Although peripheral neuropathy is the most common extraglandular manifestation of Sjögren’s syndrome, treatment of this complication is not well established.¹ Interferon alfa (IFNα) administration has been reported to alleviate Sjögren’s syndrome associated sicca symptoms, as evidenced by increased salivary flow, and also to reduce histologically evident disease activity.² So far, the effects of IFNα on any extraglandular complications such as Sjögren’s syndrome associated neuropathy (SSN) have not been reported. We describe the therapeutic effects of IFNα in three consecutive patients with two forms of SSN—two with sensory ataxic ganglionopathy and one with sensorimotor neuropathy.

CASE PRESENTATIONS

Patient 1

An otherwise healthy 46 year old man developed dysaesthesia which first involved the left foot and then progressed to affect the right foot and left hand in September 1997, finally progressing to his right foot and left hand. Over the next three years, the level of dysaesthesia gradually ascended to involve both thighs, and difficulty in walking led to hospital admission under our care. On admission, sensory examination revealed profoundly reduced vibratory and proprioceptive sensation affecting mainly the lower limbs, with slight loss of pain and temperature sensation. The heel to knee test showed marked dysmetria in both legs, particularly with the eyes closed. Deep tendon reflexes were absent in all limbs. Muscle strength was slightly decreased in the lower limbs. The patient could barely walk unassisted because of severe ataxia (sensory ataxia scale 5, table 1).³ Romberg’s test was strongly positive. No signs of autonomic dysfunction were evident. Routine haematological examination yielded normal results. Anti-SS-A/SS-B antibodies were positive at 42.0/18.0, respectively (by enzyme linked immunosorbent assay (ELISA); normal values are SS-A/SS-B <10.0/15.0). In cerebrospinal fluid (CSF), protein content was modestly raised (60 mg/dl), and the cell count was normal. Results of nerve conduction studies were normal in the upper limbs, but in the lower limbs the sensory nerve action potential (SNAP) in the sural nerve showed reduced amplitude (0.2 μV) with normal conduction velocity (CV, 48 m/s). Other nerve conduction study findings, as well as statokinesigraphy results⁴, are summarised in table 1. Somatosensory evoked potentials (SEP) could not be elicited by lower limb stimulation. Cervical spinal cord magnetic resonance imaging (MRI) on T2* weighted images showed abnormal high intensity areas in the dorsal columns, reflecting the sensory ataxic ganglionopathy.⁵ Findings on sural nerve biopsy were decreased numbers of large myelinated fibres, axonal degeneration without axonal sprouting, endoneurial oedema, and no evidence of vasculitis. The patient was treated with prednisolone (30 to 10 mg/day), cyclosporin (100 mg/day), and plasmapheresis, but without improvement. In August 2001, intravenous immunoglobulin treatment (IVIG) was given (0.4 g/kg for five days). After this treatment, dysaesthesias in the legs and left hand were reduced, and the patient was able to walk with less effort. However, he required repeated five day courses of IVIG every three to four weeks to halt the progression of the disease. Sicca symptoms developed and Schirmer’s test gave a positive result (3 mm/5 mm, right/left). A labial salivary gland biopsy specimen showed marked lymphocytic infiltration and acinar cell destruction, graded as 3 by Daniels focus scores.⁶

In November 2003, IFNα treatment was initiated (3 MIU/day, three times weekly). Over the next two months, the patient showed dramatic improvement; dysaesthesias nearly disappeared and he was able to walk without effort. Nerve conduction studies revealed improvement of SNAP amplitude in the sural nerve (table 1). Statokinesigraphy also showed significant improvement (p<0.01, table 1). Sicca symptoms resolved and lacrimation increased (Schirmer’s test; 18 mm/14 mm, right/left). Anti-SS-A/SS-B antibody titres fell to the normal range (9.1/9.2 respectively) and a follow up labial salivary gland biopsy showed fewer infiltrating lymphocytes graded as 2.⁶ The clinical and therapeutic time course of patient 1 is summarised in fig 1A.

Patient 2

A 67 year old woman with Sjögren’s syndrome developed sensory ataxic ganglionopathy over 14 years. She did not respond to prednisolone, cyclophosphamide, or plasmapheresis and required frequent (every two or three month) IVIG therapy to maintain her activities of daily living, as for patient 1. In December 2003, she was admitted to our department and treated with IFNα (3 MIU/day, three times weekly). After the initiation of IFNα, she had marked improvement in vibratory and proprioceptive sensation, leading to improvement of her activities of daily living. Nerve conduction studies showed improvement in SNAP amplitude in the sural nerve and statokinesigraphy demonstrated significantly improved balance, as in patient 1. Her clinical and neurophysiological features are summarised in fig 1B and table 1.

Patient 3

A 45 year old woman with a history of hypothyroidism developed progressive weakness and dysaesthesias in both feet in December 2000. Within six weeks she became unable to walk, and was admitted to a hospital affiliated with Nagoya University. Neurological examination indicated mild weakness, mild loss of both positional and vibratory sensation, and slight loss of pain and temperature sensation involving all limbs, especially distally. Deep tendon reflexes were absent in all limbs. No autonomic symptoms were present. Routine haematological findings were normal. Serum anti-SS-A antibody was positive while anti-SS-B antibody was negative (anti-SS-A/SS-B 47.0/8.2, respectively). CSF protein content was raised (124 mg/dl), with a normal cell count. A labial salivary gland biopsy specimen showed marked lymphocytic infiltration graded as 4 (focus 3).⁶ Nerve conduction studies showed a symmetrical sensorimotor polyneuropathy with reduced conduction velocities and the presence of temporal dispersion (table 1). The patient was treated with prednisolone (60 mg/day), cyclosporin (100 mg/day), and cyclophosphamide (100 mg/day) with no effect on the progression of disability. Plasmapheresis resulted in a slight improvement in activities of daily living lasting less than two weeks. After IVIG treatment (0.4 g/kg for five days), she had marked clinical improvement; dysaesthesias and weakness in all limbs gradually lessened, and she could walk without support. However, there were multiple relapses during the next two years. Beginning in April 2002, intervals between relapses shortened and sicca symptoms developed. A sural nerve biopsy specimen revealed subperineurial and endoneurial oedema with evidence of remyelination.

In June 2003, IFNα treatment (3 MIU/day, three times weekly) was started. Within 30 days, dysaesthesias and weakness nearly disappeared. After eight weeks, nerve conduction studies showed slight improvement (table 1). On follow up labial salivary gland biopsy specimen there were significantly fewer infiltrating lymphocytes graded as 2,⁶ and sicca symptoms resolved. The serum anti-SS-A/SS-B antibody titres also fell to within the normal range (3.1/2.6, respectively). The clinical and therapeutic time course of patient 3 is summarised in fig 1C.

DISCUSSION

Sjögren’s syndrome associated neuropathy includes a wide spectrum of manifestations such as sensory ataxic ganglionopathy, sensorimotor polyneuropathy, demyelinating polyradiculoneuropathy, multiple cranial neuropathy, and vasculitic neuropathy.⁷

Sensory ataxic ganglionopathy often develops in patients with Sjögren’s syndrome and is characterised by severe impairment of kinaesthetic sensation with no obvious motor involvement.⁸ This form of neuropathy is chronic and progressive, occasionally responding to treatment with IVIG.⁹ In our patients with this type (patients 1 and 2), IVIG treatment partially lessened the neuropathic symptoms, but repeated courses were needed and they did not improve the overall status of the Sjögren’s syndrome.

Previous reports have indicated that demyelinating polyradiculoneuropathy sometimes develops in patients with Sjögren’s syndrome, and have shown that the concurrence of Sjögren’s syndrome and demyelinating polyradiculoneuropathy is not coincidental but reflects a common underlying immunological derangement.⁶ We diagnosed patient 3 as having SSN with mainly demyelinating features (demyelinating polyradiculoneuropathy), on the basis of the clinical features and the findings on nerve conduction studies: reduced conduction velocities, presence of temporal dispersion, and histopathological evidence of remyelination.

Use of IFNα for Sjögren’s syndrome first was described in 1993.¹⁰ Since then, orally administered IFNα has been reported to be effective for the sicca symptoms of this condition, resulting in significant increases in salivary gland function and histological improvement in specimens from minor salivary glands.²

Before treatment with IFNα, all of our three patients had positive serum anti-SS-A/SS-B antibodies, characteristic salivary gland histopathological findings, and two of the following features: abnormal Schirmer’s test result, oral symptoms, or ocular symptoms. Thus they fulfilled the diagnostic criteria of the American-European Consensus Group for Sjögren’s syndrome.¹¹ After treatment, anti-SS-A/SS-B antibody titres fell dramatically to within the normal range in all patients, salivary gland lymphocytic infiltration decreased in patients with follow up specimens (1 and 3), and sicca symptoms resolved in all patients.

To our knowledge, this is the first report to show beneficial therapeutic effects of IFNα on SSN. The mechanism whereby IFNα induced marked improvement in SSN as well as in Sjögren’s syndrome itself in our patients is uncertain, but could be related to its immunomodulating effects. As IFNα caused neurological improvement in patients with two different forms of SSN, these two forms appear likely to share a common immunopathogenic mechanism responsive to IFNα treatment, irrespective of the specific form of SSN. However, as our three patients all had chronic, progressive relapsing neuropathies that responded to treatment with IVIG, IFNα effects might conceivably reflect IVIG responsive neuropathic mechanisms, even though Sjögren’s syndrome itself did not respond to IVIG.

Trials of IFNα in a variety of forms of SSN are needed to determine whether IFNα therapy represents a first line treatment.