Is Familial Lupus Different from Sporadic Lupus?: Data from LUMINA, a Multiethnic US Cohort

Abstract

Purpose

To characterize the clinical features of familial lupus, and determine its influence on damage accrual and survival using data from LUMINA, a longitudinal multiethnic US cohort.

Method

Familial lupus was defined as patients with a first degree relative with SLE. Relative risks were estimated by logistic regression; odds ratios (OR) and their 95% confidence intervals (CI) were the measure of association for familial lupus. Hazard Ratios (HR) were calculated using Cox proportional hazard adjusted for potential confounders for damage and survival.

Results

Thirty-two of 644 patients had familial and 612 had sporadic lupus; both groups were of comparable age (~ 36 years). Familial lupus patients were in decreasing order of frequency siblings, parents and children. In multivariable analyses, mucosal ulcers (OR=1.92, 95% CI 0.65–5.70), mitral valve prolapse (OR=1.74, 95% CI 0.50–6.10), cerebrovascular disease (OR=4.18, 95% CI 0.98–17.76) and oral contraceptive use (ever/never; OR=2.51, 95% CI 0.88–7.19) were more likely in familial lupus but a history of low platelet count (<150,000/mm³; OR=0.31, 95% CI 0.08–1.17) and pulmonary disease activity (OR=0.39, 95% CI 0.14–1.20) were less likely. However, none of these associations reached statistical significance. Familial lupus was not significantly associated with a shorter time to either damage accrual or death (HR=0.77, 95% CI 0.37–1.59, p = 0.4746 and HR=0.20, 95% CI 0.03–1.47, p = 0.2020, respectively).

Conclusions

Although some clinical differences were observed in patients with familial and sporadic lupus, familial lupus was not associated with a significantly greater disease burden (damage, survival) than sporadic lupus.

INTRODUCTION

Although the etiology of systemic lupus erythematosus (SLE) remains elusive, environmental and genetic factors are known to influence its onset and clinical course [1]. As with other autoimmune diseases, the genetic basis of SLE are complex with contribution from the major histocompatibility complex (MHC) genes as well as a multiplicity of other genes [2].

Lupus occurs more often within families (10–12%). Concordance rates are higher for monozygotic (24–56%) than dizygotic twins (2–5%); furthermore, a high degree of concordance in disease expression has been shown in some studies [3–6]. Despite the preponderance of lupus within families, studies conducted so far have failed to show distinctive immunological and clinical features between patients with sporadic and familial lupus [7–9] including studies conducted in populations with a high degree of consanguinity [8;10].

Nevertheless, we hypothesized that patients with familial lupus who may have a higher genetic load than patients with sporadic lupus may also have more severe disease manifestations which in turn may affect their damage accrual and their survival.

METHODS

Population

LUMINA (Lupus in Minorities, Nature versus nurture) is a longitudinal cohort of patients with SLE [11]. All patients meet the American College of Rheumatology (ACR) criteria for SLE classification [12], have disease duration of ≤5 years, are ≥16 years of age, of defined ethnicity (African American, Hispanic [from Texas and Puerto Rico], or Caucasian), and live in the geographic recruitment areas of the participating centers (the University of Alabama at Birmingham, the University of Texas Heath Science Center at Houston, and the University of Puerto Rico Medical Sciences Campus). The Institutional Review Board of these centers approved the study; written informed consent was obtained according to the Declaration of Helsinki.

Every patient had a baseline or enrollment visit (T0); follow up visits were conducted every six months during the first year (T0.5 and T1, respectively), and yearly thereafter. At each visit, the patients were interviewed and examined, and laboratory tests performed. Clinical data from missed study visits were obtained by review of all available medical records. Disease duration was defined as the time from the date the patients met four ACR criteria for SLE (TD) to T0. Duration of follow up was the period between T0 and the last visit (TL).

Familial lupus, our primary end-point, was defined as patients with any first degree relative (father, mother, siblings or children) with lupus. This diagnosis was self-reported during the patients’ study visits and verified by reviewing all available medical records. Patients without a self-reported family history of lupus were defined as having sporadic lupus.

Clinical Information

As previously reported [13], the LUMINA database includes variables from the following domains: socioeconomic-demographic, clinical, immunologic, genetic, behavioral and psychological. These variables are measured at all visits. Only the variables included in these analyses will be described.

Variables from the socioeconomic-demographic domain included were age, ethnicity, education, poverty [14], health-related behaviors, marital status and health insurance. Variables from the behavioral and psychological domains were social support [15], abnormal illness-related behaviors (IBQ) [16] and learned helplessness [17]. Clinical variables included were the number of ACR criteria at T0, onset type [acute versus insidious, if otherwise], follow-up time (T0-TL, as defined), disease manifestations, disease activity and damage, immunological variables and medications. Cumulative exposure to glucocorticoids, hydroxychloroquine, cyclophosphamide, azathioprine, mycophenolate mophetil, statins, non-steroidal anti-inflammatory drugs, low-dose aspirin, oral contraceptives, hormonal replacement therapy, warfarin and angiotensin-converting enzyme inhibitors was also recorded.

Disease activity was assessed using the Systemic Lupus Activity Measure Revised (SLAM-R) [18] at T0 and at every visit. For the purpose of this study, the average SLAM-R score for all visits (TD to TL) was calculated as a measure of disease activity over time. Damage at T0 and TL was assessed using the Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index (SDI) [19].

Laboratory Assays

Autoantibodies were obtained at T0 included antinuclear antibodies (ANA by immunofluorescence using HEp-2 cell line), anti-double-stranded DNA (anti-dsDNA, by immunofluorescence against Crithidia luciliae), anti-Smith, anti-RNP, anti-La and anti-Ro (by immunodiffusion) and antiphospholipid antibodies (APL by enzyme-liked immunoabsorbent assay and /or lupus anticoagulant by the Staclot assay). Selected HLA and FCGR3A specificities were also included. In addition the following laboratory variables were recorded at T0: non-fasting serum lipoproteins, and serum CRP, measured as high-sensitivity CRP (hs-CRP; high >9 mg/liter or the highest tertile for the distribution of our patients’ values) by immunometric assay (Immulite 2000; Diagnostic Products, Los Angeles, CA) using T0 patients’ sera.

Statistical analyses

Features from the different domains were compared between those patients with familial and sporadic lupus using Student’s t and Chi-square tests for continuous and categorical variables, respectively. Variables with p≤0.10 were entered into multivariable logistic regression models to examine their independent association with familial lupus; relative risks were estimated using odds ratios (OR) and their corresponding 95% confidence intervals (CI). The role of familial lupus in the time-to-the development of damage and death [Hazard ratios (HR)] was examined by Cox multivariable regressions adjusting for variables previously found to be associated with damage accrual [age, gender, ethnicity, highest doses of prednisone, number of ACR criteria and familial lupus] and mortality [age, gender, ethnicity, SDI baseline, SLAM baseline, poverty, familial lupus]. Statistical significance was set at p ≤ 0.05; analyses were performed using the SAS software, version 9.1 (SAS Institute, Cary, NC, USA).

RESULTS

The cohort consisted of 644 patients; 32 of them had familial lupus and 612 the sporadic form; the majority of patients in both groups were women (~90%) and their age was comparable (36.6 ± 12.5 years). All ethnic groups were represented; although there were some differences in the distribution of these groups in the two forms (Texan-Hispanic: 19% vs. 6%, Puerto Rican-Hispanic: 16% vs. 9%, African American: 37% vs. 50% and Caucasian: 28% vs. 34%), these differences were not statically significant. The mean (SD) disease duration and follow up times were also comparable: 1.7 (1.4) and 5.3 (4.0) years for familial and 1.4 (1.4) and 4.6 (3.5) years for sporadic lupus. Of the first degree relatives 50% (n=16) were siblings, 38% (n=12) parents and 12% (n=4) children.

Features associated with familial lupus

None of the socioeconomic–demographic variables examined were significantly associated with familial lupus. As shown in Table 1, oral ulcers (60.3% vs. 78.1%, p=0.043) and cerebrovascular disease (3.8% vs. 12.5%, p=0.039) were significantly associated with familial lupus. None of the patients with familial lupus presented palpable purpura, subacute cutaneous lupus, livedo reticularis or pulmonary damage as determined by the SDI.

Table 1.

Baseline Socio-demographic, Sociodemographic, Clinical, Laboratory and Genetic Features in Patients with Sporadic and Familial Lupus from the LUMINA Cohort

	Sporadic N=612	Familial N=32		P value*
Age, years, mean (SD)	36.4 (12.6)	36.7 (12.5)		0.921
Ethnicity, %
Texan Hispanic (n=117)	19.1	6.3
Puerto Rican Hispanic (n=102)	16.2	9.4		0.139
Caucasian (n=183)	28.1	34.4
African American (n=240)	36.6	50.0
Below the Poverty line†, %	34.3	21.9		0.149
ACR‡-criteria present, %
Oral ulcers	60.3	78.1		0.043
Skin manifestations, %
Palpable purpura	3.4	0.0
Subacute cutaneous lupus	1.5	0.0
Livedo reticularis	1.1	0.0
Cardiologic manifestations, %
Mitral valve prolapse	8.1	18.8		0.050
Angina	2.0	6.3		0.149
Cerebrovascular disease, %	3.8	12.5		0.039
Laboratory features, %
Platelet count (<150.000/cubic mm)	26.4	12.5		0.097
Erythrocyte sedimentation rate(> 25mm/hr)	76.8	56.3		0.008
C-reactive protein, mg/dl, mean (SD)	14.7 (32.7)	5.7 (5.5)		0.131
Triglyceride, mg/dl, mean (SD)	145.8 (87.5)	122.6 (54.5)		0.149
LDL¶, mg/dl, mean (SD)	108.1 (54.9)	122.8 (44.4)		0.142
SLAM-R§ average	8.8 (4.3)	7.5 (4.1)		0.114
SLAM-R domains, % present
Pulmonary	45.0	28.1		0.061
SDI** domains, % present
Pulmonary	8.0	0.0
Medication use, %
Cyclophosphamide	27.4	15.6		0.144
Mycophenolate mophetil	10.1	0.0
Oral contraceptive	22.5	43.8		0.006
Genetic features, %
HLA-DRB*1503	13.7	24.1		0.116
HLA-DQB*0602	31.6	44.8		0.137
ISEL (social support), mean (SD)	7.8 (1.8)	8.4 (1.5)		0.089
Patients deceased, %	15.5	3.1		0.055

^*Only P values < 0.15 are shown;

^†Poverty was determined according to US government guidelines and was adjusted for the number of persons in the household;

^‡ACR= American College of Rheumatology;

^§SLAM-R=Systemic lupus Activity Measure, Revised;

^¶LDL=Low density lipoprotein;

^**SDI=Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index.

An increased in erythrocyte sedimentation rate was associated with sporadic lupus (76.8% vs. 56.3%, p=0.008). None of the autoantibodies, lipid profile and genetic variables examined was found to be associated with familial lupus (data not shown).

The exposure to hydroxychloroquine, cyclophosphamide and azatioprine was comparable in the two groups but mycophenolate mophetil was more frequently used in the sporadic than in the familial form (10% vs. 0%). The use of glucocorticoids was comparable in both groups but oral contraceptives were used more frequently in familial lupus than in the sporadic form (44% vs. 22 %, p=0.006). Other medications such as aspirin, statins, non-steroidal anti-inflammatories, hormonal replacement therapy, warfarin, and angiotensin-converting enzyme inhibitors were used comparably in both groups (data not shown).

Independent factors associated with familial lupus

As shown in Table 2, in multivariable analyses, mucosal ulcers (OR=1.92, 95% CI 0.65–5.70), mitral valve prolapse (OR=1.74, 95% CI 0.50–6.10), cerebrovascular disease (OR=4.18, 95% CI 0.98–17.76) and oral contraceptive use (ever/never; OR=2.51, 95% CI 0.88–7.19) were more likely in familial lupus but statistical significance was not reached. The same was the case for the pulmonary domain of the SLAM-R and low platelet count (OR=0.39, 95% CI 0.14–1.20 and <150,000/mm³; OR=0.31, 95% CI 0.08–1.17; respectively) that were less likely in familial lupus.

Table 2.

Multivariable Analysis of Risk Factors Associated with Familial Lupus in Patients from LUMINA

Variable	OR	95% CI	P Value
Age	1.01	0.97 – 1.05	0.575
Gender, male	0.49	0.12 – 2.01	0.325
Ethnicity
Hispanic	0.40	0.08 – 2.16	0.287
Caucasian	0.64	0.19 – 2.18	0.470
Puerto Rican	0.56	0.13 – 2.47	0.445
African American	Reference group
Disease duration	1.08	0.97 – 1.22	0.175
Oral ulcers	1.92	0.65 – 5.70	0.240
Mitral valve prolapse	1.74	0.50 – 6.10	0.385
Cerebrovascular disease	4.18	0.98 – 17.76	0.053
Low platelets count (<150.000/cubic mm)	0.31	0.08 – 1.17	0.083
ESR† >25 mm Hg	0.88	0.29 – 2.70	0.825
SLAM-R* score average	1.07	0.92 – 1.25	0.387
SLAM-R* pulmonary domain	0.39	0.14 – 1.20	0.075
HLA-DRB1*1503	1.51	0.44 – 5.19	0.513
Oral contraceptive use (ever/never)	2.51	0.88 – 7.19	0.086
Social support score	1.24	0.92 – 1.67	0.160

^*Systemic Lupus Activity Measure-Revised;

^†Erythrocyte Sedimentation Rate

Longitudinal analysis of familial lupus

In the multivariable time to event analysis for damage accrual which included age, gender, ethnicity, highest doses of prednisone and number of ACR criteria, familial lupus was not significantly associated with either a shorter or a longer time-to-damage accrual (HR=0.77, 95% CI 0.37–1.59, p = 0.475) (data not shown).

Finally, in the multivariable time to event analysis for death which included age, gender, ethnicity, SDI baseline, SLAM baseline and poverty, familial lupus (HR 0.20, 95% CI 0.03–1.47, p=0.202) was not significantly associated with either a shorter or a longer time-to-death (data not shown).

DISCUSSION

We have shown for the first time that familial lupus exerts no negative impact in terms of either a diminished survival or damage accrual. Furthermore, we did not find a distinct clinical o serologic pattern distinguishing familial and sporadic lupus, despite the clear importance of genetic factors in the development of lupus. Even thought none of the clinical manifestation reached significance, the possible association with cerebrovascular disease and oral contraceptive use is intriguing; however, no differences in the frequency of aPL antibodies or other thrombotic events were observed in patients with familial lupus. Like in our study, an increased/decreased frequency of specific clinical and laboratory features such as photosensitivity, oral ulcers, malar rash, low platelet counts and/or anti-Ro antibodies have been reported in patients with familial lupus but none of these findings have shown to be independently associated with the development of this form of lupus [7–9] supporting the importance of environmental factors in the development of the phenotype. Furthermore, in the Chinese study of Wang et al in which 136 patients with lupus nephritis were studied, 34 of them having familial lupus, fever was the only manifestation independently associated with familial lupus but the severity of lupus nephritis, as assessed histologically, was found to be similar in both patients groups [21].

The 5% prevalence of familial lupus observed in our study is similar to frequencies previously described [8;9] but lower than in studies of patients with high degree of consanguinity [10;20].

Several studies have shown that the genetic predisposition to the development of lupus as well as its course vary according with the ethnic/racial background of the patients studied [13]. Sestak et al, for example in a study emanating from the US Lupus Family Registry and Repository stratified their patients with familial and sporadic lupus by racial/ethnic group [22]. Although a similar clinical profile was found in patients with sporadic and familial lupus of the same racial/ethnic group, there were some differences between the groups; for example, there was an increased proportion of patients with arthritis and LE cell test positivity in patients with familial lupus of European ancestry as compared to those of African and Hispanic ancestry [22].

The fact that no distinctive clinical, laboratory or genetic pattern emerged in familial lupus needs to be interpreted with caution; first, we only included first degree relatives which precluded comparing our study with others in which other relatives have also been included. Second, the sample size did not permit examining the ethnic groups separately; however, the data from Sestak et al suggest that there are no substantial differences between the racial/ethnic groups [22]. Third, we only examined MHC Class II and III genes and not others which have been associated with the disease or its different features [2]. Finally, our study did not include the full spectrum of the relatives’ clinical features to assess the co-occurrence of SLE in these families and the possible presence of sub-phenotypes according with the degree of familiarity (i.e. parents vs., siblings).

In conclusion, familial lupus, as compared to sporadic lupus, is not associated with a significantly greater disease burden as evidenced by their clinical manifestations, damage accrual and survival; thus patients with familial lupus should be treated similarly than patients with sporadic lupus and worse outcomes are not to be expected among them.

Rheumatology Key Message.

• No substantial differences in the clinical, laboratory and genetic features of patients with familial and sporadic lupus were found.

• Studies including a larger number of patients with familial lupus are necessary to establish the differences, if any, between familial and sporadic lupus as a function of racial/ethnic group.

• Familial lupus appears not to have a significant influence in damage accrual or survival.