Sjögren’s Syndrome Highlights

Alan Baer, M.D.

Abstract 476: Lymphomas in a North American Sjögren's Syndrome Population

Authors:

J Kang, L Fisher and F Vivino, University of Pennsylvania, Philadelphia, PA

Background:

Lymphoma is a potential complication of Sjögren’s syndrome (SS), occurring in up to 5% of large cohorts of affected patients. This study sought to describe lymphomas in a North American population with 1° or 2° SS.

Method:

Using a database of 527 patients with 1° or 2° SS diagnosed according to the American European Consensus Group (AECG) criteria and followed from 1/2/2003 to 2/30/2009, those with a concurrent diagnosis of lymphoma were selected and retrospectively studied.  The clinical features of these patients were ascertained through medical record review.

Results:

Nineteen patients had both lymphoma (prevalence 3.6%) and 1° or 2° SS.   The mean age was 57.3 ± 15.2 years (18 – 79 years) at the first visit.  Most patients were female (78.9%), Caucasian (94.7%), and had 1° SS (89.5%).  The patient with 2° SS had CREST.  The majority were ANA (94.7%) and SSA (68.4%) positive, with less being SSB (31.6%) positive.  Six were SSA and SSB negative (31.6%).  All had extraglandular involvement.

The mean age at lymphoma diagnosis was 56.1 ± 18.2 years (20 – 82 years).  All but one patient developed sicca symptoms prior to their lymphoma diagnosis; this patient had Hodgkin’s lymphoma diagnosed 33 years prior to the onset of sicca symptoms, and received chemotherapy and radiation.  Eighteen of 19 patients had a single course with no evidence of recurrence during follow up.  One patient had a right parotid gland MALT lymphoma, followed by recurrence on the contralateral side 6 years later. Most lymphomas presented as persistent salivary gland swelling (52.6%) or adenopathy (31.6%).  MALT lymphoma was most common (57.9%), all of which were extranodal except one.  Diffuse large B cell lymphoma occurred in 21%, and undifferentiated non-Hodgkins lymphoma in 11%. No fatalities were noted during the study period.  An average of 9.9 ± 8.1 years elapsed from the onset of sicca symptoms to the time of lymphoma diagnosis. Patients with salivary gland swelling usually had MALT lymphoma while those with localized or diffuse lymphadenopathy usually did not. 

Conclusions:

In a North American SS population defined by the AECG criteria, MALT and diffuse large B cell lymphomas were most commonly observed.  T cell lymphomas or fatalities were not noted.  Although lymphomas are a significant and at times fatal complication of SS, awareness of warning signs and prompt diagnosis may ensure a good prognosis in the North American SS population.

Editorial comment:

Sjogren’s syndrome is the autoimmune disease with the highest incidence of lymphoma. The risk is likely to be higher in a cohort of patients such as this one, with more complicated disease, receiving care at a specialized clinic in a University medical center. This study highlights several important features of this association. Lymphoma occurred in patients with extraglandular disease and was heralded by persistent salivary gland swelling or lymphadenopathy. Most were MALT lymphomas arising in extranodal sites and having a good prognosis.

Abstract 479: Manifestations of Pediatric Sjögren's Syndrome: Comparison to Adult Sjögren's Syndrome

Authors:

N Yokogawa, S Bout-Tabaku, SM Lieberman, F Alawi, JP Palazzo, L Fisher, FB Vivino, DD Sherry, University of Pennsylvania, The Children's Hospital of Philadelphia, Children's Hospital of Philadelphia, Thomas Jefferson University, Philadelphia, PA

Background:

Sjogren’s syndrome is rare in childhood. This study sought to compare the clinical and serologic features of childhood and adult Sjögren's syndrome (SS).

Methods:

A retrospective chart review was performed at the Children's Hospital of Philadelphia (CHOP) and Penn Sjögren's Syndrome Center (PSSS) to identify patients with childhood SS. The diagnosis of pediatric SS was based on: 1) Diagnosis or symptom onset at age <20 yrs 2) consensus agreement on diagnosis by pediatric and adult rheumatologists and 3) exclusion of other diagnoses. The clinical manifestations, serologic features, and SS Disease Damage Index of the pediatric patients were compared to those of adult patients from the PSSS database.

Results:

The pediatric cohort included 30 patients from CHOP and 5 from the PSSS diagnosed from 1994-2008. Twenty-eight had 1°SS and 7 had 2° SS (3 SLE, 3 UCTD, 1 JIA). Presenting manifestations included parotid swelling 43%, polyarthralgias 20%, neurological symptoms 11%, and sicca 11%.  Sicca symptoms were present in 37% at time of diagnosis and occurred in 44% during follow-up. Complications included fetal heart block during teen pregnancy (n=1) and non Hodkins B cell lymphoma (n=1).  Adult pts (n=410, diagnosis by AECG criteria) included 378 with 1°SS and 32 with 2° SS (13 CREST, 10 RA, 4 SLE, 2 UCTD, 1 Diffuse SCL, 1 MCTD, 1 DM). The mean age of the pediatric SS patients was 13 years and that of the adult patients was 55 years. Females constituted 86% of the pediatric and 93% of the adult SS groups. Caucasian ethnicity was more prevalent among the adult SS pts (84 vs 57%, p<0.001). Sicca symptoms occurred in 89% of the adult and 71% of the pediatric patients while parotitis occurred in 66% of the pediatric and only 23% of the adult patients. Extraglandular manifestations such as fever (17 vs 2%), lymphadenopathy (46 vs 15%), neurological (20 vs 9%), renal disease (14 vs 4%), cutaneous vasculitis, inflammatory eye disease (11 vs 2%) and serological abnormalities such as SSA/SSB (83 vs 56%), ANA (91 vs 69%), RF (72 vs 41%), and low C3/C4 (42 vs 23%) were all more common in pediatric than adult SS. Dry eyes, dry mouth, and leucopenia were more common in the adult SS patients. Arthritis occurred with equal frequency in the pediatric and adult groups (54-57%). The SS Disease Damage Index was not statistically different between the two groups.

Conclusions:

Like the adult disease pediatric SS can cause significant organ damage but sicca symptoms are typically absent in the majority of patients at the time of diagnosis.

Editorial comment:

This study serves to highlight a group of children, usually teenage girls, who have an autoimmune disease characterized by parotid gland swelling, arthritis, antinuclear antibodies, and SSA/SSB antibodies. Without long term followup of these children, it is uncertain whether this disease will evolve into typical systemic lupus erythematosus, or remain as a connective tissue disease best characterized as Sjogren’s syndrome.

Abstract 480: Uncoupling Salivary Gland Dysfunction From Inflammation in a Large Cohort of Primary Sjögren's Syndrome Patients

Authors:

NP Nikolov, I Alevizos, M Grisius, J Atkinson, A Cotri, BJ Baum, J Brahim, D Kleiner, L Bebris, GG Illei, NIH/NIDCR, Bethesda, MD, University of Maryland, Baltimore, MD, and NCI, Bethesda, MD

Background:

The mechanism of salivary gland hypofunction has not been delineated. While it is generally assumed that infiltration of the salivary gland by inflammatory cells is the direct cause of this dysfunction, there has been a poor correlation between the degree of this inflammation and salivary hypofunction, suggesting that other mechanistic pathways may be important. This study sought to evaluate the association between salivary gland inflammation and hypofunction in a large cohort of primary Sjögren's syndrome (pSS) patients.   
Method:

A cohort of 393 patients with pSS underwent concomitant evaluation of salivary gland inflammation and function. Salivary gland inflammation was quantified by focus score (FS) on minor salivary gland (MSG) biopsy. Salivary gland function was quantified by calculating salivary flow rates (ml/15min) after standardized collection from individual parotid, submandibular and sublingual glands in the unstimulated state (total glandular unstimulated salivary flow, USF) and after 2% citric acid stimulation (total glandular stimulated salivary flow, SSF).

Results:

A statistically significant but very weak negative correlation was observed between FS and glandular USF (r2=0.11, p<0.0001) and stimulated salivary flow (r2=0.15, p<0.0001). A similar negative correlation was observed between whole unstimulated salivary flow and FS (r2=0.21, p<0.0001, n=87). Salivary gland inflammation as measured by minor salivary gland focus score explained only 11-21% of the loss of glandular function. Multiple regression analyses showed no significant contribution of potential confounders, such as concomitant medications, duration of sicca symptoms and age, to the model.

Conclusion:

There was a statistically significant, but not clinically meaningful, negative correlation between the degree of salivary gland inflammation, as measured by FS on MSG biopsy, and glandular dysfunction, as measured by both USF and SSF, in a large cohort of pSS patients. These data suggest that salivary inflammation and dysfunction may be two different processes in the pathogenesis of SS. This observation may have significant implications for evaluating disease models, selecting appropriate outcome measures and designing clinical trials of pSS.

Editorial Comment:

This study emphasizes the poor correlation between the degree of salivary gland inflammation and the extent of saliva production. There may be multiple reasons for this. First, the extent of inflammation in the labial glands may not be an accurate reflection of that in the larger submandibular and parotid glands, responsible respectively for the bulk of basal and stimulated saliva production. Second, the dysfunction of the salivary gland may not be a direct consequence of the inflammatory infiltration. Thus, antibodies to muscarinic acetylcholine receptors and cytokine effects on acinar or cholinergic nerve function may be more important.

Abstract 487: Population-Based Prevalence of Primary Sjögren's Syndrome: A 5-Source Capture–Recapture Estimate

Authors:

C Maldini, R Seror, O Fain, R Dhote, X Mariette, MDe Bandt, J-L Delassus, G Falgarone, V Le Guern, F Lhote, O Meyer, J Ramanoelina, K Sacré, Y Uzunhan, J-L Leroux, L Guillevin and A Mahr, Hospital Cochin, Paris, France, Hospital J. Verdier, Bondy, France, Hospital Avicenne, Bobigny, France, Hospital Bicêtre, Kremlin-Bicêtre, France, Hospital R. Ballanger, Aulnay-sous-Bois, France, Hospital Delafontaine, Saint-Denis, France, Hospital Bichat, Paris, France, Hospital Montfermeil-Le Raincy, Montfermeil, France, Service médical Assurance-Maladie, Bobigny, France

Background:

The prevalence of primary Sjögren's syndrome (pSS) is uncertain. Widely divergent prevalence estimates of 0.02–3% for general populations were generated in earlier studies but were based on small numbers of identified pSS cases. The aim of this population-based study was to estimate the year 2007 prevalence of pSS.

Method:

A cross-sectional study was performed of Seine–Saint-Denis County, a Parisian suburb (France), with 1,093,515 adult inhabitants (≥15 yr old). Cases were ascertained from: 1) hospitals; 2) community-based practitioners; 3) positive anti-SSA/SSB antibody testing by private laboratories; 4) a nationwide patient support group; and 5) a national health insurance database. Recorded case diagnoses were verified against the American–European Consensus Group (AECG) criteria; to account for inconsistent availability of information on objective ocular or oral dryness, “enlarged AECG criteria” (based on the presence of ≥3/4 remaining AECG items) were defined. Five-source capture–recapture analyses (CRA) using log-linear modeling was performed to estimate the number of cases missed by any source.

Results:

The 5 sources yielded 1,385 notifications of potential pSS cases; among them, 343 charts could not be accessed for detailed examination. Excluded cases included those with non-pSS diagnoses, pSS cases diagnosed after 2007, and cases not meeting classification criteria, residing outside the study area or for whom no follow-up data were available for the study period and thereafter.  The retained cases included 122 subjects who met the AECG criteria and 183 subjects who met the enlarged AECG criteria. Numbers of missed cases calculated by CRA were 18.0 and 37.6, respectively. The year 2007 prevalence estimate, adjusted for missed cases, was 128.0 per million adults (95% CI: 107.7–150.1) for AECG criteria and 201.7 per million adults (95% CI: 176.3–229.6) for enlarged AECG criteria.

Conclusion:

This population-based study estimated pSS prevalence at 128–202/million (0.013–0.020%) adults. Although this rate must be interpreted with the possibility of undiagnosed pSS in mind, it agrees with the low prevalence obtained by the few other large population-based surveys and might represent a more accurate indication of the true pSS prevalence.

Editorial comment:

Sjogren’s syndrome is thought by some investigators to be the most common systemic rheumatic disease, occurring in up to 2% of the population. The current study suggests that the prevalence may be 100-fold less. These widely disparate estimates reflect the use of different diagnostic criteria, varying methods to identify cases in a population, and possible genetic differences in various populations. The current study also relied heavily on the recording of a physician diagnosis of Sjogren’s syndrome for inclusion in the prevalence estimate. Clearly, Sjogren’s syndrome may not be recorded as a hospital discharge diagnosis or as a relevant diagnosis for a national health insurance database, leading to an underestimate of the prevalence. Additionally, many cases of Sjogren’s syndrome remain undiagnosed unless specific testing is performed.

Abstract 499: Blocking LFA-1 with Efalizumab Promotes Inflammation, T Cell Activation and Autoimmunity in Primary Sjögren's Syndrome (pSS)

Authors:

NP Nikolov, K Nagaraju, S Ghimbovschi, M Alba, L Lapteva, D Kleiner, J Clayton, I Alevizos, M Grisius, L Bebris and GG. Illei, NIH/NIDCR, Bethesda, MD, Children's National Medical Center, Washington, DC, FDA, Silver Spring, MD, NCI, Bethesda, MD

Background:

The LFA-1/ICAM-1 interaction is important in lymphocyte migration to inflammatory sites, T cell activation and antigen presentation. In SS, activated lymphocytes have increased expression of LFA-1 whereas activated endothelial cells in the salivary and lacrimal glands have increased ICAM-1 expression. In this pilot study, the safety, clinical efficacy and biology of blocking this interaction was evaluated in primary SS (pSS) subjects with weekly efalizumab (1mg/kg), a recombinant humanized monoclonal antibody against the alpha subunit of LFA-1.

Method:

The study had three phases: Phase 1 (Week 1-12)-Randomized, Double Blinded, Placebo Controlled; Phase 2 (Week 13-24)-Open Label; and Phase 3 (Week 25-32)-Follow up. Primary outcome was measured by objective improvement in salivary and lacrimal flows and minor salivary gland (MSG) inflammation (focus score) at the end of Phase 1. Gene expression was profiled from PBMC total RNA collected at baseline and at the end of the double-blinded, open-label and follow-up phases.

Results:

After 9 pSS subjects were enrolled, 6 in the efalizumab group and 3 in the placebo group, (Table) the study was terminated due to reports of fatal JC viral infection in psoriasis patients. Of the 9 subjects, 8 were females. 7 subjects completed Phase 1, 4 in the efalizumab group, with only 1 responder in the placebo group. 6 subjects entered phase 2, but only 4 completed this phase, 2 in the efalizumab and 2 in the placebo group. There were no responders in phase 2. Repeat MSG biopsy at the end of Phase 1 showed significant increase in inflammation in EFLZ treated subjects. Two subjects developed anti-dsDNA antibodies during efalizumab treatment, one of whom developed multiple auto-antibodies and a lupus-like syndrome which resolved after efalizumab discontinuation. Serum immunoglobulins increased significantly in 3 subjects, 2 developed IgG monoclonality and one oligoclonality.

582 genes were significantly differentially expressed after efalizumab treatment. CD86-CTLA4, IL-2 and Wnt/Beta-catenin signaling pathways were the most significantly altered in efalizumab treated subjects.

Conclusion:

Efalizumab exacerbates SS and induces significant increase in local inflammation and systemic autoimmunity. The gene expression data suggest that hyperactivation of peripheral T cells may be responsible for the systemic B cell activation and auto-antibody generation. Further studies may elucidate how inhibition of LFA-1/ICAM-1 signaling drives this profound immune system perturbation.

Editorial comment:

This important study serves as a reminder that monoclonal antibody therapy beneficial for one autoimmune disease, e.g. psoriasis, may cause an exacerbation of a different autoimmune disease, Sjogren’s syndrome. The mechanisms responsible for this paradox need to be elucidated. If efalizumab therapy in Sjogren’s syndrome can cause measurable disease flares after only a few weeks, then it is hoped that other monoclonal antibody therapies can be developed that will ameliorate the disease with equal rapidity. The withdrawal of efalizumab from the market also serves to highlight the rare potential for biologic agents to increase the risk of rare infections, such as JC virus.

Abstract 508: Association of Primary and Secondary Sjögren's Syndrome in Lupus Families

Authors:

R Aggarwal, RH Scofield, University of Oklahoma Health Sciences Center, Oklahoma Medical Research Foundation, and Dept Veterans Affairs Med Center, Oklahoma City, OK

Background:

Systemic lupus erythematosus (SLE) and Sjögren’s syndrome are closely related diseases, but the relationship of primary and secondary Sjögren’s syndrome within families with SLE is not known.  This study sought to determine whether primary Sjögren’s syndrome is increased among SLE-unaffected, and whether primary and secondary Sjögren’s syndrome are associated within families with SLE.

Methods:

Families were collected through the Lupus Family Registry and Repository.  All SLE patients met the 1982 revised classification criteria for SLE, and all families had at least one member with SLE.  All subjects completed a standardized questionnaire, and had autoantibodies determined by ELISA and double immunodiffusion. Sjögren’s syndrome was considered present when subjects met the combined US-European Sjögren’s classification criteria for dry mouth as well as dry eyes and had anti-Ro/SSA.

Results:

Among a total of 2537 SLE patients, there were 505 (19.9%) SLE patients with secondary Sjögren’s syndrome. There were 65 SLE-unaffected relatives of SLE patients who had primary Sjögren’s syndrome, while none of 1304 age- and sex-matched healthy controls had Sjögren’s (χ2=12.7, p=0.02).  Seventeen (26.2%) of the 65 subjects with primary Sjögren’s had an SLE-affected relative with secondary Sjögren’s syndrome, compared to 493 (7.1%) of 6922 SLE-unaffected family members without primary Sjögren’s (χ2=34.8, p<0.00001, odds ratio=5.0).

Conclusion:

Primary Sjögren’s was present more often among relatives of SLE patients than in healthy controls.  Furthermore, primary Sjögren’s syndrome in SLE-unaffected members was associated with secondary Sjögren’s within these SLE families.  Thus, there was a familial distribution of primary and secondary Sjögren’s.  Family members affected by primary Sjögren’s were about 5 times more likely to have an SLE-affected relative with secondary Sjögren’s than family members without primary Sjögren’s. These data suggest common susceptibility factors, possibly genetic, for primary and secondary Sjögren’s syndrome.

Editorial comment:

This study provides evidence for familial factors that increase susceptibility to Sjogren’s syndrome. Importantly, these traits influenced susceptibility to the development of the Sjogren’s phenotype, either alone or in the context of SLE. Further work will be needed to determine if these traits are genetic in origin, and if so, their correlation with specific phenotypic characteristics of the Sjogren’s syndrome. Environmental factors, such as shared viral exposures, are also a possible explanation for these findings.

Abstract 1959: STAT4 Polymorphism Is a Confirmed Genetic Risk Factor for Sjögren's Syndrome That Could Be Involved in Type 1 Interferon Pathway Signalling

Authors:

C Miceli-Richard, N Gestermann, P Loiseau, , X Puéchal, E Hachulla, J-E Gottenberg, and X Mariette, Hôpital Bicêtre, Le Kremlin Bicêtre, France, Hôpital Saint-Louis, Paris, France, Centre Hospitalier du Mans, Le Mans, France, National scleroderma centre, Lille Cedex, France, University Hospital of Strasbourg, Strasbourg, France

Purpose:

Signal transducer and activator of transcription 4 (STAT4) is a transcription factor activated by interleukin 12 and to a lesser extent by type 1 IFN which promotes secretion of type 2 IFN by Th1 cells. STAT4 gene polymorphisms have been significantly associated with rheumatoid arthritis, lupus and primary Sjögren's syndrome (pSS). The functional consequences of STAT4 associated polymorphisms are poorly understood. This study aimed to confirm the association between STAT4 gene polymorphism and pSS and to assess the functional relevance of the at-risk genotypes on STAT4 mRNA level of expression (α and β subunits).

Method:

The STAT4 rs7582694 polymorphism was analyzed in an exploratory cohort of 190 pSS patients and 152 controls, and in a second independent cohort of 192 pSS patients, all of Caucasian origin. STAT4α and STAT4β are alternatively spliced isoforms of STAT4 which may differ in their functional properties. STAT4α and STAT4β mRNA levels were assessed in PBMCs from 30 pSS patients and correlated with the level of expression of PKR, MX1 and IFITM1 as surrogate markers of the type 1 IFN induction pathway.

Results:

The STAT4 rs7582694 C allele was found among 28% of pSS patients compared with 17% in controls [P=1.4x10-3 - OR 1.83 (95% CI 1.26 – 2.65)]. In the replication pSS cohort, STAT4 rs7582694 C allele frequency was comparable (30%). The overall OR for pSS in the combined cohorts was 1.92 (95% CI 1.37 – 2.68 ; P=1.10-4). This OR was even higher when considering pSS patients and controls homozygous for the rs7582694 C allele, suggesting a recessive effect of the STAT4 at-risk allele: OR 3.96 (95% CI 1.38 – 11.30.20; P=6x10-3). The association was present both in patients with and without anti-SSA/SSB antibodies. STATα and STAT4β mRNA levels in PBMCs were not significantly determined by rs7582694 genotypes (P=0.99 and P=0.17, respectively) but STAT4α was strongly correlated with the IFN type I induced genes mRNA levels: PKR (P=4x10-3), MX-1 (P=2x10-4) and IFITM-1 (P=8x10-3).

Conclusion:

These results confirm a replicated association of STAT4 polymorphism with pSS with a potential recessive effect on disease susceptibility. These results do not support the functional relevance of the at-risk genotypes on STAT4α and STAT4β mRNA level of expression. Unexpectedly, STAT4α mRNA expression was highly correlated with the level of expression of type 1 interferon induced genes, supporting a possible involvement of STAT4 not only in type 2 IFN but also in type 1 IFN production.

Editorial comment:

The identification of a genetic association of STAT4 polymorphisms with systemic lupus spurred several investigators to seek a similar association in Sjogren’s syndrome. This and other studies confirm this association in a small number of primary Sjogren’s syndrome patients, using a candidate gene approach. It thus supports the importance of type 2 interferon in the pathogenesis of Sjogren’s syndrome. Future studies will be needed to correlate STAT4 gene polymorphisms with specific phenotypic characteristics of Sjogren’s and to confirm the association in much larger populations.

AddThis Social Bookmark Button