Portal de Periódicos da CAPES

Primary Sjögren Syndrome

2002; Wolters Kluwer; Volume: 81; Issue: 4 Linguagem: Inglês

10.1097/00005792-200207000-00003

1536-5964

Mario Garcı́a-Carrasco, Manuel Ramos‐Casals, J. Rosas, Lucio Pallarés, Jaime Calvo‐Alén, Ricard Cervera, Josep Font, M Ingelmo,

Salivary Gland Tumors Diagnosis and Treatment

Introduction Sjögren syndrome (SS) is a systemic autoimmune disease that mainly affects the exocrine glands and usually presents as persistent dryness of the mouth and eyes due to functional impairment of the salivary and lacrimal glands (12). In the absence of an associated systemic autoimmune disease, patients with this condition are classified as having primary SS. The histologic hallmark is a focal lymphocytic infiltration of the exocrine glands, and the spectrum of the disease extends from an organ-specific autoimmune disease (autoimmune exocrinopathy) (50) to a systemic process with diverse extraglandular manifestations (11,19,42–44,47). The question that arises is whether epidemiologic factors (such as gender or age at onset of disease) or immunologic patterns, among other factors, are associated with particular disease expressions and define some specific primary SS subsets (10,22,41). Several studies have addressed this problem with results that are controversial due to the small number of patients studied or the different classification criteria applied (14,23,40). We analyzed the prevalence and characteristics of clinical and immunologic manifestations in a large cohort of Spanish patients with primary SS in order to define the different clinical and immunologic patterns of disease expression. (For information on the hematologic features of primary SS, see the article by Ramos-Casals and colleagues in this issue [42a].) Methods Patients Between 1994 and 2000, we investigated the clinical and immunologic characteristics of 400 consecutive patients with primary SS in our units. All patients fulfilled 4 or more of the preliminary diagnostic criteria for SS proposed by the European Community Study Group in 1993 (55) and underwent a complete history and physical examination. Diagnostic tests for SS (rose bengal staining, Schirmer test, parotid scintigraphy, and salivary gland biopsy) were applied according to the recommendations of the European Community Study Group (55), and clinical and serologic characteristics of all patients were collected in a protocol form. Salient features included in the protocol were 1) age at disease onset, defined as the initial manifestation clearly attributable to SS; 2) age at inclusion, defined as the age when the patient entered the protocol; 3) cumulative clinical manifestations during disease evolution (from the onset until protocol inclusion); 4) laboratory findings at protocol inclusion. Definition of clinical features The clinical manifestations evaluated were defined as follows: Articular involvement: arthralgia and/or nonerosive arthritis characterized by tenderness, swelling, or effusion involving 2 or more peripheral joints. We excluded patients with osteoarthritis, nonspecific painful processes, and chronic fatigue. Fever: axillary temperature >37.5 °C, not associated with infectious processes. Lymphadenopathy: enlarged nodes (>1 cm) in the cervical, axillary, or inguinal regions. Cutaneous vasculitis demonstrated by cutaneous purpura and/ or rash. Skin biopsy showing vasculitis when performed. Raynaud phenomenon: intermittent attacks of digital pallor followed by cyanosis and/or rubor of the fingers, toes, ears, nose, tongue, induced by exposure to cold, stress, or both, in the absence of any other associated disease or anatomical abnormality. Peripheral neuropathy: paresthesias, numbness, and/or motor defects of the lower/upper extremities confirmed by electromyography. Sural nerve biopsy showing vasculitis involving vasa nervorum when performed. Nephropathy: presence of persistent proteinuria >0.5 g/day, altered urinalysis (hematuria, pyuria, red blood cell casts), renal tubular acidosis, interstitial nephritis, or glomerulonephritis. Lung involvement: persistent cough and/or dyspnea, with chronic diffuse interstitial infiltrates on X-ray, altered pattern on pulmonary function studies, and/or evidence of pulmonary alveolitis/fibrosis in computed tomography scan. Autoimmune thyroiditis: altered thyroid function with positive antithyroidal autoantibodies. Immunologic studies Immunologic tests included antinuclear antibodies (ANA) (indirect immunofluorescence using mouse liver/kidney/stomach as substrates), antimitochondrial antibodies, antiparietal cell antibodies, antismooth muscle antibodies (SMA), and antiliver-kidney microsome antibodies type-1 (indirect immunofluorescence), precipitating antibodies to the extractable nuclear antigens (ENA) Sm, RNP, Ro/SS-A, and La/SS-B (ELISA), antineutrophil cytoplasmic antibodies (indirect immunofluorescence), antithyroglobulin and antithyroid peroxidase autoantibodies (ELISA), and rheumatoid factor (RF) (latex fixation and Waaler-Rose tests). Complement factors (C3 and C4) were measured by nephelometry (Behring BNA nephelometer) (1). Serum cryoglobulins were measured after centrifugation. Statistical analysis The chi-square and Fisher exact tests were applied to analyze qualitative differences. To compare quantitative parameters, the Student t-test was used for large samples of similar variance and the nonparametric Mann-Whitney U test for small samples. Values of quantitative variables are expressed as mean ± standard error of the mean (SEM). Statistical significance was established at p < 0.05. When several variables appeared to have statistical significance in the univariate analysis, an unconditional logistic regression analysis was performed by multivariate analysis to rule out possible confounding variables. Statistical analysis was performed by SPSS program. Results The cohort consisted of 373 (93%) female and 27 (7%) male patients (female: male ratio, 14: 1). Mean age at the onset of symptoms attributable to the disease was 52.7 ± 0.85 years (range, 15–87 yr) and at the time of protocol inclusion was 58.7 ± 0.72 years (range, 16–87 yr). Disease evolution at inclusion ranged from 3 to 480 months (mean, 74.9 ± 4 mo). The main glandular manifestations are summarized in Table 1. Three hundred and ninety (98%) patients presented with xerostomia; 371 (93%) xerophthalmia, 123 (31%) cutaneous dryness, 70 of 373 (19%) genital dryness, and 73 (18%) parotidomegaly. Moreover, 351 of 371 (95%) were positive according to the 1993 European criteria (55) for ocular diagnostic tests (Schirmer test and/or rose Bengal staining). The 20 patients with negative ocular tests did not report xerophthalmia. Parotid scintigraphy was positive in 253 of 328 (77%) patients, and salivary gland biopsy showed lymphocytic infiltrates (grade 3 or 4) in 169 of 228 (74%) patients.TABLE 1: Demographic and glandular features of 400 patients with primary Sjögren syndromeThe main extraglandular features were articular involvement in 147 (37%) patients, Raynaud phenomenon in 62 (16%), autoimmune thyroiditis in 61 (15%), cutaneous vasculitis in 47 (12%), pulmonary involvement in 37 (9%), peripheral neuropathy in 29 (7%), adenopathy in 28 (7%), renal involvement in 25 (6%), and fever in 24 (6%) (Table 2). Finally, the main immunologic features were ANA in 288 of 392 (74%) patients, anti-Ro/SS-A in 153 of 385 (40%), RF in 146 of 388 (38%), SMA in 102 of 293 (35%), anti-La/SS-B in 102 of 385 (26%), antiparietal cell gastric antibodies in 58 of 289 (20%), low CH50 in 35 of 302 (12%), and cryoglobulinemia in 27 of 293 (9%) (Table 3).TABLE 2: Extraglandular manifestations of 400 patients with primary SSTABLE 3: Immunologic features in patients with primary SSEpidemiologic subsets Male patients: Of the 400 patients with primary SS, 27 (7%) were men. Compared with women, men with primary SS presented a lower prevalence of Raynaud phenomenon (0% vs 17%, p = 0.013), autoimmune thyroiditis (0% vs 16%, p = 0.022), and anti-Ro/SS-A antibodies (15% vs 41%, p = 0.011) in the univariate analysis, although only anti-Ro/SS-A antibodies were a significant independent variable in the multivariate analysis (Table 4).TABLE 4: Frequency of main clinical and immunologic features according to epidemiologic characteristics*Young onset: In 63 (15%) patients, the onset of primary SS occurred before the age of 35 years. Patients with a younger onset presented a higher prevalence of fever (16% vs 6%, p =0.004), lymphadenopathy (19% vs 5%, p = 0.001), RF (53% vs 35%, p = 0.008), and anti-Ro/SS-A antibodies (68% vs 34%, p < 0.001) in the univariate analysis compared with patients with a disease onset after the age of 35 years, although only lymphadenopathy, fever, and anti-Ro/ SS-A antibodies were a significant independent variable in the multivariate analysis (see Table 4). Late onset: In 43 (11%) patients, disease onset occurred after the age of 70 years. Patients with an older onset presented a lower prevalence of positive ocular tests (87% vs 96%, p = 0.047) and anti-Ro/SS-A antibodies (23% vs 41%, p = 0.03) in the univariate analysis compared with patients with a disease onset before the age of 70 years, although none of these variables was a significant independent variable in the multivariate analysis (see Table 4). Long-term evolution: Ninety-five (24%) patients showed a long-term evolution of SS (more than 10 years from the onset of sicca symptomatology until the protocol study). Patients with a longer SS evolution showed a higher prevalence of fever (14% vs 5%, p = 0.005), lymphadenopathy (16% vs 4%, p = 0.001), pulmonary involvement (19% vs 6%, p < 0.001), peripheral neuropathy (18% vs 4%, p < 0.001), cutaneous vasculitis (24% vs 8%, p < 0.001), and positive ocular tests (99% vs 93%, p = 0.03) in the univariate analysis compared with patients with a SS evolution shorter than 10 years, although only fever and peripheral neuropathy were significant independent variables in the multivariate analysis (see Table 4). Immunologic subsets Immunonegative patients: We found negative immunologic markers (ANA, RF, Ro/SS-A, and La/SS-B) in 84 (23%) patients. This “immunonegative” subset of SS patients showed a lower prevalence of adenopathy (1% vs 9%, p = 0.02), Raynaud phenomenon (4% vs 18%, p < 0.001), and cutaneous vasculitis (4% vs 14%, p = 0.007) in the univariate analysis, although only cutaneous vasculitis and Raynaud phenomenon were significant independent variables in the multivariate analysis (Table 5).TABLE 5: Frequency of main clinical features and diagnostic tests according to immunologic pattern*Presence of ANA not related to anti-Ro/La (ANA +/ENA−): We found positive ANA (not related to positive anti-Ro/SS-A and/or anti-La/SS-B) in 141 patients. Compared with patients with negative ANA, ANA+/ENA-patients showed a higher prevalence of pulmonary involvement (14% vs 5%, p = 0.029) and Raynaud phenomenon (13% vs 2%, p = 0.001), and a lower frequency of peripheral neuropathy (2% vs 9%, p = 0.03) in the univariate analysis; all 3 of these variables were significant independent variables in the multivariate analysis (see Table 5). Presence of RF: We found positive RF in 146 (38%) patients. These patients showed a higher prevalence of articular involvement (45% vs 33%, p = 0.017), cutaneous vasculitis (19% vs 7%, p = 0.01), and Raynaud phenomenon (21% vs 12%, p = 0.028) in the univariate analysis, although only articular involvement and cutaneous vasculitis were significant independent variables in the multivariate analysis (see Table 5). Presence of anti-Ro/La antibodies: We found positive anti-Ro/SS-A and/or anti-La/SS-B in 153 of 385 (40%) patients. These patients showed a higher prevalence of adenopathy (12% vs 4%, p = 0.007), peripheral neuropathy (11% vs 5%, p = 0.03), cutaneous vasculitis (17% vs 8%, p = 0.015), Raynaud phenomenon (26% vs 9%, p < 0.001), and positive salivary gland biopsy (84% vs 69%, p = 0.01) in the univariate analysis, although only Raynaud phenomenon was a significant independent variable in the multivariate analysis (see Table 5). Presence of cryoglobulins: We found detectable cryoglobulins in 27 of 293 (9%) patients. These patients showed a higher prevalence of parotidomegaly (41% vs 7%, p = 0.02), adenopathy (23% vs 6%, p = 0.017), cutaneous vasculitis (48% vs 11%, p < 0.001), peripheral neuropathy (26% vs 6%, p = 0.003), and Raynaud phenomenon (33% vs 13%, p = 0.01) in the univariate analysis, although only adenopathy and cutaneous vasculitis were significant independent variables in the multivariate analysis (see Table 5). Other clinical subsets Sicca-limited disease: Clinically, 139 (35%) patients presented only sicca symptomatology (without extraglandular features) while the remaining 261 (65%) presented a systemic process with diverse extraglandular manifestations. When SS patients were compared according to the existence of extraglandular involvement, those with a sicca-limited disease presented a lower prevalence of ANA (66% vs 78%, p = 0.016), anti-Ro/SS-A (32% vs 43%, p = 0.039), anti-La/SS-B (19% vs 30%, p = 0.029), RF (29% vs 42%, p = 0.011), and cryoglobulinemia (4% vs 12%, p = 0.019) in the univariate analysis, although none of these variables was statistically significant in the multivariate analysis (Table 6).TABLE 6: Frequency of main clinical and immunologic SS features according to existence of extraglandular featuresFulfillment of the 1996 European criteria: All diagnostic tests included in the European criteria were performed in 298 patients. Of these patients, 253 (85%) fulfilled the more restrictive 1996 European criteria (56), while the remaining 45 patients fulfilled only the 1993 criteria (55). When SS patients were compared according to the criteria fulfilled, those who fulfilled only 1993 criteria showed a lower prevalence of peripheral neuropathy (0% vs 10%, p = 0.019), positive scintigraphy (64% vs 82%, p = 0.02), ANA (61% vs 81%, p = 0.005), RF (19% vs 48%, p < 0.001), cryoglobulinemia (0% vs 13%, p = 0.017), and hypocomplementemia (0% vs 17%, p = 0.001) in the univariate analysis, although none of these variables was statistically significant in the multivariate analysis (Table 7).TABLE 7: Frequency of main clinical and immunologic SS features according to fulfillment of SS criteria in 298 patients with all diagnostic testsDiscussion In 1965, Bloch et al (8) gave an extensive description of 62 patients with a relatively unknown entity named “Sjögren syndrome,” characterized by sicca symptomatology and articular involvement. The authors included both patients with associated autoimmune diseases (mainly rheumatoid arthritis) and patients with SS “alone” who presented more clinically apparent extraglandular features such as purpura, neuropathy and pulmonary disease. The report of Bloch et al. (8) was thus the first to define the existence of the “primary” form of SS. Thirty-seven years later, we have analyzed the clinical and serologic characteristics of 400 patients with primary SS, the largest series yet studied to our knowledge. We have confirmed that, epidemiologically, primary SS is a systemic autoimmune disease that overwhelmingly affects women (ratio 14:1) of middle age (53 years), with xerostomia, xerophthalmia, ANA, anti-Ro/SS-A, and RF being the most frequent features. Skopouli et al (47) found a similar epidemiologic profile in a series of 261 Greek patients with 96% of women (ratio 24:1) and a mean age at diagnosis of 51 years. One of the most important findings of the present study is the description of primary SS as a disease that can express in many guises depending on specific epidemiologic, clinical, or immunologic features. Epidemiologic differences Various studies have reported that age at onset modifies disease expression of some autoimmune diseases (6,17,26,48,53). In primary SS, studies of children and adolescents (5,7,14,27,28,39,45) have shown that the clinical and immunologic expression differs in various aspects from the disease in adult patients, with a higher frequency of specific clinical manifestations (recurrent parotitis or fever) and positive autoantibodies. In our study, we found a higher frequency of fever, adenopathy, and anti-Ro/SS-A antibodies in SS patients with a disease onset lower than 35 years. Haga et al (23) found that 16 (24%) of 67 SS patients with a disease onset lower than 45 years had a higher frequency of anti-ENA and RF, and Tishler et al (51) also found a higher prevalence of anti-Ro/SS-A and RF in a young-onset SS population. Thus, it seems that younger SS onset is associated with greater immunologic expression. Although the reason for this apparently age-related variability in the expression of the disease is still unclear, demographic factors and differences in genetic predisposition or responsiveness of an aging immune system may play a part (16,25). In fact, we have recently analyzed IL-10 gen promoter polymorphisms in patients with primary SS and found that patients carrying the GCC haplotype had earlier disease onset (47 yr vs 59 yr) (18). In addition, our study showed that an elderly onset of primary SS (≥70 yr) is not uncommon (11% of our patients). The clinical expression of primary SS in elderly onset patients is similar to that of younger patients, although the prevalence of some extraglandular and immunologic features was slightly lower in elderly patients (18). This lower expression of immunologic markers in older-onset patients may reflect senescence of the immune system (18). It is noteworthy that no differences in the results of diagnostic tests were observed between older and younger patients, suggesting that these tests (including salivary gland biopsy) have the same diagnostic validity in older patients as in younger ones. The clinical expression of primary SS in our male patients is similar to that in female patients, although the prevalence of specific extraglandular manifestations such as thyroiditis, Raynaud phenomenon, cutaneous vasculitis, or renal involvement was slightly lower. Similarly, the 3 studies performed in American patients (4,9,33) showed that men and women had the same risk for developing extraglandular complications, while Drosos et al (14) found a lower prevalence of arthritis and Raynaud phenomenon in Greek male patients (Table 8). Gender may also influence the immunologic manifestations of primary SS, and we found a lower frequency of anti-Ro/SS-A in male patients in the multivariate analysis. Similarly, a lower prevalence of immunologic features has been described in 2 previous reports (14,33) (see Table 8). Drosos et al (14) found a lower frequency of anti-Ro/SS-A and anti-La/SS-B antibodies in their male patients, and Molina et al (33) a lower prevalence of RF and anti-Ro/SS-A. Overall, male SS patients show a lower tendency to develop extraglandular features, although the majority of studies have not found statistical differences, and most have shown a lower immunologic expression in males, specifically a lower frequency of anti-Ro/SS-A antibodies.TABLE 8: Clinical and immunologic manifestations observed in male patients with primary SS, previous and present studiesImmunologic differences The use of the 1993 European criteria (55) permits a diagnosis of primary SS with negative immunologic markers. A “seronegative” form of primary SS was described in 1996 (37) as a subset of chronic fatigue syndrome. In our study, we found 84 patients (21%) with a negative immunologic profile (negative ANA, RF, anti-Ro/SS-A, and anti-La/SS-B). The analysis of this “immunonegative” subset of primary SS patients showed a lower frequency of extraglandular features, specifically cutaneous vasculitis and Raynaud phenomenon. Previous studies have also shown different HLA associations between immunonegative and immunopositive patients (13). However, no differences in the results of diagnostic tests in our study were detected between immunonegative and immunopositive patients, suggesting that the immunologic profile seems not to influence the outcome of diagnostic tests. Thus, it is difficult to determine if this immunonegative subset is an “incomplete” or “probable” primary SS or represents a milder form of this heterogenic autoimmune disease, as occurs in other autoimmune diseases such as systemic lupus erythematosus (SLE). We tend toward the latter view, as this subset does not show any statistical differences in the prevalence of sicca symptomatology and positivity for the diagnostic tests, including salivary gland biopsy, upon which the etiopathogenesis of primary SS is based. We identified a second immunologic subset with a noteworthy characteristic clinical profile. Of the 288 patients with positive ANA, 141 (49%) had negative anti-Ro/La antibodies. This ANA+/ENA− subset of patients showed a higher frequency of specific extraglandular features, such as pulmonary involvement and Raynaud phenomenon. A well-known characteristic of the systemic autoimmune diseases is the possibility of evolution from 1 disease to another, and it is possible that some of these ANA+/ENA− patients could evolve to a disease other than primary SS over time, such as systemic sclerosis or SLE, as suggested by Davidson et al (13). We found 146 (38%) patients with positive RF. These patients showed a higher frequency of extraglandular and immunologic features observed in the total cohort, such as articular involvement, cutaneous vasculitis, ANA, and anti-Ro/SS-A. These results support a possible role of RF in the diagnosis of primary SS, since this immunologic marker shows an independent statistical association with most of the main clinical and immunologic SS features, as suggested by previous studies (24,35,36,38). Thus, RF may be considered a useful immunologic test for the diagnosis of some subsets of patients with primary SS, such as those with extraglandular manifestations or with circulating cryoglobulins. We also found a higher prevalence of most extraglandular features in those patients with primary SS and positive anti-ENA antibodies. Previous studies have found similar results in smaller series of patients, with a close correlation between positive anti-Ro/La antibodies and clinical features (parotidomegaly [13,34], lymphadenopathy [13], cutaneous vasculitis [2,3,24,34], and neurologic involvement [2,3,34]), a focus score higher than 1 in salivary gland biopsy (30,57), and serologic markers (hypergammaglobulinemia [13,24,31,34], RF [34], and cryoglobulins [34]). The subset of patients with primary SS and anti-Ro/La antibodies probably represents the most clinically and immunologically “active” presentation of the disease. In these patients, a more exhaustive clinical follow-up could lead to earlier identification and treatment of extraglandular features. Nevertheless, we believe that the presence of Ro/La should not be considered per se a mandatory criterion for primary SS, as its presence probably identifies a specific subset of the disease. Thus, we found that some of the subsets described in our study were mainly Ro/La negative: 22 (85%) of the 26 male patients were negative, as were 33 (77%) of the 43 patients with a late SS onset and 91 (68%) of 134 patients with a sicca-limited disease. The inclusion of positive anti-Ro/La antibodies as an obligatory criterion for primary SS diagnosis limits this diagnosis to a very specific subset of patients, excluding most male patients, those with a late-onset, and those without extraglandular features. Finally, we found a higher prevalence of most of the extraglandular features in patients with cryoglobulinemia, suggesting cryoglobulins play a central role in the systemic expression of the disease. Previous studies have confirmed the role of cryoglobulins as an important prognostic marker in primary SS. Tzioufas et al (52) reported that mixed cryoglobulinemia and monoclonal RF can be used as laboratory predictive factors for lymphoma development in SS, and Skopouli et al (47) reported that the presence of purpura, hypocomplementemia, and mixed cryoglobulinemia were the strongest predictors of lymphoma and death. The role of cryoglobulinemia in the extraglandular features of patients with primary SS without hepatitis C virus infection seems to be well established, although in a previous study we found that almost 50% of patients with cryoglobulinemia and SS presented an associated hepatitis C virus infection (42). Clinical differences Two-thirds of our patients presented extraglandular features while the remaining one-third had a “sicca-limited” disease. Although the figures vary widely (Table 9), we found lower prevalences of the main extraglandular manifestations in our patients than in patients observed in most previous studies (8,15,29,32,46,47,49,54). This may be due to several factors, including the definition of each extraglandular feature (in our study we did not include patients with subclinical disease), the existence of genetic or ethnic differences, the exclusion of patients with associated hepatitis C virus (who frequently presented systemic autoimmune features), and the different criteria used for SS diagnosis. Compared with patients who fulfilled the more restrictive 1996 European criteria (56), those who fulfilled only the 1993 criteria (55) (with negative salivary gland biopsy and anti-Ro/La antibodies) showed a milder form of primary SS, with a lower frequency of extraglandular and analytic features. Nevertheless, these patients fulfilled a minimum of 4 of the 6 1993 European criteria, and their sicca symptomatology (confirmed by the corresponding diagnostic tests) was not related to pharmacologic treatments, viral infections, or non-autoimmune conditions. Thus, in these patients it is difficult to diagnose a disease other than primary SS, and only prospective studies may demonstrate a distinct clinical evolution.TABLE 9: Prevalence of extraglandular features, previous and present studiesThe different criteria currently used for the diagnosis of primary SS lead to different visions of the disease. The United States criteria (20), which are more specific, identify principally those patients with the greatest extraglandular and immunologic expression. In contrast, the 1993 European criteria (55) permit a diagnosis of SS with negative salivary gland biopsy and negative immunology. Intermediate are the 1996 modified European criteria (56), which include the existence of either positive salivary gland biopsy or positive anti-Ro/La as obligatory criteria. We believe that these modified criteria are currently the most appropriate for the diagnosis of primary SS, although in light of the results of the present study we recommend adding ANA and RF to anti-Ro/La antibodies in the obligatory immunologic criteria. We believe that a patient with xerostomia, xerophthalmia, positive ocular tests, positive parotid scintigraphy, and positive ANA/RF (but negative Ro/ La) should be classified tentatively as having primary SS in the absence of coexisting systemic autoimmune diseases (such as SLE, rheumatoid arthritis, systemic sclerosis), although this subset of patients needs a long-term follow-up to detect a possible evolution to other systemic autoimmune diseases. The low frequency of the extraglandular features observed in this study may confirm the relatively “benign” course of the disease suggested by previous studies. Kruize et al (29) reported that only 4 (19%) of 21 patients with primary SS prospectively followed for a minimum of 10 years developed new autoimmune features, and defined primary SS as an autoimmune disease characterized by a stable and relatively mild course. Skopouli et al (47) found that patients without extraglandular features were unlikely to develop them, and that the clinical picture is fairly complete at diagnosis, and Gannot et al (21) also described a slow progression of symptoms over 5–10 years. In conclusion, primary SS is clinically less severe than other systemic autoimmune diseases such as SLE and rheumatoid arthritis, and the management of most patients should be focused on the problems derived from the sicca involvement, using a multidisciplinary approach involving diverse medical specialties (rheumatology, internal medicine, oral care, dermatology, psychiatry, gynecology, family medicine). Summary In this study we analyzed the clinical and serologic characteristics of 400 patients with primary Sjögren syndrome (SS), to our knowledge the largest series to date. We confirmed that, epidemiologically, primary SS is a systemic autoimmune disease that overwhelmingly affects women of middle age; xerostomia, xerophthalmia, antinuclear antibodies (ANA), antiRo/SS-A, and rheumatoid factor (RF) are the most frequent features. We note that several other manifestations not included in the 1993 European criteria such as articular involvement (37%), antismooth muscle antibodies (35%), parotidomegaly (18%), Raynaud phenomenon (16%), autoimmune thyroiditis (15%), cutaneous vasculitis (12%), and hypocomplementemia (12%) were frequently found. One of the most important aspects of the present study is the description of primary SS as a disease that can be expressed in many guises depending on specific epidemiologic, clinical, or immunologic features. We found a greater immunologic expression in SS patients with a younger onset and a lower expression in male patients. Immunologically, the multivariate analysis showed that the presence of any 1 of 4 immunologic markers (ANA, RF, Ro/La, cryoglobulins) is statistically related to the existence of 1 or more of the main extraglandular features of primary SS, such as lymphadenopathy (cryoglobulins), articular involvement (RF), cutaneous vasculitis (RF), pulmonary involvement (ANA), and Raynaud phenomenon (ANA and anti-Ro). In addition, the immunonegative subset of primary SS patients presented a lower prevalence of the main extraglandular manifestations. In clinical practice, patients with primary SS and positive immunologic features probably represent the subset of patients needing a more specific follow-up, with special attention to the development of systemic manifestations. Clinically, we observed 2 main patterns of disease expression in our cohort of patients with primary SS: first, a subset of patients with a disease limited to glandular involvement and with a lower frequency of immunologic abnormalities and extraglandular features; second, a subset of patients with a “systemic” involvement beyond the sicca involvement with a higher presence of immunologic and extraglandular manifestations. It is noteworthy that no differences in the positivity of diagnostic tests, including salivary gland biopsy, were observed between the 2 subsets. As SS is a systemic autoimmune disease, its clinical presentation may be expressed in many forms: some patients predominantly present a classic sicca complex while others have a systemic disease with diverse analytic and extraglandular features. This second subset is easily diagnosed by the currently used sets of diagnostic criteria, while those patients with a predominantly sicca-limited disease, especially male, elderly, and immunonegative patients, are more frequently diagnosed using the 1993 European criteria. The greater heterogeneity in the clinical presentation of primary SS observed in this study shows that our understanding of this systemic autoimmune disease is still evolving. Although the 1993 European criteria probably cover best the clinical heterogeneity of primary SS, we believe the diagnosis of primary SS should include the presence of 1 of these 2 classification criteria: 1) a positive salivary gland biopsy or 2) a positive immunology test (ANA, RF, and/or anti-Ro/La autoantibodies) if salivary biopsy was negative or not performed. Thus, the 1996 modified criteria (with the inclusion of ANA and RF) would, in our judgment, be the best criteria currently available for the diagnosis of primary SS. Acknowledgment The authors thank David Buss for his editorial assistance.