Peripheral Neuropathies Associated With Primary Sjögren Syndrome

Immunologic Profiles of Nonataxic Sensory Neuropathy and Sensorimotor Neuropathy

Abstract

We conducted this study to characterize the relationship between primary Sjögren syndrome (pSS)-associated peripheral neuropathy (PN) and markers of B-cell monoclonal proliferation and chronic activation.

The cohort included 120 consecutive patients presenting with definite pSS according to the American-European Consensus Group criteria. Serum markers of chronic B-cell activation included autoantibodies and hypergammaglobulinemia. Markers of monoclonal B-cell proliferation included mixed cryoglobulin, monoclonal gammopathy, abnormal κ/λ free light chain (FLC) ratio, and B-cell non-Hodgkin lymphoma (B-NHL). Definite PN was present in 30 patients (25%) including 7 patients (23%) with sensorimotor neuropathy, 3 patients (10%) with ataxic sensory neuropathy, and 20 patients (67%) with nonataxic sensory neuropathy. Patients with a sensorimotor neuropathy differed from those without PN by higher rates of monoclonal B-cell proliferation markers, that is, mixed cryoglobulin (57% vs. 11%; p = 0.008), monoclonal gammopathy (71% vs. 17%; p = 0.004), higher FLC ratio (2.7 ± 1.5 vs. 1.7 ± 1.8; p = 0.024), and B-NHL (57% vs. 3%; p < 0.001). Patients with nonataxic sensory neuropathy were characterized by a higher age (57.5 ± 10.7 vs. 48.7 ± 14.3 years; p = 0.007), more frequent central nervous system (CNS) involvement (15% vs. 2%; p = 0.04) and a lower prevalence of chronic B-cell activation serum markers, that is, antinuclear antibodies (ANA) (60% vs. 90%; p = 0.003), anti-SSA (Ro) (40% vs. 72%; p = 0.009), anti-SSB (La) (15% vs. 41%; p = 0.039), rheumatoid factor (37% vs. 67%; p = 0.02), and hypergammaglobulinemia (35% vs. 64%; p = 0.023). In multivariate analysis, sensorimotor neuropathy was associated with the presence of B-NHL (odds ratio [OR], 39.0; p < 0.001), whereas nonataxic sensory neuropathy was associated with the presence of CNS involvement (OR, 17.0; p = 0.025) and ANA (OR, 0.07; p < 0.001).

In conclusion, we found that up to 25% of pSS patients presented with PN, predominantly sensory neuropathy. Distinctive immunologic profiles were found according to the type of SS-associated neuropathy: nonataxic sensory neuropathy was marked by a low prevalence of B-cell activation markers, and sensorimotor neuropathy was marked by a high prevalence of B-cell monoclonal proliferation markers.

Abbreviations: ANA = antinuclear antibodies, B-NHL = B-cell non-Hodgkin lymphoma, CI = confidence interval, CNS = central nervous system, ENA = extractable nuclear antigen, ENMG = electroneuromyography, FLC = free light chain, OR = odds ratio, PN = peripheral neuropathy, pSS = primary Sjögren syndrome, RF = rheumatoid factor.

INTRODUCTION

Primary Sjögren Syndrome (pSS) is a systemic autoimmune disease characterized by glandular involvement including xerophthalmia, xerostomia, and cutaneous xerosis, and by potential systemic extraglandular organ involvement. Chronic and polyclonal B-cell activation is frequent, as evidenced by the presence of hypergammaglobulinemia, rheumatoid factor (RF), and autoantibodies, mainly antinuclear (ANA), anti-SSA (Ro), and anti-SSB (La) antibodies.^12,25 This may evolve toward monoclonal B-cell proliferation, leading to the occurrence of mixed cryoglobulin and monoclonal gammopathy, together with an increased risk of B-cell non-Hodgkin lymphoma (B-NHL).^10,38

Among the extraglandular manifestations occurring in pSS, peripheral neuropathies (PNs) are frequent, with a prevalence close to 20%.^{2,13,15,16,22} In pSS, a variety of PN types, such as ataxic sensory neuropathy, nonataxic sensory neuropathy, and sensorimotor neuropathy, can be differentiated according to clinical presentation, pathophysiologic mechanisms, and electrophysiologic alterations in conventional nerve conduction study. Nonataxic sensory neuropathies include painful nonataxic sensory polyneuropathy, pure small-fiber neuropathy, and trigeminal neuropathy. Subtypes of sensorimotor neuropathies are sensorimotor polyneuropathy and multiple mononeuropathy. Other forms of PN, more rarely associated with pSS, are polyradiculoneuritis, pure autonomic neuropathy, or motor neuron disorder.^6,27 Although the prevalence and clinical features of pSS-associated PNs are relatively well defined, the relationships with specific immunologic profiles, in terms of B-cell activation or proliferation, are not well established.

In the present study we identified the different forms of pSS-associated PN in a single-center cohort of 120 patients with definite pSS. We then assessed the association of the different pSS-associated PNs with specific immunologic profiles according to the presence or absence of markers of chronic B-cell activation and monoclonal proliferation.

PATIENTS AND METHODS

Patients

We studied 120 consecutive patients fulfilling the American-European Consensus Group criteria for pSS³⁷ and followed in the Department of Internal Medicine of a university hospital between 1985 and 2009. Patients were followed for a mean of 50 months (51 mo for non-neuropathic and 46 mo for neuropathic patients). Up to 40% of patients were diagnosed with pSS after being treated as an inpatient at the university hospital; most of the patients were referred to the university hospital after the pSS diagnosis was made. Among the 30 patients with PN, 11 patients (37%) were initially followed in the Neurology Department, where the diagnosis of pSS-associated PN was made, and then referred to the Department of Internal Medicine for management.

The 120 patients were followed in an outpatient clinic at the university hospital. For each patient, all medical records obtained during the entire follow-up were reviewed and epidemiologic, clinical, laboratory, immunologic, pathologic, and electrophysiologic data were cumulatively registered in the study case report form. Cases of definite PN were secondarily classified into ataxic neuropathy, nonataxic sensory neuropathy, or sensorimotor neuropathy according to clinical and electro-neuro-physiologic features.¹¹ For all neuropathic patients, other causes of PN, such as diabetes, amyloidosis, non pSS-related systemic vasculitis, and paraproteinemic neuropathies, were ruled out at the time of the diagnosis of PN.

The extraglandular manifestations noted were cumulative over the 50-month period of observation. This cumulative registration of pSS-associated glandular and extraglandular manifestations explains the high prevalence of extraglandular manifestations (96/120, 80%), including B-NHL (9/120, 7.5%), and PN (30/120, 20%). Among the 9 patients with lymphoma, 6 had PN. In all cases, the neuropathy was diagnosed before the diagnosis of lymphoma.

Clinical Neurologic Examination

Neurologic symptoms including sensory (numbness, burning, prickling, paresthesia, dysesthesia, and allodynia) and motor symptoms were recorded. Clinical examination included a search for physical signs affecting any sensory modality (pain, warm, cold, touch, vibration, and proprioception), the motor system (weakness and atrophy), or the tendon reflexes (especially ankle jerks).

Central Nervous System Involvement

Attributing central nervous system (CNS) involvement to pSS is difficult in the absence of objective neurologic symptoms and signs. Therefore, only patients presenting with objective CNS symptoms associated with significant CNS magnetic resonance imaging (MRI) abnormalities were considered to have pSS-associated CNS involvement. T2 hyperintense lesions on MRI, which are associated with increasing age, were not considered to be related to pSS. (This may explain why few patients had CNS involvement, regardless of age, in the current study.)

Neurophysiologic Assays

Electroneuromyography (ENMG) (conventional motor and sensory nerve conduction study) was performed by the same physician (TM) in all neuropathic patients.²⁴ The test was performed in 44 patients (37%), including 30 patients with pSS-associated PN and 14 patients without pSS-associated PN. Among these 14 patients, the ENMG was normal in 11 patients and abnormal in 3 patients (2 carpal tunnel syndrome, 1 amyotrophic lateral sclerosis). For the 11 patients with a normal ENMG, the ENMG was justified by the presence of a) diffuse paresthesia for 4 patients (2 patients with normal laser-evoked potentials and 2 patients who were lost to follow-up before 2008), b) myalgia for 5 patients, and c) mechanical radiculopathy for 2 patients.

When a pure small-fiber neuropathy was suspected, additional neurophysiologic testing was performed by the same physician (JPL), including somatosensory laser-evoked potentials, quantitative sensory testing to thermal stimuli, and autonomic nervous system testing.²³ The diagnosis of small-fiber neuropathy was established according to the criteria of Devigili et al.⁷ (Neurophysiologic tests for small-fiber exploration were available only from 2008.)

Immunochemistry

Antinuclear antibodies (ANA) were detected using indirect immunofluorescence on HEp-2000 cells (Immuno Concepts, San Diego, CA), with a positive result defined as ≥1:80. Anti-extractable nuclear antigen antibodies (anti-ENA), including anti-SSA (anti-Ro 52/60) and anti-SSB (anti-La), were detected using the Line Immuno Assay INNO-LIA ANA (Innogenetics, provided by InGen) before 2006 and a multiplexed microparticle-based Luminex immunoassay (AtheNA Multi-Lyte, InGen, Antony, France) after January 2006. Doubtful results for anti-ENA assays were verified with the Ouchterlony test, which is the reference method. Serum cryoglobulin detection and immunochemical typing were performed with a validated immunoblotting method.²⁹ RF was determined by enzyme-linked immunosorbent assay (ELISA) (BMD, Marne la Vallée, France), and results greater than 20 IU/mL were considered positive. Serum free light chain (FLC) κ and λ concentrations were measured by nephelometry on a Dade-Behring BN II Analyser (Freelite, The Binding Site, Birmingham, UK). FLC results were compared with the published reference range for the FLC ratio (κ/λ: 0.26-1.65) and the proposed renal failure reference range (0.37-3.1).²¹ The dosage of serum immunoglobulins (IgG, IgA, IgM) was performed by nephelometry (Dade-Behring BN II).

Serum markers of chronic B-cell activation included the presence of autoantibodies-ANA, anti-SSA (Ro), anti-SSB (La), RF-and hypergammaglobulinemia. Monoclonal B-cell proliferation was defined by at least 1 of the following: mixed cryoglobulinemia with a monoclonal component, monoclonal gammopathy, an abnormal FLC ratio, or B-NHL that was defined according to World Health Organization criteria.²⁰

Most of the immunologic tests were performed at the time of clinical service. Only the measurement of FLCs was performed on stored sera, using the oldest available.

Statistical Analysis

The subgroups of patients with SS-associated PN (PN+) were compared with the subgroup of patients without SS-associated PN (PN−), which was considered the control group. Categorical variables were compared using the Fisher exact test, and continuous variables using the Mann-Whitney test. Based on the results of these univariate analyses, variables were included in multivariate analyses using a stepwise multiple logistic regression analysis to assess independent associations. The level of significance (p value) was set at 0.05. All analyses were performed using MedCalc software version 10.0.1.0 (Mariakerke, Belgium).

RESULTS

Characteristics of the Study Population

The main characteristics of the study population are reported in Table 1. Briefly, the 120 pSS patients included 106 female (88%) and 14 male patients (12%). The mean age at diagnosis was 50.4 ± 14 years. Xerostomia was present in 87% of patients, xerophthalmia in 93%, and a positive minor salivary glands biopsy in 80%. The B-cell activation markers included positive ANA, present in 85% of patients, anti-SSA (Ro) antibodies in 67.5%, anti-SSB (La) antibodies in 37%, RF in 61%, and hypergammaglobulinemia in 56%. A mixed cryoglobulin was present in 12.5% of patients, monoclonal gammopathy in 17.5%, and B-NHL in 7.5%.

TABLE 1: Main Features of 120 Patients With Primary Sjögren Syndrome

Thirty patients (30/120, 25%) presented with definite PN, which included 7 cases (7/30, 23%) of sensorimotor neuropathy, 3 cases (3/30, 10%) of ataxic sensory neuropathy, and 20 cases (20/30, 67%) of nonataxic sensory neuropathy. The 20 patients with nonataxic sensory neuropathy included 7 patients (7/20, 35%) with nonataxic sensory polyneuropathy, 11 patients (11/20, 55%) with pure small-fiber neuropathy, and 2 patients (2/20, 10%) with isolated trigeminal neuropathy. The clinical and laboratory features were those noted at the time of the PN diagnosis.

Fifteen of the 30 neuropathic patients received steroids after the diagnosis of pSS-associated PN. In non-neuropathic patients, 36% (32/90) received steroids during the course of the disease. The difference is not statistically significant (p = 0.2).

Features of pSS-Associated Sensorimotor Neuropathy

Compared to the patients without PN, the 7 patients with sensorimotor neuropathy had a lower prevalence of hypergammaglobulinemia (14% vs. 64%; p = 0.01) and a higher prevalence of monoclonal B-cell proliferation markers-that is, mixed cryoglobulin (57% vs. 11%; p = 0.008), monoclonal gammopathy (71% vs. 17%; p = 0.004), higher mean FLC ratio (2.7 ± 1.5 vs. 1.7 ± 1.8; p = 0.024), abnormal FLC ratio (3/4 = 75% vs. 11/46 = 24%; p = 0.06), and B-NHL (57% vs. 3%; p < 0.001). After multivariate analysis, only the presence of B-NHL remained associated with the presence of sensorimotor neuropathy (odds ratio [OR], 39.0; 95% confidence interval [CI], 5.9-255.0; p < 0.001) (Table 2).

TABLE 2: Main Features of Patients With pSS-Associated Sensorimotor Neuropathy (SMN+) and Those Without pSS-Associated Peripheral Neuropathy (PN−)

Features of pSS-Associated Nonataxic Sensory Neuropathy

Compared to the patients without PN, the 20 patients with nonataxic sensory neuropathy were older (57.5 ± 10.7 vs. 48.7 ± 14.3 yr; p = 0.007), and more often had xerostomia (100% vs. 82%; p = 0.04) and CNS involvement (15% vs. 2%; p = 0.04). They less frequently had chronic B-cell activation serum markers with a lower prevalence of ANA (60% vs. 90%; p = 0.003), anti-SSA (Ro) (40% vs. 72%; p = 0.009) and anti-SSB (La) antibodies (15% vs. 41%; p = 0.039), RF (39% vs. 67%; p = 0.02), and hypergammaglobulinemia (35% vs. 64%; p = 0.023). They also had lower serum levels of total gammaglobulins (12.4 ± 4.9 vs. 16.2 ± 7.3 g/L; p = 0.008), IgG (12.8 ± 5.0 vs. 16.8 ± 7.4 g/L; p = 0.008), and IgM (0.9 ± 0.4 vs. 1.5 ± 0.98 g/L; p = 0.0025). After multivariate analysis, the presence of ANA (OR, 0.07; 95% CI, 0.02-0.3; p < 0.001) and CNS involvement (OR, 17; 95% CI, 1.4-194.0; p = 0.025) remained associated with the presence of nonataxic sensory neuropathy (Table 3).

TABLE 3: Main Features of Patients With pSS-Associated Nonataxic Sensory Neuropathy (SN+) and Those Without pSS-Associated Peripheral Neuropathy (PN−)

DISCUSSION

In the current study, pSS-associated PN was diagnosed based on clinical and electro-neuro-physiologic data. PN was found in 25% of 120 consecutive patients with definite pSS. Nonataxic sensory neuropathies represented 67% of pSS-associated PNs-much more than sensorimotor neuropathies, which accounted for 23% of cases, and ataxic sensory neuropathies, 10% of cases. Our results are in keeping with those of previous reports where the prevalence of PN in pSS was close to 20%, ranging from 7% to 55%.^{2,3,13,15,16,22,31,36} Nonataxic sensory neuropathies represented up to 30% of pSS-associated PN, ranging from 23% to 95% of cases; ataxic sensory neuropathies, also called ganglionopathies or neuronopathies, represented 20% (range, 5%-39%); and sensorimotor neuropathies, 25% (range, 5%-72%).^{2-4,6,15,16,18,22,27,34}

The main objective of the present study was to assess the relationships between specific immunologic profiles regarding B-cell activation or monoclonal proliferation markers and subtypes of pSS-associated PN. The first immunologic hallmark of pSS is usually the presence of chronic B-cell activation, as evidenced by the frequent elevation of serum levels of gammaglobulins, mainly IgG; the presence of B-cell-dependant serologic markers of autoimmunity-that is, ANA, anti-SSA (Ro) and anti-SSB (La) antibodies and RF; and the formation of ectopic lymphoid tissue with germinal center-like structures. The second immunologic hallmark of pSS is the long-term tendency to develop an oligo and monoclonal B-cell proliferation, as evidenced by the presence of mixed cryoglobulin, monoclonal gammopathy, and/or B-NHL.¹⁹

These 2 different immunologic features are thought to be a continuum of the same underlying pathogenesis mediated by the induction of B-cell-activating factors, such as Bcl2 and translocation (14;18),^9,14,33 and the B lymphocyte stimulator (BLyS/BAFF), which is associated with serologic markers of chronic B-cell activation in pSS²⁶ and in malignant B-cell proliferation.³⁰ Thus, we wondered whether the different patterns of pSS-associated PN could have distinctive features regarding these 2 pSS-associated immunologic hallmarks.

Not surprisingly, compared to patients without PN, patients with pSS-associated sensorimotor neuropathy more frequently had positive serum markers of monoclonal B-cell proliferation, that is, mixed cryoglobulin, monoclonal gammopathy, and B-NHL (Table 2). They also had higher mean FLC ratio (2.7 ± 1.5 vs. 1.7 ± 1.8; p = 0.024), a marker associated with pSS activity¹⁷ and B- and plasma-cell proliferation.⁸ Main pSS-associated sensorimotor neuropathies have been shown to be related to a cryoglobulin-mediated vasculitic process.^27,34 Mixed cryoglobulin and monoclonal gammopathy are predictive of B-NHL in pSS patients.^1,35 Conversely, patients with pSS-associated sensorimotor neuropathy shared with patients without PN the same B-cell activation profile, with a comparable prevalence of ANA, anti-SSA (Ro) and anti-SSB (La) antibodies, and RF. The only exception was a lower prevalence of hypergammaglobulinemia, which can be explained by the high prevalence of B-NHL in this group of patients.³² One can hypothesize that patients without PN and those with pSS-associated sensorimotor neuropathy belong to the same subgroup of pSS characterized by a high level of chronic B-cell activation, which may be complicated by monoclonal B-cell proliferation, and a vasculitic process-mediated sensorimotor neuropathy. These results suggest that a benign or malignant B-cell proliferation should be assessed in case of pSS-associated sensorimotor neuropathy.

We also analyzed the relationship between pSS-associated sensory neuropathies and pSS-associated immunologic profiles. Because only 3 patients had ataxic sensory neuropathy, which is a distinctive form of pSS-associated PN, we excluded them and analyzed only those patients with pSS-associated nonataxic sensory neuropathy. These latter patients, when compared to those without PN, were characterized by a lower prevalence of positive serum markers of B-cell activation-that is, ANA, anti-SSA (Ro), anti-SSB (La), RF, and hypergammaglobulinemia (Table 3). They were older and more frequently had CNS involvement. On the other hand, although not significant, rates of markers of monoclonal B-cell proliferation were very low (Table 3).

A low prevalence of ANA and anti-SSA (Ro) antibodies has been reported previously in patients with SS-associated sensory neuropathy. Grant et al¹⁸ found ANA in only 10% of patients with pSS-associated PN, leading the authors to hypothesize that the association between isolated sicca syndrome and PN could correspond to a specific syndrome, distinctive from Sjögren syndrome. In a series of patients with SS-associated PN, Mori et al²⁷ reported that only 39% of the cases of painful sensory neuropathy had positive anti-SSA (Ro) antibodies, compared to 53% of cases of ataxic sensory neuropathy, and 64% of cases of multiple mononeuropathy. Similarly, in the study by Chai et al,⁴ 46% of patients with suspected pSS-associated small-fiber neuropathy had positive anti-SSA (Ro) antibodies compared to 55% in other forms of pSS-associated PN. Together, these results suggest that patients with pSS-associated nonataxic sensory neuropathy may belong to a distinctive subset of pSS patients, characterized by the absence or the low prevalence of immunologic markers of chronic B-cell activation. The American-European Consensus Group criteria for pSS include a seronegative subset (anti-SSA [Ro]-negative) of pSS defined by sicca syndrome and a focal lymphocytic sialadenitis.³⁷ In 2008, Ramos-Casals et al³¹ showed that the subset of anti-SSA (Ro)-negative patients was older and less often had serologic markers of chronic B-cell activation and systemic complications, including PN, confirming previous reports.^5,28 Whether this subset of anti-SSA (Ro)-negative pSS patients has a peculiar neurotropism targeting the sensory neurons remains to be determined.

In conclusion, the current study helps to demonstrate the great heterogeneity involved in pSS-associated PN, in terms of phenotypic and immunologic profiles. Sensorimotor neuropathies are associated with a high prevalence of B-cell monoclonal proliferation markers (mixed cryoglobulin, monoclonal gammopathy, B-NHL) and chronic B-cell activation. Conversely, nonataxic sensory neuropathies are marked by a low frequency of chronic B-cell activation markers-that is, ANA, anti-SSA (Ro), anti-SSB (La), RF, and hypergammaglobulinemia, suggesting that these patients may belong to a subset of pSS with a peculiar peripheral sensory neurotropism.