1628	EUSTAR n = 13,996
Female, %	89	90	89	84	80	80	82	72	81	87	86	86	86
Mean age, years	51.7 d	58.9 d	38.1 e	44.2 e	45.8 d	48.2 d	47.1 d	52.0 d	46.0 d	53.4 f	57.1 f	55.7 f	54.8 f
First non-RP to diagnosis, years	4.2	5.5	4.1	6.5	NA	NA	NA	NA	NA	7.1	10.9	10.0	8.2
Diffuse, %	22	15	28	26	30	49	35	56	50	49	24	37	36
Overlap type	SLE 23%	SLE 12%	NA	NA	NA	NA	NA	NA	NA	RA 36% SLE 54% Myositis 23% Sjogren’s 2%	RA 37% SLE 19% Myositis 16% Sjogren’s 38%	RA 24% SLE 19% Myositis 25% Sjogren’s 59%	NA
ACA, %	46	55	16	42	17	23	21	17	27	19	45	36	37
Scl70, %	26	16	47	35	24	35	15	47	27	39	14	15	33
U1RNP, %	32	12	NA	NA	NA	NA	NA	17	5	NA	NA	NA	NA
RNAP3, %	NA	NA	8	5	NA	NA	NA	NA	NA	3	13	14	5
Ro, %	22	12	NA	NA	NA	NA	NA	36	10	NA	NA	NA	NA
RP, %	97	98	NA	NA	94	91	96	100	99	83	98	97	95
Digital ulcers, %	53	39	28	37	30	26	35	36	55	22	45	60	31
Joints, %	76	76	NA	NA	NA	NA	NA	69	86	17	26	50	15
Calcinosis, %	12	26	NA	NA	21	9	29	NA	NA	14	38	42	NA
Lung, %	42	34	44 g	23	46	41	31	31	33	50	27	43	30
Suspected PH, n (%)	26/81 (32)	78/262 (30)	18	12	NA	NA	NA	11/36 (31)	70/300 (25)	NA	NA	NA	NA
PAH on RHC, n (%)	4/8 (50)	16/26 (62)	NA	NA	19/70 (27)	25/80 (31)	198/745 (27)	NA	NA	40/377 (11)	161/1569 (10)	73/1604 (5)	NA
Upper GI, %	87	85	55	63	84	69	88	78 (both upper and lower GI)	94 (both upper and lower GI)	59	88	89	NA
Lower GI, %	8	6	25	25	NA	NA	NA			Dysmotility 52% Malabsorption 16%	Dysmotility: 5% Malabsorption 5%	Dysmotility: 45% Malabsorption 20%	Malabsorption 1%
SRC, %	5	3	2	2	9	1	6	0	7	5	3	5	2
Mortality	9-year survival: 52%	9-year survival: 76%	10-year survival: 92%	10-year survival 94%	10-year survival 78%	10-year survival 75%	10-year survival 78%	NA	NA	Median survival: Incident 2.1 years Prevalent 10.9 years	Median survival: Incident 5.9 years Prevalent 15.7 years	Median survival: Incident 3.7 years Prevalent 14.3 years	Median survival: Incident 3.8 years Prevalent 12.1 years
Cause of death	PH, infection, multi-organ failure	ILD, PH, Infection	NA	NA	NA	NA	NA	NA	NA	PAH, ILD	PAH, ILD	PAH, ILD	PAH, ILD
Association with mortality	(+) Asian, older age at diagnosis, male, DcSSc, suspected PH, ILD, malabsorption		NA		NA			NA		(+) ILD and PAH
Summary of findings	Asians versus Caucasians: younger at diagnosis; More frequent DcSSc and SSc-SLE overlap, U1RNP, ILD and digital ulcers; Less frequent ACA and calcinosis; Increased mortality in Asians even after adjustment for co-variates		Asians versus Caucasians: younger at disease onset; More frequent Scl70, rapid disease onset, suspected PH, impaired FVC and DLCO; Less frequent ACA and digital ulcers; Increased mortality in Asians (disappeared on matching by centre and poor prognostic factors)		Asians versus Caucasians: More frequent DcSSc; Less frequent calcinosis, upper GI, sclerodactyly and telangiectasia; No difference in short-term survival across racial groups in unadjusted or adjusted analyses.			Asians versus Caucas-ians: Older at diagnosis; More frequent Scl70, U1RNP and Ro; Less frequent joint, GI and digital ulcers		SCORE (predominantly Asians) versus other cohorts (predominantly Caucasians): More frequent DcSSc, SSc-SLE overlap, Scl70, ILD; Less frequent ACA, Raynaud’s, digital ulcers and calcinosis; Increased mortality in Asians.

SSc: systemic sclerosis; EUSTAR: European League Against Rheumatism Scleroderma Trials and Research; ASCS: Australian Scleroderma Cohort Study; CSRG: Canadian Scleroderma Research Group; SCORE: Scleroderma Cohort in Singapore; RP: Raynaud’s phenomenon; SLE: systemic lupus erythematosus; RA: rheumatoid arthritis; ACA: anti-centromere antibody; U1RNP: U1-ribonuceloprotein; RNAP3: RNA polymerase 3; PH: pulmonary hypertension; PAH: pulmonary arterial hypertension; RHC: right heart catheterization; GI: gastrointestinal; SRC: scleroderma renal crisis; ILD: interstitial lung disease; DcSSc: diffuse cutaneous systemic sclerosis; FVC: forced vital capacity; DLCO: diffusing capacity of lungs for carbon monoxide; (+): positive association.

^aAsian includes Filipino, Chinese, Indian, Japanese, Vietnamese, Korean, Pacific islander.

^bSouth Asian includes Bangladeshi, Nepalese, Indian, Pakistani, Sri Lankan.

^cEast Asian includes Chinese, Japanese, Korean, Filipino, and Thai.

^dMean age at diagnosis.

^eMean age at onset.

^fMean age at enrolment.

^gRestrictive lung forced vital capacity < 80% predicted.

Bold denotes statistically significant comparisons within studies.

Autoantibodies

Several studies have suggested that the prevalence of SSc-related autoantibodies varies in different racial gro-ups.^{2–4,6–9,11–13} SSc cohort studies examining the autoantibody profile in Asian SSc patients living in Asian countries, including Singapore, China, Taiwan, Thailand, Korea and India, consistently showed that Asian patients had a higher prevalence (30%–80%) of anti-Scl70 antibody, with the highest frequency seen in Thai patients.^{6–9,11–13} The finding of higher frequency of anti-Scl70 antibody in Asian SSc patients also persisted in multi-racial cohorts comparing Caucasian and Asian SSc patients in the United States, Europe, Australia and Canada (24%–47% in Asians compared to 14%–35% in Caucasians).^2–4,28,29 The presence of anti-Scl70 antibody was associated with DcSSc and ILD, which were predictors of poorer survival.^7,9,13

Asian SSc patients also have a lower frequency of anti-centromere antibody (16%–46%) compared to Caucasian patients (21%–55%).^2–4,28,29 SSc patients who have anti-centromere antibody were more likely to have limited skin involvement and pulmonary arterial hypertension (PAH).^5,30

In two studies of Asian SSc patients living in the United States and Canada, the frequency of anti-U1-Ribonucleoprotein (RNP) antibody was higher among Asians (32% and 17%) compared to their Caucasian counterparts (12% and 5%, respectively).^2,29 Cohort studies of Asians living within Asia also suggested similarly high prevalence of anti-U1-RNP (24%–33%).^5,7,8,31,32 Anti-U1-RNP is often found in association with mixed connective tissue disease or overlap syndromes and may portend a more favourable outcome in Western cohorts. ³³ Systemic lupus erythematosus is more commonly seen in overlap with SSc among Asians compared to Caucasians (19%–54% vs 12%–19%, respectively), whereas Sjogren’s syndrome is more often seen in overlap among Caucasians compared to Asians (38%–59% vs 2%, respectively).^{2,4,5,7,11,29}

In the Singapore SSc cohort (SCORE) and a Canadian multi-racial cohort, Asian patients with SSc had higher frequency of anti-Ro antibody.^29,32 Anti-Ro antibody in patients with SSc was associated with increased frequency of sicca or Sjögren’s syndrome, and polymyositis. ²⁹ Anti-Ro antibody was also negatively associated with poor survival and progression to severe lung disease. ²⁹

A lower frequency of anti-RNA polymerase 3 (RNAP3) was observed in Asian (1.3%–4%) compared to Caucasian SSc patients (9%–17%). As RNAP3 antibody is strongly associated with scleroderma renal crisis (SRC), this may partially explain the lower observed prevalence of SRC among Asians.^31,34,35

Disease onset

Compared to Caucasian patients with SSc, Asian patients have a younger age of SSc onset and diagnosis. In a United States multi-racial SSc cohort, the mean age at SSc diagnosis for Asian patients was significantly younger (51.7 years) compared to Caucasian patients (58.9 years). ² This finding was similarly observed in the European Scleroderma Trials and Research (EUSTAR) cohort (38.1 years vs 44.2 years, respectively). ³ SSc studies conducted in China, Singapore, Japan, Korea, India and Iran showed that Asian SSc patients were younger by 1–2 decades compared with Caucasian SSc patients.^{5–8,11,13,14,30}

In the EUSTAR cohort, Asian patients with SSc developed Raynaud’s phenomenon at a significantly younger age compared to Caucasian patients (38.1 years vs 44.2 years, respectively). ³ After Raynaud’s onset, Asian patients also evolved the first non-Raynaud’s feature faster than Caucasian patients but slower than black patients. ³ Two years after Raynaud’s onset, 74% of Asian patients compared with 66% of Caucasian patients, had experienced their first non-Raynaud’s feature. ³ This is an interesting observation but it is uncertain whether this truly represents the rapidity of disease evolution or it may be confounded by pigmented skin masking the manifestation of overt Raynaud’s phenomenon.

Skin involvement

Diffuse skin involvement in SSc is more frequently seen in Asian patients compared to Caucasian patients as reported by multi-racial cohorts.^2,4,28,29 SSc studies in Asia, however, have reported differences in various regions of Asia. DcSSc appears to be more predominant in Thailand, China, Singapore and India.^7,9–11,13 In contrast, studies in Japan, Korea and Iran have reported a higher proportion of LcSSc.^5,6,14

In multi-racial cohort studies comparing skin scores, Asian patients had less severe or similar peak skin scores compared to their Caucasian counterparts.^3,4 Peak modified Rodnan skin score (mRSS) in DcSSc patients in the predominantly Chinese Singapore cohort (mRSS 17) was similar to the European cohort (mRSS 17), but less than the Australian (mRSS 23) and Canadian (mRSS 21) cohorts. ⁴ In the EUSTAR multi-racial cohort, Asian patients (mRSS 10) also tended to have less severe skin sclerosis than Caucasian patients (mRSS 12), and this appeared to be largely independent of geographical location. ³ In a cohort study from China, 88% of patients had mRSS less than 16. ⁸ These findings, together with differences in the rapidity of onset of disease discussed earlier, may have implications on multicentre therapeutic clinical trial study design targeted at the recruitment of early SSc patients from different racial groups.

Raynaud’s and digital ulcers

Raynaud’s phenomenon is one of the most frequent disease manifestations of SSc. Frequency of Raynaud’s phenomenon in SSc patients in Asia was comparatively lower (60%–90%) than Western cohorts that typically reported frequencies of more than 95%.^4,7,9,11,28

In the multi-racial cohort studies comparing Caucasian and Asian SSc patients in Europe, Canada and Australia, higher frequency of digital ulceration was also seen in Caucasian patients compared to Asian patients with SSc, consistent with worse vasculopathy seen in Caucasians.^3,4,29 In the United States multi-racial cohort study, however, the frequency of digital ulceration was higher in Asian compared to Caucasian SSc patients (53% vs 39%, respectively), with frequency of Raynaud’s phenomenon greater than 95% in both racial groups. ²

Pulmonary

ILD is an important cause of morbidity and mortality in SSc. The prevalence of ILD in patients with SSc ranges from 20% to 40% in Caucasians and 40% to 88% in Asi-ans.^{2–9,11,13,14,36} A recent meta-analysis of 27 studies from East Asia (Japan, China and Korea) reported that the pooled prevalence of SSc-ILD was 56%. ³⁷ A significantly higher SSc-ILD prevalence was reported in China (72%) compared to Korea (51%) and Japan (46%). ³⁷ ILD is reported to be the most common organ involvement in Chinese patients, and findings from a Chinese study reported that the proportion of Chinese patients with SSc-ILD was much higher (83% in DcSSc, 57% in LcSSc) than that in the Caucasian population (53.4% in DcSSc, 34.7% in LcSSc).^7,38 Outside of Asia, multi-racial SSc studies have also shown a higher prevalence of SSc-ILD in Asian SSc patients compared to Caucasian patients.^2–4

As SSc-ILD has been associated with the DcSSc subtype and anti-Scl70, the higher prevalence of DcSSc and anti-Scl70 in Asian SSc patients is consistent with the higher frequency of SSc-ILD seen in Asian patients.^9,37 Other risk factors implicated in SSc-ILD include longer disease duration, negative anti-centromere antibody, negative anti-U3-RNP antibody and higher erythrocyte sedimentation rate. ³⁷

While right heart catheterization (RHC) is the gold standard for the diagnosis of pulmonary hypertension (PH), a number of SSc studies reporting on PH are based on echocardiographic findings. In the EUSTAR cohort, Asian SSc patients had higher prevalence of suspected PH on echocardiography defined as systolic pulmonary arterial pressure (PASP) ⩾ 40 mmHg, compared to Caucasian patients (17.9% vs 11.6%, respectively). ³ Asian SSc patients were also more likely to have single-breath diffusing capacity for carbon monoxide < 80% of predicted, compared to Caucasian SSc patients. ³ The proportion of patients with suspected PH on echocardiography was higher in the Singapore SCORE cohort (30.4%) compared to that in the EUSTAR cohort (21.1%), the United Kingdom (18.3%) and the United States (13%).^11,38–40

A systematic review and meta-analysis looking mainly at Western countries reported that the overall prevalence of PAH by RHC was 6.4% and the overall PAH incidence was 18.2 cases per 1000 person-years. ⁴¹ This meta-analysis included one Chinese study by Huang et al. which showed a higher PAH prevalence of 15.1% and incidence of 54.3 per 1000 person-years.^41,42 SSc studies from Singapore and Thailand have also reported higher prevalence of RHC confirmed PAH (10.6% and 18.0%, respectively).^4,10 In multi-racial United States and Canadian cohorts, there were however no differences in PAH prevalence between Caucasian and Asian patients.^2,28 It is important to highlight that comparisons of PAH prevalence between cohorts should be interpreted in the context of the centre’s screening strategy. As RHC is an invasive procedure, the prevalence within each cohort is dependent on the screening strategy and PAH definitions. A lower echocardiographic screening threshold of PASP > 36 mmHg used in the study by Huang et al. would be more sensitive and potentially identify more patients with RHC confirmed PAH, compared to the Singapore cohort that used a threshold of 40 mmHg. The Thai cohort defined PH on RHC using the revised criteria of mean pulmonary arterial pressure > 20 mmHg, potentially identifying more patients with PAH. ¹⁰

Patients with PH-ILD have worse 5-year survival (54%) than those with isolated PAH (82%). ⁴³ Condliffe et al. ⁴⁴ reported that SSc patients with concomitant PH-ILD compared to isolated PAH were more likely to be non-Caucasians. The proportion of patients with suspected PH-ILD was substantial in the Singapore SCORE cohort (23%) and Chinese cohort from Beijing (14%).^8,45 In the Singapore SCORE cohort, SSc patients with PH-ILD had the highest risk of death (hazard ratio (HR) 3.8), followed by PAH (HR 3.0) and ILD (HR 1.8). ⁴⁵

Cardiac

Cardiac involvement is common in SSc, with an estimated clinical prevalence of 15%–35%. ⁴⁶ In the majority of SSc patients, however, cardiac disease often remains subclinical. ⁴⁷ In a meta-analysis of nine studies from Western countries, cardiac involvement was observed in 29% of SSc patients and was the most frequent cause of death. ⁴⁸ Due to heterogeneity in the definition of cardiac involvement in SSc, comparison across cohorts is challenging. Cardiac manifestations can include that of heart failure, myocarditis, pericarditis, arrhythmia and valvular impairment. ⁴⁷

Multi-racial cohort studies in the United States and Canada suggest no difference in the prevalence of cardiac involvement between Caucasian and Asian patients.^2,29 Interestingly, a meta-cohort study comparing SSc registries from Western and Asian countries showed that the prevalence of myocardial involvement was highest in SSc patients from Singapore (49%) followed by Canada (45%) and Australia (7.1%). ⁴ However, as the study was a meta-cohort analysis of multiple registry data, the authors acknowledged the difficulty in harmonizing variables particularly for myocardial involvement due to definitions and attribution to SSc that may confound the interpretation of prevalence across registries. SSc cohort studies conducted in Asian countries also revealed varying prevalence of cardiac involvement from 11% in Korean patients to 61% in a Chinese cohort.^6,8

Gastrointestinal

Gastrointestinal involvement is widely reported to occur in up to 90% of SSc patients based on Western cohorts. The spectrum of disease includes oesophageal dysmotility, gastroesophageal reflux disease, gastric antral vascular ectasia, malabsorption, small intestinal bacterial overgrowth and faecal incontinence.^49,50

Oesophageal involvement is frequently encountered in SSc gastrointestinal disease.^4,14,28 In a multi-racial Canadian cohort, Al-Sheikh et al. ²⁸ reported that East Asians had less frequent oesophageal dysmotility compared to Caucasian patients (69% vs 88%, respectively). In the multi-racial meta-cohort study by Proudman et al. and cohorts from Asia (Thailand, Singapore, Japan and Korea), lower prevalence of gastrointestinal involvement (33%–59%) compared to the EUSTAR (67%) and the United States (71%) cohorts was similarly reported.^{4–6,9,11,38,39}

In the Singapore SCORE cohort, although the prevalence of gastrointestinal involvement was lower than Western cohorts, malabsorption was relatively frequent (16%) compared to Caucasian patients in Canada (20%) and Australia (5.4%). ⁴ Parenteral nutrition, a surrogate for patients with gastrointestinal involvement and malnutrition, was an independent predictor of mortality. ¹¹

Renal crisis

SRC confers high risk of morbidity and mortality. ⁵¹ A recent meta-analysis of studies from 1983 to 2011 reported that the overall prevalence of SRC was 4%, of which 7%–9% occurred in patients with DcSSc, and 0.5%–0.6% in LcSSc patients. ⁵² The highest rates of SRC were reported in Australia, the United States and Europe. Risk factors for SRC included male sex, DcSSc, early SSc, RNAP3 antibody, high skin scores, rapidly progressive ILD or cardiac involvement, tendon friction rubs and steroid use. ⁵²

The prevalence of SRC in Asia is lower, ranging from 2% to 3% among the Chinese, Thai and Japanese patients.^5,7,53 In the Thai cohort, digital gangrene, current prednisolone dose ⩾ 15 mg/day, serum albumin < 3 mg/dL and cardiac involvement were independent risk factors for SRC. ⁵³ The lower prevalence of SRC may also be associated with lower prevalence of RNAP3 antibody among the Chinese and Singapore SSc patients.^31,32

Mortality

SSc carries significant mortality risks. A meta-analysis in 2014 reported a standardized mortality ratio (SMR) of 2.72 compared to the general population. ⁵⁴ Cumulative survival was estimated to be 75% at 5 years and 63% at 10 years. ⁵⁴ Most studies report survival rates based on prevalent cohorts which is likely an overestimation of the true survival rates due to survival bias. Nevertheless, survival of patients with SSc has gradually improved over the last decades, likely a result of earlier diagnosis and advances in therapeutic management of SSc. ⁵⁵

Multi-racial cohort studies found that Asian patients with SSc had lower survival rates compared to Caucasian patients.^2–4 However, whether this is due to racial differences alone or attributable to environmental and/or socioeconomic factors remain undetermined. In a United States SSc cohort, the 9-year survival for Asian patients was significantly lower than Caucasian patients (52.3% vs 75.8%, respectively) even after adjustment for co-variates. ² Likewise, in the EUSTAR cohort, Asian patients with SSc also had a higher 10-year mortality compared to Caucasian patients (7.6% vs 6.5%, respectively). ³ However, when centre- and individual-matched analyses for poor prognostic indicators were done, the observed increased mortality was no longer significant. This suggests that environmental, socioeconomic factors or differences in the healthcare system likely contributed to the increased mortality. ³ Indeed, these findings were corroborated by a Canadian multi-racial study showing no difference in survival between Asians and Caucasians. ²⁸ This study occurred in a publicly funded, universal healthcare system where all citizens and permanent residents were eligible for public health insurance. In the meta-cohort study of Western and Asian countries, survival for the prevalent and incident cohorts was reported. In the prevalent cohort, lowest median survival was seen in the Singapore cohort (10.9 years), compared to Australia (15.7 years), Canada (14.3 years) and Europe (12.1 years). ⁴ Particularly striking was the low median survival in the incident cohorts even among the Western countries (3.7 years to 5.9 years) compared to the Singapore cohort (2.1 years). ⁴ SSc-related complications accounted for more than 50% of known causes of death in all four cohorts, with PAH and ILD being the most common. ⁴

SSc cohort studies conducted in Asia have also reported elevated SMR of 2.8 and 3.2 in Japan and Taiwan, respectively.^5,12 However, in contrast to the multi-racial cohort studies above, 10-year survival of Japanese patients with SSc was slightly higher (88%) than that reported in the EUSTAR Caucasian SSc population (84%).^5,55 A study from Taiwan reported a lower mortality rate of 43 per 1000 patient-years than the EUSTAR (63 per 1000 patient-years), with male sex, older age, cancer and end-stage renal disease as predictors of mortality. ¹²

In a United States cohort, risk factors associated with higher mortality included older age at SSc diagnosis, male sex, diffuse cutaneous disease, suspected PH on echocardiogram, ILD, and malabsorption syndrome. ² Asian patients were most likely to die from PH and infections, and Caucasian patients from ILD. ² In the Singapore SCORE cohort, 57% of deaths were attributed to SSc, with PAH, ILD and gastrointestinal complications as the leading causes of death. ¹¹ Multivariate analysis of the SCORE cohort showed that smoking, SSc-overlap, baseline renal involvement, PASP ⩾ 40 mmHg on echocardiography, treatment for peripheral vasculopathy and parenteral nutrition were independent predictors of mortality. ¹¹

The differences in survival and mortality can be attributable to racial, geographical and social factors as these factors may impact on SSc disease course and mortality. Geographical background plays a role as population mortality varies in different regions and continents. ³ Access to healthcare services, treatment availability and cost, as well as racial and cultural variations in health-seeking behaviour and treatment adherence, likely contribute to differences in mortality.

Conclusion

In this review, we have examined and compared the differences in clinical characteristics, autoantibody profiles and outcomes between Asian and Caucasian SSc patients. Asian SSc patients have distinct genetic susceptibility to SSc, different antibody profile such as higher frequency of anti-Scl70 and anti-U1-RNP antibodies, younger age and rapidity of disease onset, more severe clinical phenotype and suggestion of poorer survival that is attributable to multiple contributing factors. Presence of anti-Scl70 antibody was associated with DcSSc and ILD, which were risk factors for higher mortality. Our review contributes to the existing literature that race exerts significant effects on the manifestations and outcomes of SSc. Although the influence of race in SSc is complex, understanding its impact on the disease is of increasing importance as we move into the realm of personalized medicine.