Evolution of disease burden over five years in a multicenter inception systemic lupus erythematosus cohort

Authors


Abstract

Objective

We describe disease activity, damage, and the accrual of key autoantibodies in an inception systemic lupus erythematosus (SLE) cohort.

Methods

The Systemic Lupus International Collaborating Clinics (SLICC) International Research Network, comprising 27 centers from 11 countries, has followed an inception cohort of SLE patients yearly according to a standardized protocol. Of these patients, 298 were followed for a minimum of 5 years and constitute the study population. Disease activity was assessed using the SLE Disease Activity Index 2000 (SLEDAI-2K) and damage was assessed using the SLICC/American College of Rheumatology Damage Index (SDI). Antinuclear antibody (ANA), anti-DNA, and anticardiolipin antibody (aCL) levels and lupus anticoagulant were assessed yearly. Descriptive statistics were generated and repeated-measures general linear models were used to evaluate SLEDAI-2K and SDI over time between whites and nonwhites.

Results

Of the 298 patients, 87% were women, 55% were white, 12% were African American, 14% were Asian, 16% were Hispanic, and 2% were categorized as “other.” At enrollment, the mean age was 35.3 years, the mean SLEDAI-2K score was 5.9, and the mean disease duration was 5.5 months. Mean SLEDAI-2K scores decreased in the first year and then remained low. SLEDAI-2K scores were significantly lower at each year in whites compared to nonwhites. Mean SDI scores increased progressively over 5 years; there was no significant difference between whites and nonwhites. As expected, ANA positivity was high and anti-DNA positivity was relatively low at enrollment, and both increased over 5 years. Although lupus anticoagulant increased slightly over 5 years, aCL positivity did not.

Conclusion

Disease activity in newly diagnosed patients decreases over their first 5 years, while damage increases. Antibody positivity ran variable courses over this period.

INTRODUCTION

Systemic lupus erythematosus (SLE) is a systemic autoimmune disease of unknown etiology. The burden of disease in SLE is determined both by the degree and severity of the immunologic inflammatory disease and the resultant organ damage either from the disease process itself, from comorbidities, or from treatment. Disease activity measured at an individual visit or over time, as well as accumulated damage, have been shown to predict mortality (1–7).

In SLE, autoantibodies may be diagnostic and/or markers of disease activity. Autoantibodies may be detected before the diagnosis of SLE (8). On an annual basis, the evolution of disease activity, damage, and the accrual of key autoantibodies early after diagnosis of lupus is unclear. The purpose of this study was to describe the course of disease activity, the development of damage, and the accrual of autoantibodies during the first 5 years of SLE. In addition, we aimed to determine whether there is a racial or ethnic variation in disease activity and damage.

Significance & Innovations

  • An international inception cohort of patients with systemic lupus erythematosus (SLE) with an average disease duration of 5 months followed for 5 years.

  • Disease burden changes in the course of the first 5 years of SLE.

  • While disease activity decreases, damage accumulates, especially damage related to corticosteroid therapy.

  • Although initially antinuclear antibodies are positive in 95% of the patients, the annual prevalence thereafter is <50%.

PATIENTS AND METHODS

Setting.

The Systemic Lupus International Collaborating Clinics (SLICC) group follows an international registry of newly diagnosed SLE patients, the SLICC Registry for Atherosclerosis, which was initially developed to study the prevalence and predictive factors for premature atherosclerotic vascular disease in SLE. The registry consists of a diverse population of SLE patients from 27 centers in 11 countries in North America, Europe, and Asia (1) (see Appendix A for the list of centers). SLICC inception cohort patients are followed longitudinally on an annual basis and their information is tracked on an Oracle database. Between 2000 and 2009, 1,249 patients were recruited into the SLICC inception cohort.

Patient selection and clinical assessments.

Patients are recruited into the SLICC inception cohort within 15 months of a diagnosis of SLE, based on the presence of ≥4 American College of Rheumatology (ACR) classification criteria (9). Only patients followed for at least 5 years were included in this study. Patients are assessed according to a standard protocol, which includes history, a physical examination, and laboratory assessments performed at yearly intervals. Variables necessary to calculate SLE Disease Activity Index 2000 (SLEDAI-2K) and SLICC/ACR Damage Index (SDI) are included.

Assessment of disease activity and damage.

Global SLE disease activity was measured by the SLEDAI-2K, a validated measure of disease activity in SLE (10). Cumulative damage is scored according to the SDI, a validated measure to assess damage in SLE (11, 12). Organ damage was categorized as: 1) related to corticosteroids (cataracts, osteonecrosis, and osteoporosis), 2) possibly related to corticosteroids (cardiovascular, peripheral vascular, cerebrovascular accident, and diabetes mellitus), and 3) independent of corticosteroids (renal, pulmonary, gastrointestinal, skin, gonadal failure, and malignancy).

Autoantibody assessments.

Four antibodies that are important either for classification (antinuclear antibody [ANA] and anti–double-stranded DNA [anti-dsDNA]) or monitoring disease activity (anti-dsDNA) or responsible for specific organ damage (anticardiolipin antibody [aCL] and lupus anticoagulant) were selected. ANA and anti-dsDNA antibodies are measured locally in each center. ANAs were measured by immunofluorescence in all but one center, which used an enzyme-linked immunosorbent assay (ELISA) technique. Anti-DNA was measured either by the Crithidia luciliae, ELISA, or Farr techniques. Importantly, in each center the same technique was used throughout the study. Lupus anticoagulant and aCL were measured at a single laboratory under the supervision of an author (JTM) (13).

Statistical analysis.

Descriptive statistics were used to examine trends over time in disease activity, disease damage, and presence of autoantibodies. Comparing SLEDAI-2K scores between whites and nonwhites at each year of followup was done using t-tests. In order to make the comparisons over the entire followup period, general linear models were used to adjust for repeated measures. The same approach was used to compare SDI scores. All analyses were performed using SAS software, version 9.1.

RESULTS

Patient characteristics.

A total of 298 SLICC inception cohort patients had been followed for 5 years in 2009 and were included in this study. Their demographic features are shown in Table 1. The majority were women (87%) and 55% were white (including those from the Indian subcontinent). Their mean age at enrollment was 35.3 years with a mean disease duration of 5.5 months. The mean ± SD SLEDAI-2K score at enrollment was 5.9 ± 5.5 and 93.6% of the patients were ANA positive at enrollment.

Table 1. Demographics of 298 inception cohort patients followed for 5 years*
 Value
  • *

    SLEDAI-2K = Systemic Lupus Erythematosus Disease Activity Index 2000.

Women, no. (%)258 (87)
Race/ethnicity, % 
 White55
 African American12
 Asian14
 Hispanic16
 Other2
Age at enrollment, mean ± SD years35.3 ± 14.5
Disease duration, mean ± SD months5.5 ± 4.1
Married, no. (%)125 (42)
College education, no. (%)168 (60)
SLEDAI-2K score at enrollment, mean ± SD5.9 ± 5.5

Disease activity and damage.

Yearly values for disease activity (SLEDAI-2K) are presented in Table 2. The mean SLEDAI-2K score decreases to low levels in the first year and then remains low. The same pattern is seen in whites as well as in the combined “other” ethnicities. T-tests done at every year, as well as a generalized linear model adjusting for repeated measures over time, showed that the SLEDAI-2K score was significantly lower in whites compared to nonwhites (P < 0.0001) (Table 2 and Figure 1). The difference is driven primarily by worse disease in Asian and Hispanic patients (Figure 2).

Table 2. SLEDAI-2K overall and by race/ethnicity*
Followup yearAll patients (n = 298)White (n = 164)Nonwhite (n = 134)Whites compared to nonwhites, P
  • *

    Values are the mean ± SD unless indicated otherwise. SLEDAI-2K = Systemic Lupus Erythematosus Disease Activity Index 2000.

05.93 ± 5.545.28 ± 5.426.73 ± 5.610.02
13.67 ± 4.113.03 ± 3.404.47 ± 4.740.006
23.71 ± 4.502.88 ± 3.494.71 ± 5.310.001
33.51 ± 3.643.02 ± 3.094.07 ± 4.120.02
43.30 ± 4.052.41 ± 2.724.34 ± 5.000.0002
53.37 ± 4.152.52 ± 3.274.43 ± 4.820.0001
Figure 1.

Mean Systemic Lupus Erythematosus Disease Activity Index 2000 (SLEDAI-2K) scores in 5 years by racial/ethnic group with 95% confidence intervals.

Figure 2.

Effect of ethnicity on disease activity over time. SLEDAI-2K = Systemic Lupus Erythematosus Disease Activity Index 2000.

At enrollment (of the 298 patients included in the study), 5 patients had no data on damage, 260 patients had no damage, and 33 patients (11.1%) had damage. After 5 years of followup, 115 (43.4%) of the 260 patients who had not had damage, developed damage. The mean SDI score increases progressively over the 5 years (Table 3). T-tests done at every year, as well as generalized linear model adjusting for repeated measures over time, showed that there was no significant difference between whites and patients of other ethnic/racial groups (P = 0.51).

Table 3. SDI scores overall and by race/ethnicity*
Followup yearAll patientsPatients with SDI >0, no. (%)WhiteNonwhiteWhites compared to nonwhites, P
  • *

    Values are the mean ± SD unless indicated otherwise. SDI = Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index.

00.17 ± 0.5733 (11.1)0.15 ± 0.500.20 ± 0.650.49
10.51 ± 0.9196 (32.2)0.53 ± 0.890.48 ± 0.930.62
20.64 ± 1.06110 (36.9)0.66 ± 1.010.62 ± 1.120.72
30.76 ± 1.14123 (41.3)0.76 ± 1.090.73 ± 1.210.73
40.88 ± 1.24134 (45.0)0.87 ± 1.160.89 ± 1.340.92
51.01 ± 1.34148 (49.7)0.94 ± 1.221.10 ± 1.480.32

Damage due to corticosteroids increased over the 5 years of followup, while damage independent of corticosteroid therapy remained constant throughout the period of observation. Over that period of time, the mean ± SD corticosteroid dosage decreased from 25 ± 17 mg/day to 10.6 ± 9.6 mg/day, while the mean ± SD cumulative dose increased from 2.5 ± 2.8 gm to 16.4 ± 12.0 gm at 5 years.

Autoantibodies.

Autoantibodies were evaluated in 2 ways. First, accumulation of autoantibodies over the first 5 years was analyzed. ANA positivity is high at enrollment, and the percent positivity increases by ∼4% over 5 years. Frequency of anti-DNA positivity is lower at enrollment (29%) but increases over 5 years to 55%. Lupus anticoagulant and aCL also increased over 5 years (Table 4).

Table 4. Cumulative prevalence of autoantibodies over the first 5 years*
Followup yearNo.ANAAnti-dsDNAaCL, no./total (%)LAC, no./total (%)
  • *

    Values are the number (percentage) unless indicated otherwise. ANA = antinuclear antibody; anti-dsDNA = anti–double-stranded DNA; aCL = anticardiolipin antibody; LAC = lupus anticoagulant.

0298279 (93.6)87 (29.3)66/249 (26.5)50/249 (20.1)
1298284 (95.3)112 (37.6)82/259 (31.7)75/259 (29.0)
2298284 (95.3)144 (48.3)89/274 (32.5)97/274 (35.4)
3298287 (96.3)151 (50.7)95/277 (34.3)118/277 (42.6)
4298287 (96.3)159 (53.4)100/277 (36.1)136/277 (49.1)
5298289 (97.0)165 (55.4)104/277 (37.6)144/278 (51.8)

Second, the prevalence of autoantibodies in each of the 5 years was assessed. Only anti-DNA assessment was performed in each of the 5 years in most of the 298 patients. The yearly prevalence of positive anti-dsDNA antibody studies ranged from 24–30%. Lupus anticoagulant and aCL were assessed yearly in only 25–40% of the patients in each year. In those assessed patients, the prevalence ranged from 8–25% for positive aCL and 16–26% for positive lupus anticoagulant. ANA is not repeated yearly in all patients in the SLICC inception cohort. In those who were followed, the annual prevalence of a positive test ranged from 37–47%. However, it is not discernable how many patients actually had the ANA assessment done at each visit. In a subset of 102 patients who had a yearly ANA assessment performed, 16 patients who had been ANA positive became negative. These 16 patients were not different from the rest of the group who remained persistently positive in terms of their SLEDAI-2K score, as well as their number of ACR criteria at enrollment.

DISCUSSION

Our study demonstrates that average disease activity in newly diagnosed SLE patients decreases and remains low over the first 5 years. However, despite this early disease control, disease damage increases over this time. The cumulative prevalence of ANA increases slightly over the first 5 years of SLE, while the prevalence of anti-dsDNA antibodies almost doubles by year 5, and aCL and lupus anticoagulant cumulative prevalence also increases over the first 5 years of disease.

In our study, disease activity was mild to moderate (5.9 for SLEDAI-2K score) at the first visit. Although this is an inception cohort, the patients had a disease duration at entry of 5.5 months and in many cases were already receiving hydroxychloroquine or corticosteroids. Nevertheless, the SLEDAI-2K score decreased further over the first year to 3.7 and remained at low levels over the ensuing 4 years. The European inception cohort of 200 SLE patients, with a similar age at diagnosis (34.8 years) and female preponderance (89%), reported more severe disease at presentation with an initial SLEDAI-2K score of 12.2, even though the mean duration from onset of symptoms to ACR classification was 28 months (14). However, within the first year of disease, 27.5% of these patients achieved disease quiescence and half of these maintained this state over the 4 years of their study. The LUpus in MInorities, NAture versus nurture (LUMINA) study included 542 patients with a mean age at diagnosis of 35.8 years and with 90% being women. This study used the Systemic Lupus Activity Measure (SLAM) to assess disease activity and noted that there was a sharper decrease in disease activity within the first year; the slope of decline in SLAM values decreased with increasing disease duration (15). In their multivariate analysis, whites exhibited the fastest decline in disease activity and Texan Hispanics exhibited the slowest, with African Americans following between the two. In the SLICC inception cohort study, white patients had lower disease activity at each time point in the 5-year followup compared to nonwhites, although all patients improved their disease activity over time. These differences could be based on environmental or genetic factors. When disease activity was further studied in the nonwhite patients, it was found that Asians and Hispanics influenced the difference in disease activity the most. Although patients of African origin did have slightly higher disease activity than whites, it was not statistically significant. This is similar to observations of the LUMINA cohort. Therefore, it would seem that lupus activity could be controlled early in a large proportion of patients with SLE.

Damage accumulation occurs slowly among patients with SLE. It has been shown that patients who accumulate damage within the first year of disease are at increased mortality risk compared with patients whose SDI score is 0 (4). The study by Chambers et al (5) included 232 patients followed for at least 10 years. Of these patients, 90% had no damage at year 1 postdiagnosis. However, by year 10, 50% had accrued some damage. Nossent et al (14) described an inception cohort of 200 SLE patients from 14 European centers followed up to 5 years. Forty-two percent of the patients accrued some damage during the observation period, with the highest damage accrual occurring in the first 2 years of disease. In contrast, in the current study, damage continued to accumulate over the 5 years of followup. In particular, there was a progressive increase in the damage definitely related to corticosteroids. In a previously reported inception cohort, damage increased progressively over a 15-year period (16). The percent of the SDI score related to corticosteroids increased progressively, reaching 54% by 5 years. The patients in that cohort were recruited since the 1970s and were often treated with higher doses of corticosteroids for prolonged periods, which might account for the higher progressive increase in steroid-related damage. The current study represents patients treated since the year 2000 and, although damage progressively increases over the first 5 years of followup, the percent of damage due to corticosteroids does not increase as dramatically as in previous years. This may reflect a more judicious use of corticosteroids in the current era.

Looking at the prevalence in any given year, ANA assessment is often not repeated in long-term studies as it has not been associated with disease activity. In those subjects who returned for followup, the annual prevalence was 37–47%. This may reflect negative tests or tests not performed. This may have relevance in clinical trials where inclusion criteria may require a positive ANA at inclusion. Our data would suggest that the criterion should be positive ANA at some time during the course of the disease. Anti-dsDNA antibody assessment done on a yearly basis in our study revealed an annual prevalence of 24–30%. This may be due to disease state (either improved or worsened) or treatment. Over the 5-year period, however, there was a progressive increase in the number of patients with positive anti-dsDNA antibodies such that more than 55% of the patients had been anti-dsDNA positive at some point.

In summary, disease burden changes in the course of the first 5 years of SLE. While disease activity decreases, damage accumulates, especially damage related to corticosteroid therapy.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Dr. Urowitz had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Urowitz, Gladman, Fortin, Bae, Gordon, Clarke, Bernatsky, Isenberg, Rahman, Sanchez-Guerrero, Wallace, Ginzler, Alarcón, Merrill, Bruce, Sturfelt, Nived, Steinsson, Khamashta, Petri, Manzi, Ramsey-Goldman, Dooley, van Vollenhoven, Ramos, Stoll, Zoma, Kalunian, Aranow.

Acquisition of data. Urowitz, Gladman, Fortin, Bae, Gordon, Clarke, Bernatsky, Hanly, Isenberg, Rahman, Sanchez-Guerrero, Wallace, Ginzler, Alarcón, Merrill, Bruce, Sturfelt, Nived, Steinsson, Khamashta, Petri, Manzi, Ramsey-Goldman, Dooley, van Vollenhoven, Ramos, Stoll, Zoma, Kalunian, Aranow.

Analysis and interpretation of data. Urowitz, Gladman, Ibañez, Fortin, Hanly, Wallace, Alarcón, Manzi.

Acknowledgements

We would like to gratefully acknowledge the generous donation of our patients' time and the dedication of all the fellows, research coordinators, and assistants in the SLICC group to the completion of this work. We acknowledge the contribution of Dr. P. Maddison at North West Wales National Health Service Trust, Bangor, UK.

APPENDIX A

PARTICIPATING CENTERS OF THE SLICC REGISTRY FOR ATHEROSCLEROSIS

Centre for Prognosis Studies in the Rheumatic Diseases, Hanyang University, University of Birmingham, Montreal General Hospital, Dalhousie University, University College, Instituto Nacional de la Nutricion, Cedars-Sinai Medical Center, SUNY Health Science Center, University of Alabama at Birmingham, Oklahoma Medical Research Foundation, University of Manchester, University Hospital Lund, St. Thomas Hospital, Johns Hopkins University, University of Pittsburgh, Northwestern Hospital, University of North Carolina at Chapel Hill, Karolinska University Hospital, Servicio de Enfermedades Autoinmunes, AaReha Schninzach, Lanarkshire Center for Rheumatology, University of California, Feinstein Institute for Medical Research, North West Wales NHS Trust.

Ancillary