Sjögren’s Syndrome without focal lymphocytic infiltration of the salivary glands

Abstract

Background.

Primary Sjögren’s syndrome (SS) is characterized by a focal lymphocytic infiltrate in exocrine glands. We undertook this study to describe patients that lacked this key feature.

Methods.

We evaluated subjects with sicca in a comprehensive clinic at which medical, dental and ophthalmological examinations were performed. All subjects underwent a minor salivary gland biopsy with focus score calculation. Extra-glandular manifestations were also determined. We categorized subjects as high, intermediate, or low in terms of the expression of interferon-regulated genes.

Results.

About 20% (51 of 229, 22%) of those classified as primary Sjögren’s syndrome had a focus score of zero. Compared to those with anti-Ro positivity and a focus score >1.0, the focus score zero patients (who by classification criteria must be anti-Ro positive) were statistically less likely to have anti-La (or SSB) and elevated immunoglobulin as well as less severe corneal staining. In addition, the focus score zero patients were less likely to have elevated expression of interferon-regulated genes in peripheral blood mononuclear cells than anti-Ro positive SS subjects with a focal salivary infiltrate.

Conclusions.

There are only a few clinical differences between primary Sjögren’s syndrome patients with focus score zero and those with both anti-Ro and a focus score >1.0. Nonetheless, the small subset of focus score zero patients tested did not have elevated expression of interferon-regulated genes, but did have systemic disease. Thus, extra-glandular manifestations are perhaps more related to the presence of anti-Ro than increased interferon. This may have relevance to Sjögren’s syndrome pathogenesis.

BACKGROUND

Sjögren’s syndrome is a chronic autoimmune disease characterized by lymphocytic infiltration of exocrine glands. Salivary and lacrimal gland involvement presenting as sicca syndrome is the most common manifestation of SS. Involvement of other exocrine glands like those of respiratory, gastrointestinal and genitourinary tracts can be often seen concomitantly with Sjögren’s sicca (1). The extraglandular manifestations of the disease include arthralgia, arthritis, Raynaud’s phenomena, lymphadenopathy, small airway disease, systemic vasculitis, interstitial nephritis, splenomegaly, myositis, peripheral neuropathy and lymphoma (2). The current burden of Sjögren’s syndrome in the United States is around 2–3 million making it the second most prevalent inflammatory rheumatological disease (3).

Presence of focal lymphocytic infiltration of salivary glands and anti-Ro/SSA serum autoantibodies are the cardinal features of Sjögren’s syndrome and are a part of the current classification criteria (4, 5). Focal lymphocytic infiltration is measured by focus score, which is defined as the number of mononuclear cell infiltrates containing at least 50 inflammatory cells per 4 mm² glandular section. A focus score of one or greater is considered positive.

We found a significant percentage of primary Sjögren’s syndrome subjects in our large cohort, all of whom undergo minor salivary gland biopsy and histological examination, had a focus score of zero. Thus, these patients lacked one of the two cardinal features of Sjögren’s syndrome. We undertook this study to investigate the differences (if any) between the focus score zero primary Sjögren’s subjects with those with non-zero focus scores.

METHODS

We evaluated individuals with sicca in the Oklahoma Medical Research Foundation (OMRF) Sjögren’s Syndrome Research Clinic, which has previously been described in detail (6–9). Subjects were evaluated by a rheumatologist, an ophthalmologist (or optometrist) and a dentist. The dentist performed an oral examination consisting of measurement of stimulated and timed whole unstimulated salivary flow (WUSF), a lip biopsy and collection and storage of saliva. Participant evaluation did not include sialography or scintigraphy. The ocular specialist performed ocular surface staining with Lissamine green and fluorescein, an unanesthetised Schirmer’s I test, and collection and storage of tears. The ocular vital dye score was determined using the quantitative dot-counting method according to both the van Bijsterveld (10) and ocular staining score methods (11), rather than by descriptive features. A physician completed a detailed history and physical examination, including general medical, rheumatological and neurological evaluations. Blood samples were collected for extraction and storage of DNA, RNA and serum. Anti-Ro (or SSA) and anti-La (or SSB) autoantibodies were determined by multiple methods, as described (6). Additionally, all patients were tested for rheumatoid factor (RF), antinuclear antibodies (ANA), autoantibodies associated with other connective tissue disorders, hepatitis C serology, complete blood count (CBC) with differential, immunoglobulin profile and urinalysis. If patients gave a history of a past diagnosis of rheumatoid arthritis, mixed connective tissue disease, systemic sclerosis, myositis, primary biliary cirrhosis, multiple sclerosis, or systemic lupus erythematosus, classification criteria for these illnesses were specifically ascertained by history, medical record review and testing for the corresponding autoantibodies. Minor salivary gland pathology was determined by a dental pathologist with calculation of focus score following the method of Daniels (12) as follows: number of lymphocytic foci adjacent to normal appearing acini with >50 cells observed ÷ area studied (mm²) × 4 mm². As delineated in the recent guidelines (13), which we followed, focus score calculation does not take into account non-focal lymphocytic infiltration, which may be present despite a focus score of zero.

Subjects were classified according to the AECG and ACR/EULAR criteria for Sjögren’s syndrome (4, 5, 14, 15), both of which require one of anti-Ro/La or a focus score >1.0. Thus, all subjects classified as Sjögren’s syndrome with a focus score <1.0 must have anti-Ro under the ACR/EULAR criteria or either anti-Ro or anti-La under the AECG criteria. Clinical diagnosis of Sjögren’s syndrome is by expert opinion and does not rely on these classification criteria, which are formulated for research purposes only. All subjects had a clinical diagnosis of Sjögren’s syndrome made at the time of the evaluation in the OMRF Sjögren’s Syndrome Research Clinic.

We determined mRNA expression by microarray, as we have previously described (16, 17). We categorized individual subjects as having high, intermediate or low expression of interferon-regulated gene expression by unsupervised hierarchical clustering using centroid linkage with uncentered correlation, again using techniques previously described (18). We used Cluster 3.0 (19) for clustering of gene expression and visualized these results in Java TreeView (18).

We compared the focus score zero primary Sjögren’s syndrome patients in our cohort with those with focus scores of less than one and greater than or equal to one, in terms of classification criteria and extraglandular manifestations, using Statistical Analysis System software for data analysis. We used chi-square or Fisher’s exact test for comparison between the groups. Understanding the need for correction for multiple comparisons we agreed upon an alpha of 0.005 for statistical significance, so p values less than 0.005 were considered statistically significant. Since this was an investigative study we used the alpha of 0.005 rather than Bonferroni correction, which is more conservative in its estimation, and therefore more likely to result in false negative results. Given the a priori hypothesis for the interferon-regulated gene expression studies and the small number of statistical tests, no correction for multiple comparisons was made here.

All procedures were approved by the Oklahoma Medical Research Foundation Institutional Review Board (approval # 11–03). Each participant provided written informed consent prior to entering the study.

RESULTS

Among 229 subjects classified as primary Sjögren’s syndrome, we found 51 focus score zero subjects (FS=0, 22%), 167 focus score positive subjects (FS≥1, 73%) and 11 subjects with focus score between 0 to 1 (0<FS<1, 15%). No patient with a focus score of zero had a gland replaced by fibrosis, and all subjects had adequate tissue for examination and calculation of focus score. All subjects with a focus score less than 1 had anti-Ro/SSA in their sera. This is, of course, by definition of the criteria, which require either a focus score ≥1.0 or autoantibodies. For focus score positive patients, we only included those with positive anti-Ro in the study in order to match with the focus score zero patients, all of whom have anti-Ro. Nonetheless, FS=0 subjects were statistically less likely to have anti-La (or SSB) detected than among those with FS≥1 subjects (10/51, 19.6% versus 98/167, 58.7%, p<0.001). There was only a statistical trending for the 0<FS<1 subjects to have less anti-La comparing to FS≥1 subjects (Table 1).

Table 1.

Classification criteria among primary Sjögren’s syndrome subjects with a focus score of >1 (FS>1), a focus score of zero (FS=0), and a focus score between zero and one (0<FS<1). All subjects answered at least one dry question and one dry mouth question from the AECG criteria positively. WUSF = whole unstimulated salivary flow

	FS>1	FS=0	0<FS<1	P values
	n=167	n=51	n=11	>1 v 0	>1 v <1
Anti-La/SSB	98 (59%)	10 (20%)	4 (36%)	<0.0001	0.145
WUSF	103† (63%)	36 (71%)	6 (55%)	0.309	0.160
Schirmer’s	85** (52%)	18* (36%)	6 (55%)	0.05	0.140
Lissamine Green	115‡ (72%)	25 (50%)	8 (73%)	0.0032	0.152

^*not performed in 1 subject

^†not performed in 4

^**not performed in 3

^‡not performed in 7

We next examined the presence or absence of the classification criteria (Table 1). When considering the AECG sicca criteria, all subjects answered at least one dry mouth and one dry eye question positively. Thus, there was no statistical difference in the presence of sicca symptoms among the groups. Likewise, there was no difference between the presence of an abnormal Schirmer’s test among the 3 groups; however, abnormal Lissamine green corneal staining was statistically more common among those subjects with FS ≥1.0 compared to those with a FS of zero (p = 0.0032; Table 1). Whole unstimulated salivary flow did not differ between the three groups (Table 1).

When comparing manifestations of the disease not captured in the classification criteria but part of the Sjögren’s Syndrome Disease Activity Index (SSDAI) (20, 21), we found only a few differences between the groups. An elevated serum IgG, which is a criterium within the Biological Domain of the SSDAI, was significantly higher in focus score positive subjects as compared to those with focus score zero (p = 0.0003; Table 1). Otherwise, neither salivary gland enlargement nor other extraglandular manifestations differed among the three groups (Table 2).

Table 2.

Manifestations from the European Sjögren’s syndrome Disease Activity Index among primary Sjögren’s syndrome subjects with a focus score of >1 (FS>1), a focus score of zero (FS=0), and a focus score between zero and one (0<FS<1).

	FS>1	FS=0	0<FS<1	p values
	n=167	n=51	n=11	>1 v 0	>1 v <1
Parotid	36	9	0	0.558	0.150
Submandibular	21	9	0	0.349	0.150
Lymphadenopathy	8	4	0	0.179	0.448
Constitutional	50	22	3	0.07	0.174
Peripheral Neuropathy	51	14	4	0.690	0.230
Persistent Cough	50	17	3	0.62	0.264
Arthralgia	105	42	8	0.0094	0.232
Arthritis†	12	4	1	0.231	0.424
Arthritis‡	21	6	0	0.889	0.293
Raynaud’s‡	52	14	2	0.633	0.260
Leucopenia	19	5	1	0.763	0.420
Neutropenia	0	0	0	NA	NA
Lymphopenia	7	2	1	0.311	0.370
Thrombocytopenia	2	1	0	0.414	0.822
Hyper IgG*	61	5	3	0.0003	0.114
Hyper IgA	26	7	3	0.759	0.177
Hyper IgM	28	9	1	0.883	0.311
Low C3	0	0	0	NA	NA
Low C4	7	2	0	0.311	0.674
Other Antibodies	51	10	3	0.133	0.253
Hypo IgG	5	2	0	0.307	0.674

^†Documented in history and physical by physician

^‡Self-reported on questionnaire

^*We used 20 grams/liter as the demarcation for an elevated value because in the Biological Domain of the SSDAI a serum IgG level >20g/L is considered positive. That is, the Biological Domain is considered active if the serum IgG level is above this concentration.

Previously published results show that increased expression of interferon-regulated genes in peripheral blood cells is highly correlated with the presence of anti-Ro among primary Sjögren’s syndrome subjects (16, 17, 22). We hypothesized that this might not be the case among those with anti-Ro but no salivary gland mononuclear cell infiltrate. In fact, we found that focus score zero Sjögren’s syndrome subjects were less likely to have high expression of interferon-regulated gene than anti-Ro-positive subjects with a focus score ≥1.0 (Table 3). We had peripheral blood gene expression data on 6 of the 47 Sjögren’s subjects with a focal score of zero and none clustered with the high interferon group. Among those with a focal score ≥1 and anti-Ro positivity, 31 of 47 were in the high interferon group by the cluster analysis (p=0.001 by Fisher’s exact test). Four of the focus score zero, anti-Ro-positive Sjögren’s subjects had low interferon-regulated gene expression and 2 had intermediate by the cluster analyses. Meanwhile, among the comparison group of anti-Ro-positive, focus score ≥1 Sjögren’s subjects, only 15 were intermediate and 5 low in expression of interferon-regulated genes in peripheral blood cells (Table 3).

Table 3.

Interferon-regulated gene expression categorized by unsupervised hierarchical clustering among primary Sjögren’s syndrome – some with a focus score >1 on minor salivary gland pathology and some with a focus score of zero. All had anti-Ro. Gene expression was performed on a random sample from FS=0 and FS>1 subjects.

	Interferon-regulated gene expression
	High	Intermediate	Low
Anti-Ro, FS=0 (n=6)	0*†	2	4
Anti-Ro, FS≥1 (n=47)	31	15	5

FS= focus score

^*p=0.001 comparing high vs low/intermediate

^†p=0.02 comparing distribution across high, intermediate and low interferon regulated

DISCUSSION

Focus score is the quantitative measure of focal sialadenitis; and, along with serum anti-Ro antibodies, is the cardinal feature of Sjögren’s syndrome. Presence of one of the two is required to classify patients in a research setting, while a diagnosis of the disease is made on clinical grounds. We found that the two subgroups with and without focal infiltration were not clinically different, other than higher degree of corneal staining with Lissamine green, serum anti-La antibodies and elevated IgG in FS≥1 patients. Hypergammaglobulinemia. Elevated IgG. and serum anti-La antibodies are highly associated with the presence of anti-Ro (23). Higher Lissamine green staining without significantly different Schirmir’s test in these patients points to greater degree of corneal involvement even in the absence of markedly different tear production. This may be due to alteration in protein secretion in tears causing a qualitative loss of function.

There are few previous study that have analyzed Sjögren’s patients without focal sialadenitis. These studies have found associations of high focus score with clinical manifestations of the disease, but generally these studies have parsed Sjögren’s syndrome patients as focus score above 3 or 4 and less than 3 or 4. For instance, a recent study found that a focus score >4.0 was associated with interstitial lung disease (24). Other work has also found associations of focus scores above 3 with lymphoma (25). A longitudinal study found focus score ≥1.0 was associated with worsening Schirmer’s test results (26). A study of 265 Sjögren’s subjects showed an association of positive focus score with stimulated salivary flow as well as decayed/missed/filled (DMF) teeth (27). We did not examine stimulated salivary flow or DMF teeth. Carubbi and colleagues (28) examined a large group of Sjögren’s patients (n=794), of whom 72 (19%) had a focus score of zero. These investigators found a number of differences between focus score zero subjects and those with a focus score ≥1.0 including xerostomia, salivary gland enlargement, hematological involvement, central nervous system involvement, and hypergammaglobulinemia. However, no corrections for multiple comparisons was made in this study, which, given the lack of an a priori hypothesis concerning any differences between the groups, we thought necessary to perform. When examining the results from Carubbi, et al. we find that only hypergammaglobulinemia and hematological involvement remain statistically significant after applications of the correction for multiple correction that we employed. Carubbi, et al. found focus score ≥1 to be associated with lymphoma using univariate logistic regression (28). We did not have enough patients with lymphoma (only one in our cohort) to make any assessment. Thus, our study and the previous ones comparing those with and without a positive focus score are largely in agreement, especially when considering that hypergammaglobulinemia and elevated IgG are generally correlative.

Our findings do have certain implications for clinical diagnosis of Sjögren’s syndrome. The diagnosis is inherently difficult as it requires multiple specialties, and it is often difficult to find expertise for Sjögren’s syndrome in a dentist, ophthalmologist and pathologist at one center. Since the FS≥1 and the FS=0 groups are not remarkably different in terms of clinical manifestations, in the presence of sicca symptoms with serum anti-Ro antibody, a clinical diagnosis of SS can be made for all intents and purposes without the need for tissue biopsy. There will, however, remain the subset of seronegative Sjögren’s patients who may require a salivary gland biopsy for clinical diagnosis. At our center, we would rarely make a diagnosis without one of anti-Ro or focal lymphocytic infiltration of salivary glands. Others may not take this approach, and instead make a clinical diagnosis without salivary gland pathology even in the absence of anti-Ro. The research classification will continue to require one of salivary gland biopsy or serum autoantibodies in concomitance with the clinical findings.

These clinical findings along with the difference in expression in interferon-related genes have even greater implications regarding the pathogenesis of the disease. Firstly, manifestation of sicca syndrome without focal infiltration of exocrine glands points towards another pathology of gland dysfunction without gland destruction, perhaps antimuscarinic 3 receptor autoantibodies or altered expression of aquaporin molecules. Clearly, these patients without focal infiltrates do not have immune-mediated destruction of the salivary glands.

Secondly, the difference in interferon regulated gene expression between the two groups delineates differences in underlying pathogenesis. Previous work has found that increased expression of interferon-regulated genes in Sjögren’s syndrome is highly related to anti-Ro positivity. However, we did not find high expression of this set of genes among the patients with anti-Ro but no focal salivary gland infiltrate.

Unsupervised hierarchical clustering analyses showed a statistically significant lower expression of interferon-regulated genes among anti-Ro positive, focus score zero Sjögren’s syndrome compared to anti-Ro positive with a focal infiltrate (Table 3). One possibility is that cells infiltrating the salivary gland are the source of interferon that is driving expression of genes in peripheral blood cells. The data are clear that the salivary glands are a site of development of lymphocytes and that B cells infiltrating the salivary gland produce anti-Ro (29, 30). These cells may also serve as antigen presenting cells and produce interferon. Alternatively, interferon produced in some other location could be critical to the production of anti-Ro as well as to development of the lymphocytic salivary gland infiltrate that characterizes the disease. Finally, there could be factors, such as genetics of genes such as OAS1 (17) that influence all three of anti-Ro, interferon production, and infiltration of the gland.

Sjögren’s syndrome can be divided into those patients with only exocrine gland features versus those who also have systemic or extra-glandular features. Data suggest that the presence of anti-Ro identifies the latter group (31). The FS=0, anti-Ro positive patients had a similar degree of extra-glandular manifestations as the group with focus score ≥1.0 and anti-Ro. Thus, extra-glandular disease associates anti-Ro regardless of the presence focal sialadenitis. However, this was not the case for increased expression of interferon-regulated genes, which was found not among FS=0 patients despite the presence of extra-glandular manifestations. Thus, anti-Ro, not expression of interferon-regulated genes, is the correlate of extra-glandular manifestations. Since almost all subjects did not have a previous diagnosis of Sjögren’s syndrome, medications that might alter interferon, including hydroxychloroquine, were not prescribed.

There are several limitations of our study. We had only a small number of subjects in whom we studied expression of interferon-regulated genes in peripheral blood cells. That is, this was a sample of convenience. So, our results and conclusions must be considered preliminary. We do not have longitudinal data. Furthermore, we do not have reliable data concerning the onset of disease, which we find is difficult to ascertain when talking to patients, many of whom have had sicca for years. Perhaps focal score zero subjects have early disease and will develop focal lymphocytic infiltration over time. Our study does not address whether pathological findings might change over course of the illness, although studies of sequential biopsies have not shown changed pathology (32). Minor salivary gland histopathology is variable in a given subject (33), more extensive examination of multiple tissue sections might have shown focal infiltration in some FS=0 subjects. Finally, pathology in the minor salivary glands may not reflect pathology found in the major salivary glands. We have not examined the major salivary glands in our cohort.

Conclusions

A significant fraction of patients coming to a comprehensive sicca evaluation clinic who were classified as primary Sjӧgren’s syndrome had no focal infiltrate upon pathological examination of minor salivary gland biopsy specimens. These subjects had only a few clinical differences when compared to those with both anti-Ro and a focus score greater than 1. There were less signs of severe dry eyes among the anti-Ro positive focus score negative patients. In addition, we found statistically less anti-La as well less elevated immunoglobulin levels in these subjects. These serological findings are of interest because both are highly associated with the presence of anti-Ro. In addition, we found the primary Sjӧgren’s syndrome subjects with no focal infiltrate but anti-Ro positivity did not have elevated expression of interferon-regulated genes, another characteristic highly correlated with anti-Ro. Thus, these focus score negative, anti-Ro positive patients differ from focus score >1, anti-Ro positive patients in interesting ways that might demonstrate aspects of Sjӧgren’s syndrome pathogenesis.