Clinical manifestations

Presentation and course

Sicca symptoms are one of the most frequent causes of ocular and oral complaints. Sjögren syndrome may occur in conjunction with other autoimmune disorders, including rheumatoid arthritis and systemic lupus erythematosus. In this setting, it is known as associated Sjögren or Sjögren-overlap syndrome. Moreover, Sjögren syndrome is a serious disease with an excess of mortality mainly due to hematological cancer. Sjögren syndrome is more frequent than may be thought, affecting an estimated 2 to 4 million people in the United States and with a prevalence in Europe of 0.6% to 3.3%. Sjögren syndrome overwhelmingly affects middle-aged women, although children, men and the elderly are also affected (101).

Systemic involvement in Sjögren syndrome includes chronic fatigue affecting as many as 50% of patients; arthralgias present in 53%; hematological abnormalities including anemia and leukopenia (33%), increased erythrocyte sedimentation rate (22%), hypergammaglobulinemia (22%); myalgias (22%); and skin lesions such as “burning skin” (18%), cutaneous vasculitis (10%), papular lesions, and annular erythema (63; 105). Lymphoma may develop in 5% to 7% of patients with primary Sjögren syndrome, often associated with cutaneous purpura (140). Less common manifestations include lung involvement with lymphocytic alveolitis, lymphocytic interstitial pneumonitis, fibrosis, pulmonary pseudolymphoma, and pulmonary hypertension. Other less common manifestations include pericarditis, vascular lesions manifested by Raynaud phenomenon, renal tubular lesions, interstitial nephritis and glomerulonephritis, malabsorption due to lymphocytic infiltrates of the intestine, mild pancreatitis, and hepatitis. Exclusion criteria for the diagnosis of Sjögren syndrome include infections by HIV, HTLV-1, or hepatitis C virus (104).

Classification Criteria are updated as new tools are validated. The most recent classification criteria for Sjögren syndrome were finalized in 2016 and approved by the American College of Rheumatology (ACR) Board of Directors and the EULAR Executive Committee (120). Criteria are based on the following:

First, a patient must either be suspected of having Sjögren syndrome based on the EULAR Sjögren Syndrome Disease Activity Index (ESSDAI), with at least one domain being positive OR respond positively to at least one of the following questions about symptoms:

Table 1a. ACR-EULAR 2016 Classification Criteria for Sjögren Syndrome

Table 1b. Revised Rules for Classification

The American College of Rheumatology, in consensus with the Sjögren’s International Collaborative Clinical Alliance (SICCA) investigators, proposed provisional criteria to improve the criteria’s specificity for enrollment in trials for biological agents and other therapies. These criteria require at least two of the following three findings:

These criteria were found to have a sensitivity of 93% and a specificity of 95% (121).

A number of neurologic manifestations have been reported in about 20% of patients with Sjögren syndrome (range 6% to 70%) (71; 32; 29; 86). In most cases, neurologic symptoms precede the diagnosis (10; 86). Historically, the most common symptoms include involvement of the peripheral nervous system. The prevalence of peripheral nervous system involvement in primary Sjögren syndrome has been reported to range from 2% to over 50%. Bias in study designs, including low numbers of patients and unclearly defined rheumatologic and neurologic diagnoses, as well as changes in classification criteria with the individualization of small fiber neuropathy could explain such variability (20). More recent reports emphasize the involvement of the central nervous system that went underrecognized in the past, often mimicking the clinical symptoms of primary progressive or relapsing-remitting multiple sclerosis. An observational, single center study from Italy found that 5.8% of patients presenting to rheumatology clinic with a diagnosis of Sjögren syndrome had central nervous system involvement (78) whereas neurologic manifestations in a French cohort were present in 74 of 392 (18.9%) patients, including 63 (16%) with PNS manifestations and 14 (3.6%) with CNS manifestations (23). Table 2 lists some of the most common neurologic manifestations of primary Sjögren syndrome.

Table 2. Neurologic Manifestations of Primary Sjögren Syndrome

Peripheral nervous system

Most studies report that the frequency of peripheral nervous system disease is between 5% and 20% (12). Mori and colleagues reviewed 92 patients (86% women, mean age 60 years) with Sjögren syndrome and found the following types of neuropathy: sensory ataxia (39%); sensory painful neuropathy (20%); trigeminal neuropathy (17%); multiplex mononeuropathy (12%); multiple cranial neuropathies (5%); autonomic neuropathy (3%); and radiculoneuropathy (4%) (86). Brito-Zerón and colleagues analyzed 563 consecutive patients with suspected peripheral nervous system involvement and found a 10% prevalence of Sjögren syndrome-related peripheral neuropathy, including axonal sensorimotor polyneuropathy (n=24), pure sensory neuronopathy (n=15), multiplex mononeuropathy (n=15), and demyelinating polyradiculoneuropathy (n=1). Survival was significantly reduced in patients with peripheral neuropathy (especially in those with mononeuropathy multiplex and axonal polyneuropathy) in comparison with the control group (log rank = 0.001) (16).

Dyck described the following nerve fiber lesions in these cases (35). In the ataxic variety, large primary afferent neurons or fibers are selectively affected as suggested by inflammatory infiltrates found in spinal ganglia. In trigeminal neuropathy, all classes of sensory neurons or fibers are involved. In sensory and autonomic neuropathy affecting limb or trunk, all classes of sensory neurons or fibers are affected, but symptoms suggest predominant small fiber sensory involvement with either tactile (allodynia) or thermal hypersensitivity (hyperalgesia), as well as sudomotor abnormalities such as decreased sweating in toes and feet or in an asymmetrical, radicular pattern. These sensory and autonomic neuropathies may occur with minimal or no evidence of sicca symptoms (35).

Axonal neuropathies. Between 17% and 39% of patients with Sjögren syndrome have minor symptoms of peripheral neuropathy (10; 35). The most common axonal forms are symmetrical sensory neuropathy and sensorimotor neuropathy (10; 71; 32; 35; 86). Vibration perception threshold and thermal perception are sensitive and useful methods of monitoring peripheral nerve function in these patients (10). The course is generally slowly progressive. Autonomic symptoms are rare. A common complication is carpal tunnel syndrome. Motor and sensory action potentials in the involved nerves are markedly reduced. Sural nerve biopsies have shown vasculitis of vasa nervorum, perivascular cellular infiltration, and necrotizing vasculitis along with focal or multifocal axonal degeneration of large and small myelinated fibers and minor demyelination.

Painful sensory neuropathy (without sensory ataxia). Painful sensory neuropathy can present as acute, subacute, or chronic onset of burning dysesthesias in the toes, feet, or hands, usually in either one limb or over the entire body, including the trunk and the face (86). Sensory loss in a stocking distribution is typically associated with cryoglobulinemic vasculitis (04). Dysautonomia may also be present, although there is no evidence of dorsal column involvement. These patients had preserved motor and vibratory or proprioceptive function, with impairment only in pinprick and temperature sensation on neurologic exam. Sural nerve biopsies show small-fiber loss suggesting predominant impairment of small sensory neurons with preservation of large-diameter sensory neurons (24). The measurements of intraepidermal nerve fiber densities in skin punch biopsy specimens show less than 3.4 fibers/mm, which is consistent with the morphological criteria for small-fiber neuropathy (45).

Painful small-fiber neuropathy. Small-fiber involvement is common in Sjögren syndrome, occurring either as a “pure,” painful small-fiber neuropathy or in combination with large sensory fiber (LSF) involvement. However, even patients with “mixed” large- and small-fiber sensory neuropathy often have small-fiber neuropathy symptoms (eg, subjective sensory symptoms, mostly painful, including burning, numbness, prickling, paresthesia, and dysesthesia) as a chief complaint, with large-fiber involvement detected only on clinical examination or electrophysiology. The diagnosis of a pure small-fiber neuropathy requires specific histologic or neurophysiologic examinations. The histologic investigation is based on a skin biopsy and the evaluation of the intraepidermal nerve fiber density (IENFD) after immunostaining with anti-neuropeptide (peptide gene product 9.5) antibodies (34). Among 317 patients with Sjögren syndrome, Lacout and colleagues found large- and small-fiber neuropathies in 22 (6.9%) and 17 (5.4%) patients, respectively (70). In the study by Seeliger and colleagues, there were no observed differences in patients with small fiber neuropathy who exhibited sensory symptoms (116). However, 24 out of 97 individuals (25%, median age 48.5 years, 75% female) were diagnosed with Sjögren's syndrome. Additionally, patients with both small fiber neuropathy and Sjögren's syndrome had significantly lower intraepidermal nerve fiber densities (IENFD) (mean 2.6 ± 1.2/mm) compared to patients with idiopathic small fiber neuropathy (mean 3.2 ± 1.5/mm; p = 0.048). Sene and colleagues compared 40 patients with small-fiber neuropathy to 100 without peripheral neuropathy and found that patients with small-fiber neuropathy were older and more frequently had xerostomia and arthralgias and a lower prevalence of serum immunological markers (antinuclear antibodies, anti-Ro, and anti-La, rheumatoid factor, and hypergammaglobulinemia) (35% vs. 62%; p=0.005) (119). Three additional studies reported lower frequencies of anti-Ro antibodies in patients with Sjögren syndrome presenting with small fiber neuropathy in comparison with those with large fiber neuropathy: 43% versus 55% (24), 39% versus 64% (86), and 5.4% versus 6.9% (p=0.002) (70), respectively. This type of neuropathy is more frequently reported in males and in those without hypergammaglobulinemia or positive autoantibodies (15).

Sensory ataxic ganglionopathy. This is a distinctive type of sensory neuronopathy with a reported association with ganglion neuronal antibodies in one study (90). The antigenic target of these antibodies is not known, and there are no posterior studies that confirmed this finding. The neuropathology of the sensory neuronopathy is characterized by lesions of the dorsal root ganglia with neuronal loss and presence of mononuclear cell infiltrates (ganglionitis) but without vasculitis (48; 90). Clinically, it is characterized by a profound loss of proprioception and vibratory perception leading to sensory ataxia, with a positive Romberg sign and global areflexia, without motor involvement. The sensory symptoms are usually asymmetrical, segmental, or multifocal, including trigeminal nerve involvement (86). Birnbaum and colleagues described dorsal root ganglion enlargement and T2 hyperintensity in early dorsal root ganglionitis (14), whereas Yoshida and colleagues detected atrophy of the dorsal root ganglion in chronic sensory ataxic ganglionopathy (149).

There is usually autonomic involvement manifested by abnormal pupillary responses including Adie pupils, anisocoria, and oval pupils; orthostatic hypotension decreases in 123I-MIBG cardiac uptake, hypohidrosis, and segmental anhidrosis of the trunk. Sensory potentials are absent, contrasting with normal motor potentials and normal peripheral nerve conduction velocities. Somatosensory evoked potentials are usually absent. Sural nerve biopsies show a selective loss of large, myelinated fibers, with minimal inflammation or vasculitis. Cervical spinal cord MRI may show hyperintense T2-weighted signals corresponding to involvement of the fasciculus cuneatus and fasciculus gracilis in the dorsal columns, proportional to the severity of the damage (87). This form of neuropathy is chronic and progressive and is usually only minimally responsive to treatment (12). Some authors have estimated a prevalence of sensory neuronopathy around 5%. Sensory neuronopathy is probably less frequent than painful axonal neuropathy. Although less frequent than other forms of peripheral neuropathies, sensory neuronopathy causes greater handicap. According to Pereira and colleagues, sensory neuronopathy is generally distinguishable from other forms of peripheral neuropathy in Sjögren syndrome because it tends to be asymmetrical and to predominate in the upper limbs (94). In their series, sensory neuronopathy preceded or coincided with Sjögren syndrome diagnosis. The most common neurologic findings were ataxia and areflexia followed by paresthesia and pain. Lower limbs were more affected than upper limbs, neurologic deficits were often symmetric, and cranial nerves were affected in three patients. Nine patients had positive ANA (69%) or anti-SS-A (38%). No patient presented with sphincter or autonomic dysfunction. Three differentiated clinical courses have been reported: subacute progression in less than 1 month (7%), late acceleration of pure sensory neuropathy 2 to 4 years after an initial indolent onset (20%), and a very long-term insidious, chronic evolution (73%) with a poor response to treatment, although stabilization of symptomatology for long periods may be observed (38).

Cranial neuropathies. The most common cranial nerve affected in Sjögren syndrome is the trigeminal nerve, which is discussed in more detail in the next section. The facial nerve is the most commonly targeted motor nerve and can present bilaterally. Other symptoms include recurrent diplopia due to oculomotor and trochlear nerve involvement, swallowing problems due to compromise of the glossopharyngeal and vagus nerves, and less commonly, simultaneous lesions in multiple cranial nerves (86). Neurosarcoidosis should be included in the differential diagnosis in patients with multiple cranial nerve involvement (134).

Trigeminal neuropathy. Trigeminal neuropathy is a common complication of Sjögren syndrome that often presents with bilateral involvement. There is facial numbness, decreased corneal reflexes, and loss of pin prick and soft touch perception in the trigeminal nerve distribution, without motor trigeminal nerve involvement. Immune-mediated damage of Gasserian ganglion neurons is suspected to be the most plausible cause of the neuropathy. Simultaneous sensory neuropathy of the trigeminal, glossopharyngeal, and vagus nerves has been reported (135).

Hearing loss. Sensorineural hearing loss, due to lesion of the vestibulocochlear nerve manifested by sudden or progressive deafness, has been reported in Sjögren syndrome. Retamozo and colleagues analyzed the frequency of non-ESSDAI features in The Big Data Sjögren Project Consortium, an international multicenter registry, and found hearing loss in 35 of 6331 patients (0.6%) (103). In some cases, hearing loss was associated with the presence of antiphospholipid antibodies (88).

Seeliger and colleagues conducted a study on auditory neuropathy in neurologic patients with Sjögren syndrome (115). The study included 30 patients (57% females, median age 59 years) who showed signs of peripheral polyneuropathy and met the current ACR/EULAR classification criteria for Sjögren syndrome. Results revealed that 33% of patients with Sjögren syndrome and polyneuropathy had definite hearing loss. Further pathological test results on audiometric testing were found in other patients, resulting in a total of 80% of patients with hearing dysfunction. The severity of hearing loss, as detected on pure tone audiometry, ranged from mild in most patients (60%) to severe in 10% of patients in the cohort. Interestingly, 15 out of 19 patients with evidence of retrocochlear auditory dysfunction showed pathological auditory brainstem response potentials. Hence, auditory brainstem response potentials are crucial in identifying retrocochlear hearing impairment.

Autonomic neuropathy. The peripheral sympathetic nervous system is severely involved in this form of neuropathy. These patients commonly present with sicca syndrome, orthostatic hypotension with syncope, anhidrosis, Adie pupil, neurogenic bladder, erectile dysfunction, as well as gastrointestinal symptoms such as abdominal pain, constipation, or diarrhea (32; 35; 86). In addition to the above, examination reveals reduced cardiac 123I-MIBG uptake, no plasma norepinephrine response after standing, and hypertensive response to minimal doses of norepinephrine (86). These patients usually have elevated titers of nicotinic ganglionic acetylcholine receptor autoantibody, a putative effector of autoimmune cholinergic dysautonomia (114).

Radiculoneuropathy. A demyelinating polyradiculoneuropathy is the most uncommon peripheral nervous system manifestation in patients with Sjögren syndrome. Patients present with progressive sensory disturbances in a glove-and-stocking distribution, muscle weakness, and sensory ataxia but without autonomic symptoms. CSF protein is elevated, and nerve conduction studies show reduced conduction velocities and temporal dispersion. Biopsies show histopathological evidence of remyelination. According to Mori and colleagues, the primary lesion in these patients is an inflammatory radiculoneuropathy (86).

Skeletal muscle

Inflammatory myositis. Myalgias and diffuse musculoskeletal pain, without elevation of muscle enzymes or electromyographic changes, have been reported in 30% to 44% of patients with Sjögren syndrome (75); many of them fulfill criteria for fibromyalgia. Fibromyalgia is a diagnosis of exclusion for these patients, and vitamin D deficiency as well as thyroid disease should be ruled out prior to making the diagnosis (12). Inflammatory myopathies are uncommon in Sjögren syndrome, reported in 1% to 2% of cases of peripheral nervous system manifestations. The ASSESS cohort analyzed the frequency of myositis (defined according to the 2017 EULAR/ACR criteria) in 395 patients with primary Sjögren syndrome (37). Myositis was suspected in 38 patients but was confirmed in only 4 [1.0% (95% CI: 0.40, 2.6)]. Patients with suspected, but not confirmed, myositis had higher patient-reported scores and more frequent articular and peripheral nervous involvement than others. By contrast, disease duration in patients with confirmed myositis was 3-fold longer than without myositis. Among the four patients with confirmed myositis, two fulfilled criteria for sporadic inclusion body myositis. Muscle biopsy results on this small subset of patients show signs of inflammation in 50% to 72% and evidence of polymyositis (inflammation combined with degeneration and regeneration of muscle fibers) in 47% (33). Some cases have symmetrical inclusion body myositis diagnosed by electron microscopy (75; 33). Myopathy due to medications such as steroids must be considered in the differential diagnosis.

Central nervous system

Multiple sclerosis-like manifestations. Alexander and colleagues first reported the frequent occurrence of a relapsing remitting syndrome resembling multiple sclerosis associated with cutaneous vasculitis in patients with Sjögren syndrome (03). All cases had confirmed primary Sjögren syndrome by lip biopsy, and all cases met diagnostic criteria for clinically definite multiple sclerosis. Eighty nine percent of the patients had oligoclonal bands in the CSF, but occasionally the bands were lowered with corticosteroid therapy. Fifteen out of 19 patients had positive visual evoked potential testing. Many of these cases presented a combination of optic neuropathy and chronic myelopathy that could be confused with neuromyelitis optica or multiple sclerosis, as well as brainstem and cerebellar symptoms. Sjögren syndrome and multiple sclerosis share common immunologic dysregulation, as a dominant feature of the immunopathogenesis in Sjogren syndrome is B cell dysfunction and excessive BAFF (B cell activating factor). Self-reactive B cells as well as BAFF itself can stimulate T cells, which are implicated in the pathogenesis of multiple sclerosis (26). Peripheral nervous system involvement was seen in 55% of these patients, and perhaps this could help differentiate between multiple sclerosis and CNS Sjögren syndrome. Neuroimaging studies were carried out by Akasbi and colleagues and disclosed white matter abnormalities in 49% of patients with primary Sjögren syndrome with suspected neurologic involvement. White matter abnormalities were classified as vascular pathological changes in 21 patients: 10 had multiple small focal lesions, 7 had beginning confluence of lesions, and 4 had diffuse involvement of the entire region. White matter abnormalities were classified as inflammatory or demyelinating lesions (multiple sclerosis-like) in four patients who fulfilled the MRI Barkhof criteria for multiple sclerosis. Patients with inflammatory or demyelinating lesions were younger and had a lower frequency of hypertension and altered glomerular filtration rate in comparison with patients with vascular lesions. The multivariate age-sex adjusted model identified hypertension and HDL-c levels as independent predictors of white matter abnormalities in patients with primary Sjögren syndrome (01). There is no consensus in the medical literature to determine whether a patient has a relapsing remitting demyelinating disease of the CNS as a result of Sjögren syndrome or if Sjögren syndrome and multiple sclerosis can present independently. A cross sectional study of 68 patients with Sjögren syndrome and 68 healthy controls found no difference in white matter hyperintensity load or distribution between the two groups (51). However, in other autoimmune disorders such as lupus and antiphospholipid syndrome, cerebral white matter hyperintensities are fairly common.

The prevalence of Sjögren syndrome in patients with multiple sclerosis ranges from 0% to 15% to 17% (30; 95), although the largest epidemiological study (a population-based study carried out in Taiwan) has estimated a prevalence of 2.4% (36).

Neuromyelitis optica spectrum disorder (NMOSD). The combination of optic neuritis with concurrent myelitis is included in the diagnostic criteria for neuromyelitis optica, also known as Devic disease (125). More than 95% of neuromyelitis optica cases test positive to the disease-specific serum aquaporin-4 autoantibody (144). Antinuclear autoantibodies or other systemic autoimmune diseases are often found in patients with neuromyelitis optica, including Sjögren syndrome (97), and 7.7% to 12% of patients with neuromyelitis optica have anti-Ro/SS-A or anti-La/SS-B antibodies (53). In Korea, Kim and colleagues studied eight patients with Sjögren myelopathy and demonstrated that most patients exhibited clinical, radiological, and immunological characteristics of neuromyelitis optica, including positive aquaporin-4 autoantibody testing (67). Min and colleagues studied 12 women with primary Sjögren syndrome and recurrent cerebral manifestations (84). MRI showed lesions characteristic of neuromyelitis optica in the third and fourth ventricles and in the posterior limb of the internal capsule, along with cerebral or cerebellar lesions larger than 3 cm in size and with cavity-like formations. Aquaporin-4 autoantibody was positive in six of eight patients tested, and all the seropositive patients showed lesions with increased MRI diffusion, suggestive of vasogenic edema. Birnbaum and colleagues analyzed the relationship between primary Sjögren syndrome and neuromyelitis optica and compared frequencies of anti-AQP-4 and Sjögren syndrome–associated antibodies in 109 patients (13). They found that anti-AQP-4 antibodies were seen exclusively in patients with Sjögren syndrome with neuromyelitis optica (72.7%) but not in those without neuromyelitis optica (p < 0.01). In contrast, anti-Ro 52, anti-Ro 60, and other autoantibodies were not more prevalent in patients with Sjögren syndrome with neuromyelitis optica versus those without.

Two large, population-based studies carried out in patients with NMOSD from China and Taiwan have reported a prevalence of Sjögren syndrome of 7% to 8% (36; 130).

Javed and colleagues tested patients with neuromyelitis optica for concomitant Sjögren syndrome with labial biopsies. Although less than 20% of patients had symptoms that were suggestive of Sjögren syndrome, labial biopsy was positive for 9 of 12 patients with confirmed neuromyelitis optica and for 7 of 8 patients with longitudinally extensive transverse myelitis (60). These authors recommend testing for aquaporin-4 autoantibody in all patients with Sjögren myelitis, followed by early aggressive immune therapy in positive patients.

Spinal cord involvement. Following an initial report (142), at least 60 cases of spinal cord involvement in Sjögren syndrome have been reported, indicating that this neurologic complication is not uncommon (68; 145; 09; 52; 54; 129; 05; 95; 138; 41; 28; 98; 127; 132; 125; 65; 67; 118; 72; 82). Furthermore, patients with primary progressive multiple sclerosis have a prevalence of Sjögren syndrome (16.6%) that is much higher than expected in the general population (1% to 5%) (30).

Acute transverse myelitis is the most frequent form of spinal cord involvement in Sjögren syndrome (142). The symptoms of acute transverse myelitis develop abruptly, usually with severe neck and thoracic back pain followed by sensory and motor deficits below the level of the lesion. In Sjögren syndrome, this form of myelopathy is thought to be due to vasculitis and carries a high morbidity and mortality.

Spinal cord involvement may also present as a progressive myelitis. A diagnosis of Sjögren myelopathy requires a high index of suspicion and should be considered especially in women over 45 years of age with progressive spastic paraparesis and abnormalities on spinal cord MRI, even with negative antiextractable nuclear antigen antibodies (Ro/SS-A or La/SS-B) (95). The presence of autoantibodies against fodrin also helps in differentiating myelopathy in Sjögren syndrome from primary progressive multiple sclerosis (31). A positive test has 70% sensitivity, 86.7% specificity, 63.6% positive predictive value, and 89.6% negative predictive value (31).

Less common is the combined involvement of pyramidal tracts and anterior horn cells resembling amyotrophic lateral sclerosis. Postmortem study of two cases confirmed in one patient a lower motor neuron syndrome combined with flaccid bladder and rectum, and in the other patient, unilateral hearing loss, sensory neuronopathy, Adie pupils, upper motor neuron signs, and autopsy-proven anterior horn cell degeneration (66). These cases demonstrate the wide multisystem neuronal involvement that may occur in Sjögren syndrome. A similar syndrome called pseudo-amyotrophic lateral sclerosis has been reported in patients with HTLV-1 infection (110).

Focal encephalic manifestations. Focal manifestations may present acutely, with stroke-like features such as aphasia, hemiplegia, or numbness, consistent with a focal vasculitis (32). Intracerebral or subarachnoid hemorrhage may signal the presence of vasculitis. Recurrent or relapsing CNS symptoms may mimic multiple sclerosis as discussed above.

The pooled prevalence of cerebellar ataxia in primary Sjögren syndrome is estimated to be 1.5% (95% CI: 0.3%–6.8%) (73). The subacute development of cerebellar ataxia and tremor may be caused by demyelination. In one patient, an MRI showed T2-hyperintensities in the cerebellar white matter and the pons; severe necrotic lesions were found in the cerebellar white matter bilaterally, with several foci of perivenous demyelination in the periphery of the lesions and similar demyelinated areas in the pons; there was minimal granulomatous angiitis (56). Yang and colleagues analyzed the clinical features of cerebellar involvement in 13 patients with primary Sjögren syndrome (147). Nine (69.2%) patients went to the clinic because of ataxia, and primary Sjögren syndrome was not suspected until accidental screening for autoantibodies. Dysarthria (7, 59.8%), limb tremor (4, 30.8%), and nystagmus (2, 15.4%) were the other symptoms related to the cerebellum. Of the patients, 81.8% (9/11) had abnormal cerebrospinal fluid findings, and 11 patients (84.6%) had cerebellar atrophy on brain MRI. Anti-Ro/SSA antibody was positive in 12 (92.3%) patients and anti-La/SSB in 6 (46.2%) patients.

Tetsuka and colleagues conducted a literature review to assess the clinical characteristics, diagnostic methods, and therapeutic strategies used for patients with primary Sjögren syndrome who had cerebellar degeneration (128). They found that all patients in the literature review, as well as their own patient, were positive for anti-Ro/SSA antibodies, although some patients were negative for anti-La/SSB antibodies. The average age of onset was 48.9 ± 19.4 years, and the male to female ratio was overwhelmingly female, with three men and 11 women. Cerebellar atrophy was observed with brain MRI in the majority of patients as the disease progressed (12/14 cases). Three main symptoms suggested cerebellar disturbances: ataxia (14/14 cases), nystagmus (6/14 cases), and dysarthria (10/14 cases). In some cases, cerebellar degeneration was the first manifestation of primary Sjögren syndrome. Anti-Ro/SSA antibodies were observed in both serum and CSF, but anti-Ro/SSA antibodies were negative in the CSF of patients with primary Sjögren syndrome without CNS involvement. Cerebellar atrophy was observed, and sequelae remained in the majority of patients. Autopsy findings indicated a selective loss of Purkinje cells.

Other focal neurologic manifestations described in Sjögren syndrome include: internuclear ophthalmoplegia, nystagmus, dystonia, athetosis, and intention tremor; or aseptic meningitis with confusion, cerebellar involvement, and spastic tetraparesis (89). A rigid form of parkinsonism with preponderant akinesia but without tremor and resistance to L-DOPA treatment has been described (141). There are also rare cases of generalized chorea (137). Focal and generalized seizures may occur during relapses of cerebral vasculitis.

Sjögren syndrome may also present as limbic encephalitis. A case series of three patients describes limbic encephalitis occurring after the diagnosis of Sjögren syndrome was made, with or without other neurologic complications such as motor axonal neuropathy (27). Antineuronal antibodies were tested and were negative, thus, suggesting Sjögren syndrome was the cause of these symptoms.

Osmotic demyelination syndrome. Osmotic demyelination syndrome is a rare noninflammatory demyelinating disorder of the CNS that primarily affects the white matter tracts of the basilar pons (central pontine myelinolysis) and, in some cases, extrapontine regions, such as the basal ganglia. In a literature review of patients with both Sjögren syndrome and renal tubular acidosis (SS-RTA), Sandhya found multiple cases of osmotic demyelination syndrome (112). This is intriguing because osmotic demyelination syndrome is generally considered a noninflammatory disease and has not been associated with autoimmune disorders, including Sjögren syndrome. In response, Sandhya and colleagues performed a detailed literature review of osmotic demyelination syndrome in patients with Sjögren syndrome and noted the characteristics of these patients, including the presence of renal tubular acidosis (113). The authors compared the demyelinating autoimmune CNS disorder most closely associated with Sjögren syndrome, neuromyelitis optica spectrum disorders (NMOSD), to osmotic demyelination syndrome. They found that both of these inflammatory and noninflammatory demyelinating disorders of the CNS are often accompanied by Sjögren syndrome.

The authors identified 15 patients (all women, median age 40 years) with osmotic demyelination syndrome and comorbid renal tubular acidosis. These patients showed heterogeneous clinical manifestations and outcomes. The most common symptom was quadriparesis, seen in 14 of the 15 patients. Eleven of the 15 patients had one of the following features, either alone or in combination: worsening of the sensorium, extensor plantar response, dysphagia/dysarthria, and facial palsy. Ocular palsy was seen in only four of the 15 patients and was a late manifestation. One patient who had extensive long-segment myelitis and subsequent osmotic demyelination syndrome died, but most patients recovered without significant sequelae. None had hyponatremia, whereas all patients had hypokalemia and/or hypernatremia. The authors suggest that hypokalemia causing nephrogenic diabetes insipidus followed by a rapid rise in sodium and the resultant osmotic stress could potentially explain the occurrence of osmotic demyelination syndrome in SS-RTA. They recommend that when a patient with flaccid quadriparesis does not respond to the correction of potassium or develops additional neurologic features along with a rise in sodium, osmotic demyelination syndrome should be suspected in the setting of SS-RTA.

Meningoencephalitis. Meningoencephalitis is a relatively common neurologic complication of Sjögren syndrome (32). It begins with headache, myalgias, confusion, and meningeal signs without fever (111); in some cases, sensorineural deafness may occur. Focal neurologic signs may be present; brain MRI shows hyperintense multifocal inflammatory changes in the cerebral white mater and cortex. The CSF profile is consistent with an aseptic lymphocytic meningitis, with up to 900 cells/mm3. Recurrences occur, and changes associated with vasculitis may be demonstrated by angiography.

Psychiatric and cognitive disorders. Central nervous system involvement precedes the symptoms of Sjögren syndrome in 52% to 80% of patients (76). In many patients, symptoms of depression and anxiety may antedate the diagnosis of Sjögren syndrome or become chronic accompaniments of the disease (136). Also common are cognitive changes with poor concentration and memory and abnormalities in neuropsychiatric testing, including executive dysfunction and frontal lobe deficits. The pathogenesis of these changes is not clearly understood, but it may be part of the so-called “vascular depression-vascular cognitive disorder,” resulting from ischemic interruption of prefrontal circuits important for mood and behavior (109). Sicca symptoms are more common in patients with cognitive disturbances and low inflammatory disease activity. A study comparing the risk of Alzheimer disease in patients with primary Sjögren syndrome and healthy controls revealed a 2.68-fold higher risk in the patients with primary Sjögren syndrome after 10 years (74).

Optic neuropathy. Patients with Sjögren syndrome can present with bilateral visual loss secondary to retrobulbar optic neuropathy (03; 32). In some cases, blindness secondary to bilateral optic neuropathy was the first manifestation of Sjögren syndrome (133; 99; 11). In about 12% to 15% of patients, the diagnosis was revealed by abnormal visual evoked potentials (32). The pathogenesis of optic nerve involvement in Sjögren syndrome is postulated to result from a combination of ischemic vasculitis and demyelination. The differential diagnosis should include multiple sclerosis and neuromyelitis optica.

Prognosis and complications

Primary Sjögren syndrome can be associated with neurologic and cognitive involvement, negatively affecting patients' quality of life. Goulabchand and colleagues conducted a nationwide retrospective study using the French Health insurance database to assess whether patients with Sjögren syndrome are at higher risk of hospitalization for neurologic diseases (47). The study selected patients hospitalized with new-onset Sjögren syndrome between 2011 and 2018 and compared the incidence of hospitalization for dementia, multiple sclerosis, encephalitis, and peripheral neuropathy with an age- and sex-matched (1:10) hospitalized control group. Adjusted hazard ratios (aHR) considered confounding factors, particularly socioeconomic status and cardiovascular diseases. The analysis included 25,661 patients hospitalized for Sjögren syndrome and 252,543 matched patients. The incidence of hospitalization for dementia, as well as the incidence of hospitalization for multiple sclerosis, encephalitis, and inflammatory polyneuropathies, were significantly higher in patients with Sjögren syndrome.

In a study, patients with Sjögren's syndrome and neurologic involvement had a different clinical phenotype than patients with Sjögren's syndrome without neurologic involvement (117). The study included 512 patients treated for Sjögren's syndrome without neurologic involvement (pSS) / neurologic manifestations of primary Sjögren's syndrome (pSSN) between April 2018 and July 2022 (238 pSSN patients [46%] vs. 274 pSS patients [54%], cross-sectional design). The study found that male sex, older age at disease onset, hospitalization at first presentation, lower IgG levels, and higher eosinophil values were independent predictors of neurologic involvement in Sjögren's syndrome. Additionally, univariate regression showed older age at diagnosis, lower prevalence of rheumatoid factor, SSA(Ro)/SSB(La) antibodies, higher white blood cell count, and CK levels (treatment-naïve) in pSSN. Therefore, these data support the hypothesis that neurologic involvement in Sjögren's syndrome has been underestimated until now. Physicians must be aware of these neurologic risks to choose the most appropriate diagnostic work-up.

Clinical vignette

A 63-year-old woman developed paraplegia, urinary frequency, urgency, and incontinence; she had been experiencing sensory symptoms in her legs for the last 10 months. Neurologic examination revealed spastic paraparesis, lower limb hyperreflexia, and flexor plantar responses. No sensory level was noted, but she had hypoesthesia to pinprick in her left buttock and posterior thigh. MRI demonstrated abnormal focal signal intensities on T2-weighted images, enhancing with gadolinium on T1-weighted images, at the C5-6 and T8-9 levels and on the conus medullaris and cauda equina. MRI of the brain was unremarkable. CSF showed lymphocytic pleocytosis with 0.011 x109 cells/L, few atypical lymphocytes, a mildly elevated protein level (0.85 g/L), a nonreactive VDRL test, and absent oligoclonal bands. A second CSF had 0.006 x109 cells/L, a protein level of 0.72 g/L, and normal cytology. Serum and urine protein electrophoresis showed neither monoclonal gammopathy nor elevation of the angiotensin-converting enzyme level; serum B12 levels were normal. Antibodies to HIV and HTLV-1 were nonreactive. Chest x-ray was normal. A gallium body scan demonstrated increased activity in the lachrymal glands, regional lymph nodes, and right paratracheal region; there was no evidence of sarcoidosis.

Over the next 3 months, her paraparesis progressed to near paraplegia associated with a sensory level at the umbilicus and with left arm paresthesias. Antinuclear antibody ratio was 1:640; the erythrocyte sedimentation rate was 41 mm/h. Mild sicca symptoms (xerostomia and keratoconjunctivitis) were found on further questioning. Neurologic examination showed a left afferent pupillary defect, spastic paraplegia with hyperreflexia, crossed adductor responses, ankle clonus, and bilateral Babinski sign. Motor strength was 5/5 (MRC scale) in the upper extremities and 0/5 in the lower extremities, except for 1/5 for ankle dorsiflexors and plantar flexors. A T10 sensory level was found, along with decreased anal sphincter tone. MRI of the cervical, thoracic, and lumbar spine showed multifocal areas of increased T2-signal intensity, enhanced by gadolinium at the C5-7, T3-5, and T8-11 levels and on the conus medullaris. A third CSF sample had 0.035 x109 cells/L, a glucose level of 2.9 mmol/L (53 mg/dL), and a protein level of 0.62 g/L. Oligoclonal bands and myelin basic protein were absent, and the IgG index was normal (0.63). A second antinuclear antibody test was positive, with a titer of 1:2560; Ro/SS-A antibody was positive with a speckled pattern. Complement C3 was 1.78 g/L (reference range, 0.86 to 1.84 g/L) and C4, 0.23 g/L (reference range, 0.20 to 0.59 g/L). Visual evoked potentials were prolonged in the left eye. Examination of a minor salivary gland biopsy specimen showed chronic sialadenitis, with a focus score of 3. NMO-IgG antibody was negative.

The patient was diagnosed with CNS Sjögren syndrome because of the presence of Sicca syndrome, abnormal serological test results, and the salivary gland biopsy results, which fulfilled the American-European Consensus Criteria for Sjögren syndrome. The patient received a pulse dose of intravenous methylprednisolone sodium succinate, 1 g/d for 3 days, followed by oral prednisone, 60 mg/d. Rheumatology consultants recommended treatment with intravenous cyclophosphamide, 0.75 g/m2, followed by equal monthly doses for 6 months, along with oral prednisone, 20 mg/d. With this treatment, the patient experienced marked improvement in the strength of most of the muscles in her lower extremities, going from 0-1/5 to 3/5 (MRC scale). Furthermore, sensory complaints, particularly subjective paresthesias in the left upper extremity, slowly abated during the ensuing several weeks (142).

Biological basis

Etiology and pathogenesis

The pathogenesis of Sjögren syndrome has been reviewed by Fox (39). The exact mechanisms are still unknown, but the most prominent immunoregulatory alterations are B-cell hyperreactivity and enhanced levels of B-cell-activating factor/B-lymphocyte stimulator. The pathogenesis is multifactorial, whereby environmental factors activate glandular endothelial or epithelial cells, triggering inflammation in individuals with a genetic predisposition (HLA DR). The inappropriate B-cell activation can follow various stages of evolution, leading in extreme cases to malignant transformation of B cells. Patients with Sjögren syndrome have a 6- to 44-fold increased risk for developing non-Hodgkin lymphoma (12).

Polyclonal B-cell hyperreactivity in Sjögren syndrome accounts for the hypergammaglobulinemia, circulating immune complexes, and multiple autoantibodies directed against both organ- and nonorgan-specific autoantigens. Clinically, the most important and best characterized are the autoantibodies anti-Ro (SS-A) and anti-La (SS-B) directed against cellular heterogeneous ribonucleoprotein complexes consisting of antigenic proteins (61). The antibodies recognize autoantigens, which bind to ribonucleoprotein particles consisting of a 60kD SS-A/Ro RNA binding protein and hY1 RNAs and 48 KD RNA binding protein, which facilitates maturation of RNA polymerase III transcripts, such as precursors to tRNA and 5S-RNA (69). These antibodies are found in approximately 50% of the patients with Sjögren syndrome and tend to be associated more with severe glandular and extraglandular manifestations (61; 69; 39).

There are also alterations of cellular immunity in Sjögren syndrome. Mononuclear cells (primarily T-lymphocytes) infiltrate salivary and lachrymal glands with partial destruction of acinar and ductal structures. The T-lymphocytes and also the glandular cells cause the release of cytokines (especially interleukin-1, interleukin-6, and tumor necrosis factor alpha) (39). These cytokines, along with autoantibodies and metalloproteinases, cause a decreased release of neurotransmitters and a diminished response of the residual glandular cells to available neurotransmitters, resulting in the symptoms commonly seen in Sjögren syndrome (40). Interaction between constitutional factors (hormones and major histocompatibility complex) and environmental factors (most likely viruses) are thought to be important in the etiology of Sjögren syndrome. Females are affected in a ratio of 9:1 compared to males. Hormones such as estrogens, reactive hypothalamic and hypophyseal peptide hormones and dehydroepiandrosterone may play a role. Sjögren syndrome is associated with HLA-DR 3 and linked genes B8, DQ 2, and the C4 null gene in about 50% of the patients. Postulated infectious agents capable of triggering the immune process include herpes viruses (particularly Epstein Barr virus, cytomegalovirus, and human herpesvirus-6), H pylori and human retroviruses, in particular HTLV-1.

HTLV-1 and Sjögren syndrome. Transgenic mice expressing the HTLV-1 tax gene develop an exocrinopathy similar to that seen in patients with Sjögren syndrome. The expression of sequences homologous to the HTLV-1 tax gene has been found in labial salivary glands of patients with Sjögren syndrome (126). Kompoliti and colleagues described a patient with Sjögren syndrome, lymphocytic pneumonitis, and HTLV-1 myelitis (68). In Japan, Nakamura and colleagues investigated the presence of Sjögren syndrome in patients with HTLV-1-associated myelopathy, including a histological examination of labial salivary glands; definite Sjögren syndrome was found in 65% (13/20) of patients (91). More severe inflammatory cell infiltration in labial salivary glands was found in patients with HTLV-1-seropositive Sjögren syndrome than in seronegative controls. Pot and colleagues reported a spectacular radiological and clinical recovery of a patient with HTLV-1 myelitis and Sjögren syndrome who was treated with combined antiretroviral drugs (lamivudine and tenofovir) plus immunosuppressant therapy with prednisone and mycophenolate mofetil (98). Based on the aforementioned findings, it is advisable to test for HTLV-1 antibodies in all patients with Sjögren syndrome who develop a myelopathy.

Epidemiology

The reported incidence and prevalence of Sjögren syndrome varies according to both the study design and the classification criteria used. The pooled prevalence rates in studies that used the 1993 European Classification Criteria was 12­fold higher than the pooled prevalence rates in studies that used the 2002 American-European Consensus Group (AECG) criteria, whereas the pooled prevalence rates reported in population­based epidemiological studies were slightly lower than that calculated in the total population. The incidence of Sjögren syndrome ranges between 3 and 11 cases per 100,000 individuals, whereas the prevalence ranges between 0.01% and 0.72%. It is also likely that there are asymptomatic cases that never were diagnosed. Only one study has evaluated the influence of race or ethnicity on the prevalence of Sjögren syndrome. This study found a significant two-fold higher prevalence among individuals with a non­European background than in those with a European background in the general population of the Greater Paris area. At presentation of disease, the epidemiological profile of Sjögren syndrome is typical, which might aid early diagnosis.

Sjögren syndrome predominantly affects women. In fact, Sjögren syndrome has the most unbalanced gender ratio out of all systemic autoimmune diseases; nearly a 10:1 female:male ratio was reported in a big data study of more than 14,000 patients with Sjögren syndrome.

Although Sjögren syndrome can occur at all ages, it is mainly diagnosed between 30 and 50 years of age. Sjögren syndrome is rare in children, and the female:male ratio is less evident in children than in adults. Sjögren syndrome and other autoimmune diseases frequently coincide (associated Sjögren syndrome) in daily practice. In patients with systemic autoimmune diseases, the proportion of patients with concomitant Sjögren syndrome differs between conditions; 14% to 18% of patients with lupus, 7% to 17% with rheumatoid arthritis, and 12% with systemic sclerosis also have associated Sjögren syndrome. In clinical practice, the management of associated Sjögren syndrome should be the same as the management of primary Sjögren syndrome (17).

Neurologic manifestations have been reported traditionally in about 20% of patients with Sjögren syndrome (range 6% to 70%) (71; 32; 25). Goransson and colleagues performed a population-based study in Sweden and found that 27% of patients with Sjögren syndrome had neuropathy, including 31% with motor neuropathy, 13% with sensory neuropathy, and 11% with sensorimotor neuropathy (45). Central nervous system involvement ranges from 2% to 25% in hospitalized patients (44).

Ye and colleagues reviewed a large cohort of patients with primary and secondary Sjögren syndrome in a cross-sectional study and found that low complement (C3) levels, xerophthalmia, positive ANA, cardiac involvement, and labial salivary gland histologic results were useful in predicting the risk of neurologic complications (148).

Prevention

Sjögren syndrome is a chronic disease that cannot be prevented. However, screening approaches that target specific subsets of patients with Sjögren syndrome might either prevent or ensure timely treatment of the main complications. The presence of autoantibodies (especially anti­Ro/SS-A antibodies) can be used for the early diagnosis of Sjögren syndrome in these patients. However, the frequency of anti­Ro/SS-A or anti­La/SS-B autoantibodies is often lower in patients with Sjögren syndrome who only have neurologic involvement than in patients with Sjögren syndrome who have no neurologic involvement, and these patients often require a salivary gland biopsy showing the presence of focal lymphocytic sialadenitis for the early diagnosis of Sjögren syndrome (17).

Differential diagnosis

Depending on the form of neurologic presentation, the differential diagnosis of neurologic Sjögren syndrome include many processes that effect either the peripheral or central nervous system. As mentioned above, demyelinating diseases such as multiple sclerosis, neuromyelitis optica, and transverse myelitis are commonly on the differential for central nervous system Sjögren syndrome. Other multisystemic autoimmune disorders should be considered, as well as antineuronal antibody encephalitis and infectious processes (especially HLTV-1). The onset of transverse myelitis with thoracic back pain and rapid loss of motor function is reminiscent of nucleus pulposus herniation (131). Serum levels of vitamin B12 and methylmalonic acid are diagnostic for subacute combined degeneration.

Peripheral nervous system presentations vary widely and, therefore, the diagnostic work up should be focused on the patient’s symptoms, dependent on the clinical presentation.

The differential diagnosis of Sjögren syndrome includes a past history of head and neck radiation therapy, hepatitis C infection, AIDS, preexisting lymphoma, sarcoidosis, graft-versus-host disease, IgG4­related disease (IgG4­RD), and the recent use of anticholinergic drugs.

Diagnostic workup

The diagnostic workup of Sjögren syndrome is outlined in Tables 1a and 1b. Sjögren syndrome often has a variable course and a wide spectrum of clinical manifestations, making the diagnosis difficult or delayed. Early recognition of Sjögren syndrome may prevent complications, and it allows for clinical surveillance of the development of serious extraglandular systemic manifestations. The diagnosis of primary Sjögren syndrome is strongly suggested in patients who present with signs and symptoms of oral and ocular dryness and who test positive for antibodies to the anti-SS-A or anti-SS-B antigen, or who have a positive salivary gland biopsy (17).

There is no gold standard test for the diagnosis of neurologic Sjögren syndrome. However, it is important to note that inflammatory CSF changes are rarely found and if so, a pathognomic pattern does not exist. CSF analysis is often used to exclude other autoimmune and infectious processes (93).

Management

The therapeutic management of Sjögren syndrome has not changed substantially in recent decades: treatment decisions remain challenging in clinical practice, without a specific therapeutic target beyond the relief of symptoms as the most important goal. In the 2020 EULAR guidelines, the first line of therapy for ocular dryness should be volume replacement and lubrication using artificial tears and ocular gels, whereas severe or refractory keratoconjunctivitis sicca might require topical cyclosporine or autologous serum eye drops (19; 102). With respect to oral dryness, nonpharmacological glandular stimulation is the preferred first-line therapeutic approach in patients with mild glandular dysfunction, using gustatory or mechanical stimulants, whereas in those with moderate glandular dysfunction, pharmacological stimulation (pilocarpine and cevimeline) with muscarinic agonists may be considered (102).

With respect to extraglandular manifestations of Sjögren syndrome, only a few clinical trials have investigated the effect of systemic treatments (17). In addition, the lack of head-to-head studies comparing the efficacy and safety profile of immunosuppressive agents (leflunomide, methotrexate, azathioprine, mycophenolate, cyclophosphamide) does not permit a recommendation on the use of one agent over another, except when patient characteristics or comorbidities are considered with respect to the safety profile. Direct and indirect B cell blockade (rituximab, belimumab) seems to be the most promising approach in terms of biological therapies for the management of Sjögren syndrome, although other therapeutic targets are under investigation. However, the disappointing results of two large randomized controlled trials have opened a debate about rituximab use in Sjögren syndrome. These trials used a composite outcome based on the subjective evaluation of dryness, fatigue, and pain. The strong influence of personal and environmental factors on the intensity of this triad of symptoms could explain the lack of significant differences. In addition, inadequate patient selection, the influence of concomitant drugs, and the heterogeneity of diagnostic tests could have contributed to the lack of significant differences between placebo and therapeutic groups (17).

As a summary about the therapeutic management of systemic Sjögren syndrome, the 2020 EULAR guidelines recommend to limit the use of systemic therapies (glucocorticoids, antimalarials, immunosuppressive agents, intravenous immunoglobulins, and biologics) to patients with active systemic disease, always after a careful organ-by-organ evaluation of both severity and organ damage and following a schedule consisting of a 2-stage sequential regimen: first, an intensive immunosuppressive approach targeted to restore organ function as soon as possible (induction of remission), and a second therapeutic course aimed at maintaining the initial therapeutic response (maintenance of remission).

Isolated studies are exploring other drugs, such as fingolimod (an immunomodulatory drug) in patients with multiple sclerosis­like CNS involvement and bortezomib (59; 122).

Vitamin B12 (cobalamin) deficiency. Andres and colleagues documented the common occurrence of vitamin B12 deficiency in 80 patients with primary Sjögren syndrome (06). Serum B12 levels below 200 pg/mL were found in 8.8% (7/80) of patients, and 56.2% of patients had B12 levels between 200 and 300 pg/mL, for an overall prevalence of B12 deficiency of 65%. In comparison, they found B12 deficiency in 5.3% of patients in an Internal Medicine clinic. In the general population, B12 deficiency occurs in 15% of people over 60 years of age. Cobalamin is first released from proteins in food by pepsin and stomach acid; then cobalamin is bound to the salivary vitamin B12 R-binder protein before it can be attached to intrinsic factor. Andres and colleagues postulated that B12 deficiency from food-cobalamin malabsorption is the result of a lack of saliva, which is typical of Sjögren syndrome (08).

Vitamin B12 deficiency causes a number of neurologic manifestations (108) that could worsen the neurologic complications of Sjögren syndrome. The most common manifestations of cobalamin deficiency include: peripheral sensory neuropathy; subacute combined degeneration of the spinal cord presenting with sensory ataxia and pyramidal tract involvement with bilateral Babinski sign (07); cerebellar syndromes; cranial nerves neuropathies, including optic neuritis and optic atrophy; urinary or fecal incontinence; stroke and atherosclerosis from hyperhomocysteinemia with cognitive decline or dementia (42); parkinsonian syndromes; and depression. In practice, it is advisable to exclude and treat B12 (cobalamin) deficiency in patients with Sjögren syndrome and psychiatric or neurologic symptoms, including the several types of neuropathy and myelopathy described.

The treatment of neurologic complications in Sjögren syndrome is dictated by the clinical symptomatology, the clinical course, and the implicated pathogenetic mechanism.

Peripheral nervous system. The common symmetric, distal axonal sensory and sensorimotor neuropathy is believed to be caused by perivascular cellular infiltration and necrotizing vasculitis. These axonal neuropathies usually follow a slowly progressive and insidious course that is often treated with the usual symptomatic therapy of Sjögren syndrome (salicylates, nonsteroidal agents, hydroxychloroquine, oral corticosteroids). IVIg has been shown to improve sensorimotor or nonataxic sensory neuropathies, but not in ataxic neuropathies (106). In cases with mononeuritis multiplex, multiple cranial neuropathies, pseudo-amyotrophic lateral sclerosis forms, or with lesions suggestive of severe vasculitis, first-line treatment appears to be intravenous corticosteroid therapy; however, when the patient’s course fails to improve or deteriorates, a nonsteroidal immunosuppressant agent should be considered (81). Patients with vasculitis- or cryoglobulinemia-related neuropathies have had good response to rituximab treatment (80).

The sensory ataxic ganglionopathy with ganglion neuron antibodies and the demyelinating polyradiculoneuropathy of Sjögren syndrome respond poorly to triple treatment with prednisolone, cyclosporine, and cyclophosphamide (64). There is a case report of a patient with sensory neuronopathy treated with infliximab (22). There was marked improvement in clinical and neurophysiologic deficits associated with the neuronopathy using a dose of 3 mg/kg given at weeks 0, 2, 6, and every 12 weeks thereafter. Plasmapheresis and rituximab may produce improvement, and there is occasional response to treatment with IVIG (0.4 g/kg for 5 days) (57). Yamada and colleagues reported excellent results with IFN-alpha treatment (3 MIU/day, 3 times weekly), resulting also in marked improvement of the clinical and laboratory manifestations of Sjögren syndrome (146). This treatment could potentially be useful also for predominantly autonomic neuropathies with elevated titers of ganglionic acetylcholine receptor autoantibody, trigeminal neuropathy, and for the painful neuropathy without sensory ataxia of Sjögren syndrome.

The EULAR-Sjögren Syndrome Task Force Group developed a therapeutic algorithm for first-, second-, and third-line treatments (102). They recommended general sequential use of the three main categories of immunosuppressive agents in Sjögren syndrome based on a similar approach to that reported for other systemic autoimmune disease, such as systemic lupus erythematosus or vasculitis, with no controlled studies supporting this approach in Sjögren syndrome. The immunosuppressive approaches traditionally used in multineuritis and axonal polyneuropathy (predominantly vasculitis related) are high dose of glucocorticoids as first-line treatment and oral immunosuppressive agents or rituximab as second-line. In refractory cases, cyclophosphamide or plasma exchanges are recommended as rescue therapy. Azathioprine and mycophenolate mofetil may be used in lieu of cyclophosphamide or following cyclophosphamide as “remission maintenance” agents, although neither medication has been evaluated for this purpose in formal clinical trials. In cases in which the axonal polyneuropathy is sensory, first-line treatment will be symptomatic, but in refractory cases, motor involvement, ganglionopathy, and or chronic inflammatory demyelinating polyneuropathy, EULAR Task Force Group recommend intravenous immunoglobulins 0.4 to 2 g/kg for 5 days as first-line treatment and methylprednisolone pulses as second-line or cyclophosphamide pulses 0.5 g/15 day (maximum six pulses) as rescue therapy.

Skeletal muscle. Myalgias in patients with Sjögren syndrome may respond to hydroxychloroquine (6 to 8 mg/kg daily) (39). Additional treatments may include corticosteroids, azathioprine, and methotrexate.

Encephalic manifestations. Meningoencephalitis, focal manifestations, or seizures consistent with vasculitis should be treated with intravenous steroids (55). In severe cases, intravenous corticosteroids may be used in conjunction with cyclophosphamide, rituximab, azathioprine, cyclosporine, or methotrexate (92; 62). Plasmapheresis, or treatments with IVIG, are probably indicated, but there are minimal reported data with these therapeutic modalities (21).

Spinal cord involvement. Rogers and colleagues have reviewed this topic (107), but these data are based on anecdotal reports. Nonetheless, it seems clear that treatment of Sjögren syndrome myelopathy with intravenous steroids alone is not adequate (143; 107; 28). Combination therapy with agents such as cyclophosphamide, rituximab, azathioprine, cyclosporine, or methotrexate must be considered.

Cyclophosphamide is the historical agent of choice, given intravenously in pulse dosing (142; 28). Current treatment regimens consist of a monthly intravenous bolus of drug for 6 to 12 months (28); then, if needed, every 3 to 6 months for the second year. Doses used are 0.75 to 1.0 gm/m2. Caution is required due to the high frequency of lymphoma in Sjögren syndrome. The most common severe adverse effects associated with cyclophosphamide are hemorrhagic cystitis and myelosuppression. Hemorrhagic cystitis can be prevented by adequate hydration and use of mercapto ethane sulfonate sodium, a sulfhydryl compound that binds to the toxic metabolite formed by cyclophosphamide, thus, preventing it from binding to the bladder wall and causing damage. Myelosuppression may be dose-limiting, with the white blood cell nadir occurring 7 to 10 days after treatment. Owing to the aforementioned side effects, the use of mycophenolate mofetil is being explored as an alternative to cyclophosphamide in the treatment of vasculitis. Rituximab is being used more often now because of the more favorable side effect profile and similar efficacy to cyclophosphamide.

Azathioprine has been used for relapses or intolerance to cyclophosphamide therapy (52). In uncomplicated Sjögren syndrome, azathioprine by itself was ineffective in a dose of 1 mg/kg/d for 6 months. The most serious side effects are dose-related myelosuppression, gastrointestinal disturbances, stomatitis, and hepatotoxicity.

Cyclosporine in combination with corticosteroids may be effective in the treatment of optic neuritis. Cyclosporine is a potent immunomodulator that decreases cytokines involved in T-cell activation and has direct effects on B-cells and macrophages. The onset of effect of cyclosporine is 1 to 3 months. It is only available in an oral dosage form. Serious adverse effects are hypertension, hyperglycemia, nephrotoxicity, tremor, gastrointestinal intolerance, hirsutism, and gingival hyperplasia.

Methotrexate at a dose of 0.2 mg/kg/week tends to improve subjective oral and ocular symptoms, without objective changes. It is an antineoplastic and antiinflammatory agent. Serious adverse effects include thrombocytopenia, gastrointestinal intolerance, stomatitis, hepatotoxicity, and pulmonary toxicity.

Other nonsteroidal immunosuppressant agents should be considered, especially when lack of efficacy or intolerance to cyclophosphamide appears. Other symptoms may benefit from concomitant immunosuppressant treatment. In addition, plasmapheresis appears to be useful when acute relief of symptomatology is needed, such as in a patient with rapid deterioration who is waiting for an immunosuppressant to work. IVIG has been successfully used in Sjögren CNS vasculitis (21). The dose used in the patient was 400 mg/kg/d for 5 days along with corticosteroids, with additional monthly doses of IVIG for 6 months and then every 2 to 5 months as indicated by the patient’s neurologic symptoms.

Rituximab, a chimeric monoclonal antibody directed against the B cell differentiation marker CD20, has been used with some success in the treatment of primary Sjögren syndrome (58; 02). Most of the evidence consists of case reviews for treatment of extraglandular manifestations of Sjögren syndrome. To our knowledge, there are no randomized placebo-controlled clinical trials of rituximab in the treatment of neurologic manifestations associated with Sjögren syndrome.

The EULAR-Sjögren Syndrome Task Force Group therapeutic algorithm for central nervous system involvement is based on the type of neurologic manifestation (102). The authors suggested that corticosteroids plus cyclophosphamide could be the first- and second-line therapeutic approach in patients with CNS vasculitis and neuromyelitis optica spectrum disorder but that corticosteroids should be rapidly tapered with the aim of decreasing the risk of severe infection. Rituximab or/plus plasma exchanges may also be recommended for the treatment of this manifestation, especially in patients with severe or refractory disease, but its use is currently limited by the lack of US Food and Drug Administration or European Medicines Agency approval for these indications. When lymphocytic meningitis does not respond to symptomatic treatment, the same therapeutic algorithm should be followed.

Unfortunately, no controlled data were identified to support a differentiated organ-guided therapeutic approach for multiple sclerosis-like manifestations in Sjögren syndrome, and EULAR Task Force Group recommend specific multiple sclerosis therapies, with an individualized therapeutic approach being preferable.

Outcomes

Although today the true prevalence of CNS and PNS involvement in patients with Sjögren syndrome is still not known exactly, the analysis of the main studies carried out in the past 25 years suggests a figure between 5% and 10%. It is important to highlight the great variability of its neurologic symptoms and signs, which means that it can be confused with other neurologic diseases with a similar clinical spectrum. In clinical practice, the spectrum of neurologic manifestations in patients with Sjögren syndrome is broad, ranging from asymptomatic forms with the exclusive finding of demyelinating lesions in the white matter to forms of focal or diffuse cerebral involvement of spinal cord or epilepsy. The situations that most frequently induce greater complexity in the differential diagnosis of CNS involvement in patients with Sjögren syndrome are cerebrovascular disease and demyelinating lesions. There is no standardized treatment for CNS involvement in Sjögren syndrome. It will fundamentally depend on the type of neurologic involvement, the symptoms, and the possible coexistence of other associated processes. In severe forms of neurologic involvement, intravenous immunosuppressive treatment with pulses of methylprednisolone and cyclophosphamide should be used to reduce the number of neurologic sequelae; in refractory or life-threatening cases, the use of immunoglobulins, plasma exchanges, and, above all, the possible future role of biological therapies directed against B lymphocytes, such as rituximab, should be considered.

Special considerations

Pregnancy

In a questionnaire-based study, primary Sjögren syndrome had no impact on pregnancy outcome before disease onset (49). The most important condition associated with primary Sjögren syndrome in anti-SS-A-positive mothers was congenital heart block in the offspring. Autoimmune congenital heart block is an immune-mediated acquired disease that is associated with the placental transference of maternal antibodies specific for Ro and La autoantigens. The disease develops in a fetal heart without anatomical abnormalities that could otherwise explain the block; it is usually diagnosed in utero but also at birth or within the neonatal period. Autoantibody-mediated damage of fetal conduction tissues causes inflammation and fibrosis and leads to blockage of signal conduction at the atrioventricular node. Irreversible complete atrioventricular block is the principal cardiac manifestation of congenital heart block, although some babies might develop other severe cardiac complications, such as endocardial fibroelastosis or valvular insufficiency, even in the absence of cardiac block (18). This condition is associated with high morbidity and mortality, and these patients should receive multidisciplinary care in a neonatal intensive care unit (96).

Childhood-onset primary Sjögren syndrome

Pediatric onset of the disease is rarely reported, and there is no information on the frequency of pediatric presentation. In addition, a clear view of how the disease presents in children is difficult to obtain due to the scarce and heterogeneous available data. The main series that are currently published are characterized by a heterogeneous methodology using a variable definition of the pediatric age (from 14 to 18 years) and include up to 50% of patients who do not fulfill the current classification criteria as well as a mix of primary and associated forms of the disease (102).

Among the 12,083 patients included in the Sjögren Big Data Registry, 158 (1.3%) patients were diagnosed at an age younger than 19 years. There were 136 (86%) girls and 22 (14%) boys, with a mean age at first sign or symptom suggestive of the disease of 13.2 (SD 3.2) years and of 14.2 (SD 3.5) years at the time of diagnosis of primary Sjögren syndrome (100). With respect to systemic disease, the highest mean ESSDAI score was observed in patients with childhood onset, who also showed the highest frequencies of systemic disease in five (constitutional, lymphadenopathy, glandular, cutaneous, and hematological) out of the 12 ESSDAI domains and the lowest frequencies in four domains (articular, pulmonary, peripheral nerve, and central nervous system). A younger diagnosis was associated with an increased risk of presenting activity at diagnosis in some clinical domains (constitutional, lymphadenopathy, glandular, cutaneous, renal) but a decreased risk of presenting activity in others (pulmonary and neurologic domains). Although the nervous system was the organ-specific ESSDAI domain reported less frequently, some isolated cases of neurologic involvement have been reported in children with primary Sjögren syndrome (43). They mainly consisted of cerebellar ataxia and acute encephalitis, which are features that are not included in the neurologic ESSDAI domain. Some of these cases were even reported with other concomitant neurologic disorders (mental retardation, psychosis) (123; 79; 50).