Familial Lupus Erythematosus

Introduction Systemic lupus erythematosus (SLE) is a heterogeneous systemic autoimmune disease characterized by the production of a wide range of autoantibodies mainly directed to nuclear antigens and immune complexes that may lead to tissue injuries (12). Although the pathogenesis of the disease remains poorly understood, genetic predisposition is probably the greatest risk factor for SLE (2,7). Evidence for a genetic susceptibility to SLE in humans is based on the high concordance rate observed in identical twins (29%–57%) and on the relatively high incidence of familial cases (10%–12%) (6,26). The contribution of genetic susceptibility to discoid lupus erythematosus (DLE), the most common form of chronic cutaneous lupus erythematosus, is not known since familial cases of DLE have been reported only rarely (25). It is admitted that SLE, like many other autoimmune diseases, is a complex genetic trait with contribution from multiple genes (7,26). Until recently, genetic studies in humans have focused almost exclusively on major histocompatibility complex (MHC) class II alleles and complement gene deficiencies, which probably have a weak contribution to SLE susceptibility (5,18). During the last 3 years, both association and linkage studies in multiplex families have led to significant advances in the identification of new non-MHC candidate genes or regions (9,16,20,23). However, since in most recent genetic studies clinical data are lacking, and no large clinical study on familial SLE has been reported to date except for a review of the literature by Arnett (2), many questions about familial lupus are still unanswered. Whether clinical expression and immunologic characteristics of familial lupus differ from those of sporadic lupus is not clearly established, and the most frequent pattern of intrafamilial aggregation among lupus multiplex families is not known. Moreover, the frequency of antiphospholipid syndrome, an entity closely related to SLE, has never been evaluated in lupus familial cases. To answer these questions, we analyzed the clinical and immunologic features of 125 French lupus multiplex families recruited through a national survey. To better estimate the prevalence of familial lupus in the French Caucasian population, 137 lupus patients seen consecutively in our department and whose familial status was unknown were first investigated using a standardized procedure. The main characteristics of 90 multiplex lupus probands were then compared with those of 100 patients with sporadic cases sharing the same French Caucasian origin. Patients and Methods Patients and family enrollment Consecutive lupus patients: In order to establish the degree of familial clustering in a French homogeneous population, 109 unselected Caucasian patients with SLE and 28 Caucasian patients with DLE without features of systemic involvement, whose familial status was unknown, were first investigated. All patients were seen consecutively over a 6-month period at the Internal Medicine Department of Pitié-Salpêtrière hospital, a tertiary referral center in Paris. Definite SLE required at least 4 of the revised 1982 American College of Rheumatology criteria (21), and drug-induced lupus was excluded from the study. In the absence of acknowledged criteria, erythemato-squamous coin-shaped cutaneous lesions with atrophic scarring in the central portion were considered as highly suggestive of DLE. Diagnosis was assessed by an experienced dermatologist (CF) and confirmed by a skin biopsy if necessary. Pedigrees extending to second-degree relatives were established for each family with the help of the probands. Both affected and unaffected family members aged above 18 years who gave their informed consent were interviewed using a standardized procedure, and then had blood drawn. Medical interviews of presumably unaffected relatives were particularly focused on SLE-and DLE-related manifestations or any other medical history suggestive of an autoimmune disorder. Preexisting medical charts of relatives were collected whenever justified, especially when SLE or DLE was claimed or suspected. Lupus multiplex families: In the second part of the study, we started in July 1997 to register nationwide, in collaboration with the Société Nationale Française de Médecine Interne, and the Société Française de Rhumatologie, families in which a diagnosis of SLE and/or DLE had been made in 2 or more members (first-or second-degree related). Again, pedigree was established with the help of the proband for each family, and all availaible family members were interviewed by the same investigator using the same standardized procedure. Complete data from medical records for each SLE or DLE patient were collected, reviewed, and registered by the same physician (MM). Laboratory methods Blood samples were obtained for both affected and unaffected family members who were available and willing to be studied. Antinuclear antibodies (ANA) were determined by indirect immunofluorescence (IF) on Hep2 cell monolayers. For screening, sera were diluted at 1:40 and, in case of positive fluorescence, at 1:80, 1:160, 1:320, and 1:640. A titer equivalent to or greater than 1:80 on Hep2 cells was considered significant. Specificity of positive samples was determined by searching for anti-ds DNA by Farr assay (Ortho Clinical Diagnostics, Amersham, UK) and antibodies to soluble nuclear antigens (anti-Ro/SSA, anti-La/SSB, anti-Sm, and anti-nRNP) by double immunodiffusion using human spleen and calf thymus extract as antigens. This technique was carried out using identity reaction with a reference serum obtained from the Centers for Disease Control and Prevention (Atlanta, GA). Stastistical analysis Ninety unrelated probands with familial SLE whose clinical data have been fully reviewed, and 100 unselected probands with sporadic SLE sharing the same French Caucasian origin have been compared for the following items: sex ratio, age at inclusion, age at disease onset (defined by the age at the first SLE-related manifestation), malar rash, photosensitivity, discoid rash, oral ulcers, arthritis, serositis, neurologic involvement, renal involvement, thrombocytopenia, hemolytic anemia, ANA with a titer of 1/80 or more, anti-DNA antibodies of IgG isotype according to the 1982 American College of Rheumatology classification criteria (20), associated antiphospholipid syndrome (10), and Raynaud phenomenon. Prevalence of positive ANA in unaffected first-degree relatives of respectively multiplex and sporadic SLE probands has been compared to that observed in 100 French healthy controls matched for age distribution. First-degree relatives of multiplex probands were considered clinically unaffected if they had no clinical evidence of either SLE or DLE. Quantitative data were analyzed using the Student t-test. Qualitative data were analyzed using the 2-sided Fisher exact test, and p values were corrected by multiplying the actual value by the number of multiple comparisons. Values of p less than 0.05 were considered significant. Ethics This procedure has been approved by local advisory and ethical committees (CCPPRB). Results Prevalence of familial lupus and of familial autoimmune diseases Among the 137 French Caucasian patients (109 with SLE and 28 with DLE) who were consecutively investigated, 11 (8%) (9 SLE probands and 2 DLE probands) had at least 1 first-degree relative with a definite diagnosis of either SLE or DLE. In 1 pedigree, the DLE proband had 1 sister with DLE and another with a definite diagnosis of primary antiphospholipid syndrome. When autoimmune diseases besides lupus were considered among relatives, the prevalence of “autoimmune families” rose to 23% (31 of 137). Autoimmune thyroiditis (10 cases) and rheumatoid arthritis (3 cases) were preferentially observed among lupus relatives. Moreover, 15 of 137 lupus probands (11%) were found to have another distinct associated autoimmune disease: autoimmune thyroiditis in 9, relapsing polychondritis in 2, vitiligo in 2, rheumatoid arthritis in 1, and premature ovarian failure in 1 (the presence of coexisting antiphospholipid syndrome was not considered as a distinct disease for analysis). Features of multiplex families From July 1997 to April 1998, 125 multiplex families were recruited through the national survey. One hundred families contained 2 affected relatives and 25 families contained 3 or more affected relatives, including 1 extended pedigree in which either SLE or DLE was present in 6 relatives (Family A). Ninety-six families were of French Caucasian origin, 11 families were from the French West Indies, 10 from northern Africa, 4 of Sepharadic Jewish origin, 2 from Asia, 1 from Iraq, and 1 from Brazil. The total number of affected patients in our multiplex families was 282 (mean, 2.3 per family), of whom 86.5% were female. Fifteen of 282 affected individuals had died before the beginning of the study but their diagnosis was assessed retrospectively by rewieving their medical records. Familial occurrence of both SLE and DLE was observed in 26 multiplex families, while pure DLE was observed in both affected relatives in only 3 of 125 multiplex families. In 15 families (12%), at least 1 of the affected relatives had a diagnosis of antiphospholipid syndrome. Genetic inheritance The most frequent mode of familial intra-aggregation of the disease was affected sib pairs (45% of all pedigrees) (Table 1). Although no obvious mode of inheritance could be suspected in most of those pedigrees, few of them suggested a Mendelian trait. One extended pedigree with 6 affected members was consistent with an autosomal dominant mode of inheritance (pedigree A, Figure 1). A recessive trait was suspected in 5 pedigrees with 3 affected sibs each (for example, see pedigree I, Figure 1); parents were consanguineous in 2 of those 5 pedigrees. A complement component deficiency (C2 heterozygous deficiency detected by polymerase chain reaction) was ascertained in only 1 of those 5 pedigrees, in which 2 brothers had DLE and 1 sister had SLE. Figure 1 represents some of the various modes of intrafamilial aggregation found in our large group of multiplex families.TABLE 1: Relationships among familial cases of SLE and/or DLE in 125 multiplex familiesFig. 1: Ten of the 125 French lupus multiplex pedigrees with different modes of intrafamilial relationships suggesting a genetic heterogeneity: Family A and B suggest a dominant mode of inheritance; Family I suggests a recessive mode of inheritance. Family J is an example of a multiplex pedigree including different autoimune diseases (systemic lupus erythematosus: n = 2, immune thyroiditis: n = 1, Sjögren disease: n = 1). Abbreviations: AID = autoimmune disease.Comparison of familial and sporadic SLE In order to analyze the main clinical and biologic features of familial SLE, we compared the characteristics of 90 unrelated probands with familial SLE and available charts with those of 100 unselected consecutive patients with sporadic SLE sharing the same French Caucasian origin. Sex-ratio and mean age at disease onset and at inclusion were similar in both groups (Table 2). Moreover, after correction for multiple comparisons, no significant differences were found between familial and sporadic SLE (see Table 2).TABLE 2: Features in multiplex and sporadic SLE patients of French Caucasian originPrevalence of ANA in first-degree unaffected relatives One hundred eighty-four clinically unaffected first-degree relatives belonging to multiplex SLE families of French Caucasian origin (mean age, 46 yr) were typed for autoantibodies and compared with 165 clinically unaffected first-degree relatives from sporadic families (mean age, 45 yr) and 100 age-matched healthy controls of French Caucasian origin (Table 3). ANA (dilution 1:80) were positive in 35 of 184 (19%) unaffected first-degree relatives from multiplex families compared with 9% of healthy controls (p < 0.03). On the other hand, the prevalence of ANA observed in unaffected first-degree relatives from sporadic families did not differ from that observed in healthy controls. The rates of unaffected relatives with positive anti-DNA, anti-Ro/SSA, and anti-Sm or -RNP antibodies did not differ in multiplex and sporadic families (see Table 3). Moreover, when titers of positive ANA were compared in both multiplex and sporadic unaffected relatives, no difference was found in terms of distribution. Among the 35 multiplex relatives with positive ANA, the titer was at 1/80 in 10 individuals (29%), at 1/160 in 13 (37%), and at 1/320 or higher in 12 (34%), compared with 22% (1/80), 36% (1/160), and 40% (1/320 or higher) among unaffected relatives in sporadic families.TABLE 3: Autoantibodies in unaffected first-degree relatives from multiplex and sporadic SLE probands and in healthy controlsDiscussion Recent advances in genetic methods have allowed significant progress in the identification of new susceptibility genes or regions in multifactorial complex diseases such as SLE (27). In the last 3 years, different groups have identified promising new candidates genes or loci for SLE susceptibility mainly, but not exclusively, located on the long-arm of chromosome 1 (9,14,16,20,23,24). To perform accurate linkage studies and genome-wide scans in multifactorial diseases, the recruitment and analysis of a large number of multiplex families is an essential step (27). However, few studies focusing on clinical and immunologic features of familial SLE have been performed since 1983, when Reveille et al (18) reported immunogenetic studies in 8 SLE multiplex families. In addition, the previous analysis of 201 familial cases of SLE (including monozygotic twin pairs) made by Arnett (2) based on a literature review did not take into account discoid lupus erythematosus and antiphospholipid syndrome, 2 diseases closely related to SLE, or immunologic abnormalities in clinically unaffected relatives. In the present report, we describe the main features of 125 lupus multiplex families recruited from all over France by the same investigator. We first checked that the prevalence rate of 8% of familial lupus found in a population of unselected SLE patients of French Caucasian origin was in keeping with that generally admitted in literature (2). This finding suggests that only a few if any of the families investigated by our standardized procedure were erroneously classified as being sporadic. It is noteworthy that, in this population of consecutive patients, we also found that 11% of the lupus probands had another autoimmune disorder and that the prevalence of familial cases rose to 23% when autoimmune disorders other than lupus were considered among relatives. This suggests that both organ-and nonorgan-specific autoimmune diseases may share some susceptibility genes, and families with multiple cases of various autoimmune disorders should certainly be considered in future linkage analysis. Moreover, the coexistence of SLE, DLE, and primary antiphospholipid syndrome cases in some of our multiplex families suggests that a common genetic background may predispose to those clinically closely related disorders. To evaluate the hypothesis that a stronger genetic background could influence disease expression, we then compared the main SLE-related manifestations in multiplex and sporadic cases. Indeed, while some genetic studies (4,13) performed in other autoimmune disorders have pointed out distinct clinical features in familial cases, it has not been well established whether familial SLE cases share different characteristics or not (3,10). Since it is known that the lupus phenotype may be influenced by ethnic origin (1), we considered only patients of French Caucasian origin for comparisons in our study and did not observe any significant difference between multiplex and sporadic SLE cases (see Table 2). Unexpectedly, there was even a trend for a higher prevalence of renal and neurologic involvement among sporadic SLE (see Table 2). However, since all sporadic SLE probands came from the same tertiary referral center, and multiplex patients were recruited in different departments throughout the country, a recruitment bias cannot be excluded. Another potential bias is that SLE might have been diagnosed earlier in some multiplex probands since 58% of them had a prior case of lupus in their family. A possible clustering of end-organ phenotypes within individual families was not evaluated in this study. When intra-aggregation of disease was analyzed in our multiplex families, affected sib pairs were preeminent. They represented nearly 50% of all multiplex families (see Table 1), whereas in the literature review of 201 SLE familial cases reported by Arnett (2), parent-offspring pairs were slightly more frequently observed than affected sib pairs. The diversity of familial relationships together with the finding of large extended pedigrees suggesting a Mendelian pattern of inheritance emphasizes the genetic heterogeneity of SLE. Moreover, when the presence of positive ANA was taken into account among clinically unaffected relatives, the penetrance of an “immunologic phenotype” became strikingly higher. Although the significance of positive ANA in clinically healthy individuals and the cutoff point of 1:80 serum dilution are both still controversial (21), one can hypothesize that clinically unaffected relatives with positive ANA may share some lupus susceptibility gene(s) with their clinically affected relatives. However, since only a small fraction of ANA-positive healthy individuals also harbor anti-DNA antibodies, several other loci probably have to be exercised in order to generate anti-DNA antibodies. Thus, we believe that in linkage analysis currently in progress, clinically unaffected relatives should be analyzed differently depending on whether they have an “autoimmune phenotype” or not. Another important issue that needs to be addressed is how lupus extended pedigrees might be better analyzed in genetic studies. Indeed, in the last few years, nonparametric methods have been developed to allow genetic studies of multifactorial diseases in which the mode of inheritance is not known. Although such methods seem helpful for analyzing a large number of small pedigrees, such as affected sib pairs (24,27), extended pedigrees with 3 or more affected relatives probably would benefit more from classical parametric methods of linkage (19). In conclusion, analysis of the 125 lupus multiplex families we have recruited to date suggests a genetic heterogeneity that may reflect the clinical heterogeneity of lupus. In most of the recent genetic linkage studies performed by several groups on large series of SLE multiplex families, clinical data about affected family members are not provided, and clinically unaffected relatives sharing an “autoimmune phenotype” are not considered. Since it is known in murine SLE models that a single susceptibility gene may contribute to a particular phenotype (15), we believe that an accurate analysis of human lupus multiplex families will be helpful in the future to estimate the relative contribution of newly identified susceptibility genes or loci to disease occurrence and expression. Summary Evidence for a genetic susceptibility to systemic lupus erythematosus (SLE) in humans is based on the high concordance rate observed in identical twins and on the relatively high incidence of familial cases. Although recent genetic studies have lead to significant advances in the identification of new susceptibility genes in SLE, no large clinico-pathologic study of familial SLE has been reported to date. In the present study, we describe the main clinical and immunologic features of 125 lupus multiplex families including at least 2 cases of SLE and/or discoid lupus erythematosus (DLE), recruited through a French national survey starting in July 1997. Medical records of all affected members were reviewed by the same investigator, all available family members were interviewed using the same standardized procedure, and blood was drawn for autoantibodies typing. Clinical and immunologic features of 90 probands from multiplex SLE families were compared with those of 100 sporadic SLE patients sharing the same French Caucasian origin. The 125 lupus multiplex families included 282 affected members (2.3 patients per family); of the 125 families, 96 were of French Caucasian origin. One hundred multiplex families included 2 affected relatives, while 25 included 3 or more affected individuals. The relationship between affected members was sibs (45%), parent-offspring (31%), and second-degree (24%). An autosomal dominant mode of inheritance was strongly suggested in 1 extended pedigree with 6 clinically affected members, and a recessive pattern was suspected in 5 other families. No obvious mode of inheritance could be suspected in most of the remainder. Among French Caucasians, sex ratio, mean age at onset, and clinical and biologic SLE-related manifestations were not significantly different in multiplex compared with sporadic SLE cases. The analysis of these 125 multiplex families suggests a genetic heterogeneity that should be considered for ongoing genomic screening.