Ethnicity and environmental factors could be involved in the heterogeneity of systemic lupus erythematosus (SLE). We conducted this study to define clinical and serologic correlations and autoantibody clusters in SLE patients in South China. We retrospectively reviewed the records of 917 patients with SLE admitted to our hospital between January 2005 and June 2008.
We found the following associations between autoantibodies and clinical manifestations to be statistically significant: anti-double-stranded DNA (anti-dsDNA) with higher prevalence of renal disorder, leukopenia, and anemia; anti-Sm with higher prevalence of malar rash/discoid rash, pericarditis, and leukopenia; anti-ribonucleoprotein (anti-RNP) with higher prevalence of Raynaud phenomenon and photosensitivity; anti-deoxyribonucleoprotein (anti-DNP) with higher prevalence of arthritis and lower prevalence of renal disorder; anti-Scl-70 with higher prevalence of anemia and Raynaud phenomenon; anti-Jo-1 with higher prevalence of pericarditis; and anti-centromere with higher prevalence of Raynaud phenomenon. Three autoantibody clusters were identified: Cluster 1 (anti-Ro, anti-Sm, and anti-RNP [Ro/Sm/RNP], with a significantly lower percentage of elderly SLE and higher prevalence of photosensitivity, malar rash/discoid rash, Raynaud phenomenon, and leukopenia); Cluster 2 (anti-Ro [Ro], with a lower percentage of pediatric SLE); and Cluster 3 (the absence of anti-extractable nuclear antigen antibodies [ENA ve], with a lower percentage of adult SLE and lower prevalence of alopecia).
In summary, this study not only confirms both anti-dsDNA and anti-Sm as specific markers for classifying SLE, but also demonstrates that photosensitivity is not associated with anti-Ro but with anti-RNP, and a negative association is found between renal disorder and anti-DNP in patients in South China. These results are different from results found in other populations. The higher prevalence of anti-dsDNA and renal disorder results in less difference in the prevalence of anti-dsDNA and renal disorder among the 3 autoantibody clusters in SLE patients in South China, which could be related to ethnicity and widespread industrial pollution in South China.
Abbreviations: ANA = anti-nuclear antibody, anti-DNP = anti-deoxyribonucleoprotein, anti-dsDNA = anti-double-stranded DNA, anti-RNP = anti-ribonucleoprotein, CI = confidence interval, ENA = extractable nuclear antigen antibodies, ENA ve = the absence of anti-extractable nuclear antigen antibodies (Venezuelan cluster), OR= odds ratio, RP = Raynaud phenomenon, SLE = systemic lupus erythematosus.
From Department of Dermatology (XT, XZ), The First Affiliated Hospital, Sun Yat-sen University, Guangzhou, Guangdong; and Zhongshan School of Medicine (YH, WD, LT, WW); and Sun Yat-sen University, Guangzhou, Guangdong, China.
Received August 24, 2009, and in revised form October 24, 2009.
Accepted for publication November 16, 2009.
Reprints: Xingqi Zhang, MD, PhD, Department of Dermatology, The First Affiliated Hospital, Sun Yat-sen University, Guangzhou, Guangdong, China (e-mail: email@example.com).
Systemic lupus erythematosus (SLE) is an autoimmune disease characterized by a diverse autoantibody profile and clinical manifestations. It has been reported that certain autoantibodies are associated with a certain spectrum of clinical manifestations, such as anti-double-stranded DNA (anti-dsDNA) and renal disorder,24 anti-Ro and photosensitivity,9,14 anti-ribonucleoprotein (anti-RNP) and Raynaud phenomenon.23 Autoantibody clustering can help us to differentiate between various subsets of SLE, allowing the prediction of subsequent disease course and organ damage.25
Ethnicity and environmental factors such as occupation, pollutants, socioeconomic status, and behavior can affect the course and outcome of SLE.18 It was estimated that the prevalence of SLE in workers in a textile factory in Shanghai in China was 70.41/100,000 population,6 which was higher than in other populations.27 We conducted the current study to determine if, compared to reported data, there exist any specific clinical and serologic correlations and autoantibody clusters in SLE patients from Guangdong (Canton) province. This province is one of the most economically developed provinces in South China, with widespread manufacturing factories, such as electronics and textile factories, with industrial dust from silica and synthetic fiber that could function like immunologic adjuvant such as incomplete Freund's adjuvant in the pathogenesis of SLE.3
PATIENTS AND METHODS
The patients included in this study were those admitted to The First Affiliated Hospital, Sun Yat-sen University, Guangzhou, China, from January 2005 to June 2008. All patients fulfilled 4 or more classification criteria for the diagnosis of SLE as defined by the American College of Rheumatology, that is, the criteria revised by American Rheumatology Association in 1982.21
All data reported in this retrospective cohort study were based on hospital records. We collected the following data: sex; age; overlapped autoimmune diseases; concomitant diseases, such as tuberculosis, hypertension, and diabetes; thyroid disorder; and familial SLE. The main clinical manifestations were defined based on previous studies.1,4 Anti-extractable nuclear antigen antibodies (anti-ENA) were qualified with immunoblotting. Anti-deoxyribonucleoprotein (anti-DNP) was detected with a semi-quantitative latex slide test. Anti-nuclear antibody (ANA) and anti-dsDNA were quantitated with enzyme-linked immunosorbent assay. The first values of these tests were used in the study. Data from investigations on the prevalence of hepatitis B virus surface antigen assayed in 765 cases and anti-hepatitis C virus antibody assayed in 748 cases were also included in the study.
A descriptive analysis was performed using SPSS software version 13.0 to analyze the frequency of overlapped rheumatic diseases; concomitant diseases such as tuberculosis, hypertension, and diabetes; familial SLE; positive serum hepatitis B surface antigen; and anti-hepatitis C virus antibody. Associations between autoantibodies and clinical manifestations were analyzed by chi-square tests. Odds ratios (ORs) and 95% confidence intervals (CIs) were estimated using logistic regression with adjustment for potential confounders. The OR was calculated for assessing the risk of appearance of each variable. A 95% CI lower limit of >1.0 was taken to indicate statistical significance in the case of positive association, and an upper limit <1.0 in the case of negative association. To identify subsets of SLE patients with similar autoantibody patterns, 4 anti-ENA antibodies (anti-Sm, anti-Ro, anti-La, and anti-RNP) were selected for cluster analysis using the K-means cluster analysis (non-hierarchical clustering) procedure.22 The p value was 2-tailed, and a p value < 0.05 was considered significant.
General Characteristics of the Study Population
The study cohort comprised 917 patients with SLE: 788 female patients (85.9%) and 129 male patients (14.1%). The female: male ratio (F:M) was 6.11:1. The number of cases in each of 3 different age-groups was 81 in the pediatric SLE group (aged <15 yr), 725 in the adult SLE group (aged 15-49 yr), and 111 in the elderly SLE group (aged >49 yr). The F:M ratio was 5.75:1 in pediatric SLE, 7.33:1 in adult SLE, and 2.70:1 in elderly SLE. Overlapped autoimmune diseases were 13 cases with Graves disease (1.4%), 11 Hashimoto thyroiditis (1.2%), 14 rheumatoid arthritis (1.5%), 11 Sjögren syndrome (1.3%), 4 scleroderma (0.4%), 4 polymyositis (0.4%), and 3 with dermatomyositis (0.3%). Concomitant diseases included 47 cases with hypertension (5.0%) and 15 with tuberculosis (2.0%). Sixteen cases (1.7%) had a familial history of SLE. The prevalence of hepatitis B surface antigen in 765 cases assayed was 3.7%, and the prevalence of anti-hepatitis C virus antibody in 748 cases assayed was 1.2%.
Relationship Between Autoantibodies and Clinical Manifestations
Anti-dsDNA was found in 693 of the 917 patients (75.6%) and was associated with a higher prevalence of renal disorder (p = 0.006; OR, 1.561; 95% CI, 1.129-2.160), leukopenia (p < 0.001; OR, 2.136; 95% CI, 1.471-3.100), and anemia (p < 0.001; OR, 1.741; 95% CI, 1.272-2.382) (Table 1). Anti-Sm was found in 278 patients (30.3%), with a higher prevalence of malar rash/discoid rash (p = 0.003; OR, 1.774; 95% CI, 1.215-2.590), pericarditis (p = 0.002; OR, 1.875; 95% CI, 1.231-2.857), and leukopenia (p = 0.001; OR, 1.564; 95% CI, 1.156-2.116). Anti-RNP was present in 425 patients (46.3%), with a significantly increased prevalence of Raynaud phenomenon (p = 0.001; OR, 3.119; 95% CI, 1.531-6.355) and photosensitivity (p = 0.003; OR, 2.004; 95% CI, 1.264-3.178). Anti-DNP was found in 116 patients (12.6%), with a higher prevalence of arthritis (p = 0.024; OR, 1.585; 95% CI, 1.069-2.350) and a lower prevalence of renal disorder (p = 0.028; OR, 0.639; 95% CI, 0.424-0.965). Twenty-five patients (2.7%) were found to be anti-SCL-70 positive, with a higher prevalence of anemia (p = 0.017; OR, 4.015; 95% CI, 1.184-13.617) and Raynaud phenomenon (p = 0.004; OR, 4.201; 95% CI, 1.345-13.123). Twenty-six patients (2.8%) were found to be anti-Jo-1 positive, with a higher prevalence of pericarditis (p < 0.001; OR, 4.284; 95% CI, 1.839-9.982), and only 17 patients (1.9%) were found to be anti-centromere positive, with a higher prevalence of Raynaud phenomenon (p = 0.007; OR, 4.513; 95% CI, 1.216-16.746).
Autoantibody Clusters and Related Demographic, Serologic, and Clinical Features
Defining subsets of SLE patients with distinct characteristics is important for clinical management and prognosis. To be meaningful clinically, the prevalence of the autoantibody should be statistically different between clusters.25 When we analyzed anti-dsDNA and anti-ENA antibodies together, we could not differentiate the prevalence of anti-dsDNA antibody between clusters. By clustering 4 anti-ENA antibodies (anti-Ro, anti-La, anti-Sm, and anti-RNP) into 3 clusters, we identified 3 distinct groups of patients with distinctly different autoantibody profiles: Cluster 1 (anti-Ro, anti-Sm, and anti-RNP [Ro/Sm/RNP]), Cluster 2 (anti-Ro [Ro]), and Cluster 3 (the absence of ENA antibodies, ENA Venezuelan cluster [ENA ve]). In this group of 917 patients, 350 patients were assigned to Cluster 1, 333 patients were assigned to Cluster 2, and the other 234 patients were assigned to Cluster 3. The frequency of demographic, clinical, and serologic features in each cluster is shown in Table 2. Patients in Cluster 1, when compared with those in Cluster 2 and Cluster 3, had a significantly higher prevalence of photosensitivity, and when compared with those in Cluster 3, had a significantly higher prevalence of malar rash/discoid rash, Raynaud phenomenon, and leukopenia, and a lower percentage of elderly SLE patients. Patients in Cluster 2 had a significantly lower percentage of pediatric SLE patients, and a higher prevalence of anti-DNP when compared with those in Cluster 1 or Cluster 2. Patients in Cluster 3 had a significantly lower percentage of adult SLE patients, and a lower prevalence of alopecia and ANA when compared with those in Cluster 1 and (or) Cluster 2.
Coexistence of Other Autoimmune Diseases and Hepatitis Viral Infection in SLE
Similar to a European study,4 in the current study the most common overlapped autoimmune disease in SLE patients was autoimmune thyroid disease. The reported prevalence of Hashimoto thyroiditis and Graves disease in Hungarian SLE patients is 2.3% and 2.9%, respectively,2 slightly higher than we found in the current study (1.2% and 1.4%, respectively). Autoimmune thyroid disease may have been underestimated in retrospective study, because a prospective study performed in Portuguese patients revealed that the prevalence of subclinical hypothyroidism and clinical hypothyroidism in SLE was 11% and 13%, respectively.26 It is noteworthy that the prevalence of autoimmune thyroid disease in patients with other autoimmune diseases was also higher when compared with the prevalence in the general population, that is, patients with systemic sclerosis, polymyositis-dermatomyositis, Sjögren syndrome, and rheumatic arthritis.2 Therefore, it may be necessary to screen for autoimmune thyroid disease in patients with SLE and other autoimmune diseases.
Researchers in Israel reported an unexpectedly lower percentage of core antibodies of hepatitis B in patients with SLE compared with healthy matched donors (2.5% vs. 10.7%, respectively).19 Therefore, hepatitis B virus could have a protective role in SLE, contrary to the ubiquitous and pathogenetic Epstein-Barr virus. This was similar to our data: the positive finding of hepatitis B virus surface antigen was much lower in our SLE patients compared with that in the general population aged over 3 years old in China8 (3.7% vs. 9.07%, respectively). The prevalence of hepatitis C virus was also slightly lower in our study compared with that in the general population in China (1.2% vs. 1.55%, respectively).29 This suggests that hepatitis C virus could also have a protective role in SLE.
Comparing SLE Patients in the Present Report and Previous Large Cohort Studies
The female: male ratio in the current study (6.1:1) was similar to the ratio reported previously in white patients (6.6:1), but lower than that in other large cohort studies4,10,13,17 (average, 9.4:1) (Table 3). The prevalence of malar rash, photosensitivity, oral ulcer, arthritis, and neurologic disorder recorded in other studies (average, 53.6%, 44.3%, 36.0%, 81.9% and 28.2%, respectively) was higher than that found in our study (13.5%, 9.2%, 9.6%, 47.1%, and 15.6%, respectively). The prevalence of discoid rash in the current study (5.7%) was lower than that found in Latin American, African American, and white American patients (average, 14.2%). The prevalence of renal disorder in our study (73.2%) was similar to that found in India (73%), but higher than that found in other previous large cohort studies (average, 44.3%).
As for autoantibody profile, the prevalence of anti-Ro in our study (66%) was similar to that found in Hong Kong (57%), but almost 2-fold higher than that found in other previous large cohort studies (average, 35.6%). It has been shown that residents who were exposed to oil field waste such as mercury had an increased prevalence of SLE with a higher prevalence of neurologic symptoms, respiratory symptoms, and several cardiovascular problems including stroke and angina.5 Guangdong is an economically developed region in China with many industrial zones, which may be related to the higher prevalence of anti-dsDNA antibody and renal disorder; the lower prevalence of malar rash, photosensitivity, oral ulcer, and arthritis; and the lower female: male ratio, compared to previous studies carried out in Anhui province of China16 and Hong Kong,12 where manufacturing plants are far fewer. A similar situation was discovered in an epidemiologic study6 of a textile factory in Shanghai province in 1982, when a much higher prevalence of SLE was found (70.41/100,000 population, compared to 11.67/100,000 in the general population in Guangzhou28). Among these SLE patients, 7 of 8 had renal disorder. However, further concurrent studies are needed in those different districts to confirm the linkage between industrial pollution and the above special features of SLE.
Clinical and Serologic Correlations in SLE
In the current study, we found that anti-Sm was significantly associated with a higher prevalence of a set of clinical manifestations, that is, malar rash/discoid rash, leukopenia, and pericarditis. The result was similar to another study which revealed that anti-Sm antibody was a highly specific marker of SLE and was related to active SLE.16 Anti-RNP was associated with Raynaud phenomenon, which was confirmed in our study. No significant association between anti-Ro and photosensitivity was found in our study, which could be related to the lower prevalence of photosensitivity and the higher prevalence of anti-Ro in Chinese SLE patients;11,16 however, anti-RNP antibody was found to be significantly associated with photosensitivity, which was rarely reported previously. Anti-dsDNA as a specific marker in SLE was associated with disease activity and renal disorder, which was confirmed in the current study. Anti-Jo1 was reported to be associated with interstitial lung disease, myositis, polyarthritis, fever, Raynaud phenomenon, and mechanic hands.7 In this study, we found anti-Jo 1 to be associated with pericarditis but not Raynaud phenomenon and other features, which could be because only SLE patients were included in our study. Anti-DNP was found to be positively associated with arthritis, which was rarely reported previously,15 and negatively associated with renal disorder, which was contrary to a previous study.20 The current study confirms most previously reported associations between autoantibodies and clinical manifestations in SLE patients, but also presents some peculiar associations in patients in South China.
Cluster Analysis of Autoantibody in SLE
SLE is an entity with a widely varied phenotype clinically and serologically. In the current study, we classified SLE patients into 3 autoantibody clusters: Ro/Sm/RNP, Ro, and ENA ve, which was different from previous studies.22,25 Patients in the Ro/Sm/RNP cluster were less likely to be elderly and had more cutaneous manifestations of SLE, such as malar rash/discoid rash, photosensitivity, and Raynaud phenomenon, which was similar to patients in the Sm/RNP cluster with absence of anti-Ro reported previously.25 Moreover, we did not find the phenomenon that prevalence of renal disorder was increased in the absence of ENA antibody, which was different from previous reports in other populations.22,25 This may be due to the higher prevalence of anti-dsDNA antibody (77.9%) and renal disorder (73.2%) in our study. Autoantibody clusters in South China presented with distinct predominant age-groups and certain distinct clinical manifestations. The 3 clusters were different from those in other populations in that there was not a significantly different prevalence of renal disorder and anti-dsDNA between clusters with and without anti-ENA antibody.
In conclusion, distinct clinical and serologic correlations and autoantibody clusters are present in SLE patients in South China, such as a positive association between anti-RNP and photosensitivity, a negative association between anti-DNP and renal disorder, and less difference in the prevalence of anti-dsDNA and renal disorder between autoantibody clusters. In addition to ethnicity, these distinctions could be associated with environmental risk factors in South China. More attention and funding should be paid to understand the necessary and sufficient environmental exposures in the pathogenesis of SLE, which would be helpful in minimizing the new onset of SLE eventually.
1. The American College of Rheumatology nomenclature and case definitions for neuropsychiatric lupus syndromes. Arthritis Rheum
2. Biro E, Szekanecz Z, Czirjak L, Danko K, Kiss E, Szabo NA, Szucs G, Zeher M, Bodolay E, Szegedi G, Bako G. Association of systemic and thyroid autoimmune diseases. Clin Rheumatol
3. Bondanza A, Zimmermann VS, Dell'Antonio G, Cin ED, Balestrieri G, Tincani A, Amoura Z, Piette JC, Sabbadini MG, Rovere-Querini P, Manfredi AA. Requirement of dying cells and environmental adjuvants for the induction of autoimmunity. Arthritis Rheum
4. Cervera R, Khamashta MA, Font J, Sebastiani GD, Gil A, Lavilla P, Domenech I, Aydintug AO, Jedryka-Goral A, de Ramon E, et al. Systemic lupus erythematosus: clinical and immunologic patterns of disease expression in a cohort of 1,000 patients. The European Working Party on Systemic Lupus Erythematosus. Medicine (Baltimore)
5. Dahlgren J, Takhar H, Anderson-Mahoney P, Kotlerman J, Tarr J, Warshaw R. Cluster of systemic lupus erythematosus (SLE) associated with an oil field waste site: a cross sectional study. Environ Health
6. Epidemiological survey of systemic lupus erythematosus in textile workers in Shanghai. J Med Res
7. Imbert-Masseau A, Hamidou M, Agard C, Grolleau JY, Cherin P. Antisynthetase syndrome. Joint Bone Spine
8. Liang X, Chen Y, Wang X. A study on the sero-epidemiology of hepatitis B virus in Chinese population aged over 3-years old. Chinese Journal of Epidemiology
9. Maddison PJ, Mogavero H, Reichlin M. Patterns of clinical disease associated with antibodies to nuclear ribonucleoprotein. J Rheumatol
10. Malaviya AN, Chandrasekaran AN, Kumar A, Shamar PN. Systemic lupus erythematosus in India. Lupus
11. Mok CC, Lau CS, Chan TM, Wong RW. Clinical characteristics and outcome of southern Chinese males with systemic lupus erythematosus. Lupus
12. Mok CC, Lau CS. Lupus in Hong Kong Chinese. Lupus
13. Mok CC, Tang SS, To CH, Petri M. Incidence and risk factors of thromboembolism in systemic lupus erythematosus: a comparison of three ethnic groups. Arthritis Rheum
14. Mond CB, Peterson MG, Rothfield NF. Correlation of anti-Ro antibody with photosensitivity rash in systemic lupus erythematosus patients. Arthritis Rheum
15. Natali PG, Ashton M. Soluble deoxyribonucleoprotein (sNP) complexes in the pleural effusion of rheumatoid arthritis. Clin Immunol Immunopathol
16. Ni JD, Yao X, Pan HF, Li XP, Xu JH, Ye DQ. Clinical and serological correlates of anti-Sm autoantibodies in Chinese patients with systemic lupus erythematosus: 1,584 cases. Rheumatol Int
17. Pons-Estel BA, Catoggio LJ, Cardiel MH, Soriano ER, Gentiletti S, Villa AR, Abadi I, Caeiro F, Alvarellos A, Alarcon-Segovia D. The GLADEL multinational Latin American prospective inception cohort of 1,214 patients with systemic lupus erythematosus: ethnic and disease heterogeneity among "Hispanics." Medicine (Baltimore)
18. Pons-Estel GJ, Alarcon GS, Scofield L, Reinlib L, Cooper GS. Understanding the epidemiology and progression of systemic lupus erythematosus. Semin Arthritis Rheum
. 2009.[Epub ahead of print.]
19. Ram M, Anaya JM, Barzilai O, Izhaky D, Porat KB, Blank M, Shoenfeld Y. The putative protective role of hepatitis B virus (HBV) infection from autoimmune disorders. Autoimmun Rev
20. Robitaille P, Tan EM. Relationship between deoxyribonucleoprotein and deoxyribonucleic acid antibodies in systemic lupus erythematosus. J Clin Invest
21. Tan EM, Cohen AS, Fries JF, Masi AT, McShane DJ, Rothfield NF, Schaller JG, Talal N, Winchester RJ. The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum
22. Tapanes FJ, Vasquez M, Ramirez R, Matheus C, Rodriguez MA, Bianco N. Cluster analysis of antinuclear autoantibodies in the prognosis of SLE nephropathy: are anti-extractable nuclear antibodies protective? Lupus
23. ter Borg EJ, Groen H, Horst G, Limburg PC, Wouda AA, Kallenberg CG. Clinical associations of antiribonucleoprotein antibodies in patients with systemic lupus erythematosus. Semin Arthritis Rheum
24. ter Borg EJ, Horst G, Hummel EJ, Limburg PC, Kallenberg CG. Measurement of increases in anti-double-stranded DNA antibody levels as a predictor of disease exacerbation in systemic lupus erythematosus. A long-term, prospective study. Arthritis Rheum
25. To CH, Petri M. Is antibody clustering predictive of clinical subsets and damage in systemic lupus erythematosus? Arthritis Rheum
26. Viggiano DP, da Silva NA, Montandon AC, Barbosa VS. [Prevalence of thyroid autoimmune disease in patients with systemic lupus erythematosus]. Arq Bras Endocrinol Metabol
27. Wallace DJ, Hahn BH. Dubois' Lupus Erythematosus. In: Rus V, Maury EE, Hochberg MC, eds. The Epidemiology of Systemic Lupus Erythematosus
. Philadelphia: Lippincott Williams & Wilkins; 2007:37.
28. Xu D, Li B, Chen H, Ruan Z, Zeng Y, Yang X. Epidemiological survey of systemic lupus erythematosus in Guangzhou. Chinese Journal of Dermatology
29. Zhu W, Sun F, Fan E. Epidemiological survey and analysis of serum hepatitis B virus and hepatitis C virus infection in 938 hospitalized patients. Journal of Norman Bethune University of Medical Science